Académique Documents
Professionnel Documents
Culture Documents
T I T L ES IN T H E SE R IES
Live Long and Prosper: Aging in East Asia and Pacific
Toward Gender Equality in East Asia and the Pacific: A Companion to the World Development
Report
Putting Higher Education to Work: Skills and Research for Growth in East Asia
1 2 3 4 18 17 16 15
This work is a product of the staff of The World Bank with external contributions. The fi ndings, interpre-
tations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its
Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the
accuracy of the data included in this work. The boundaries, colors, denominations, and other information
shown on any map in this work do not imply any judgment on the part of The World Bank concerning the
legal status of any territory or the endorsement or acceptance of such boundaries.
Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and
immunities of The World Bank, all of which are specifically reserved.
This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) http://
creativecommons.org/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free
to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following
conditions:
Attribution—Please cite the work as follows: World Bank. 2016. Live Long and Prosper: Aging in
East Asia and Pacifi c. Washington, DC: World Bank. doi:10.1596/978-1-4648-0469-4. License: Creative
Commons Attribution CC BY 3.0 IGO
Translations—If you create a translation of this work, please add the following disclaimer along with the
attribution: This translation was not created by The World Bank and should not be considered an offi cial
World Bank translation. The World Bank shall not be liable for any content or error in this translation.
Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the
attribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed
in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed
by The World Bank.
Third-party content—The World Bank does not necessarily own each component of the content con-
tained within the work. The World Bank therefore does not warrant that the use of any third-party-owned
individual component or part contained in the work will not infringe on the rights of those third parties.
The risk of claims resulting from such infringement rests solely with you. If you wish to re-use a component
of the work, it is your responsibility to determine whether permission is needed for that re-use and to obtain
permission from the copyright owner. Examples of components can include, but are not limited to, tables,
figures, or images.
All queries on rights and licenses should be addressed to the Publishing and Knowledge Division, The
World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@
worldbank.org.
Cover photo: © Qilai Shen / Panos Pictures. Used with the permission of Qilai Shen / Panos Pictures. Further
permission required for reuse.
Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv
Acknowledgments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii
Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xix
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Understanding the rapid demographic and epidemiological transition . . . . . . . . . . . . . . . . . 3
Situation and sources of support of older people . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Managing the effect of aging on growth and its drivers . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Developing pension, health, and long-term care systems for aging societies . . . . . . . . . . . . 13
Addressing cross-cutting challenges in aging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
I The Demographic Transition and Well-Being of Older People in East Asia
and Pacific. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
1. The Demographic and Epidemiological Transitions in East Asia and Pacific . . . . . . . . . . . 23
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
The demographic transition in East Asia and Pacific . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
The epidemiological transition in East Asia and Pacific . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
2. The Welfare and Labor Market Behavior of Older People in East Asia and Pacific . . . . . . 47
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Poverty among older people in East Asia and Pacific . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Other measures of well-being . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Living arrangements and access to support among older people . . . . . . . . . . . . . . . . . . . . . 54
v
vi CONTENTS
Boxes
1.1 Uncertainty and assumptions in United Nations population data . . . . . . . . . . . . . . . . 30
1.2 Alternative measures of dependency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
1.3 Is longevity accompanied by more good health or less? . . . . . . . . . . . . . . . . . . . . . . . . 35
1.4 Very high prevalence of smoking and related death rates in the region . . . . . . . . . . . . 41
1.5 Indonesia: Increasing and often unequal exposure to risks . . . . . . . . . . . . . . . . . . . . . 41
1.6 Malaysia: Undiagnosed noncommunicable disease risks and modest social
economic status differentials in management of risk . . . . . . . . . . . . . . . . . . . . . . . . . . 42
1.7 Adverse childhood socioeconomic conditions and increased mortality
risks later in life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
3.1 The scope for improving skills of an aging workforce . . . . . . . . . . . . . . . . . . . . . . . . . 88
3.2 A labor productivity paradox? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
4.1 Fertility and public policy in East Asia and Pacific and the OECD . . . . . . . . . . . . . . 113
4.2 The lump of labor fallacy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
4.3 Employment services targeted to older workers in Japan and the Republic of Korea. . . 117
4.4 Reforming seniority wage systems in the Republic of Korea, Singapore, and Japan . . . 120
4.5 East Asia and Pacific’s distinctive approach to retirement age and continued work . . . . 121
4.6 Assumptions underlying labor force scenarios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
4.7 Second Basic Employment Promotion Plan (2012–16) for the aged in
the Republic of Korea. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
5.1 Automatic adjustments for changes in life expectancy in defined benefit schemes . . . 170
5.2 Integrating noncontributory and contributory pensions: Chile’s solidarity pillar. . . . 174
6.1 Poor qualifications of staff to provide effective diagnosis and treatment . . . . . . . . . . 190
6.2 Limited supply-side readiness to manage and treat noncommunicable
diseases at the primary health care level in Indonesia . . . . . . . . . . . . . . . . . . . . . . . . 192
6.3 The role of technological change and interactions with aging in health
care cost escalation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
viii CONTENTS
Figures
O.1 East Asia and Pacific has more older people than any other region . . . . . . . . . . . . . . . . 3
O.2 East Asian and Pacific economies are aging more rapidly than
economies elsewhere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4
O.3 Three distinct aging patterns are apparent in East Asia and Pacific . . . . . . . . . . . . . . . . 5
O.4 In the richer economies of East Asia and Pacific, the working-age population
is projected to shrink between 2010 and 2040. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
O.5 Labor force participation rates in East Asia and Pacific vary, but across
the region, rural people work significantly longer than urban people . . . . . . . . . . . . . . 7
O.6 Work remains the primary source of old-age support in most countries in
East Asia and Pacific . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
O.7 Attitudes on the preferred source of fi nancial support in old age and on the expected
source of personal care are shifting in East and Southeast Asian economies . . . . . . . . . 9
O.8 East Asia and Pacific’s developing economies are getting old before getting rich . . . . . 10
O.9 Greater participation by women, older people, and migrants can mitigate the
projected decline in the labor force in richer East and Southeast Asian economies . . . 12
O.10 Across APEC, pensions will eat up a larger and larger share of GDP in
coming decades without reforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.1 East Asia and Pacific is a diverse region in terms of the share of older people . . . . . . . 24
1.2 East Asia, led by China, has more old people than any other region . . . . . . . . . . . . . . 25
1.3 East Asian and Pacific economies are aging more quickly than other
economies in the past . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
1.4 The share of elderly population in East Asia and Pacific will converge
with that of OECD countries by 2060 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
1.5 The population of “oldest old” will increase more in East Asia and
Pacific between 2010 and 2060 than in any other region . . . . . . . . . . . . . . . . . . . . . . . 27
1.6 Three distinct aging patterns are apparent among East Asian and
Pacific economies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
1.7 People in aging East Asian and Pacific countries see aging as a major
challenge for their countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
1.8 Fertility rates have declined significantly across all East Asian and
Pacific economies but are projected to flatten in orange and red economies. . . . . . . . . 29
B1.1.1 UN population projections for East Asia and Pacific have been subject to
significant uncertainty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
CONTENTS ix
1.9 Declines in age-specific mortality rates in East Asia and Pacific in recent decades
have been substantial, particularly in the early years of life and for women . . . . . . . . 31
1.10 The typology of East Asian and Pacific economies is robust in scenarios
with low-, medium-, and high-fertility assumptions. . . . . . . . . . . . . . . . . . . . . . . . . . . 31
1.11 The working-age population will shrink over coming decades in richer
East Asian and Pacific economies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
1.12 Total dependency ratios conceal contrasting trends in youth and elderly
population shares . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
1.13 Life expectancy and health-adjusted life expectancy in East Asia and
Pacific both increased between 1990 and 2010 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
1.14 Large differentials in healthy years of life are evident across the region,
but men generally live shorter, less healthy lives than women . . . . . . . . . . . . . . . . . . . 37
1.15 NCDs have increased sharply as leading causes of disease burden in
East Asia and Pacific in recent decades . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
1.16 NCDs and injuries dominate the leading causes of mortality in China . . . . . . . . . . . . 39
1.17 In the Republic of Korea both men and women 65 years of age and older
are more likely to have multiple chronic conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
1.18 In China, in 2006, socioeconomic status had an inverse relationship
with noncommunicable diseases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
1.19 NCD-related risk factors grew in East Asia and Pacific between 1990 and 2010 . . . . 40
2.1 Poverty rates increase with age in some East and Southeast Asian countries,
but elderly poverty has fallen over time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.2 Poverty among the rural elderly remains significantly higher than for urban elderly . . . . 49
2.3 Difficulties with physical function and instrumental activities of daily
living increase with age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
2.4 Poorer and less educated elderly are more likely to face difficulty
performing instrumental activities of daily living . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
2.5 Rates of depression tend to increase with age and are higher among
poor households . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
2.6 Elderly co-residence rates are generally high in East Asia and Pacific but
vary across countries and by gender and location within countries . . . . . . . . . . . . . . . 55
2.7 Co-residence rates of older adults with adult children have declined
over time in China and Thailand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
2.8 As countries get richer, elderly co-residence rates tend to decline. . . . . . . . . . . . . . . . . 56
2.9 Labor is the leading income source for people ages 60 to 85 in various
East and Southeast Asian countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
2.10 Source of fi nancial support can affect poverty rates in various East and
Southeast Asian countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
2.11 The relative roles of public and private transfers and of savings and
assets across the life cycle varies notably across East and Southeast Asian
economies and between East and Southeast Asia and other regions . . . . . . . . . . . . . . 61
2.12 Attitudes on the preferred source of fi nancial support in old age
and on the expected source of personal care are shifting in East and
Southeast Asian economies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
2.13 Labor force participation rates vary by location and gender in various
East and Southeast Asian countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
2.14 Better-educated people tend to withdraw from the labor force earlier in
East and Southeast Asia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
2.15 Self-employment is an increasingly important source of work as people
age in East Asia and Pacific . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
x CONTENTS
2.16 Those who continue to be employed in older age work long hours until
advanced ages in East and Southeast Asia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
2.17 Pension availability and mandatory retirement rules have strong effects on
labor force behavior of older Chinese workers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
2.18 Labor force participation is influenced by pension receipt in select
East and Southeast Asian countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
2.19 Older people in East and Southeast Asia are a significant source of
volunteer work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
3.1 East Asia and Pacific’s developing economies are getting old before getting rich . . . . . 82
3.2 Aging’s influence on GDP per capita is the product of the employment rate,
the working-age share of the population, and GDP per employed worker . . . . . . . . . . 84
3.3 Older women in East Asia and Pacific will be able to work later in life
by virtue of rising health-adjusted longevity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
3.4 Prevalence of stunting among children under age five could undermine
investments in education and affect cognitive skills in East Asian and
Pacific economies of the green group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
3.5 Productivity depends on the age distribution of the population, but
mitigation channels exist . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
3.6 Although in many countries, entrepreneurship seems to decline with
worker age, business ownership in East Asian countries is significant
among older workers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
3.7 Compared with household savings rates in other regions, the rate in
East Asia and Pacific is relatively high and stable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
3.8 Projected increases in pension spending to 2030 point to a significant
increase in China and Malaysia, with more modest increases for Indonesia,
the Philippines, and Thailand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
3.9 Estimated increases in East and Southeast Asian pension spending for 2014 to
2050 fi nd significant fiscal risks from pensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
3.10 Pension spending projections by stylized groupings for APEC economies,
2010–70, illustrate rapid increase in pension spending in a no-reform scenario . . . . . 97
3.11 Pension spending increases can be substantially reduced with key reforms . . . . . . . . . 98
3.12 Projected increases in public spending on health caused by aging, though
significant, are not as substantial as for pensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
3.13 Projected increases in public health spending reflect the importance of
policy choices in sustainability of health care systems . . . . . . . . . . . . . . . . . . . . . . . . . 99
3.14 Projected increases in public long-term care spending indicate additional
fiscal pressure in countries where the aging effect is strong . . . . . . . . . . . . . . . . . . . . . 99
3.15 Schematic overview of the links between demographics and fiscal accounts . . . . . . . 101
3.16 Initial debt and spending levels and future debt dynamics have
important interactions with demographics that can sharply increase
the fiscal impacts of aging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
4.1 The Kurumin mark is given to fi rms in Japan that pursue child-friendly policies . . . 112
4.2 The relationship between women’s work and fertility has reversed sharply
in recent decades in most member countries of the OECD . . . . . . . . . . . . . . . . . . . . 113
4.3 Japanese workers show a desire to work in old age well above
OECD counterparts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
B4.2.1 In OECD countries, higher employment rates of older workers are associated
with modestly higher rates of youth employment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
4.4 Tax force to retire is strongly correlated with labor force participation of
older men in OECD countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
CONTENTS xi
4.5 The potential channels for mitigating labor force decline vary across
aging countries in East and Southeast Asia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
4.6 Greater participation by women, older people, and migrants can mitigate
the projected decline in the total labor force in richer East and Southeast Asian
economies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
4.7 Migration leads to a 5 percent decline in the total labor force size of the
migrants’ origin countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
4.8 Japan and the Republic of Korea show low rates of on-the-job training
for middle-aged and older workers relative to countries such as the
United States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
5.1 East Asian and Pacific economies can be divided into globally average
adopters, late adopters, and global outliers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
5.2 East Asian and Pacific economies currently spend less on pensions than
other economies after taking into account their demographics . . . . . . . . . . . . . . . . . 148
5.3 Coverage of contributory pension schemes is around par globally for
income levels and demography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
5.4 Elderly pension coverage is low in the bottom income deciles in China,
Indonesia, and the Republic of Korea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
5.5 Increasing contributory pension coverage among the poorer segments
of populations is difficult . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
5.6 Elderly pension coverage is higher for men than for women . . . . . . . . . . . . . . . . . . . 153
5.7 A wide range of design choices in the pension benefit is available . . . . . . . . . . . . . . . 153
5.8 In most economies, target replacement rates are significantly higher for
lower-income workers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
5.9 In most economies, actual replacement rates in DB schemes tend to be
lower than would be expected . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
5.10 Real returns on DC accounts and growth of real wages . . . . . . . . . . . . . . . . . . . . . . 155
5.11 Returns on funds in DC accounts in China have barely exceeded inflation and
have lagged far behind wage growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
5.12 Social pension benefits are modest in East Asia and Pacific . . . . . . . . . . . . . . . . . . . . 156
5.13 Older economies have higher pension contribution rates than do younger economies . . . 157
5.14 Retirement ages in national pension schemes are relatively low in East Asia
and Pacific compared with life expectancy at age of retirement . . . . . . . . . . . . . . . . . 159
5.15 Ability to work at older ages is not closely linked to labor force
participation rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
5.16 Pension spending is closely linked to labor force participation rates . . . . . . . . . . . . . 160
5.17 Only higher-income economies have pension assets that are large
relative to GDP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
5.18 Pension agencies in East Asia and Pacific generally have conservative
investment practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
5.19 Except in Japan, contribution rates required to pay for defi ned benefits
are significantly higher than actual rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
5.20 Retirement ages for men and women are converging in OECD countries . . . . . . . . . 170
5.21 Expanding coverage has become difficult in recent years . . . . . . . . . . . . . . . . . . . . . . 171
5.22 The increase in coverage of contributory pension systems in East Asia and
Pacific is unlikely to keep pace with population aging . . . . . . . . . . . . . . . . . . . . . . . . 172
5.23 In the Republic of Korea, a dramatic change in coverage took place . . . . . . . . . . . . . 173
B5.2.1 Design of the Chilean pension system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
6.1 Health care in East Asia and Pacific faces major health system and service
delivery challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
xii CONTENTS
6.2 Hospital share of health expenditures is often greater than nonhospital share . . . . . 187
6.3 Tertiary hospitals experienced the fastest growth of inpatient and
outpatient visits in China . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
6.4 Average length of stay in East Asian and Pacific acute care hospitals is
higher than or comparable to OECD averages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
6.5 Inpatient admission rates increase with mean age of household . . . . . . . . . . . . . . . . . 189
6.6 Health share of GDP has been rising in East Asia and Pacific with
income growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
6.7 Age-specific expenditure profi le exhibits J-shape for Japan . . . . . . . . . . . . . . . . . . . 197
6.8 Spikes in death rates vary with the speed of aging across East Asia and Pacific . . . . . 197
6.9 Years lived with disability are increasing in East Asia and Pacific as life
expectancy increases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
6.10 Total and public health spending has increased more rapidly than GDP
in much of East Asia and Pacific in the 1990s and 2000s . . . . . . . . . . . . . . . . . . . . . 199
6.11 Age-specific expenditure profi les vary for OECD countries . . . . . . . . . . . . . . . . . . . . 200
6.12 The rising tide of cancer in East Asia and Pacific . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
6.13 A health system is transformed from a pyramidal model into a primary
care–driven coordinated care model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
6.14 PHARMAC has positively affected drug expenditure over time in New Zealand . . . 218
B6.8.1 National Health Security Office’s decision process for new technologies
involves extensive criteria in Thailand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
7.1 Long-term care users in OECD countries are expected to increase with
the aging population . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
7.2 In China, ADL and IADL disabilities may increase rather than decline
as in OECD countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
7.3 Care gaps are greater for the rural elderly, the female elderly, and those
living in poverty in East Asian and Pacific countries . . . . . . . . . . . . . . . . . . . . . . . . . 230
7.4 Long-term care expenditures as a share of GDP are relatively low but
growing faster than health expenditures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
7.5 Long-term care services include both social and health care provided
in a variety of settings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
B7.1.1 Average length of stay in long-term care hospitals in the Republic of
Korea has declined only marginally . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
7.6 The health and long-term care systems in England are complex and
riddled with perverse incentives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
Tables
2.1 Lower elderly poverty rates when older people receive pensions . . . . . . . . . . . . . . . . . 50
2.2 Private transfers to elderly people are not significantly affected by level of
transfers from public sources in East and Southeast Asia. . . . . . . . . . . . . . . . . . . . . . . 60
2.3 Poor health results in lower probability of work at older ages in East and
Southeast Asia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
2.4 Hours spent providing care for children and elderly among Chinese
cohorts ages 45 and above . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
2.5 Influence of employment of a spouse on probability of own employment . . . . . . . . . . 73
3.1 The educational attainment of future generations in East Asia and
Pacific will be considerably higher than for the current working-age
population (secondary and above) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
CONTENTS xiii
A
s a region, East Asia and Pacific is and its citizens, which will influence policy
aging rapidly. The region is home to responses.
over a third of the global population This report finds that East Asia and Pacific
ages 65 and older—mostly in China—and is well positioned to manage the risks from
to more old people than any other region. aging. First, people in the region already have
More significantly, the region is aging more long working lives. Second, entitlements to
rapidly than any region in history—a trend pensions, health care, and long-term care
driven both by sharp declines in fertility and for most of the population are still modest,
by steady increases in life expectancy—and and there is scope to act now to put in place
in many countries, aging is occurring at systems that can be sustainable in the future.
relatively low-income levels. Nevertheless, Third, household savings in the region are
the pattern and pace of aging across East already high, and people tend to save until
Asia and Pacific is diverse, ranging from later in life, suggesting that they may be
aged richer countries such as Japan and the better prepared for old age. Finally, people in
Republic of Korea; to rapidly aging middle- East Asia and Pacific have in recent decades
income countries such as China, Thailand, seen a steady increase in the number of years
and Vietnam; to younger and poorer coun- lived in good health.
tries such as the Lao People’s Democratic The key message of the report is that
Republic, which will only start to age rapidly it will be possible to manage rapid aging
two or three decades from now. in East Asia and Pacific while sustaining
Across these countries, aging raises economic dynamism. This effort will require
policy challenges and economic and social politically difficult policy choices, including
risks. Policy makers express concerns about dealing with associated fi scal risks. Several
the impact that shrinking and aging labor complementary policy reforms are required
forces will have on economic growth. More to manage these risks. In the labor market,
immediately, there are fiscal challenges to a range of measures are needed (a) to encour-
providing sustainable health, pension, and age increased female labor force participa-
long-term care systems in the face of rapid tion while stemming the sharp decline in
aging. Aging is occurring in a wider context fertility, (b) to ensure that older workers in
of shifting relationships between the state the formal sector do not retire too early as
xv
xvi FOREWORD
a result of social security and wage-setting that are already experiencing or are about
systems, (c) to accommodate older workers in to experience rapid aging. In the meantime,
the workplace through greater use of flexible countries with younger populations will need
work arrangements, and (d) to open up aging to prepare now for rapid aging in the future
labor markets to greater inflows of young by setting up sustainable health and pension
immigrants. systems while maximizing the demographic
Pension systems also need reform. Formal dividend from their still-expanding labor
sector pension schemes will need to be forces.
more fiscally sustainable, which will require As we have seen in the richer economies
reforms such as gradual increases in the in East Asia and Pacific and elsewhere in
retirement age. Such changes are needed to the world, the political economy of reform
create fiscal space for expanding pension in aging societies can be difficult. Societal
coverage to the large and currently uncovered expectations change rapidly, especially as
informal sector—an effort that will require incomes grow, and people expect more sup-
greater use of noncontributory approaches port and better services from the state in
such as social pensions. old age. Managing aging is not only about
In the health sector, the impacts of aging “old people” or about “the old” versus
will likely be significant, particularly given “the young.” Rather, it requires a life-cycle
the increased incidence of noncommunicable perspective on policy design that recognizes,
diseases. This trend underscores the urgency for example, that the health of children
to shift from a hospital-centric model to one affects their health as adults, that saving
in which primary care plays a bigger role for old age needs to start early, and that
and in which the treatment of older patients the burden of taxation to provide services
with chronic conditions is managed afford- and benefits needs to be spread fairly across
ably at the right levels of the system. As in generations.
the case of pensions, the financing model for This report argues that impending demo-
health care will need to rely more on budget graphic changes are important and will
financing to achieve universal health cov- become even more so in the countries of
erage as the needs of an aging population East Asia and Pacific in coming decades. But
grow. Aging is also creating a rapidly grow- demography is not destiny. Through the right
ing demand and market for long-term care. mix of policies, governments can help soci-
Innovation is urgently needed in this area as eties adapt to rapid aging and improve the
traditional family networks become increas- well-being of citizens.
ingly stretched. I hope this report contributes to a consid-
Given the diverse stages of demographic ered debate of the policy choices needed to
transition across the region, lessons can be achieve healthy and productive aging across
drawn from the experience of richer East the region.
Asian and Pacific economies. Their expe-
riences, as well as those of high-income Axel van Trotsenburg
countries in other regions, highlight that Regional Vice President
this policy agenda will be challenging for East Asia and Pacific Region
developing East Asian and Pacific countries The World Bank
Acknowledgments
T
his report has been prepared as part Production and editing of the report were
of the World Bank East Asia and coordinated by the World Bank’s Publish-
Pacific Regional Report series by a ing and Knowledge Department under the
team consisting of Philip O’Keefe (team task direction of Susan Graham, Mark Ingebret-
leader), Nithin Umapathi, Aparnaa Somana- sen, and Patricia Katayama.
than, Robert Palacios, John Giles, Thomas The team would like to acknowledge a
Flochel, Ralph Van Doorn, Juan Pradelli, number of people who have provided inputs
Yang Huang, Harry Moroz, Shonali Sen, and guidance during the course of report
Malathi Velamuri, Naoko Miake, Çaglar preparation. Professor Albert Park and par-
Özden, Mauro Testaverde, Minna Han ticipants in a joint workshop held in 2014
Tong, Yuqing Hu, Risa Nakayama, Andrew by the World Bank and Hong Kong Uni-
Beath, and Yuki Ikeda. It has been prepared versity of Science and Technology in Hong
under the initial oversight and manage- Kong SAR, China, provided useful feedback.
ment of Bert Hofman (then regional chief Professor John Campbell provided invalu-
economist, East Asia and Pacific region) able insights on long-term care and Japan’s
and Xiaoqing Yu (then sector director for health system. Professors Hidehiko Ichimura
Human Development, East Asia and Pacific and Yasuyuki Sawada also gave important
region) and more recently under the guid- input. Professor David Canning shared his
ance of Sudhir Shetty (regional chief econo- insights on aging in East Asia and Pacific.
mist, East Asia and Pacific region) and Jehan Richard Jackson shared recent work on atti-
Arulpragasam (practice manager for Social tudes toward old-age support in Asia. The
Protection and Labor, East Asia and Pacific team would also like to acknowledge assis-
region). The peer reviewers for the report tance provided by Yuki Murakami of the
were Professor Andrew Mason, University Organisation for Economic Co-operation
of Hawaii at Manoa and East-West Center; and Development and by Eduardo Klien and
Professor John Piggott, University of New his colleagues at the HelpAge International
South Wales, director of the ARC Centre Regional Office for Asia. The team also ben-
of Excellence in Population Aging Research efited from insights from regional staff at the
(CEPAR); and Emily Sinnott, senior econo- United Nations Population Fund, the United
mist, Poverty Global Practice, World Bank. Nations Economic and Social Commission
xvii
xviii ACKNOWLEDGMENTS
for Asia and the Pacific, and the International in Chiangmai, Thailand, gave insights on
Labour Organization. Professor John Knodel work from the report that were most help-
participated in upstream discussions of the ful. The team also acknowledges the assis-
study in Bangkok, Thailand. Yvonne Sin of tance provided by John Beard of W HO
Towers Watson provided useful input on the Geneva and by Rafal Chomik of CEPAR.
pensions chapter, and Professor Atsushi Seike Many World Bank colleagues provided
gave generously of his time in Japan and valuable assistance, including Axel van Trot-
beyond, as did Professor Soonman Kwon, senburg, Arup Banerji, Yasusuke Tsuka-
Professor Rachel Lu, Professor Winnie Yip, goshi, Rocio Castro, Omar Arias, Makoto
Kate Mandeville, and Katie Leach-Kemon. Nakagawa (now returned to the government
The team’s official counterparts from of Japan), Dewen Wang, Elena Glinskaya,
C h i na , T ha i la nd , a nd Viet na m gave Mitchell Wiener, Mark Dorfman, Toomas
useful guidance to the team as it framed Palu, Ajay Tandon, Gerard La Forgia, Eko
the research agenda for the study, and Pambudi, Sutayut Osornprasop, Johannes
offi cials from the Republic of Korea pro- Koettl, Brooks Evans, Michele Gragnolati,
vided insights on Korea’s labor market Maurizio Bussolo, Hans Timmer, Victoria
and social security reforms. Offi cials from Kwakwa, Joyce Msuya, Carl Hanlon, Alek-
the Ministry of Finance and the Ministry sandra Posarac, Nga Nguyet Nguyen, Reena
of Health, Labor and Social Welfare of Badiani-Magnusson, Lars Sondergaard,
Japan facilitated a study visit of the team Nikola Spatafora, and participants in a joint
in late 2013. Participants in World Health workshop on aging and the macroeconomy
Organization (WHO) conferences in Kobe held by the East Asia and Pacific and Europe
and Hanoi and in a HelpAge conference and Central Asia regions of the World Bank.
Abbreviations
xix
xx ABBRE VIATIONS
1
2 LIVE LONG AND PROSPER
but also new expectations from citi- 5. Although commonalities exist, initial
zens. The speed of all of these changes conditions vary enormously across the
matters. region, and emerging policy responses
2. Several middle-income countries in the already show considerable diversity.
region are growing old before growing Some countries are already very old; oth-
rich. Unlike most Organisation for ers have decades of potential demo-
Economic Co-operation and Development graphic dividend to realize. However,
(OECD) member countries, which aged whether aging is a current, emerging, or
at a gradual pace as they grew wealthy, more distant reality, nearly all societies in
several East Asian and Pacifi c countries the region need to frame public policy
are experiencing—or will experience in with the realization that rapid aging will
coming decades—accelerated aging at come at some point.
relatively low levels of per capita income. 6. Responding to aging is not just about old
This pattern has major implications for people. Effective responses require policy
the welfare of older people and the possi- and behavioral change across the life
bilities for public policy and institutions cycle of individuals. Actions to effect
to help manage the aging process. these changes must consider fertility and
3. Aging is occurring in a wider context of child care policies; education and skills
shifting relationships between state and systems; tax, social security, and transfer
citizens in many East Asian and Pacifi c systems; labor market policies across the
countries. In much of the region, the life cycle; and policies directed at older
state has traditionally provided more people themselves. At the most basic
limited direct support than in other level, evidence increasingly points toward
parts of the world. Traditional values the importance of early childhood health
have placed greater emphasis on the sup- and nutrition as predictors of adult and
port role of families and other networks. old age health.
Although such values have strong resid- 7. Aging is multidimensional and requires
ual influence, citizens are increasingly new modes of policy coordination. The
expecting more from the state. In rapidly fact that aging cuts across all dimensions
aging countries in the region, citizens of people’s lives and public policy, whereas
view aging as a major challenge: about public institutions tend to operate in silos,
80 percent of the people in the Republic creates new challenges to craf ting
of Korea and two-thirds of the people in responses that are more than the sum of
China regard aging as a “major prob- their parts. The expansion of national
lem” for their country (Pew Global aging policies and cross-cutting institu-
Attitudes Survey 2013). tions in East Asia and Pacific is an
4. Demographics are largely a given (at acknowledgment of the challenge.
least for the foreseeable future), but pol-
icy and behavioral responses are within The overview is organized as follows.
the control of governments and citizens. The first section reviews the demographic
However, the behavioral changes required and epidemiological transitions in East Asia
of people, employers, and societies do not and Pacific. It is followed by a discussion of
happen overnight. In a number of cases, the living standards and other indicators of
they go to the heart of sociocultural welfare of older people in the region, such
beliefs about gender and the obligations as sources of support and living arrange-
of generations toward one another. Policy ments. The third section discusses the effects
changes to address aging and realize the of aging on growth, the labor force, savings,
potential of older people are challenging and government budgets, and the ways in
and often unpopular. which countries can manage them. The fourth
OVERVIEW 3
section looks at challenges facing pension, FIGURE O.1 East Asia and Pacific has more older people than any
health, and long-term care systems. This sec- other region
tion is followed by a short discussion of cross- Number of people ages 65 and above by region and East Asian and Pacific
cutting challenges. The final section provides subregion
some concluding thoughts.
OECD countries
Understanding the
China
rapid demographic and
epidemiological transition South Asia
Although much of East Asia and Pacific has ECA emerging
still to experience its most rapid period of countries
aging, it already has the largest regional pop- LAC developing
ulation of people ages 65 and older, primar- countries
ily in China. In 2010, East Asian and Pacific EAP high-income
economies
countries had about 185 million people ages
EAP developing
65 and older, about 36 percent of the global
countries
population of this age group (figure O.1).
China alone had almost 114 million old peo- Sub-Saharan Africa
ple, more than any developing region. MENA developing
More significantly, East Asia and Pacific is countries
aging more rapidly than any region in history.
40
0
0
20
60
80
10
12
14
The phenomenon is driven primarily by a
rapid decline in fertility but also by increased Number of people age 65 and older, 2010 (millions)
longevity. Nearly all middle-income countries
Source: World Bank estimates based on data from UN 2013.
in the region are in the midst of or will soon Note: EAP = East Asia and Pacific; ECA = Eastern Europe and Central Asia; LAC = Latin America
experience a pace of aging that is unprec- and the Caribbean; MENA = Middle East and North Africa; OECD = Organisation for Economic
Co-operation and Development. EAP developing countries exclude China.
edented, transitioning from young to old
societies in 20 to 25 years—a transition that
took 50 to more than 100 years in OECD
countries (figure O.2). Although the starting Declining fertility has combined with steady
points vary, nearly all East Asian and Pacific increases in both life expectancy and health
countries move from aging to aged societies life expectancy, in contrast to regions such as
relatively quickly when the aging threshold is Eastern Europe and Central Asia.
reached (exceptions are the Philippines and The demographic diversity of East
Papua New Guinea). Asian and Pacific economies suggests three
The main driver of aging in East Asia and groupings. The first group (red)—comprising
Pacific has been the sharp decline in fertility, Hong Kong SAR, China; Japan; the Republic
which fell from 5.91 children per woman in of Korea; and Singapore—includes the
1960 to 2.46 in 2005. This decline was signifi- wealthiest economies. People ages 65 and
cantly faster than the global decline (from 5.51 older represented 14 percent of the total
children per woman to 3.03). Fertility rates in population of this group in 2010. The second
several East Asian economies are now among group (orange) includes China, Indonesia,
the lowest in the world, with richer economies Malaysia, Mongolia, Thailand, and Vietnam.
in the region averaging only 1.28 children These countries are aging very quickly. The
per woman in 2010. In major middle-income third group (green) comprises Cambodia, the
countries, such as China, Thailand, and Lao People’s Democratic Republic, Myanmar,
Vietnam, the fertility rate is well below replace- Papua New Guinea, the Philippines,
ment fertility and continues to fall (UN 2013). Timor-Leste, and the Pacific Island countries
4 LIVE LONG AND PROSPER
FIGURE O.2 East Asian and Pacific economies are aging more rapidly than economies elsewhere
Years to move from 7 to 14 percent population share 65 years and older and the start and end years of transition
Vietnam (15)
Lao PDR (20)
Malaysia (20)
Indonesia (20)
Brazil (20)
Thailand (20)
Korea, Rep. (20)
Timor-Leste (25)
Mongolia (25)
Myanmar (25)
Cambodia (25)
Turkey (25)
China (25)
Singapore (25)
Japan (25)
Hong Kong SAR, China (30)
Philippines (35)
Papua New Guinea (40)
South Africa (40)
United Kingdom (45)
Russian Federation (50)
United States (69)
France (115)
1860 1885 1910 1935 1960 1985 2010 2035 2060 2085
Sources: World Bank estimates based on data from UN 2013 and Kinsella and He 2009.
Note: Figure shows starting and ending year for transition from 7 percent (aging) to 14 percent (aged) of population ages 65 and older. Aging and aged
thresholds are based on United Nations definitions. East Asia and Pacific economies rounded to five-year increments.
(not shown in figure O.3). These countries are ratios. The record of demographic projec-
still young—just 4 percent of the population tions in the region is mixed, suggesting that
was age 65 or older in 2010—but they will projections should be treated with caution.
begin to age quickly in coming decades. Fertility decline, for example, has tended
Demographics drive divergent trends in to consistently outpace projections in the
working-age populations. Several East Asian region; assumptions about future reversals
and Pacific countries are projected to expe- in fertility for richer East Asian and Pacific
rience sharp declines in their labor forces economies risk extrapolating from unstable
(figure O.4). Between 2010 and 2040, the short-run dynamics.
share of the working-age population will The basic challenges are further compli-
fall by more than 15 percent in Korea and cated by the speed of urbanization, internal
by at least 10 percent in China, Japan, and and foreign migration, and other factors
Thailand. In contrast, in green and some that will inevitably affect demographic
orange countries, the working-age share of trajectories. Moreover, even in cases where
the population is not expected to shrink until aggregate projections are reliable, rethinking
after 2040. In absolute terms, the Philippines traditional measures such as elderly depen-
and Indonesia will account for the largest dency ratios will be needed. The static defini-
share of the regional increase. tion of the working-age population as people
Although the directions of the demo- ages 15 to 64, for example, is increasingly
graphic transition in East Asia and Pacific are being questioned. Alternative dependency
clear, considerable uncertainty remains about measures that have been proposed (such as
the speed of the transition and the way to years of remaining life or ratios based on age
think about key measures such as dependency profiles of actual labor force participation)
OVERVIEW 5
FIGURE O.3 Three distinct aging patterns are apparent in East Asia and Pacific
Percentage of population age 65 or older in East Asian and Pacific economies
40
36%
35
Share of population age 65 and older (%)
30
24%
25
20
14%
13%
15
10
6%
5
4%
0
50
60
70
80
90
00
10
20
30
40
50
60
70
80
90
00
19
19
19
19
19
20
20
20
20
20
20
20
20
20
20
21
High-income economies, aged population
Middle- and lower-middle-income countries, aging population
Low-income countries, young population
may be more relevant for policy purposes as bulk of disability-adjusted life years among
people live longer and young people spend people ages 60 and over in 2010, with
more time acquiring education. comorbidities complicating approaches to
Together with the demographic transition, treatments and care. Underlying these trends
East Asia and Pacific is experiencing a rapid is growing exposure to risk factors, to which
epidemiological transition toward noncom- the poor are most vulnerable. Between 1990
municable diseases (NCDs), in part driven and 2010, premature death and disability
by population aging. Healthy life expectancy associated with NCD-related risk factors
in the region increased between 1990 and (such as high salt intake, blood pressure,
2010, albeit by slightly less than total life total cholesterol, and blood sugar) increased
expectancy. At the same time, the burden everywhere in the world. The increase was
of disease related to NCDs has grown rap- particularly pronounced in East Asia and
idly, with initial onset increasingly occurring Pacific. As a result of earlier onset of lifestyle-
in middle age. NCDs caused 76 percent of related conditions such as hypertension and
deaths in the region in 2008, and the share obesity, in many countries in the region, the
is projected to rise to 85 percent by 2030 current cohort of middle-aged people is more
(IHME 2013). Cardiovascular disease, can- vulnerable to poor health in old age than
cer, and diabetes already accounted for the were earlier cohorts (IHME 2013).
6 LIVE LONG AND PROSPER
FIGURE O.4 In the richer economies of East Asia and Pacific, the Co-residence of older people with adult
working-age population is projected to shrink between 2010 and children is high in East Asia and Pacific,
2040 although it varies across countries and has
Percentage change in population ages 15–64 between 2010 and 2040, declined significantly over time in some coun-
selected economies
tries. It generally increases between middle
age and the mid-60s. For people ages 60
Hong Kong SAR, China and older, co-residence rates range from 25
Korea, Rep. to 30 percent to more than 80 percent, with
Singapore significant variation by gender and by urban
Thailand and rural location. Rates are particularly high
Japan in low-income countries, and, consistent with
China
global patterns, lower at higher-income levels
within countries. Co-residence with children
Vietnam
has been declining rapidly in China, Korea,
Mongolia
and Thailand. In China, the co-residence rate
Malaysia for people ages 65 to 70 fell from almost
Myanmar 66 percent in the early 1980s to roughly
Indonesia 43 percent by 2011. In Korea, co-residence of
Cambodia people ages 65 and older fell from more than
Philippines 80 percent in 1980 to well below 30 percent
Papua New Guinea by 2010 (Giles and Huang 2015).
Older people in East Asia and Pacific often
Lao PDR
work until very old age. However, significant
Timor-Leste
differences exist between urban and rural areas
–20 –15 –10 –5 0 5 10 15 and between men and women (figure O.5).
Percentage change On average, people work until late in life, but
many people in urban areas stop working rel-
Source: World Bank estimates based on data from UN 2013. atively early, and some groups (such as urban
Chinese female workers) retire very early (only
about one-third are still working at age 60).
Situation and sources of support Withdrawal of urban people from work is
of older people clearly correlated with access to a formal sec-
The situation of older people in East Asia and tor pension, though caregiving responsibilities
Pacific is diverse across and within countries. for grandchildren and other elderly also play a
Across developing countries, older people role in the early withdrawal of urban women
shared in the significant decline in poverty, from formal work. Nearly everywhere, urban
but poverty in a number of the region’s coun- female workers are the least likely to work
tries tends to rise with age after age 45, before into old age, and gender gaps in participation
flattening or even declining at the oldest ages. are substantial. Among people who continue
As with the general population, older people working into old age, self-employment is the
with higher levels of education (and people dominant form of work, and self-employment
with children with higher levels of educa- rates of 90 percent and higher are common
tion), access to a formal sector pension, and in rural areas of the region (Giles, Hu, and
better health status tend to be better off than Huang 2015).
their peers. In contrast to the situation of the Older people in East and Southeast Asia
younger adult population, however, employ- tend to continue working long hours, particu-
ment status has a more complex relation- larly in rural areas. In the economies analyzed
ship with poverty among the elderly, in part for this report, men who continue to work past
because many poorer, often rural, older peo- age 65 work, on average, 40 hours or more a
ple have to “work till they drop.” week, and women work 30–45 hours a week.
OVERVIEW 7
FIGURE O.5 Labor force participation rates in East Asia and Pacific vary, but across the region, rural people work
significantly longer than urban people
Percentage of population participating in the labor force, by age, gender, and urban or rural location, selected countries
% of cohort working
% of cohort working
70 70
60 60
50 50
40 40
30 30
20 20
10 10
0 0
40 50 60 70 80 40 50 60 70 80
Age Age
70 70
60 60
50 50
40 40
30 30
20 20
10 10
0 0
40 50 60 70 80 40 50 60 70 80
Age Age
70 70
60 60
50 50
40 40
30 30
20 20
10 10
0 0
40 50 60 70 80 40 50 60 70 80
Age Age
Sources: Giles, Hu, and Huang 2015, based on data from CHARLS 2011; IFLS 2007; KLoSA 2010; World Bank East Asia and Pacific Standardized Household Surveys, various years; and
ThaiSES 2011.
8 LIVE LONG AND PROSPER
FIGURE O.6 Work remains the primary source of old-age support in most countries in East Asia and Pacific
Percentages of total income of people ages 60 and above, in urban and rural areas, from labor, public and private
transfers, and other sources, selected countries
a. Urban
100
90
80
% of total income
70
60
50
40
30
20
10
0
p.
a
sia
te
lia
m
s
d
in
di
ne
Re
es
na
an
go
ne
Ch
bo
pi
-L
a,
et
ail
on
do
ilip
m
or
re
Vi
Th
M
Ca
In
Ko
Tim
Ph
b. Rural
100
90
80
% of total income
70
60
50
40
30
20
10
0
p.
a
sia
lia
te
m
s
d
in
di
ne
Re
es
na
an
go
ne
Ch
bo
pi
-L
a,
et
ail
on
do
ilip
m
or
re
Vi
Th
M
Ca
In
Ko
Tim
Ph
Sources: Giles and Huang 2015, based on data from CHARLS 2011; IFLS 2007; KLoSA 2010; ThaiSES 2011; and World Bank East Asia and Pacific Standardized
Household Surveys, various years.
In rural areas, men who continue to work do support of their families, especially in rural
so for 30–40 hours a week even at age 75, and areas. Labor income plays a critical role in old-
women that age work 20–35 hours a week. In age support in the region, with work income
addition to what is formally counted as work, dominating other sources of support in rural
older people in East Asia and Pacific often areas in nearly all countries and in urban areas
bear significant caregiving responsibilities, a in most countries (figure O.6). In most coun-
social and economic role that is not formally tries, private transfers to elderly people are also
recognized in national accounts. more significant than public transfers. Familial
The roles of the state, families, and indi- support is even more pronounced when one
viduals in old-age support in developing East factors in nonfinancial support.
Asian and Pacific countries are distinctive. In Diversity is also growing across and within
developing regions such as Latin America and countries, rooted in initial conditions such
the Caribbean and Eastern Europe and Central as labor market formality, urban and rural
Asia, the state plays an important role through shares, and formal social security coverage.
generous and widespread pension systems. In Public transfers play an important role in
contrast, older people in East Asia and Pacific urban China and in Mongolia, for example,
rely more heavily on their own labor and the but only a negligible role in Cambodia and
OVERVIEW 9
FIGURE O.7 Attitudes on the preferred source of financial support in old age and on the expected source
of personal care are shifting in East and Southeast Asian economies
Percentage of adults reporting government, selves, or family as likely sources of support, selected economies
70
60
% of respondents
50
40
30
20
10
0
p.
m
na
ia
sia
re
d
ne
in
Re
ys
an
na
po
hi
ne
Ch
pi
ala
ail
et
,C
a,
ga
do
ilip
re
Vi
Th
M
an
Sin
In
Ko
Ph
iw
Ta
Indonesia, where informality is high and pen- Thailand provides a case in point.
sion and social assistance systems are under- Although it has enjoyed solid growth in
developed. Within China, differences between recent decades, the pace of aging is outstrip-
rural and urban areas are sharp. Although ping per capita income growth. It is following
public transfers play a relatively modest role a path closer to that of Poland, which is now
in old-age support in most of the region’s struggling with fiscal pressures from aging.
countries, social expectations of the relative Mongolia is experiencing stagnant GDP per
roles of the state, families, and retirees them- capita and rapid aging.
selves in postretirement support are changing The demographic dividend has played a
rapidly, with opinion surveys showing expec- substantial role in the region’s remarkable
tations of future state (rather than familial) growth story in recent decades. However,
support rising sharply (figure O.7). its younger countries will need to deepen
reforms if they are to enjoy the same growth
effects as older countries did at similar
Managing the effect of aging on stages of the demographic transition. They
growth and its drivers can adopt productivity-enhancing reforms
Developing countries in East Asia and Pacific that will result in more manageable elderly
are getting old before getting rich. In par- dependency and GDP trajectories. Around
ticular, orange (largely aging middle-income) one-third of the high per capita income
countries have been aging—in some cases growth in East Asia from 1960 to the 1990s
rapidly—at low levels of gross domestic prod- can be attributed to the demographic divi-
uct (GDP) per capita (figure O.8). This trend dend, with some estimates of the effect as
boosts the pressure to increase productivity high as 44 percent of growth in per capita
growth in the region’s developing countries. income (Bloom, Canning, and Malaney
Even with sustained productivity growth, 2000; Bloom, Canning, and Sevilla 2003;
however, orange countries will not reach the Bloom and Williamson 1998; Kelley and
income levels of red (richer and older) East Schmidt 2005; this literature generally
Asian and Pacific economies or OECD and excludes the Pacific). In most East Asian
middle-income countries such as Mexico at countries, improvements in educational
similar points in the demographic transition. attainment, capital investment, and other
10 LIVE LONG AND PROSPER
FIGURE O.8 East Asia and Pacific’s developing economies are getting old before getting rich
Elderly dependency ratio by GDP per capita (2005 PPP), various economies, 1980–2010
60,000
50,000
GDP per capita (US$2005 PPP)
40,000
30,000
20,000
10,000
Sources: GDP data in 2005 US$PPP for 1980 to 2010 from the World Bank’s World Development Indicators database; elderly dependency ratios calculated as
ratio of population over age 65 to population ages 15–64, using data from UN 2013.
Note: PPP = purchasing power parity.
factors compounded the pure demographic to growth are real, even rapidly aging coun-
effect. For younger East Asian and Pacific tries have significant scope to act. Managing
countries currently in the midst of their the growth and other macroeconomic effects
youth bulges, for which rapid aging will of aging will require public policy action and
come only in future decades, the key is to wider behavioral change in a range of areas
make productivity-enhancing reforms and that span the life cycle. Priorities differ across
investments now, so that they both maximize country groupings:
the demographic dividend and are prepared
for rapid aging when it comes. • For red economies, slowing the structural
The tailwind to growth provided by demo- decline in the size of the labor force will be
graphics in older East Asian and Pacific coun- the most critical challenge, because most
tries has raised fears of a headwind as the size economies are closer to the productivity
of the labor force declines and aging accel- frontier and therefore lack easy options
erates, but policy and behavioral responses for increasing labor productivity sig-
can mitigate these forces. Although the risks nificantly. A sharper focus on increasing
OVERVIEW 11
female labor force participation, extend- in a variety of ways. Various labor mar-
ing productive working lives, and, in ket institutions and policies, as well as
some countries, increasing immigration of social attitudes, constrain productive labor
younger workers from within the region force participation of groups that need
will be vital. to be mobilized to mitigate the structural
• Orange countries will need to sustain high decline in the workforce. The experience of
productivity growth and undertake struc- red economies points to a distinctive East
tural reforms of social security, health and Asian approach to increasing labor force
long-term care, and labor market policies. participation that combines incentives and
Even if they make such changes, however, mandates for employers. Channels for
these countries will become aged societies mitigating the effects of aging on the labor
at much lower levels of income than the force include the following:
red economies. They will need to mitigate
the labor supply and fiscal effects of rapid • Extending productive working lives
aging through ongoing reforms of pension through labor market and social secu-
and health systems and labor policies to rity policies. Countries could gradually
extend the working lives of their urban raise retirement ages and reverse incen-
and formal sector workers as they undergo tives for early retirement; incentivize
rapid urbanization. There are reasons for workplace adjustments to accommodate
cautious optimism, because more edu- older workers (through, for example, flex-
cated cohorts will be better prepared for ible and part-time work arrangements
the prospect of longer working lives than and midcareer skill upgrading); and pro-
previous generations. mote attitudinal change among employ-
• Green countries will enjoy favorable ers. They could also reform labor market
demographics as the youth bulge con- institutions such as seniority wage sys-
tinues to feed their labor forces for the tems, which create barriers to productive
next few decades. The priorities are to employment of older workers.
establish conditions to realize maximum • Raising female labor force participa-
GDP growth from the demographic tion, with an emphasis on making child
dividend (through investments that care available and providing subsidies
raise productivity and maximize youth to make it affordable. The experience of
employment) and to avoid adopting poli- richer East Asian and Pacific economies
cies in areas such as pensions and health that adopted measures to promote female
that are affordable now but would rap- labor force participation demonstrates the
idly become unsustainable when aging challenges of shifting strong gender norms
accelerates. These countries’ success in in the region.
doing so will determine their readiness to • Increasing immigration, which will
manage aging when it accelerates over the require proactive efforts in both sending
coming two to three decades. and receiving countries. The different tim-
ing of aging across East Asia and Pacific
Aging has potentially deleterious effects on offers opportunities for “chronological
the labor force, capital formation and savings, arbitrage.” Singapore and Hong Kong
and the budget, each discussed in turn below. SAR, China, offer examples of the eco-
Policies are needed to mitigate these effects. nomic benefits of more open immigration
policies. Other countries are exploring
ways to overcome traditionally limited
Mitigating the effects on the labor force
reliance on immigration. Although this
East Asian and Pacifi c countries can help area is politically challenging, the benefits
reduce constraints to labor force par- for sending and receiving countries are
ticipation and thus mitigate the demo- clear in the context of rapid but asynchro-
graphic effect of aging on labor force size nous aging across the region.
12 LIVE LONG AND PROSPER
FIGURE O.9 Greater participation by women, older people, and More flexible employment arrangements,
migrants can mitigate the projected decline in the labor force in support for parents through subsidized child
richer East and Southeast Asian economies care services, and broader shifts in gender and
Projected changes in labor force size of destination economies with family relations are more likely to affect work
changes in female, elderly, and migrant labor force participation
and fertility decisions.
Efforts to mitigate the shrinking size of the
130
labor force will be supplemented by improve-
ments in labor force quality as a result of
expanded access to higher levels of educa-
120 tion in recent decades and ongoing efforts to
improve educational quality. Capital deepen-
Millions of people
100
The effects on savings and capital
formation
90 Debates on aging often focus on poten-
2010 2015 2020 2025 2030 2035 2040 2045 2050 tial declines in savings rates. The effects on
Baseline household savings depend on the relative
Female increased participation
strength of two offsetting effects: (a) the
compositional effect from a higher share of
Elderly increased participation
older people in the population and (b) the
Permanent migration behavioral effect as people save more to
Temporary migration finance a longer expected period of retire-
ment. Predictions on which of these effects
Source: Özden and Testaverde 2014. will dominate in East Asia and Pacific are
Note: Destination economies are Hong Kong SAR, China; Japan; the Republic of Korea; Malaysia; and
Singapore. Sending economies are Cambodia, Indonesia, Lao PDR, Mongolia, Myanmar, Papua New mixed, but on balance they suggest that
Guinea, the Philippines, Thailand, Timor-Leste, and Vietnam. Green line shows increase in labor concerns about the effects of aging on sav-
force relative to baseline when women participate at same level as men. Red line shows increase
given increased labor force participation by the elderly. Orange line shows increase given migration ings and capital formation in the region
of 10 percent of labor force ages 25–35 (migrants stay permanently). Light blue line shows increase may be overstated, for three reasons. First,
given temporary increase in migration of 20 percent of labor force ages 25–35 (migrants stay
10 years). Dark blue line shows the baseline. household and corporate savings rates are
high in the region and provide a more solid
foundation than they do in other regions.
Second, survey evidence points to flatter sav-
The relative importance of these chan- ings and age profiles in the region than in
nels varies across countries. In the richest other parts of the world and to an increase
East Asian and Pacific economies, the effect in savings rates at all ages in the region in
is likely to be greatest from higher female recent decades. Third, significant inefficien-
labor force participation and, in some cases cies in financial markets in the region’s devel-
(for example, Japan and Korea), migration. In oping countries suggest that scope exists for
some middle-income countries, the potential more efficient mediation of savings into capi-
role of increased participation by the elderly tal formation and increased productivity.
labor force is also significant (figure O.9).
In contrast to these measures, policies to
The effects on fiscal balances
increase the birth rate through direct finan-
cial incentives such as birth grants or favor- Even if East Asia and Pacific avoids severe
able tax treatment do not appear to have adverse growth effects from rapid aging,
been effective in stemming fertility decline. the fiscal risks of business as usual in core
OVERVIEW 13
programs such as pensions, health, and long- FIGURE O.10 Across APEC, pensions will eat up a larger and larger
term care are substantial. Major fiscal risks share of GDP in coming decades without reforms
from aging are already manifesting them- Projected pension spending as a percentage of GDP in APEC economies
selves in the region and will require policy (no-reform scenario), 2009–75
leadership to mitigate.
Projections of pension and health spend- 18
16
ing reveal significant fiscal pressure in
14
coming decades in the absence of reform.
12
Stylized projections by broad country type in
% of GDP
10
Asia-Pacific Economic Cooperation (APEC) 8
member countries through 2070 show 6
economies converging to ratios of pension 4
spending to GDP that are 8 to 12 percent- 2
age points higher than current levels (figure 0
O.10). Cash flow deficits are expected to 2009 2019 2029 2039 2049 2059 2069 2079
emerge in most defined benefit schemes in High-income economies, mature pension system
developing East Asian and Pacific countries, Middle-income economies, maturing pension system
with deficits of 1.4 to 4.5 percent of GDP
High-income economies, maturing pension system
by 2040. Health care spending projections
also point to higher public spending (though Low-income economies, developing pension system
lower than increases in pension spending). Source: Hinz and Zviniene 2011.
Aging accounts for about one-third of the Note: APEC = Asia-Pacific Economic Cooperation.
projected increase in public health spending
in several developing countries in the region, broadening pension coverage, and explor-
and the effect is even higher in China (Hinz ing the role of the state in long-term care.
and Zviniene 2011). A growing commitment to shifting some
A more complex (and uncertain) picture of the burden of old-age financial sup-
of fiscal pressures from aging emerges when port and health care costs from families
debt dynamics are incorporated into fiscal and individuals to the state through pool-
projections. Where increased age-related ing mechanisms can be seen. The gradual
spending drives up deficits and they are shift of financing from private to public
funded through debt, the potential for an sources exerts significant pressure on pub-
upward spiral in public spending is substan- lic spending.
tial. The risk is higher where significant shares • The financing models for pensions and
of debt are denominated in foreign currency. health are increasingly mixed, illustrating
The importance of factoring in demographics the limitations of contributory financing
in scenarios of public debt should therefore models in settings with high labor mar-
not be underestimated. ket informality. More nuanced blends of
contributory and general revenue financ-
ing are emerging that also aim to priori-
Developing pension, health, and tize inclusion of the poor and near-poor.
long-term care systems for aging For health care coverage in particular,
societies the emphasis has been almost entirely on
expanding coverage through general rev-
The key areas of pensions, health, and long-
enue financing.
term care face several common reform
• Pension and health systems face the chal-
challenges:
lenges of balancing coverage, adequacy,
• Most orange and some green countries and sustainability. To date, the emerging
are expanding policy-driven coverage— approach in developing East Asian and
achieving universal health coverage, Pacific countries has been to prioritize
14 LIVE LONG AND PROSPER
coverage expansion, albeit often with The good news is that considerable inno-
shallow financial protection. vation is already evident within East Asia and
• Pension systems in many East Asian and Pacific. In response to the persistent failure of
Pacific countries are fragmented, with contributory pension schemes in closing the
separate schemes for the civil service, the coverage gap, Chile (2008), China (2011),
military, and state-owned enterprises; Korea (2004), Mexico (2012), and Thailand
often fragmented pooling of contribu- (the 2000s) all rapidly expanded their social
tions, which leads to variable benefits and pensions and budget-financed matching con-
undermines the insurance function; and tribution schemes for informal sector work-
distinct rural and urban systems in China. ers (see Holzmann, Robalino, and Takayama
Health financing is characterized by simi- 2009 for regional and global experience with
lar challenges. social pensions; Hinz et al. 2013 for experi-
ence with matching defined contributions).
Countries in the region have also expanded
Pensions health insurance coverage, through heavily
subsidized health insurance premiums. Use of
Pension systems in East Asia and Pacific are consumption taxes in Japan and budget sup-
quite heterogeneous. The pension challenges port for the contributory pension scheme in
facing countries in the region vary not only China are driven partly by reluctance to raise
because of their disparate demographic payroll tax rates further. The trend is not uni-
situations, but also because of past policy form, however, as evidenced by the introduc-
choices and the nature of their labor and tion of a social insurance scheme in Lao PDR
capital markets. Richer (red) economies have and legislation in Indonesia and Myanmar
achieved wide coverage of their systems with that make these countries among the last in
better financial sustainability than other rich the world to rely on contributory approaches
economies, but they face challenges of pen- to expanding pension provision. The expe-
sion adequacy and hence old-age financial rience of dozens of countries that had once
protection. Middle-income (largely orange) faced the same choice suggests that it may not
countries are grappling with the dual chal- be the appropriate path.
lenges of improving the sustainability of Reform should be based on the premise
their existing or legacy systems while try- that the redistributive element of the pen-
ing to expand coverage to large uncovered sion system should be financed from general
informal sectors. For poorer and younger revenues and clearly distinguished from the
(green) countries, the key issues are choosing insurance or savings component, which ide-
a pension system model that can achieve sig- ally should be financed by individuals on
nificant coverage over time. These countries an actuarially fair basis (that is, identified
need to avoid overreliance on contributory sources of financing are sufficient to pay
pension models, which have proven ineffec- the expected expenditures). The simplest
tive in extending coverage to informal sector approach would be to eschew the mandated
workers. scheme altogether and instead broaden the
Addressing the coverage gap, particularly coverage of a social pension (for example,
in countries with high levels of informality, by making it universal or means tested) and
will require policies that move away from sole complement it with a voluntary retirement
reliance on payroll taxes to finance pensions. savings scheme. In New Zealand, which
This paradigm shift in the way pension sys- adopted this approach, the opt-in default
tems have historically been financed is moti- for a voluntary defined contribution scheme
vated by the limitations of a traditional social led to relatively high participation rates.
insurance approach, which risks losing the This option is not likely to be realistic in
race between pension coverage expansion and countries that already have contributory
rapid societal aging. mandates, however, and it would be difficult
OVERVIEW 15
even for countries such as Myanmar and in healthy life expectancy in assessing aging’s
Timor-Leste, which have nascent contribu- effects over time is important.
tory systems. These countries, along with The structural reform challenges fac-
Indonesia, should aim for actuarially fair ing health delivery systems in the region
and modest schemes, leaving redistribution will become more acute as populations age
to be financed through general revenues or because the increased incidence of chronic
through social assistance or social pensions. NCDs with age means both that the share
The wider group of East Asian and Pacific of people living with these conditions will
countries with mature, but low-coverage man- increase and that larger shares of people will
dated contributory systems should also con- be living with multiple chronic NCDs. Health
sider shifting the balance of pension system delivery systems in East Asia and Pacific are
financing toward general revenue–financed ill prepared for the NCD epidemic, especially
redistribution and modest target benefit lev- in countries with rising shares of older peo-
els, but in a way that does not compromise ple, for several reasons. First, these systems
incentives to participate in the contributory tend to provide low-value, high-cost services
scheme. Doing so requires strong coordina- that are likely hospital-centric; preventive
tion between social and contributory pension and primary care services tend to be inef-
design and implementation. The most feasible fective and of poor quality. Too often, bet-
path for such countries is likely to mix resid- ter health is incorrectly attributed to more
ual mandatory contributory systems with consultations, admissions, drugs, and pro-
subsidized voluntary contributory approaches cedures. Second, too little attention is paid
and social pensions targeted to wider or nar- to the prevention, early diagnosis, treatment,
rower populations, depending on the level and control of health conditions. When a
of coverage achieved through contributory diagnosis is made, care is rarely coordinated
mechanisms. Countries such as China, Korea, across provider levels, resulting in service
and Thailand provide distinctive models of duplication and lack of continuity. In addi-
this approach. Although the adequacy of tion, health care is often sought too late,
pensions for the majority is limited currently, leading to high-cost treatment in expensive
sustained growth should allow for deepening acute care hospitals. Third, the lack of effec-
financial protection over time. tive referrals, gatekeeping, and postdischarge
care contributes to costly (and avoidable)
readmissions. Fourth, incentives embedded
Health
in provider payment systems encourage phy-
In the health sector, the effects of aging are sicians to overprovide services or provide
likely to be significant. In East Asia and Pacific unnecessary care.
and globally, age-specific profiles of expen- Health system shortcomings and wel-
ditures are typically a J-shape, with health fare consequences are more severe for older
spending rising steeply toward the end of people, who are much more likely to receive
life. Consequently, a common assumption is inpatient care (often unnecessarily), use
that aging will generate substantial increases more pharmaceuticals than average (and
in health care spending. In fact, the effects hence are more affected by poor pharmaceu-
of aging on health expenditure are com- tical purchasing and prescription practices),
plex. Globally, technology and policy choices and experience substantially higher out-of-
regarding the financing and delivery of health pocket spending and higher rates of cata-
care, rather than aging itself, are the biggest strophic health spending. In China in 2011,
drivers of health spending. Aging does add to about 32 percent of households with a mean
health costs (accounting for about one-third age of 60 or more incurred out-of-pocket
of spending increases in several East Asian health payments of more than 25 percent of
and Pacific countries) (Somanathan 2015). total nonfood spending; in Vietnam in 2012,
However, taking into account improvements the figure was 30 percent of households
16 LIVE LONG AND PROSPER
(Somanathan 2015). Older people have Health systems will need to spend more
higher comorbidity rates, which require a efficiently if the benefits of system transforma-
mixed set of services and more complex tion are to be realized and afforded. Fee-for-
management of movement between levels service provider payments will need reform
of the health system. They also experience to reduce overprovision of services and over-
functional and cognitive decline, which puts charging for care. Global experience with
new demands on health systems. For all of payment reforms points toward case-based or
these reasons, countries with aging popula- diagnosis-related group approaches for sec-
tions need to create health systems in which ondary and tertiary care, mixed-method pay-
prevention and treatment are offered in a ments for primary care, and global budgets to
coordinated manner over a sustained period contain costs. Having a large single purchaser
of time while enabling individuals to assume of health services with strategic purchasing
greater responsibility for managing their capacity has also proven critical for cost con-
own care. tainment (Somanathan 2015).
To reorient health systems to meet the rein- Many countries in the region will need to
forcing challenges of universal health care, continue reforms of pharmaceutical procure-
NCDs, and aging, countries will need to pur- ment to rein in the high prices paid by govern-
sue several reforms in parallel that together ments and patients. These countries can learn
will take at least a generation to achieve. from experiences both within the region and
An overarching need is to transform health globally in purchasing and other practices.
delivery systems by strengthening primary Practices to control health technology
care services, shifting care away from acute diffusion, such as health technology assess-
care hospitals, reducing overprovision, and ments, are also crucial to control health
improving coordination among providers. spending growth. East Asian and Pacific
The quality of the health workforce also countries could begin by instituting trans-
needs to be substantially improved. These parent and evidence-based processes for
reforms will help contain costs in the long prioritizing new technologies and drugs,
term, but they will require additional spend- using the findings from more established
ing in the medium term. health technology assessment agencies to
Because mobilizing new resources to meet inform the prioritization processes. This
all of the additional requirements of service challenging and interrelated set of reforms
delivery reforms would simply not be sus- is necessary even without aging. Accelerated
tainable, generating efficiency savings to aging makes reform even more essential and
gain better value for money from existing urgent.
health systems is imperative in East Asia and
Pacific. Although the reformed service deliv-
Long-term care
ery model will generate significant efficiency
savings over time, the initial investment Formal long-term care (LTC) systems in
costs are likely to be substantial. Meanwhile, developing East Asian and Pacific coun-
other policy priorities—not least of which is tries remain nascent. However, a growing
the policy commitment to achieve universal number of aging countries are grappling
health coverage—will continue to place signif- with the appropriate role of the state in an
icant demands on East Asian and Pacific gov- area that has traditionally been the domain
ernments’ budgetary allocations to the health of families, communities, and the health
sector. Health system inefficiencies are already system.
exerting pressure on tax and social health Rapid aging and social change have
insurance revenues. Service delivery reforms exposed the limitations of traditional infor-
are thus critical to ensure the sustainability of mal modes of LTC for frail elderly people
health systems for aging populations and jus- in the region. Part of the response has been
tify additional investment. default reliance on health systems, but this
OVERVIEW 17
approach is costly and complicates health better placed than some others institutionally.
reform efforts. Proactive policy choices in the Its consolidated health, labor, and welfare
LTC domain are therefore important. They ministry brings a critical mass of age-related
require careful planning with respect to their policies and programs under a single body.
interaction with informal care systems and
formal health and welfare systems. Whatever
The political economy of reform and
the design choice, countries need to craft
behavioral change
explicit (rather than costly default) LTC strat-
egies, with a strong emphasis on home- and Many of the policy reforms needed to pro-
community-based care. “Aging in place” is mote healthy and productive aging are politi-
important because it not only provides posi- cally difficult and may challenge societal
tive experiences for older people, but also is norms. Extending working lives by raising
more sustainable than other options as the the retirement age has proven politically
elderly population grows. challenging in many parts of the world.
Increasing female labor force participation
involves deeply rooted cultural norms and,
Addressing cross-cutting in some cases, trade-offs (grandparents who
challenges in aging work late into life will not be available to
In addition to sector-specific challenges, a provide the informal child care and aged care
range of cross-cutting considerations under- on which working-age households currently
pins policy choices and societal responses. rely, for example).
Such considerations will require more inte- The performance on immigration from
grated ways of approaching not only public younger to older countries in the region is
policy and service delivery systems but also mixed. Health system reform is notoriously
a broader societal need to forge consensus difficult, for both providers and the wider
within and across generations. population, and changing ingrained lifestyle
habits that drive the NCD epidemic is a gen-
erational challenge. All these measures will
The multisectoral nature of aging
require patient building of social consensus.
One of the key challenges in shaping public Intergenerational equity and political econ-
policy responses to aging is its inherently mul- omy is also a wider challenge. In other parts
tisectoral nature. Like areas such as early child- of the world, older people have successfully
hood, aging presents inherent challenges to the lobbied for their interests, at times at the cost
institutional and policy setups of governments of younger generations. Red economies in
because it cuts across many sectors and stages East Asia and Pacific seem to have balanced
of life. Governments are typically not orga- the interests of different generations and
nized to respond multisectorally. Although a shared the costs and benefits of an aging pop-
number of East Asian and Pacific countries ulation more equitably than have countries
have recognized this challenge and established in other parts of the world. The evolution of
national aging commissions to promote more the political economy dynamics around aging
holistic thinking on aging, these agencies tend will be an important consideration in shaping
to be primarily advocacy bodies that may lack policy responses.
the political and bureaucratic clout to drive
whole-of-government approaches.
The roles of public policy and market
Some richer countries have sought to bring
forces
a wider view to sectoral policies. One exam-
ple is the periodic intergenerational reports Another broad issue is the appropriate roles of
produced for the Australian Treasury, which public policy and market forces in shaping the
help set long-term expenditure and policy response to rapid aging. Proactive public pol-
priorities for age-related programs. Japan is icy will be required, but it will not necessarily
18 LIVE LONG AND PROSPER
be the correct response to every challenge. A hands of policy makers, communities, and
careful balance of mandates, incentives, and citizens. Demography is a powerful force
reliance on market forces will be required. in development, but it is not destiny. Policy
The experience of red economies in East makers in the region have the potential to
Asia and Pacific in extending working lives shape responses that increase the chances for
illustrates how supportive public policies are healthy and productive aging and promote
gaining more traction as the market forces of societies in which the compact between gen-
supply and demand have increased the appeal erations is one that is fair and realizes people’s
of older workers. The most effective public potential at all ages.
policies will sometimes involve a retreat from In this process, East Asian and Pacific
state direction where policies (such as rigid economies can learn from the experiences—
seniority wages or unnecessary restrictions good and bad—of richer and older coun-
on women’s work) hinder the play of market tries, as well as from one another. Crafting
forces that would be expected to increase the appropriate policies will inevitably involve
employment of older people and women. experimentation and course correction. It
will also require strong leadership in the face
of inevitable reluctance to embrace change.
Knowledge gaps on aging
Fortunately, the region’s strong economic and
Much remains unknown about aging, in East cultural traditions place it in a favorable posi-
Asia and Pacific and globally. Understanding tion to create environments in which citizens
long-term trends and policy effects is often will live long and prosper.
speculative. Significant knowledge gaps
remain—on the interaction of labor and
savings market behavior and rapid social References
security reforms; the societal and individ- Bloom, David E., David Canning, and Pia N.
ual behavioral change in the face of rapid Malaney. 2000. “Population Dynamics and
growth, urbanization, and changing family Economic Growth in Asia.” Population and
structures; the way to reverse ultralow fertil- Development Review 26 (suppl.): 257–90.
ity rates in richer East Asian and Pacific econ- Bloom, David E., David Canning, and Jaypee
omies; the productivity of older workers and Sevilla. 2003. The Demographic Dividend:
the cost-effectiveness of policies to extend A New Perspective on the Economic
working lives; the cross-country dimensions Consequences of Population Change. Santa
Monica, CA: RAND.
of aging in the region with respect to migra-
Bloom, David E., and Jeffrey G. Williamson. 1998.
tion, capital flows, and other factors; and “Demographic Transitions and Economic
more broadly, the roles of market forces and Miracles in Emerging Asia.” World Bank
public policy in promoting healthy and pro- Economic Review 12 (3): 419–55.
ductive aging. CHARLS (China Health and Retirement
Even an understanding of the most funda- Longitudinal Study). 2011. National School of
mental question of the trajectory of fertility Development, Peking University. http://charls
and aging has been imperfect. The speed of .ccer.edu.cn/en.
aging in the region and its diverse cultural Giles, John, Yuqing Hu, and Yang Huang. 2015.
and political economy contexts suggest that “Understanding the Retirement Decision in
caution is therefore needed in predicting Aging East Asia.” Background paper for the
East Asia and Pacific regional report on aging,
future trends.
World Bank, Washington, DC.
Giles, John, and Yang Huang. 2015. “Are the
Conclusion Elderly Left Behind in a Time of Rapid
Demographic and Economic Change?
Aging will fundamentally reshape societies A Comparative Study of the Poverty
and economies in East Asia and Pacific, but and Well-Being of East Asia’s Elderly.”
the shape of the future is very much in the Background paper for the East Asia and
OVERVIEW 19
Pacific regional report on aging, World Bank, Kelley, Allen C., and Richard M. Schmidt. 2005.
Washington, DC. “Evolution of Recent Economic-Demographic
Heston, Alan, Robert Summers, and Bettina Aten. Modeling: A Synthesis.” Journal of Population
2012. Penn World Table Version 7.1. Center Economics 18 (2): 275–300.
for International Comparisons of Production, Kinsella, Kevin, and Wan He. 2009. An Aging
Income, and Prices, University of Pennsylvania, World: 2008. U.S. Census Bureau, International
Philadelphia. Population Report P95/09-1. Washington, DC:
Hinz, Richard, Robert Holzmann, David Tuesta, U.S. Government Printing Office.
and Noriyuki Takayama, eds. 2013. Matching KLoSA (Korean Longitudinal Study of
Contributions for Pensions: A Review of Ageing). 2010. Korea Labor Institute, Seoul.
International Experience. Washington, DC: http://www.kli.re.kr/klosa/en/about/introduce
World Bank. .jsp.
Hinz, Richard, and Asta Zviniene. 2011. Özden, Çaglar, and Mauro Testaverde. 2014.
“The Potential Fiscal Effects of Population “International Migration in Aging Societies:
Aging: Simulation of Aggregate Pension Impacts and Benefits in East Asia and Pacific
and Health Expenditure for Four Stylized Countries.” Background paper for the East
Countries at Various Levels of Economic Asia and Pacific regional report on aging,
Development.” Social Protection and Labor World Bank, Washington, DC.
Practice, World Bank, Washington, DC. Pew Global Attitudes Survey. 2013. Pew
Holzmann, Robert, David A. Robalino, and R e s e a r c h C e n t e r, Wa s h i n g t o n , D C .
Noriyuki Takayama, eds. 2009. Closing the http://www.pewglobal.org/search/global
Coverage Gap: The Role of Social Pensions +attitudes+survey+2013/.
and Other Retirement Income Transfers. Somanathan, Aparnaa. 2015. “Health and Health
Washington, DC: World Bank. Care in an Ageing East Asia and Pacific.”
IFLS (Indonesia Family Life Survey). 2007. RAND, World Bank, Washington, DC.
Santa Monica, CA. http://www.rand.org/labor ThaiSES (Thai Household Social and Economic
/FLS/IFLS.html. Surveys). 2011. National Statistical Office,
IHME (Institute for Health Metrics and Bangkok. http://web.nso.go.th/survey/house
Evaluation). 2013. The Global Burden of _seco/socio.htm.
Disease: Generating Evidence, Guiding Policy. UN (United Nations). 2013. “World Population
Seattle, WA: IHME. Prospects: The 2012 Revision, Highlights
Jackson, Richard, and Tobias Peter. 2015. “From and Advance Tables.” Working Paper ESA/P/
Challenge to Opportunity: Wave 2 of the WP228, Department of Economic and Social
East Asia Retirement Survey.” Global Aging Affairs, Population Division, United Nations,
Institute, Alexandria, VA. New York.
The Demographic Transition
and Well-Being of Older People
in East Asia and Pacific
I
The Demographic and
Epidemiological Transitions 1
in East Asia and Pacific
This chapter is based on background papers by Thomas Flochel, Yuki Ikeda, Harry Moroz, and Nithin
Umapathi (2014) and by Aparnaa Somanathan (2015), with inputs from Ying Ho.
23
24 LIVE LONG AND PROSPER
To provide context for the remainder of modest shares of elderly similar to those found
the report, this chapter discusses both the in the Middle East and North Africa and in
demographic and the epidemiological tran- South Asia. Overall, in East Asia and Pacific,
sitions taking place in East Asia and Pacific. the share of the population ages 65 and above
The next section describes the demographic accounted for an average of around 7 percent
transition. It highlights aging trends across the of the total population in 2010 (figure 1.1).
region and compares the transition to those in This share is similar to that in Latin America
other regions of the world. It also examines and the Caribbean (6 percent) and somewhat
the factors underlying such aging trends and higher than the shares in the Middle East and
discusses trends in working-age population North Africa and in South Asia (5 percent).
and dependency ratios. The remainder of the However, the regional average for East Asia
chapter focuses on the epidemiological tran- and Pacific conceals enormous diversity:
sition, providing more in-depth discussion high-income countries have shares of older
of trends in life expectancy and the growing people that are close to the OECD average
burden of NCDs. (around 14 percent, which means that they
are aged societies according to the United
Nations definition), whereas many young and
The demographic transition in poorer countries have very modest shares of
East Asia and Pacific older people (for example, only 3 percent of
East Asia and Pacific is a diverse region in the populations of Papua New Guinea and
terms of the share of older people across coun- Timor-Leste are 65 years old and above—the
tries, ranging from levels found in member same as the average for Sub-Saharan Africa).
countries of the Organisation for Economic Despite the mixed picture on shares of
Co-operation and Development (OECD) to older people, in terms of absolute numbers
East Asia and Pacific has easily the largest
regional population of individuals who are
65 years old and above. China alone is home
FIGURE 1.1 East Asia and Pacific is a diverse region in terms of the
share of older people
to more old people than any other developing
Share of population ages 65 and above, 2010, selected regions and East region. In 2010, East Asian and Pacific econo-
Asian and Pacific subregions mies were home to around 187 million people
ages 65 and above (around 150 million of
18 them in developing countries). This number
Share of population ages 65 and older (%)
16
represents around 36 percent of the global
population in this age group. China is respon-
14
sible for the bulk of this elderly population,
12
with nearly 114 million people ages 65 and
10 older (figure 1.2).
8 East Asia and Pacific is also aging more
6 quickly than any other region in history.
4 The population ages 65 and over in many
2 countries in the region will increase from
0
7 percent to 14 percent of the total popula-
tion in just two or three decades—a change
CD rie e
em a
ia
r g
tri ng
P h tr ing
es
a
tri ng
OE unt om
in
LA unt pin
ric
As
co s
de ies
es
ig ies
tri
un pi
So es
EA un g
Af
co -inc
co er
un
co lo
co velo
EA uth
co lo
e
an
ev
h
v
de
Pd
A
h
EC
C
A
Sa
EN
M
Su
Central Asian countries in each 20-year FIGURE 1.2 East Asia, led by China, has more old people than any
period from 1955 to 2055. Between 2015 other region
and 2034, five-year growth rates of the Number of people ages 65 and above by region and East Asian and Pacific
population ages 65 and above will aver- subregion
age 22 percent in East Asia, second only to
rates in the Middle East and North Africa. OECD countries
As a result, by 2060, the average share of
population ages 65 and over in East Asia China
(22 percent) will approach that of the
OECD countries (24 percent) and Eastern South Asia
Europe and Central Asia (26 percent), with ECA emerging
the share in high-income East Asian and countries
Pacific economies and in China being even LAC developing
higher (figure 1.4). Whereas in 2010 only 1 countries
of the 25 “oldest” economies (by share of EAP high-income
economies
population ages 65 and older) was in East
EAP developing
Asia (Japan), by 2060 five East Asian econ-
countries
omies are projected to be among the 25.3
Although most of the region will age Sub-Saharan Africa
quickly, the onset of aging varies, and some MENA developing
countries will experience especially rapid countries
aging. Figure 1.3 shows the different points
40
0
0
20
60
80
10
12
14
in time and speeds at which East Asian and
Pacific economies have moved or will move Number of people age 65 and older, 2010 (millions)
from being classified as aging to aged (that
Source: World Bank estimates based on data from UN 2013.
is, moving from 7 percent of the population Note: EAP = East Asia and Pacific; ECA = Eastern Europe and Central Asia; LAC = Latin America
ages 65 and older to 14 percent). At one and the Caribbean; MENA = Middle East and North Africa; OECD = Organisation for Economic
Co-operation and Development. EAP developing excludes China.
extreme, Japan completed its transition to
an aged society decades ago. Others, includ-
ing China, Korea, and Thailand, are already
aging, and the poorer societies in the region the population across all world regions,
have decades to go before they cross that including in East Asia and Pacific. However,
threshold (although Cambodia and Myanmar the region’s average population of those
will do so much sooner). In terms of speed of ages 80 years and above will increase by
aging, the typical East Asian and Pacific econ- 6.2 percentage points on average between
omy shown in figure 1.3 moves from aging to 2010 and 2060, more than in any other
aged in around 25 years, but countries such region. For high-income East Asian and
as Indonesia, Korea, and Vietnam are notable Pacific economies, both the rate and ulti-
for their even more accelerated demographic mate level are even more pronounced, rising
transitions. Equally notable is that nearly all by 14 percentage points to 17 percent of
East Asian and Pacific countries move from the population. This group is important for
aging to aged societies relatively quickly future trends in demand for long-term care,
once the 7 percent threshold is reached (with with 50 and 64 percent of long-term care
exceptions such as Papua New Guinea and clients in Korea and Japan, respectively,
the Philippines). being 80 years old and above.
The population of those ages 80 and The demographic diversity of econo-
above—the “oldest old”—will also grow mies in East Asia and Pacific suggests three
faster in East Asia and Pacific than in any groups, which are used for much of the analy-
other region (figure 1.5). In 2010, the old- sis in this report.4 Hong Kong SAR, China;
est old represented a very small portion of Japan; Korea; and Singapore—the wealthiest
26 LIVE LONG AND PROSPER
FIGURE 1.3 East Asian and Pacific economies are aging more quickly than other economies in the past
Years to move from 7 to 14 percent population share 65 and older and the start and end years of transition
Vietnam (15)
Lao PDR (20)
Malaysia (20)
Indonesia (20)
Brazil (20)
Thailand (20)
Korea, Rep. (20)
Timor-Leste (25)
Mongolia (25)
Myanmar (25)
Cambodia (25)
Turkey (25)
China (25)
Singapore (25)
Japan (25)
Hong Kong SAR, China (30)
Philippines (35)
Papua New Guinea (40)
South Africa (40)
United Kingdom (45)
Russian Federation (50)
United States (69)
France (115)
1860 1885 1910 1935 1960 1985 2010 2035 2060 2085
Sources: World Bank estimates based on data from UN 2013 and Kinsella and He 2009.
Note: Figure shows starting and ending year for transition from 7 percent (aging) to 14 percent (aged) of population ages 65 and older. Aging and aged
thresholds are based on United Nations definitions. East Asia and Pacific economies rounded to five-year increments.
FIGURE 1.4 The share of elderly population in East Asia and Pacific will converge with that of OECD
countries by 2060
Share of population 65 and older by region and East Asian and Pacific subregion, 1950–2060
40
Share of population ages 65 and older (%)
35
30
25
20
15
10
0
50
60
70
80
90
00
10
20
30
40
50
60
20
19
19
19
19
19
20
20
20
20
20
20
economies in the region—are called the “red” FIGURE 1.5 The population of “oldest old” will increase more
economies. Red economies have the highest in East Asia and Pacific between 2010 and 2060 than in any other
number of people of advanced age, with an region
average of 14 percent of their total popula- Percentage of population 80 years old and above by region and East Asian
and Pacific subregion, 2010 and 2060
tions 65 years old and above (figure 1.6).
China, Indonesia, Malaysia, Mongolia,
tri e
ric n
de Asia
tri ng
ig tries
CD ries g
tri ng
tri ng
in
un om
still quite young—an average of 4 percent of
Af ara
t n
es
a
de es
un pi
es
es
un gi
un pi
un pi
Ch
n
h
co nc
co velo
EN uth
co mer
co velo
co velo
EA cou
Sa
i
h-
o
b-
e
e
S
Su
Pd
Ph
A
EC
C
in 2010—but they will begin to age in com-
OE
EA
LA
M
ing decades. This typology also holds for the 2010 2060
population 80 years of age and above: by
2060, the oldest old population will average Source: World Bank estimates based on data from UN 2013.
Note: EAP = East Asia and Pacific; ECA = Eastern Europe and Central Asia; LAC = Latin America
17 percent in red economies, 7 percent in and the Caribbean; MENA = Middle East and North Africa; OECD = Organisation for Economic
orange economies, and just 3 percent in green Co-operation and Development. EAP developing excludes China.
economies.
The pace and progression of aging are decline from 5.51 to 3.03 children per
reflected in the relative strength of concerns woman. Fertility rates in several East Asian
about aging among East Asian and Pacific countries are now among the lowest in the
populations. This comparison can be seen world. The decline in fertility can be seen
in figure 1.7, which shows findings of the across the region but has occurred unevenly
2013 Pew Global Attitudes Project survey across countries. The older East Asian coun-
on the share of people reporting aging as tries have extremely low total fertility rates
a “major problem” in their countries (Pew (TFRs), averaging 1.28 children per woman
Research Center 2014). Three countries (figure 1.8) in 2010. Although younger coun-
in the region easily rank highest, and even tries in the region have also experienced
Indonesia (with relatively young demo- significant declines in TFR, their rates are
graphics) shows results similar to those of more than twice as high (3.45 children per
the United States. In one sense, the results woman). The orange economies fall between
are positive in terms of the societal aware- the two extremes with an average TFR of
ness of the challenges of aging populations, 1.93. In this intermediate group, 2010 TFRs
but in another sense they reflect genuine had fallen to 1.66 in China, 1.41 in Thailand,
concerns about the readiness of countries to and 1.75 in Vietnam. Fertility rates are pro-
manage aging effectively. jected to continue to decline in the youngest
In most countries in East Asia and countries and to flatten in the middle ones.
Pacific, the decline in fertility rates has Notably, the United Nations (UN) projects a
been the major driver of population aging. modest increase in fertility rates in the oldest
Fertility rates declined dramatically from countries. However, as the UN itself recog-
5.91 children per woman in 1960 to 2.46 nizes, the recovery of TFRs from extremely
in 2005—significantly faster than the global low levels in the oldest countries in the region
28 LIVE LONG AND PROSPER
FIGURE 1.6 Three distinct aging patterns are apparent among East Asian and Pacific economies
Percentage of population 65 years old and above
40
36%
35
Share of population ages 65 and older (%)
30
24%
25
20
14%
13%
15
10
6%
5
4%
0
50
60
70
80
90
00
10
20
30
40
50
60
70
80
90
00
19
19
19
19
19
20
20
20
20
20
20
20
20
20
20
21
High-income economies, aged population
Middle- and lower-middle-income countries, aging population
Low-income countries, young population
is far from certain and is driven to a signifi- years in the red economies and is currently
cant extent by the convergence assumption much higher than that of their peers in the
used in the UN models (see box 1.1 regarding orange or green economies. Average gains
population projections).5 in life expectancy in the green, or youngest,
Although not as dramatic, changes in economies have been much less dramatic, so
age-specific mortality are also a notable the gap in life expectancy at age 60 between
part of the aging story, with sharp improve- the region’s youngest and oldest economies
ments in early life conditions increasing life is projected to grow from around 3 years in
expectancy in East Asia and Pacific faster 1950 to 7 years in 2010 and to 10 years by
than in any other region of the world. 2060. The reverse is true of the gap in life
Dramatic gains in life expectancy at birth in expectancy at birth between the green and
the region, from less than 45 years in 1950 red economies: the gap is projected to nar-
to 74 years today, mirror the decline in mor- row from 19 years in 1950 to 15 years in
tality rates (figure 1.9). The three groups of 2010 and to 14 years by 2060. This inver-
East Asian and Pacific economies exhibit sion is consistent with the youngest econo-
distinct patterns of life expectancy. Life mies becoming younger because mortality
expectancy of individuals ages 60 and 80 rates are lower at younger ages, whereas
has increased much more quickly in recent the oldest economies become older through
T H E D E M O G R A P H I C A N D E P I D E M I O LO G I C A L T R A N S I T I O N S I N E A S T A S I A A N D PAC I F I C 29
extensions in life expectancy at the oldest FIGURE 1.7 People in aging East Asian and Pacific countries see
ages. Bloom, Canning, and Finlay (2010) aging as a major challenge for their countries
calculate that the proportion of individuals Percentage of people by country who rate aging as a “major problem” for
0 to 5 years of age in East Asia and Pacific their country, selected countries
would have been 7.0 percentage points
higher in 2005 had fertility rates remained 100
at 1960 levels rather than declined. This 90
contrasts with a 0.1 percentage point 80
decline in the same population had age- 70
specific mortality rates remained at 1960 60
Percent
levels.6 In the future, the low fertility and 50
mortality rates that evolved between 1960
40
and 2005 will continue to drive population
30
aging as those birth cohorts move through
the age distribution, even as the future rates 20
themselves are projected to stabilize. 10
These uncertainties and assumptions 0
mean that all population figures and projec-
p.
a
n
ce
sia
y
ey
il
es
in
in
pa
an
Re
az
do
an
at
rk
ne
nt
Ch
Ja
Br
rm
ng
St
a,
Tu
Fr
tions should be treated with caution. The
ge
do
re
Ge
d
Ki
Ar
In
Ko
ite
d
selection of fertility scenarios, for instance,
ite
Un
Un
affects conclusions about the evolution
of population aging in the region. This Source: Pew Research Center 2014, based on the 2013 Pew Research Center Global Attitudes Project survey.
report—in line with common practice—uses
the medium fertility scenario throughout.
FIGURE 1.8 Fertility rates have declined significantly across all
Figure 1.10 shows the effect of different East Asian and Pacific economies but are projected to flatten in
TFR assumptions on the population share of orange and red economies
people 65 years of age and older by country Total fertility rates, 1950–2060
group. Although the division of East Asian
and Pacific economies into three typologies
is largely robust in the fertility scenarios, the 8
variation between scenarios within the red 7
and orange economies is nearly as large as
the variation across the two groups, showing 6
the significant uncertainty about projections
Total fertility rate
5
further into the future.
The UN extrapolations of fertility rates 4
are particularly controversial in the East
Asian context. The lack of parallelism in the 3
projected recovery of East Asian TFR, as
2
occurred in Europe, is debated for the fol-
lowing reasons. First, the impact of relax- 1
ing China’s one-child policy is not known.
Evidence shows that fertility norms have 0
50
60
70
80
90
00
10
20
30
40
50
60
19
19
19
19
20
20
20
20
20
20
20
The population data used in this chapter, which are tion of fertility rates, which range from low to high
drawn from the 2012 revision of the United Nations fertility. It makes additional assumptions in a sixth
(UN) World Population Prospects, involve important scenario that projects population with constant mor-
uncertainties and assumptions that affect long-term tality rates, and in a seventh scenario that projects
projections (UN 2013). First, historical fertility and population with no migration. An eighth scenario
life expectancy rates are estimated with significant projects population holding fertility and mortality
error rates, introducing imprecision into current constant. Although the UN invests significant effort
population figures. This uncertainty means that in providing the best projections possible, the com-
even past fertility, mortality, and population data plexity of the endeavor results in projections that
should be used with caution. Especially for devel- can be considerably inaccurate. This can be seen for
oping countries, the UN draws from a variety of selected East Asian and Pacific countries in figure
sources to develop its estimates. Second, UN projec- B1.1.1, which compares 1978 projections with 2000
tions require nontrivial assumptions about fertility, outcomes, highlighting both the significant average
mortality, and international migration. In response, margins of error and the variability of the errors
the UN provides five different scenarios for evolu- across the age distribution.
FIGURE B1.1.1 UN population projections for East Asia and Pacific have been subject to significant
uncertainty
Comparison of 1978 projections with 2000 outcomes
100
80
60
40
20
Percent error
–20
–40
–60
–80
–100
+
4
9
0–
5–
–1
–1
–2
–2
–3
–3
–4
–4
–5
–5
–6
–6
70
10
15
20
25
30
35
40
45
50
55
60
65
Age group
Indonesia Malaysia Philippines Singapore Thailand Vietnam
Source: UN 2013.
T H E D E M O G R A P H I C A N D E P I D E M I O LO G I C A L T R A N S I T I O N S I N E A S T A S I A A N D PAC I F I C 31
often care of the elderly (Jones 2011). The FIGURE 1.9 Declines in age-specific mortality rates in East Asia
third reason is the UN’s practice of project- and Pacific in recent decades have been substantial, particularly in
ing fertility, which is strongly influenced by the early years of life and for women
the small increases in total fertility observed Percentage of decline in age-specific mortality rates in East Asia and Pacific
by gender, 1970–2010
in recent years. Some research has suggested
that the recent increases were driven by
90
important but short-lived effects of auspi-
cious animal years (for example, the Golden 80
20 9
40 9
70 9
+
4
10 9
15 4
30 9
50 9
55 4
60 59
9
25 4
35 4
75 4
65 4
45 4
–1
–3
–6
1–
5–
–1
–2
–4
–5
–7
–2
–3
–7
80
–6
–4
–
experience sharp declines in their labor
forces in coming years. Figure 1.11 shows Age group
the change in shares and the total size of Male Female
the working-age population between 2010
and 2040. As expected, several economies Source: IHME and World Bank 2013.
FIGURE 1.10 The typology of East Asian and Pacific economies is robust in scenarios with low-, medium-,
and high-fertility assumptions
Effect of different total fertility rate assumptions on share of people 65 years of age and older by country group
0.45
0.4
Share of population ages 65 and older (%)
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
90
10
20
50
60
50
60
70
80
00
30
40
19
20
20
20
20
19
19
19
19
20
20
20
FIGURE 1.11 The working-age population will shrink over coming will experience substantial declines in their
decades in richer East Asian and Pacific economies shares of working-age population: over 15
Percentage change and change in absolute terms in population, ages percent in Korea and 10 percent or above
15–64, in selected economies, 2010–40 in China, Japan, and Thailand. Whereas in
relative terms China fares somewhat better
a. Percentage change
than richer East Asian and Pacific econo-
Hong Kong SAR, China mies, in absolute terms China dwarfs all
Korea, Rep. other economies, with an expected reduction
Singapore of 90 million people of working age between
Thailand 2010 and 2040. In contrast, in lower-income
Japan green economies and some orange econo-
China mies with younger populations, the share of
Vietnam
the working-age population is not expected
to shrink until after 2040. In absolute terms,
Mongolia
Indonesia and the Philippines will account
Malaysia
for a large share of the regional increase,
Myanmar
but in relative terms, Timor-Leste, Lao PDR,
Indonesia
Papua New Guinea, and the Philippines will
Cambodia lead the way.
Philippines The three groups of East Asian and
Papua New Guinea Pacific economies also exhibit distinct pat-
Lao PDR terns of dependency ratios. As shown in
Timor-Leste figure 1.12, changes in total dependency
–20 –15 –10 –5 0 5 10 15 ratios (TDRs) will be driven largely by ris-
ing elderly dependency ratios in rich and
Percentage change
middle-income economies of the region, and
both rising elderly and falling youth depen-
dency ratios will drive changes in younger
b. Absolute numbers and poorer East Asian and Pacific econo-
China mies. In the youngest economies, TDRs have
Japan been declining and are projected to continue
Thailand to decline until 2045. TDR in the oldest red
Korea, Rep. economies, in contrast, started increasing
Hong Kong SAR, China sharply from 2010 and will reach 94 depen-
Singapore dents for every 100 working-age individu-
Timor-Leste als by 2060. TDR has also begun a steep
Mongolia
increase in orange economies and will reach
Lao PDR
67 dependents for every 100 working-age
individuals by 2060. The rates of change
Papua New Guinea
in TDRs in East Asia and Pacific are much
Cambodia
steeper than those experienced by OECD
Myanmar
and emerging Eastern European and Central
Vietnam
Asian economies.
Malaysia
The classic measures of dependency in
Philippines terms of the rigid definition of working
Indonesia age have important limitations. In short,
–100 –80 –60 –40 –20 0 20 40 60 standard dependency ratios conflate age
Working-age population (millions) with dependency, both for those 65 years
of age and above and for those under
Source: World Bank estimates based on data from UN 2013. age 15. Economically, use of this measure
T H E D E M O G R A P H I C A N D E P I D E M I O LO G I C A L T R A N S I T I O N S I N E A S T A S I A A N D PAC I F I C 33
FIGURE 1.12 Total dependency ratios conceal contrasting trends in youth and elderly population shares
Elderly, youth, and total dependency ratios in three groups of East Asian and Pacific economies, 1950–2060
90 90
80 80
70 70
60 60
50 50
40 40
30 30
20 20
10 10
0 0
50
60
70
80
90
00
10
20
30
40
50
50
50
60
70
80
90
00
10
20
30
40
60
60
19
19
19
19
19
20
20
20
20
20
20
20
19
19
19
19
19
20
20
20
20
20
20
20
c. Green economies
their sum to the population ages 15–64 (%)
100
Ratio of population ages 0–14 and 65+ and
90
80
70
60
50
40
30
20
10
0
50
60
70
80
90
00
10
20
30
40
50
60
19
19
19
19
19
20
20
20
20
20
20
20
is justified for the young who are truly work, savings and assets, and other factors.
dependent, in most cases—though whether Some alternative definitions of dependency
15 years of age remains an appropriate are discussed in box 1.2.
threshold in societies where the majority of
young people complete secondary school
and many go on to postsecondary studies
The epidemiological transition in
is a question. However, for individuals 65
East Asia and Pacific
years of age and older, it presents a prob- Together with the demographic transition,
lem, given the variation in actual depen- East Asia and Pacific is experiencing a rapid
dency as a result of differential patterns of epidemiological transition, driven in part by
34 LIVE LONG AND PROSPER
Many attempts have been made to rethink depen- remaining life expectancy as the threshold. Empirical
dency ratios to make them more relevant to policy evidence linking this particular 15-year threshold to
making (Sanderson and Scherbov 2005, 2007, 2013; dependency is limited. Another alternative depen-
Spijker and MacInnes 2013). In a recent example, dency ratio makes additional progress in overcoming
Spijker and MacInnes (2013) propose using 15 years the arbitrary dependency cutoff. This measure uses
or less of remaining life expectancy as a proxy for the life-cycle pattern of consumption described in
dependency, arguing that remaining life expectancy is National Transfer Accounts to directly compute the
closely linked to health and active behaviors in addi- ratio of people weighted by age-specific earnings and
tion to being an important “second” component of age-specific labor force participation rates to people
age. The authors also propose the employed popu- weighted by age-specific consumption (Prskawetz
lation as a more accurate indicator of who works and Sambt 2014). The results are the opposite of
than the more traditional definition of working-age those found in Spijker and MacInnes (2013): in the
population as 15 to 64 years of age. The results of future, the Prskawetz and Sambt measure declines
applying this alternative old-age dependency ratio to more (that is, implies more dependency) than the tra-
the United Kingdom and several member countries ditional economic support ratio on which it is based,
of the Organisation for Economic Co-operation and which divides the working-age population by overall
Development are striking. The traditional old-age population and uses arbitrary age cutoffs.
dependency ratio results in far fewer working-age The contrasting results of Spijker and MacInnes
individuals for every person 65 years of age and older (2013) and Prskawetz and Sambt (2014) illustrate
and implies that this trend will continue. The Spijker that measures of aging should be approached with
and MacInnes measure, in contrast, finds that depen- an open mind. Nonetheless, both these alternatives
dency fell in the past, will continue to fall in the near are improvements on the traditional measure because
future, and will rise only gradually in the long term. they acknowledge that dependency ratios have policy
As the authors put it, “over the past four decades the implications: dependency is not only about demo-
population, far from aging, has in fact been getting graphic structure but also about behavior that policy
younger” (Spijker and MacInnes 2013, 21). can influence. These refined dependency ratios sug-
However, the bright picture of aging presented in gest potential policy remedies—such as incentives to
Spijker and MacInnes has its own issues. Although increase labor force participation of older people,
their alternative dependency ratio incorporates investments in health that are associated with length-
changes in education, female labor force participa- ier working lives, and adjustments to pension eligi-
tion, and early retirement, it still relies on an arbi- bility age—when policy makers consider dependency
trary cutoff for dependency, setting 15 years of ratios to be too high.
population aging but also by factors such as health, takes into account age-specific mortal-
higher incomes and urbanization. Apart from ity, morbidity, and functional health status.7
aging itself, a vital element of understanding The main hypotheses regarding how healthy
changing demographics is knowing whether life expectancy evolves with the increase in
people in the region are experiencing healthy life spans, and the empirical evidence to date,
aging and how the disease profile is chang- are summarized in box 1.3.
ing. “Healthy aging” implies not only living Health-adjusted life expectancy (HALE),
longer but also spending those additional or healthy life expectancy, increased in East
years of life in good, active health. Thus, Asia and Pacific between 1990 and 2010,
examining healthy life expectancy is useful. although slightly less than total life expec-
This measure, which estimates the number of tancy (figure 1.13). Each year of increase in
years that a person can expect to live in good life expectancy at birth has been accompanied
T H E D E M O G R A P H I C A N D E P I D E M I O LO G I C A L T R A N S I T I O N S I N E A S T A S I A A N D PAC I F I C 35
In the past 100 to 150 years, medical technologies and prevalence of chronic illnesses and morbidity
have been effective at saving lives from a range of are compressed toward the later period of life at a
fatal diseases. Often, however, this success has meant faster pace than reductions in mortality. People do
that the life years saved are spent in poor health and not just live longer; they live longer in better health.
characterized by chronic illness and disability. As • The dynamic equilibrium hypothesis (Manton
preventive health care throughout the course of life 1982). Under this hypothesis, the effects of com-
becomes more widespread, and as medical technolo- pression and expansion of morbidity offset one
gies improve in their ability to reduce mortality and another, with, on one hand, decreasing prevalence
morbidity, the additional years of life saved may be or incidence of chronic illness, and on the other
spent in good health. Better health goes along with hand, decreasing fatality rates of diseases, leading
fewer health care needs and may drive down use of to longer prevalence of disability.
health services and health expenditure (Rechel et al.
The empirical research is not clear on whether
2009). Therefore, an important way to measure
compression or expansion of morbidity occurs
progress is to determine if longevity is accompanied
in countries that have already experienced rapid
by more good health or less.
aging. For example, higher levels of some disabling
Three hypotheses have been put forward to pre-
conditions (dementia and musculoskeletal diseases)
dict possible future interaction between the evolu-
go along with falling rates of prevalence of other
tion in life expectancy and changes in prevalence of
conditions (cardiovascular and chronic respiratory
disability and bad health. This demonstrates that
diseases). The growing consensus from Organisa-
predicting changes in health status and morbidity is
tion for Economic Co-operation and Develop-
problematic, not least because epidemiological data
ment countries is as follows: the prevalence of
are subject to greater uncertainty. The three hypoth-
severe disability has been reduced substantially,
eses are as follows:
the prevalence of mild to moderate disability has
• The expansion of morbidity hypothesis (Gruen- increased, and overall, health status has improved
berg 1977; Olshansky et al. 1991; Verbrugge (Christensen et al. 2009).
1984). Under this hypothesis, the decline in mor- Trends in disability may reflect underlying trends
tality is largely due to a reduction in disease fatality in other domains, such as the rising use of assistive
rather than a reduction in prevalence or incidence technology, improved accessibility at home and in
of those diseases. Consequently, falling mortality public spaces, and changes in social perceptions of
is accompanied by an increase in morbidity and disability. The decline in disability might also reflect
disability. improvements in living conditions, not only of those
• The compression of morbidity hypothesis (Fries currently elderly, but also of their childhood living
1980, 1989). Under this hypothesis, the incidence conditions (Christensen et al. 2009).
by an 11-month increase in healthy life expec- region at age 50 by country and gender
tancy at birth. The biggest improvements in and shows the gender gap. The difference
healthy life expectancy in the region were between life expectancy and healthy life
in low-income countries, and the slowest expectancy can be interpreted as the aver-
improvements were in middle-income coun- age number of years of healthy life lost to
tries (IHME and World Bank 2013). poor health. HALE at age 50 is also a use-
Large differentials exist across East Asian ful indicator in thinking about appropriate
and Pacific countries in the number of years retirement ages in pension systems, as dis-
that people age 50 can expect to live in cussed in chapter 5. A significant gap exists
good health. Figure 1.14 compares healthy between total and healthy life expectancy
life expectancy with life expectancy in the in the region, with important implications
36 LIVE LONG AND PROSPER
FIGURE 1.13 Life expectancy and health-adjusted such as Indonesia and China. At the age of
life expectancy in East Asia and Pacific both 50, men and women in Japan can expect
increased between 1990 and 2010 to live free of disability approximately 10
Years of life expectancy and health-adjusted life years more than their counterparts in Papua
expectancy, by gender
New Guinea. The gender gap also is signifi-
cant in all East Asian and Pacific countries,
a. Male with women spending more years free of
80 disability than men. The policy and imple-
mentation implications of those systems are
70 discussed in chapter 6.
60
The leading causes of mortality and dis-
ability in East Asia and Pacific have shifted
Number of years
a. Life expectancy and health-adjusted b. Life expectancy and health-adjusted c. Gender gap at age 50
life expectancy for women at age 50 life expectancy for men at age 50
Papua New Guinea Kiribati Timor-Leste
Solomon Islands Papua New Guinea Fiji
Kiribati Solomon Islands Papua New Guinea
Marshall Islands Marshall Islands New Zealand
Micronesia, Fed. Sts. Mongolia Solomon Islands
Vanuatu Micronesia, Fed. Sts. Indonesia
Fiji Vanuatu Marshall Islands
Myanmar Myanmar Lao PDR
Lao PDR Fiji Australia
Timor-Leste Lao PDR Vanuatu
Mongolia Philippines Singapore
Cambodia Cambodia Samoa
Samoa Tonga Micronesia, Fed. Sts.
Tonga Samoa Thailand
Indonesia Timor-Leste Malaysia
Philippines Korea, Dem. People’s Rep. Cambodia
Korea, Dem. People’s Rep. Indonesia Myanmar
Malaysia Malaysia Korea, Dem. People’s Rep.
Thailand Vietnam Tonga
Vietnam Thailand Japan
China China Taiwan, China
New Zealand Taiwan, China China
Australia Korea, Rep. Korea, Rep.
Taiwan, China New Zealand Vietnam
Singapore Australia Kiribati
Korea, Rep. Singapore Philippines
Japan Japan Mongolia
0 10 20 30 40 0 10 20 30 40 0 2 4 6 8
Years Years Years
FIGURE 1.15 NCDs have increased sharply as leading causes of hearing loss (WHO 2014). Musculoskeletal
disease burden in East Asia and Pacific in recent decades conditions, which include lower back pain
Percentage change in total DALYs from 1990 to 2010 for East Asia and and neck pain, are major sources of dis-
Pacific by condition ability both globally and in East Asia and
Pacific for those ages 45 years and above.
Stroke The regional pattern of increased preva-
Ischemic heart disease lence of NCDs is mirrored in individual
Road injury
countries. Figure 1.16 shows the changes in
COPD
Low back pain
the leading causes of death in China from
Major depressive disorder 1990 to 2010 as an example of how NCDs
Lung cancer have increased as causes of mortality. Only
Lower respiratory infections
one communicable disease remains among
Diabetes
Liver cancer the leading causes of death in China, with
Tuberculosis 83 percent of total mortality attributable to
Falls NCDs and injuries (Bloom et al. 2014; IHME
Neck pain
Other musculoskeletal disorders
and World Bank 2013).
Preterm birth complications As people age and develop chronic
Self-harm illnesses, they are increasingly likely to have
Stomach cancer
several comorbid conditions, as is observed
Congenital anomalies
Neonatal encephalopathy
in East Asia and Pacific. Comorbidities can
Diarrheal diseases interact with each other and complicate
Cirrhosis treatment. Among the elderly in Korea, for
Iron-deficiency anemia
example, over half of all women and one-
HIV/AIDS
Chronic kidney disease third of men had three or more chronic
Drowning conditions (figure 1.17). Similarly, findings
–80 –60 –40 –20 0 20 40 60 80 100 120 140 160 180 200 from a 2008 Australian study reveal the
% change in total disability-adjusted life years, 1990–2010 extent of the problem among the elderly:
Communicable, newborn, nutritional, Noncommunicable diseases over half of elderly patients with arthritis
and maternal diseases also had hypertension, 20 percent had car-
Injuries
diovascular disease, 14 percent had dia-
betes, and 12 percent had mental health
Source: IHME and World Bank 2013. problems. Of those with cardiovascu-
Note: COPD = chronic obstructive pulmonary disease; DALYs = disability-adjusted life years; NCDs =
noncommunicable diseases. lar disease, 60 percent also had arthritis,
20 percent had diabetes, and 10 percent
had asthma or mental health problems
Timor-Leste, communicable diseases still (Caughey et al. 2008).
account for a significant share of prema- The burden of disease caused by chronic
ture mortality and morbidity, resulting in a conditions is also characterized by large
double burden of disease. inequalities based on social determinants
Sensory organ diseases (for example, such as poverty and lack of education.
cataracts) and musculoskeletal disorders In China, for instance, men and women
(for example, osteoarthritis) also feature from the poorest quintiles of the popula-
prominently among major causes of dis- tion were three to four times more likely to
ease burden in the region, especially in have a chronic illness in 2006 than those in
higher-income countries with a greater pro- the top three quintiles (figure 1.18) (World
portion of older people. The demographic Bank 2012). These inequalities persist even
shift toward older age groups and projected in high-income countries that have made
increases in longevity will inevitably lead significant progress in reducing the burden
to an increase in disability caused by sen- of cardiovascular disease. For instance, in
sory organ diseases, such as cataracts and Australia (Brown 2013) and New Zealand
T H E D E M O G R A P H I C A N D E P I D E M I O LO G I C A L T R A N S I T I O N S I N E A S T A S I A A N D PAC I F I C 39
(Wang et al. 2014), people from ethnic FIGURE 1.16 NCDs and injuries dominate the leading causes of
minorities are at significantly higher risk of mortality in China
experiencing cardiovascular disease than the Percentage change in total disability-adjusted life years (DALYs) by cause,
rest of the population. 1990–2010
The burden of NCDs across East Asia
Stroke
and Pacific reflects growing exposure to
Ischemic heart disease
risk factors, which are quite pronounced COPD
compared with the rest of the world Lung cancer
(figure 1.19). These risk factors include Liver cancer
the very high prevalence of smoking (see Stomach cancer
50
such as Malaysia and Thailand, the NCD-
40
related risk factors were the dominant causes
30
of DALYs, whereas in low-income and
lower-middle-income countries, iron defi- 20
ciency and childhood undernutrition were 10
dominant causes. Country-level evidence, 0
None One Two Three and more
although somewhat limited, points to signifi- Chronic conditions
cant socioeconomic gradients in exposure to Total Men Women
and treatment of NCD-related risks (boxes
1.5 and 1.6). Source: Sunwoo 2013.
40 LIVE LONG AND PROSPER
FIGURE 1.18 In China, in 2006, socioeconomic status had an Adverse initial conditions in life, including
inverse relationship with noncommunicable diseases low birthweight, have a significant impact on
Proportion of population with a chronic illness, by gender and income mortality and morbidity later in life, a find-
quintile ing that has significant implications for East
Asia and Pacific. First, current and future
40
elderly cohorts in low- and middle-income
Share of population (%)
BOX 1.4 Very high prevalence of smoking and related death rates in the region
East Asia and Pacific has the highest prevalence of smoking among men ages 15–69, at 54 percent,
smoking in the world, and smoking is among the is among the highest in the world. However, the
top three risk factors in both lower- and middle- number of daily cigarettes consumed per smoker
income countries. Two-thirds of the world’s tobacco in Asian countries is low compared with that in
users live in 15 countries (WHO 2011b), and 5 of developed countries (WHO 2011a). Despite the
these countries with high-risk burdens are in Asia: lower rates of consumption, smoking death rates
China, India, the Philippines, Thailand, and Viet- per smoker appear to be as extreme as those seen in
nam. China alone is the world’s largest producer the high-income countries, where smoking has been
and consumer of tobacco: prevalence of tobacco prevalent for decades.
• High body mass index (BMI) has increased. In correlation with the probability of having either
2007, around 31 percent of elderly women had high total cholesterol or low HDL.
a BMI of 25 or over, more than double the per- • Levels of hypertension are substantial but have
centage in 1993. Among elderly men in 2007, not changed over time. Among men ages 45 and
17 percent were overweight, compared with over older, around 44 percent had hypertension in
8 percent in 1993. 2007, the same percentage as in 1997. Similarly
• BMI increases with both education and income. among women, 53 percent were hypertensive,
Completing primary school (for women) and and the number did not change much over the
junior high school (for men) was positively asso- years. These are considered high prevalence rates
ciated with increased BMI, as was per capita at the population level. Overall, a strong posi-
income. tive relationship exists between age and being
• Waist circumference has also increased, lead- hypertensive.
ing to higher cardiovascular risks. For women • Education and income are associated with a
more than men, mean waist circumference has higher probability of the hypertension being diag-
increased while BMI has remained constant. nosed and of adherence to treatment. Probability
• Cholesterol levels are not correlated with educa- of underdiagnosis is larger for lower-income and
tion or income. Socioeconomic status has little uneducated individuals, particularly women.
importance of better understanding the house- already have proved technically and politi-
hold and individual behavior and preferences cally challenging, both in global contexts and
of the region’s older populations to inform in richer economies in the region. The fol-
policy choices. The “big picture” trends pre- lowing chapter moves from the big picture
sented in this chapter point to the need for and aims to provide insights into these policy
reforming policies in a number of areas that reforms.
42 LIVE LONG AND PROSPER
BOX 1.6 Malaysia: Undiagnosed noncommunicable disease risks and modest social
economic status differentials in management of risk
• Diabetes is largely undiagnosed or is not con- among men. Prevalence is higher among better-
trolled. Only 47 percent of diabetics in Malaysia off men, but no socioeconomic status gradient
are currently diagnosed. Of the 7 percent who exists for women.
reported being diagnosed as diabetic, only 41 • Significant socioeconomic status disparities can
percent (that is, 3 percent of the population) had be seen in the rates of diagnosis and treatment
controlled blood glucose levels. of hypercholesterolemia. A person in the richest
• Little variation by social economic status is quintile is 2.2 times more likely than a person in
evident in diabetes prevalence and treatment. the poorest quintile to have been diagnosed and
Overall patterns of actual diabetes prevalence and 2.1 times more likely to be successfully treated.
the likelihood of being diagnosed do not vary by • Hypertension remains largely undiagnosed. Over
income quintile. 32 percent of the population was found to have
• Hypercholesterolemia (high cholesterol in the blood hypertension, compared with the self-reported
stream) remains largely undiagnosed. Prevalence rate of 13 percent.
among adults in 2011 was 35 percent, with the • Modest socioeconomic status disparities can be
rate higher among females (40 percent) than males seen in the rates of diagnosis and treatment of
(30 percent). Only 8 percent had been advised by hypertension. A person in the richest quintile is
medical personnel about having the condition. 1.2 times more likely than a person in the poor-
• Large variation by socioeconomic status is evi- est quintile to have been diagnosed, and 1.5 times
dent in prevalence of hypercholesterolemia more likely to be successfully treated.
BOX 1.7 Adverse childhood socioeconomic conditions and increased mortality risks later in life
edited by Takatoshi Ito and Andrew Rose, IHME (Institute for Health Metrics and Evaluation)
61–89. Chicago: University of Chicago Press. and World Bank. 2013. The Global Burden of
Brown, Alex. 2013. “Essential Service Standards Disease: Generating Evidence, Guiding Policy.
for Equitable National Cardiovascular Care Seattle, WA: IHME.
for Aboriginal and Torres Strait Islander Jones, Gavin W. 2011. “Recent Fertility Trends,
People: An Exemplar Approach to Closing the Policy Responses, and Fertility Prospects in
Gap.” Presentation made at the World Heart Low Fertility Countries of East and Southeast
Federation’s World Congress of Cardiology, Asia.” Expert Paper 2011/5, Population
Melbourne, Australia, May. Division, Department of Economic and Social
Caughey, Gillian, Agena Vitry, Andrew Gilbert, Affairs, United Nations, New York. http://
and Elizabeth Roughead. 2008. “Prevalence of www.un.org/esa/population/publications
Comorbidity of Chronic Diseases in Australia.” / expertpapers/2011-5_Jones_Expert-Paper
BioMed Central Public Health 8: 221. http:// _FINAL_ALL-Pages.pdf.
www.biomedcentral.com/1471-2458/8/221. Kinsella, Kevin, and Wan He. 2009. An Aging
Choe, Minja Kim, and Robert D. Retherford. World: 2008. International Population Reports
2009. “The Contribution of Education to PS95/09-1. Washington, DC: U.S. Government
South Korea’s Fertility Decline to ‘Lowest- Printing Office.
Low’ Level.” Asian Population Studies 5 (3): Manton, Kenneth G. 1982. “Changing Concepts
267–87. of Morbidity and Mortality in the Elderly
Christensen, Kaare, Gabriele Doblhammer, Population.” Milbank Memorial Fund
Roland Rau, and James Vaupel. 2009. “Aging Quarterly 60: 183–244.
Populations: The Challenges Ahead.” Lancet Mundy, Simon. 2014. “South Koreans Alarmed
374 (9696): 1196–208. by Prospect of Extinction by 2750.” Financial
Finch, Caleb E., and Eileen M. Crimmins. 2004. Times, August 25, 2014.
“Inflammatory Exposure and Historical Olshansky, S. Jay, Mark A. Rudberg, Bruce Carnes,
Changes in Human Life Spans.” Science 305 Christine K. Cassel, and Jacob A. Brody. 1991.
(5691): 1736–39. “Trading Off Longer Life for Worsening
Flochel, Thomas, Yuki Ikeda, Harry Moroz, and Health” Journal of Aging and Health 3 (2):
Nithin Umapathi. 2014. “Macroeconomic 194–216.
Implications of Aging: Demographics, Labor Pew Research Center. 2014. “Attitudes about
Markets, and Productivity in East Asia Pacific.” Aging: A Global Perspective.” Pew Research
Background paper for the East Asia and Center, Washington, DC.
Pacific regional report on aging, World Bank, Prskawetz, Alexia, and Jože Sambt. 2014. “Economic
Washington, DC. Supports Ratios and the Demographic Dividend
Fries, James F. 1980. “Aging, Natural Death, and in Europe.” Demographic Research 30 (34):
the Compression of Morbidity.” New England 963–1010.
Journal of Medicine 303 (3): 130–35. Rechel, Bernd, Jonathan Erskine, Barrie
———. 1989. “The Compression of Morbidity: Dowdeswell, Stephen Wright, and Martin
Near or Far?” Milbank Memorial Fund McKee. 2009. “Capital Investment for Health:
Quarterly 67 (2): 208–32. Case Studies from Europe.” Observatory Study
Galobardes, Bruno, John Lynch, and George Davey 18, World Health Organization on behalf of
Smith. 2008. “Is the Association between European Observatory on Health Systems and
Childhood Socioeconomic Circumstances and Policies, Copenhagen.
Cause-Specific Mortality Established? Update Sanderson, Warren, and Sergei Scherbov. 2005.
of a Systematic Review.” British Medical “Average Remaining Lifetimes Can Increase as
Journal 62 (5): 387–90. Human Populations Age.” Nature 435 (7043):
Gruenberg, Ernest M. 1977. “The Failure of 811–13.
Success.” Milbank Memorial Fund Quarterly ———. 2007. “A New Perspective on Population
55 (1): 3–24. Aging.” Demographic Research 16 (2):
Gu, Baochang, and Yong Cai. 2009. “Fertility 27–58.
prospects in China.” UN Document UN/POP/ ———. 2013. “The Characteristics Approach
EGM-FERT/2009/6, Population Division, to the Measurement of Population Aging.”
Department of Economic and Social Affairs, Population and Development Review 39 (4):
United Nations, New York. 673–85.
T H E D E M O G R A P H I C A N D E P I D E M I O LO G I C A L T R A N S I T I O N S I N E A S T A S I A A N D PAC I F I C 45
Somanathan, Aparnaa. 2015. “Health and Health “Comparing Contemporary Coronary Artery
Care in an Ageing East Asia and Pacific.” Bypass Grafting across Six Main Ethnic
Background paper for the East Asia and Groups of New Zealand: Another Example
Pacific regional report on aging. World Bank, of Inequalities in Cardiovascular Disease.”
Washington, DC. Presented at the World Heart Federation’s
Spijker, Jeroen, and John MacInnes. 2013. World Congress of Cardiology, Melbourne,
“Population Ageing: The Time Bomb That Australia, May 4–7.
Isn’t?” British Medical Journal 347 (7933): World Bank. 2012. “Toward a Healthy and
20–22. Harmonious Life in China: Stemming the Rising
Sunwoo, Duk. 2013. “Economic Costs and Tide of Non-communicable Diseases.” Report
Opportunities for Aging and Health Care.” 62318-CN, World Bank, Washington, DC.
Presented at the Forum on Innovation for WHO (World Health Organization). 2011a.
Ageing Populations, Kobe, Japan, December MPOWER: A Policy Package to Reverse the
10–12. Tobacco Epidemic. WHO, Geneva.
UN (United Nations). 2013. “World Population ———. 2011b. Noncommunicable Diseases
Prospects: The 2012 Revision, Highlights, Country Profiles 2011. Geneva: WHO.
and Advance Tables.” Working Paper ESA/P/ ———. 2012. “Global Health and Aging.” WHO,
WP228, Department of Economic and Social Geneva.
Affairs, Population Division, UN, New York. ———. 2014. The Health of Older People in
Verbrugge, Lois M. 1984. “Longer Life but Selected Countries of the Western Pacific
Worsening Health? Trends in Health and Region. Geneva: WHO.
Mortality of Middle-Aged and Older Persons.” Yeung, Gary Y. C., Gerard J. van den Berg,
Milbank Memorial Fund Quarterly 62 (3): Maarten Lindeboom, and France R. M.
475–519. Portrait. 2014. “The Impact of Early-Life
Wang, Tom Kai Ming, Tharumenthiran Economic Conditions on Cause-Specific
Ramanathan, Sue Crengle, Greg Gamble, Mortality during Adulthood.” Journal of
Ralph Stewart, and Harvey White. 2014. Population Economics 27 (3): 895–919.
The Welfare and Labor Market
Behavior of Older People in 2
East Asia and Pacific
This chapter is largely based on two background papers, John Giles, Yuqing Hu, and Yang Huang (2015)
and John Giles and Yang Huang (2015), both of which contain fuller bibliographical references.
47
48 LIVE LONG AND PROSPER
FIGURE 2.1 Poverty rates increase with age in some East and Southeast Asian countries, but elderly
poverty has fallen over time
Percentage below poverty line by consumption and by income, by age, selected countries
40
Poverty rate (%)
40
35 35
30 30
25 25
20 20
15 15
10 10
5 5
0 0
25–45 46–60 61–70 71–80 81+ 25–45 46–60 61–70 71–80 81+
Age group Age group
Indonesia 2008 Mongolia 2007 Indonesia 2011 Mongolia 2011
Thailand 2006 Vietnam 2006 Thailand 2011 Vietnam 2012
40 40
35 35
30 30
25 25
20 20
15 15
10 10
5 5
0 0
25–45 46–60 61–70 71–80 81+ 25–45 46–60 61–70 71–80 81+
Sources: Giles and Huang 2015 based on data from CHNS 2006, 2011; ThaiSES 2011; World Bank East Asia and Pacific Standardized Household Surveys,
various years; and VHLSS 2012.
Note: Poverty is based on per capita daily income less than US$1.25 (2005 dollar purchasing power parity).
T H E W E L FA R E A N D L A B O R MA R K E T B E H AV I O R O F O L D E R P E O P L E I N E A S T A S I A A N D PAC I F I C 49
notable because it occurred during the period For Mongolia and Thailand, this difference is
when the social pension was universalized. relatively small, whereas for China, Indonesia,
In addition, a comparison of the profiles across the Philippines, and Vietnam with regard to
consumption (panels a and b) and income consumption poverty, the rural-urban divide
(panels c and d) shows that consumption- is much more pronounced. This variation
based poverty rates are generally smoother deserves more exploration but may reflect fac-
across the life cycle, as one would expect given tors such as the generally low level of poverty
the reliance on savings and assets. and much wider coverage of social pensions in
As one would expect, poverty among the Mongolia and Thailand, and perhaps factors
rural elderly is higher than for urban elderly such as relative differences in patterns of land
in nearly all cases (figure 2.2). Although and asset ownership across countries.
higher rural poverty among the elderly mir- The correlates of old age poverty in East
rors the spatial pattern of the general popula- Asia and Pacific are not surprising but provide
tion, the size of the difference in rural-urban grounds for optimism regarding the future of
elderly poverty rates across countries varies. poverty among older people. The correlates
FIGURE 2.2 Poverty among the rural elderly remains significantly higher than for urban elderly
Percentage below poverty line by age and rural or urban location, selected countries
Consumption poverty
a. Consumption-based poverty rates by age, urban, 2011–12 b. Consumption-based poverty rates by age, rural, 2011–12
40 40
35 35
30 30
Poverty rate (%)
Poverty rate (%)
25 25
20 20
15 15
10 10
5 5
0 0
25–45 46–60 61–70 71–80 81+ 25–45 46–60 61–70 71–80 81+
Income poverty
c. Income-based poverty rates by age, urban, 2009–2012 d. Income-based poverty rates by age, rural, 2009–2012
40 40
35 35
Poverty rate (%)
30 30
Poverty rate (%)
25 25
20 20
15 15
10 10
5 5
0 0
25–45 46–60 61–70 71–80 81+ 25–45 46–60 61–70 71–80 81+
Sources: Giles and Huang 2015 based on data from CHNS 2006, 2011; ThaiSES 2011; World Bank East Asia and Pacific Standardized Household Surveys, vari-
ous years; and VHLSS 2012.
Note: Poverty is based on per capita daily income less than US$1.25 (2005 dollar purchasing power parity).
50 LIVE LONG AND PROSPER
of poverty for China, Indonesia, Thailand, also declined sharply in both Thailand
and Vietnam are as follows: and Vietnam, where increases in average
educational attainment of adult children
• Not surprisingly, older people with high are associated with declines in the prob-
school education or above are less likely ability that both urban and rural residents
to be poor. Educated older people will will be income poor or consumption poor.
have earned higher incomes over their • Pension receipt is associated with signifi-
lives than their less educated peers and cant poverty reduction in both rural and
will thus have higher savings and other urban areas. Table 2.1 shows the reduc-
sources of wealth to support consump- tion in income poverty rates when older
tion. Reflecting this, attainment of sec- people in a household are receiving pen-
ondary education is associated with a sions. Reductions in income poverty are
4.0 percentage point reduction in the greater in rural areas of all countries,
probability that consumption falls below where the probability of falling into
the poverty line in rural China, and 8.4 poverty is greater and pension receipt is
and 11.5 percentage point reductions in lower. As one might expect, the effect on
urban and rural Indonesia, respectively. poverty is lower in Thailand given that
Significant reductions in consumption the large majority are receiving modest
poverty among the elderly are also associ- social pensions. Research conducted to
ated with educational attainment in urban identify a causal relationship finds that
and rural Thailand and rural Vietnam. the New Rural Pension Program in China
If one considers the higher educational contributes to poverty reduction among
attainment of current working-age adults, rural residents over age 60 (Zhang, Giles,
the relationship between own education and Zhao 2014).
and likelihood of falling into poverty sug- • Unlike the general adult population, one
gests that the elderly are likely to be better should not expect a priori to observe
off in the future. Although returns among positive or negative relationships between
older workers may decline with increases employment status and poverty for older
in the elder share with high school edu- people. Older workers with few resources
cation or more, the more highly educated may be unable to retire, and continued
future elderly will likely be employed in work may be important for staying out
less physically demanding jobs and are of poverty. Alternatively, some elderly
likely to adjust more readily to new work may continue to be poor even if they
environments and tasks than will less edu- are actively farming. In urban China
cated workers. and Indonesia, increases in the number
• Apart from own education, older people
who have had more children and more
TABLE 2.1 Lower elderly poverty rates when older
educated children are less likely to be
people receive pensions
poor. Declining family size with demo- Percent reduction in poverty rate among elderly
graphic transition may lead to fewer receiving pensions by rural or urban location, selected
sources of financial support for the countries
elderly, but increases in average educa-
tion promise to more than offset the Country Urban Rural
consequences of demographic change. China −10.5 −12.7
Earlier work demonstrates this in urban Indonesia −8.8 −13.9
China and suggests that declines in sup- Thailand −0.6 −1.2
port as a result of the One Child Policy in Vietnam −5.6 −22.1
China will not necessarily lead to finan- Sources: Giles, Hu, and Huang 2015, based on data from CHARLS 2011, IFLS
2007, ThaiSES 2011, and VHLSS 2012.
cial impoverishment of the elderly (Cai, Note: Results are significant at 1 percent level, except Thailand at 5 percent
Giles, and Meng 2006). Family size has level.
T H E W E L FA R E A N D L A B O R MA R K E T B E H AV I O R O F O L D E R P E O P L E I N E A S T A S I A A N D PAC I F I C 51
FIGURE 2.3 Difficulties with physical function and instrumental activities of daily living increase with age
Instrumental activity of daily living z-scores among elderly, selected countries
a. Men b. Women
1.0 1.0
Instrumental activities of
0.8 0.8
Instrumental activities of
daily living (z-score)
0.6 0.6
0.4 0.4
0.2 0.2
0.0 0.0
−0.2 –0.2
−0.4 –0.4
50 60 70 80 50 60 70 80
Age (years) Age (years)
Sources: Giles, Hu, and Huang 2015 based on data from CHARLS 2011, IFLS 2007, JSTAR 2012, and KLoSA 2010.
52 LIVE LONG AND PROSPER
FIGURE 2.4 Poorer and less educated elderly are more likely to face difficulty performing instrumental
activities of daily living
Instrumental activity of daily living z-scores among elderly, by education and income, selected countries
Instrumental activities of
0.8 0.8
daily living (z-score)
0.6 0.6
0.4 0.4
0.2 0.2
0.0 0.0
−0.2 −0.2
−0.4 −0.4
Elementary Elementary and High school Elementary Elementary and High school
middle school and above middle school and above
Education Education
China Indonesia
Japan Korea, Rep.
Instrumental activities of
0.8 0.8
daily living (z-score)
0.6 0.6
0.4 0.4
0.2 0.2
0.0 0.0
−0.2 −0.2
−0.4 −0.4
1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8
(poorest) (richest) (poorest) (richest)
Quintile Quintile
China Indonesia Korea, Rep.
Sources: Giles and Huang 2015 based on data from CHARLS 2011, IFLS 2007, JSTAR 2011, and KLoSA 2010.
T H E W E L FA R E A N D L A B O R MA R K E T B E H AV I O R O F O L D E R P E O P L E I N E A S T A S I A A N D PAC I F I C 53
about 10 percent for those living indepen- and age is much stronger in countries such
dently, 25 percent for those with concur- as China and the Republic of Korea than in
rent chronic diseases, and even higher for Indonesia and Japan. In addition to being a
those in institutional care (McDougall et al. type of disability, depression is also a cause
2007). 4 Surveys in the region show that of higher suicide rates among the elderly
rates of depression tend to increase with (Reynolds and Kupfer 1999). The sharp
age and are greater among poor and less increase in depression rates with old age in
educated elderly people (figure 2.5).5 Rates Korea is especially notable because suicide
also tend to be higher for older women than rates among elderly Koreans are some of the
older men, in part perhaps because of the highest in the world at 80 per 100,000 in
higher likelihood of women living alone. 2011: the rate has increased fivefold in the
The correlation between depression rates past two decades (Eyres 2014). Suicide rates
FIGURE 2.5 Rates of depression tend to increase with age and are higher among poor households
Percentage depressed, according to the CES-D 10 measure, by age, gender, and income, selected countries
50 50
40 40
30 30
20 20
10 10
0 0
50 60 70 80 50 60 70 80
Years Years
China Indonesia
Japan Korea, Rep.
50 50
40 40
30 30
20 20
10 10
0 0
1 2 3 4 5 1 2 3 4 5
(poorest) (richest) (poorest) (richest)
Quintiles Quintiles
China Indonesia Korea, Rep.
Sources: Giles and Huang 2015 based on data from CHARLS 2011, IFLS 2007, JSTAR 2011, and KLoSA 2010.
Note: CES-D 10 = Center for Epidemiologic Studies Depression Scale.
54 LIVE LONG AND PROSPER
are also high among the elderly in Japan Consistent with descriptive results, individu-
and are linked to loneliness and income- als with secondary education and above in
related stressors. China, Indonesia, and Korea are less likely
We a k e n i n g c o g n i t i v e f u n c t i o n , to be depressed, as are the elderly with ter-
particularly dementia, is another aspect tiary education or greater in Japan. Whether
of old age and a major source of disability because of a source of pride and a sense of
and dependence. Dementia is not easy success in raising their children or simply
to diagnose, and estimates of prevalence as a reflection of the prospect of higher pri-
vary considerably because of measurement vate transfers, the elderly in both China and
issues. In 2000, just under 7 percent of peo- Indonesia are happier when they have more
ple ages 65 and above in Organisation for educated children.6
Economic Co-operation and Development In addition to their income support role
(OECD) member countries were estimated for the elderly, public pensions also have
to be affected by dementia, with incidence the potential to support other dimensions
of the disease rising sharply with age: fewer of well-being. Receipt of a pension is asso-
than 3 percent of those ages 65–69 years, ciated with higher body mass index in rural
but almost 30 percent of those ages 85–89 Indonesia and with significant reductions
years (WHO 2012). Alzheimer’s disease, the in the likelihood of depression among the
most common form of dementia, accounted elderly in both rural and urban areas of China
for between two-fifths and four-fifths of all and Indonesia.
dementia cases in the OECD study. Evidence
on dementia in East Asia and Pacific is more
limited. In Korea, the prevalence of demen-
Living arrangements and access
tia among elderly people was estimated to
to support among older people
be 9.4 percent in 2013, of which over two- Co-residence of older people with adult chil-
thirds were described as having mild to mod- dren is significant, although it varies consid-
erate dementia. In Indonesia, analysis shows erably within East Asia and Pacific. Across
a strong negative relationship between age the developing world, co-residence with adult
and memory for men and women (for Korea, children is frequently viewed as an impor-
Sunwoo 2013; for Indonesia, Witoelar, tant indicator of whether the elderly have
Strauss, and Sikoki 2012). With life expec- access to both instrumental and financial
tancy increases, a substantial rise in preva- support. In East Asia and Pacific, co-residence
lence of dementia is anticipated, although of older people is quite high and gener-
there is some evidence of compression of ally increases with age between age 50 and
cognitive decline at older ages in OECD mid-60s (figure 2.6). For the elderly ages 60
countries (Christensen et al. 2009). and above, co-residence rates vary signifi-
If one looks at multivariate findings, cantly across countries in the region and by
living in poverty, having low education gender and urban-rural location within coun-
levels, and being female are associated with tries. Rates are particularly high in countries
deterioration in the subjective well-being of such as the Lao People’s Democratic Republic
older people (higher CES-D 10 [Center for and Timor-Leste and, consistent with the
Epidemiologic Studies Depression Scale] global pattern, tend to be lower in countries
scores) in urban and rural areas of China, at higher per capita income levels (Evans and
Indonesia, Japan, and Korea. In rural areas of Palacios 2015).
China, Indonesia, and Korea, one observes a At the same time, in a number of East Asian
significant association between increasing age and Pacific countries, elderly co-residence has
and decreasing subjective well-being. In addi- declined significantly over time in the face of
tion, women tend to have a higher incidence smaller families, higher incomes, and factors
of depression than men in rural areas of such as migration (figure 2.7). Compounding
China and Indonesia, in Japan, and in Korea. the demographic facts, privacy may be viewed
T H E W E L FA R E A N D L A B O R MA R K E T B E H AV I O R O F O L D E R P E O P L E I N E A S T A S I A A N D PAC I F I C 55
FIGURE 2.6 Elderly co-residence rates are generally high in East Asia and Pacific but vary across
countries and by gender and location within countries
Share of elderly by age and gender living with an adult child, selected countries
100 100
90 90
80 80
Share of population (%)
70 70
60 60
50 50
40 40
30 30
20 20
10 10
0 0
40 50 60 70 80 40 50 60 70 80
Age (years) Age (years)
100 100
90 90
80 80
Share of population (%)
70 70
60 60
50 50
40 40
30 30
20 20
10 10
0 0
40 50 60 70 80 40 50 60 70 80
Sources: Giles and Huang 2015 based on data from IFLS 2007; World Bank East Asia and Pacific Standardized Household Surveys, various years; and VHLSS 2012.
Note: Adult child is defined as child age 24 and above.
as a normal good by both older and younger which finds both declining co-residence as
generations. When the elderly and the young countries get richer (figure 2.8) and lower
have the ability to live separately, they may co-residence among the richer elderly across
simply choose to do so, without any con- countries (median co-residence in a sam-
sequence for well-being, as suggested by ple of 61 developing countries in the mid-
research from China (Giles and Mu 2007; Lei 2000s was over 83 percent for people ages
et al. 2012). This view is also consistent with 60 and above in the poorest quintile but fell
global evidence from developing countries, to only 64 percent for the richest quintile)
56 LIVE LONG AND PROSPER
FIGURE 2.7 Co-residence rates of older adults with adult children have declined over time in
China and Thailand
Percentage by age of older adults living with adult children, various years
a. China b. Thailand
100 100
90 90
80 80
Share of population (%)
Sources: Giles and Huang 2015 based on data from CHARLS 2011; China Census 1982, 1990, and 2000; and ThaiSES, various years.
Note: Adult children are defined as ages 24 and above.
FIGURE 2.8 As countries get richer, elderly co-residence rates tend by 2011, while in Korea, co-residence of
to decline those ages 65 and above decreased from over
Share of older people co-residing with non-elderly and GNI per capita, 80 percent in 1980 to well under 30 percent
latest available year by 2010 (see Lee 2009). The speed of declines
in co-residence contrasts with trends in a
18,000
number of developed countries, such as the
GNI per capita (US$ PPP-adjusted)
16,000
United States, where elderly co-residence was
14,000
72 percent in 1880 and fell below 30 percent
12,000 only during the 1960s (Costa 1998). Most
10,000 East Asian and Pacific countries also have
8,000 notable variations between different groups
6,000
R2 = 0.509
of older people by rural-urban location and
4,000 gender. Cai, Giles, and Meng (2006) also find
2,000 that adult children in China are more likely
0 to reside with older parents who have higher
40 50 60 70 80 90 100 incomes and more housing wealth.
Share of co-residing elderly (%) Although co-residence trends are impor-
tant, they need to be interpreted with caution
Source: Evans and Palacios 2015. in terms of potential effects on the welfare
Note: GNI = gross national income; PPP = purchasing power parity. Elderly are defined as people
60 years and over, living with working-age adults whether or not related. of older people. On the one hand, evidence
shows that co-residence has welfare benefits
in East Asia and Pacific. In both China and
(Evans and Palacios 2015). Since the mid- Indonesia, for example, poverty and depres-
1980s, co-residence with adult children in sion rates and ADL scores are higher for older
Thailand has declined steadily to only 56 people who live separately from children liv-
percent in 2011 (Knodel 2014b). Rates of co- ing in the same area than for those who live
residence have also been declining in China with adult children or have migrant children
and Korea. In China, the co-residence rate for living farther away. On the other hand, the
those ages 65 to 70 fell from nearly 66 per- differences are not dramatic and, nota-
cent in the early 1980s to roughly 43 percent bly, separate residence does not necessarily
T H E W E L FA R E A N D L A B O R MA R K E T B E H AV I O R O F O L D E R P E O P L E I N E A S T A S I A A N D PAC I F I C 57
mean a decline in intergenerational support, income for urban and rural people ages 60 to
as demonstrated in China by research using 85, respectively.7 With the notable exceptions
CHARLS (China Health and Retirement of urban China and Mongolia, labor contin-
Longitudinal Study) and noted by researchers ues to be the dominant source of income for
in Southeast Asia (Giles, Lei et al. 2015; older people across the region, even in OECD
Knodel 2014a). countries and in Korea (owing in part to its
still immature pension system). Labor income
also maintains its overall importance with
Sources of financial support and regard to the elderly ages 70 and above in
labor market behavior of older the region (see discussion below). In contrast,
workers public transfers are a relatively minor share
Labor income remains the dominant income of elderly income in most cases, including in
source for people ages 60 and above in nearly rural China where the relative importance
all East Asian and Pacific countries, and even of labor and transfers is very different from
more so in rural areas in most of those coun- that in urban areas. The relative importance
tries. Figure 2.9 shows the main sources of of public and private transfers is also notable
FIGURE 2.9 Labor is the leading income source for people ages 60 to 85 in various East and Southeast
Asian countries
Percentage of total income of people ages 60 and above, in rural and urban areas, from labor, public and private
transfers, and other sources, selected countries
a. Urban
100
90
80
% of total income
70
60
50
40
30
20
10
0
p.
a
sia
te
lia
m
s
nd
in
di
ne
Re
es
na
go
ne
Ch
bo
ila
pi
-L
a,
et
on
do
lip
a
m
or
re
Vi
Th
M
Ca
In
Ko
Tim
Ph
b. Rural
100
90
80
% of total income
70
60
50
40
30
20
10
0
.
a
ia
lia
ste
m
ep
d
in
di
ne
es
na
an
go
,R
e
Ch
bo
pi
n
-L
et
ail
on
ea
do
ilip
m
or
Vi
Th
r
M
Ca
In
Ko
Tim
Ph
Sources: Giles and Huang 2015 based on data from CHARLS 2011; IFLS 2007; KLoSA 2010; ThaiSES 2011; World Bank East Asia and Pacific Standardized
Household Surveys, various years; and VHLSS 2012.
58 LIVE LONG AND PROSPER
in China, including very low private trans- additional public transfers for older people
fers in several cases, but significant in other interact with private transfers and whether
countries like Korea and the Philippines, such transfers simply displace private trans-
and generally plays a slightly more signifi- fers at different income levels of older peo-
cant role in rural areas in most countries. ple (so-called crowding out). Analysis of the
Overall, the commonly held assumption responsiveness of private transfers to low
that familial support is the dominant source incomes in households with elderly residents
of income for elderly people in East Asia for this report shows the following:
and Pacific is not borne out by the evidence,
• For low-income elderly in China and
although such support clearly remains impor-
Indonesia, private transfers are on aver-
tant, especially in settings like rural China
age insufficient to lift older people out of
and the Philippines where remittances are
poverty.
significant.8
• Transfers from adult children tend to be
The role of private and public transfers
significantly higher when some children
in reducing elderly poverty varies across the
are migrants (around twice the level on
region, although private transfers remain
average in China and Thailand for elderly
important in nearly all East Asian and Pacific
households at half the poverty line).
countries. To gauge the relative importance
• A larger number of adult children is
of public and private transfers for older peo-
associated with higher private transfers
ple in the region, we follow an approach used
to older people, but higher educational
by Park, Lee, and Mason (2012) to examine
attainment of children is also associated
the likely rates of poverty in the absence of
with higher transfers, suggesting that
public and private transfers.9 Figure 2.10
smaller family size can be offset in part by
shows poverty rates by age, including with
better-educated (and paid) children.
all earnings and transfers, without pub-
• Most important, in many cases the effect
lic transfers, without private transfers, and
of increased transfers from nonfamily
without all transfers. For China, Korea, and
sources (including the state) on transfers
Mongolia, public transfers play a more sig-
from adult children is modest.
nificant role in reducing poverty. In China,
for example, eliminating public transfers These findings can be seen in table 2.2,
would lead to an increase in the poverty rate which shows the elasticities of private transfers
of a 70-year-old person by nearly 20 percent- to older family members to increased transfers
age points (a 70 percent increase in poverty). from other sources at different income levels
In contrast, elimination of private transfers of the elderly household. For both China and
would lead to only a five percentage point Indonesia, the evidence of crowding out is
increase in the elderly poverty rate. In the weak, with a reduction of only 3 to 9 percent
Philippines, Thailand, and Vietnam, private in private transfers to the poorest elderly
transfers play a significant role in reducing households in response to increased transfers
poverty, though public transfers in Thailand from other sources, including social pensions.
and Vietnam also play an important role The pattern also holds for elderly households
in supporting the elderly. However, not all at and above the poverty line in both coun-
elderly in the region benefit from substantial tries and is robust for households with and
private and public transfers. In Cambodia without adult migrant children. Thailand is
and Indonesia, both public and private trans- distinguishable in terms of the scale of crowd-
fers play a relatively minor role in poverty ing out at low incomes, though in most cases,
alleviation. the reduction in private transfers is no more
Available evidence from East Asia and than one third. These findings have important
Pacific suggests that the risk of public transfers implications for pension and transfer policies
significantly crowding out private transfers as countries throughout the region seek to
is modest. Policy makers must consider how expand coverage of public pension programs.
T H E W E L FA R E A N D L A B O R MA R K E T B E H AV I O R O F O L D E R P E O P L E I N E A S T A S I A A N D PAC I F I C 59
FIGURE 2.10 Source of financial support can affect poverty rates in various East and Southeast Asian countries
Percentage below the income poverty line with and without public and private transfers, selected countries
70 70
60 60
50 50
40 40
30 30
20 20
10 10
0 0
40 50 60 70 80 40 50 60 70 80
Age (years) Age (years)
Sources: Giles and Huang 2015 based on data from CHARLS 2011; IFLS 2007; KLoSA 2010; ThaiSES 2011; World Bank East Asia and Pacific Standardized Household Surveys, various years;
and VHLSS 2012.
Note: Poverty for China is defined as daily per capita income less than Y 6.3 (official poverty line). Poverty line for the Republic of Korea is defined as half of daily mean national per
capita income in 2009: 20.8 won. Poverty lines for other countries are defined as per capita daily income (2005 dollars purchasing power parity) less than US$1.25.
60 LIVE LONG AND PROSPER
TABLE 2.2 Private transfers to elderly people are not significantly affected by level of transfers from public
sources in East and Southeast Asia
Elasticities of private transfers to the elderly to increased transfers from other sources (percent change in private
transfers), selected countries
Country 0.5 × Poverty line Poverty line 2.0 × Poverty line 4.0 × Poverty line
China
With migrants −0.027 −0.103 −0.046 −0.058
Without migrants −0.062 −0.049 −0.009 −0.006
Indonesia
With migrants 0.018 −0.086 0.114 0.035
Without migrants −0.092 0.010 0.010 −0.022
Thailand
With migrants −0.177 −0.144 0.048 0.156
Without migrants −0.194 −0.278 −0.126 −0.034
Vietnam
With migrants −0.379 −0.067 −0.105 −0.060
Without migrants −0.196 −0.077 −0.042 −0.041
Sources: Giles and Huang 2015 based on data from CHARLS 2011, IFLS 2007, and ThaiSES 2011.
Another way to assess the relative impor- 2004 and precede the major expansion of
tance of public and private transfers and self- social pensions, and data for the Philippines
support is to look across the life cycle, using are from 1999 and precede the introduction
the life-cycle deficit approach pioneered of a major conditional cash transfer pro-
within the National Transfer Accounts gram that now covers around 20 percent of
(NTA). NTA is a system of national accounts households.) Notably, NTA does not capture
decomposed by age profile. It reflects a cohort differences.
whole-of-life approach that captures the net Although public transfers currently play
contribution of public and private trans- a relatively modest role in old age support in
fers and asset-based reallocations (through East Asia and Pacific relative to other regions,
channels such as housing, land, household social expectations of the relative roles of
durables, asset income, and savings) to the the state, families, and retirees themselves in
life-cycle deficit. The system is being more postretirement support are changing rapidly.
widely adopted in Asia and globally and Policy makers and citizens in the region point
provides a rich tool for cross-country com- to the important role of family support net-
parisons.10 As figure 2.11 shows, asset-based works, though the earlier analysis shows limi-
reallocations are the dominant source of tations of such networks in providing elderly
support in Indonesia, the Philippines, and financial support. However, opinion surveys
Thailand, whereas public transfers assume a also show that expectations are changing
more significant role in richer Japan; Korea; rapidly, as shown in figure 2.12, where the
and Taiwan, China, as one might expect. expected roles of self-support and the state
The relative role of public transfers in East in financial support easily dominate expec-
Asian and Pacific middle-income countries tations of support from adult children and
is minor compared to other middle-income family. The results also indicate that expecta-
countries such as Brazil, although famil- tions of the state are considerably higher in
ial transfers are more significant except in the region’s middle-income countries than
Japan. The results are a useful confirma- in its higher-income countries like Korea
tion of the broad patterns seen in the analy- and Singapore where societal aging is more
sis above, though some differences appear, advanced. For caregiving support, the expec-
most likely driven by the time period of the tations of the state are also growing (and
data. (For example, Thai NTA data are from are already higher in richer East Asian and
T H E W E L FA R E A N D L A B O R MA R K E T B E H AV I O R O F O L D E R P E O P L E I N E A S T A S I A A N D PAC I F I C 61
FIGURE 2.11 The relative roles of public and private transfers and of savings and assets across the life
cycle varies notably across East and Southeast Asian economies and between East and Southeast Asia and
other regions
Percentage of average life-cycle deficit for public and familial transfers and asset-based reallocations
130
120
110
100
90
% of average life-cycle deficit
80
70
60
50
40
30
20
10
0
–10
–20
–30
p.
s
es
sia
na
an
ico
en
nd
il
ne
an
di
Re
az
at
hi
p
ne
ed
pi
In
ila
ex
rm
Br
Ja
St
a,
,C
ilip
do
Sw
a
M
re
Ge
d
Th
an
In
ite
Ko
Ph
iw
Un
Ta
FIGURE 2.12 Attitudes on the preferred source of financial support in old age and on the expected
source of personal care are shifting in East and Southeast Asian economies
Percentage of adults reporting government, selves, or family as likely sources of support, selected economies
a. Who, ideally, should be most responsible for providing income to retired people?
70
60
% of respondents
50
40
30
20
10
0
p.
m
na
ia
sia
re
nd
ne
in
Re
ys
na
po
hi
ne
Ch
ila
pi
ala
et
,C
a,
ga
do
a
ilip
re
Vi
Th
M
an
Sin
In
Ko
Ph
iw
Ta
60
50
% of respondents
40
30
20
10
0
p.
m
na
ia
ia
re
d
ne
in
Re
ys
es
an
na
po
hi
Ch
pi
ala
on
ail
et
,C
a,
ga
ilip
re
Vi
d
Th
M
an
Sin
In
Ko
Ph
iw
Ta
and elderly with high school or college educa- people in work as they age. East and Southeast
tion become evident at somewhat older ages, Asian countries commonly have self-employed
although in Vietnam, the gaps across educa- rates of 90 percent for older rural men and
tion groups are not as pronounced. These pat- women, most of whom work in agriculture or
terns contrast with the EU13,11 where partici- related activities. In urban areas, an increase in
pation rates of those with college education the self-employed share around age 60 reflects
are 20 to 25 percentage points higher than the fact that self-employed urban workers,
those with middle school education at age 60 often in commerce and trade, remain in the
(Giles, Lei et al. 2015). workforce while employees from the formal
As figure 2.15 shows, self-employment sector retire. In countries with a higher share of
accounts for an increasing share of older urban residents in the informal self-employed
T H E W E L FA R E A N D L A B O R MA R K E T B E H AV I O R O F O L D E R P E O P L E I N E A S T A S I A A N D PAC I F I C 63
FIGURE 2.13 Labor force participation rates vary by location and gender in various East and Southeast Asian countries
Percentage in labor force by age, gender, and urban or rural location, selected countries
% of cohort working
% of cohort working
70 70
60 60
50 50
40 40
30 30
20 20
10 10
0 0
40 50 60 70 80 40 50 60 70 80
Age Age
70 70
60 60
50 50
40 40
30 30
20 20
10 10
0 0
40 50 60 70 80 40 50 60 70 80
Age Age
70 70
60 60
50 50
40 40
30 30
20 20
10 10
0 0
40 50 60 70 80 40 50 60 70 80
Age Age
Sources: Giles, Hu, and Huang 2015 based on data from CHARLS 2011; IFLS 2007; KLoSA 2010; World Bank East Asia and Pacific Standardized Household Surveys, various years; and
ThaiSES 2011.
64 LIVE LONG AND PROSPER
FIGURE 2.14 Better-educated people tend to withdraw from the labor force earlier in East and Southeast Asia
Employment rate by age, gender, and urban or rural location, selected countries
80 80
60 60
40 40
20 20
0 0
40 50 60 70 80 40 50 60 70 80
Age (years) Age (years)
c. Indonesia, men d. Indonesia, women
100 100
Employment rate (%)
60 60
40 40
20 20
0 0
40 50 60 70 80 40 50 60 70 80
Age (years) Age (years)
e. Thailand, men f. Thailand, women
100 100
Employment rate (%)
80 80
60 60
40 40
20 20
0 0
40 50 60 70 80 40 50 60 70 80
Age (years) Age (years)
g. Vietnam, men h. Vietnam, women
100 100
Employment rate (%)
80 80
60 60
40 40
20 20
0 0
40 50 60 70 80 40 50 60 70 80
Age (years) Age (years)
Middle school or less High school College and above
Sources: Giles, Hu, and Huang 2015 based on data from CHARLS 2011, IFLS 2007, ThaiSES 2011, and VHLSS 2012.
T H E W E L FA R E A N D L A B O R MA R K E T B E H AV I O R O F O L D E R P E O P L E I N E A S T A S I A A N D PAC I F I C 65
FIGURE 2.15 Self-employment is an increasingly important source of work as people age in East Asia
and Pacific
Percentage of self-employed as a share of total in work, by age, gender, and rural or urban location, selected countries
a. China b. Indonesia
100 100
90 90
who are self employed (%)
Share of population
70 70
60 60
50 50
40 40
30 30
20 20
10 10
0 0
40 50 60 70 80 40 50 60 70 80
Age (years) Age (years)
c. Thailand d. Timor-Leste
100 100
90 90
who are self employed (%)
80 80
Share of population
Share of population
70 70
60 60
50 50
40 40
30 30
20 20
10 10
0 0
40 50 60 70 80 40 50 60 70 80
Age (years) Age (years)
Urban, men Urban, women Rural, men Rural, women
Sources: Giles, Hu, and Huang 2015 based on CHARLS 2011; IFLS 2007; ThaiSES 2011; and World Bank East Asia and Pacific Standardized Household Survey
for Timor-Leste, various years.
sector—Timor-Leste is an extreme case—one 30–45 hours per week depending on the coun-
does not see as big an increase in the self- try. The situation in rural areas is even more
employed share of older urban residents. striking, with rural men continuing to work in
For those who continue working at older most countries 30–40 hours per week even at
ages in East Asia and Pacific, hours of work age 75 and women working between 20 and
remain quite high. In developed countries, 35 hours per week at the same age. Given the
retirement is often a gradual process: workers nature of most rural work, such long hours and
may reduce work hours at current employers, high participation rates confirm the impression
move to work arrangements requiring fewer of many rural people in the region having to
hours per week, or even transition in and out “work till they drop.”
of retirement. In contrast, in much of East Asia Differences exist across countries also.
and Pacific, hours of work for older people con- In urban China and Vietnam, hours of work
ditional on working remain quite high in both for elderly still working (primarily the self-
rural and urban areas (figure 2.16). For people employed) tend to decline little and then drop
working at age 65 in the six countries shown, off sharply at age 70 and age 65, respectively,
urban men work on average 40 hours or more in contrast to the more gradual decline in
per week, and all women work on average hours worked in other countries. In rural
66 LIVE LONG AND PROSPER
FIGURE 2.16 Those who continue to be employed in older age work long hours until advanced ages in
East and Southeast Asia
Hours of work per week among those working, by age, gender, and rural or urban location, selected countries
60 60
Hours worked per week
50 50
40 40
30 30
20 20
10 10
0 0
45 50 55 60 65 70 75 80 85 45 50 55 60 65 70 75 80 85
Age (years) Age (years)
China Indonesia Korea, Rep. Vietnam Philippines
Sources: Giles, Hu, and Huang 2015 based on data from CHARLS 2011, FIES 2009, IFLS 2007, JSTAR 2011, KLoSA 2010, and VHLSS 2012.
Note: Employment is defined as those who have worked for more than one hour weekly on average.
Vietnam, the decline in work hours exhibits person and his or her spouse, (c) demands for
a steeper slope, and hours worked tend to be caring for grandchildren or older parents, and
well below those of other countries. In con- (d) work status of the spouse. Each driver is
trast, in China men and women still working discussed in turn.
in agriculture put in long hours at relatively First, availability of a pension appears to
old ages, raising the question whether the have a strong association with lower labor
absence of migrant adult children contributes force participation when the pension pro-
to a scarcity of labor and higher labor input vides adequate financial protection. For
of rural elderly. example, after controlling for age, educa-
What are the primary drivers of the diverse tion, health status of respondent and spouse,
labor market behavior of older people in East wealth, and family demographic characteris-
Asia and Pacific, both within and across coun- tics, we found that receiving a formal sector
tries? Potential drivers include (a) availability pension is associated with 52 and 54 percent
of a pension, (b) health status of the older reductions in the probability that men and
T H E W E L FA R E A N D L A B O R MA R K E T B E H AV I O R O F O L D E R P E O P L E I N E A S T A S I A A N D PAC I F I C 67
women, respectively, are working in urban most likely because of the immaturity of the
China. The behavioral response of older pension system and modesty of benefits for
workers to mandatory retirement provisions the current elderly cohort;12 and in Japan,
can be seen in the hazard rates in figure 2.17, because availability of disability and survivor
which contrasts the smooth glide into with- benefits is likely to have larger effects. For
drawal from work in rural areas with spikes the Japanese, however, receipt of a pension is
in retirement at official pension ages in urban associated with 3.0 hours less work per week
areas (ages 50 and 55 for blue-collar and for those who work, and receipt of disabil-
white-collar women, respectively, and age ity benefits is associated with 6.0 hours less
60 for men). Although China has no explicit work per week for men and 3.3 hours less
provisions against working beyond retire- for women.
ment age, the relative generosity of the urban As would be expected, the generosity of
employee pension combined with manda- pensions is another important factor in labor
tory retirement makes them much less likely. market effects. The continued work after
Similarly, one sees a pronounced (though retirement of residents receiving social pen-
weaker) negative effect on employment of sions in China, the Philippines, and Thailand
men in urban and rural Indonesia, with 19 and the high labor supply rates of workers
and 25 percent reductions, respectively, in over age 55 in Japan and Korea suggest that
probability of working. In contrast, in Japan retirement benefits are simply not generous
and Korea, there is a negligible effect of enough to lead to exit from work. For exam-
pension receipt for either gender: in Korea, ple, figure 2.18 shows that over 80 percent
FIGURE 2.17 Pension availability and mandatory retirement rules have strong effects on labor force
behavior of older Chinese workers
Retirement hazard rates by gender, age, and urban or rural location, China, 2011
60
40
20
0
45 50 55 60 65 70 75 80
Retirement age (years)
60
40
20
0
45 50 55 60 65 70 75 80
Retirement age (years)
Male Female
Sources: Giles, Lei et al. 2015 based on data from CHARLS 2011.
68 LIVE LONG AND PROSPER
FIGURE 2.18 Labor force participation is influenced by pension receipt in select East and Southeast Asian
countries
Percentage of people in labor force by age, gender, and urban or rural location, and percentage eligible for
pensions, selected countries
80 80
60 60
40 40
20 20
0 0
45 50 55 60 65 70 75 80 85 45 50 55 60 65 70 75 80 85
Age (years) Age (years)
80 80
60 60
40 40
20 20
0 0
45 50 55 60 65 70 75 80 85 45 50 55 60 65 70 75 80 85
Age (years) Age (years)
Pension eligible Working total
T H E W E L FA R E A N D L A B O R MA R K E T B E H AV I O R O F O L D E R P E O P L E I N E A S T A S I A A N D PAC I F I C 69
FIGURE 2.18 Labor force participation is influenced by pension receipt in select East and Southeast Asian
countries (continued)
Percentage of people in labor force by age, gender, and urban or rural location, and percentage eligible for
pensions, selected countries
80 80
Share of population (%)
40 40
20 20
0 0
50 55 60 65 70 75 80 50 55 60 65 70 75 80
Age (years) Age (years)
Receiving other social assistance Receiving old age pension Working total
80 80
60 60
40 40
20 20
0 0
45 50 55 60 65 70 75 80 85 45 50 55 60 65 70 75 80 85
Age (years) Age (years)
80 80
60 60
40 40
20 20
0 0
45 50 55 60 65 70 75 80 85 45 50 55 60 65 70 75 80 85
Age (years) Age (years)
Pension eligible Working total
FIGURE 2.18 Labor force participation is influenced by pension receipt in select East and Southeast Asian
countries (continued)
Percentage of people in labor force by age, gender, and urban or rural location, and percentage eligible for
pensions, selected countries
100 100
80 80
60 60
40 40
20 20
0 0
+
9
9
9
+
–5
–4
–5
–6
–6
–7
–7
–5
–4
–5
–6
–6
–7
–7
80
80
55
45
50
60
65
70
75
45
50
55
60
65
70
75
100 100
80 80
60 60
40 40
20 20
0 0
9
59
49
64
+
–4
–5
–6
–6
–7
–7
80
–5
–5
–6
–7
–7
80
–
–
45
50
55
60
65
70
75
45
50
55
60
65
70
75
Sources: Giles, Hu, and Huang 2015 based on data from CHARLS 2011, IFLS 2007, JSTAR 2011, KLoSA 2010, ThaiSES 2011, and VHLSS 2012.
T H E W E L FA R E A N D L A B O R MA R K E T B E H AV I O R O F O L D E R P E O P L E I N E A S T A S I A A N D PAC I F I C 71
TABLE 2.3 Poor health results in lower probability of work at older ages in East and Southeast Asia
Percentage of reduction in employment rate when in poor health, by gender and urban or rural location
TABLE 2.4 Hours spent providing care for children and elderly among Chinese cohorts ages 45 and above
lower, suggesting that couples frequently 40 and above globally are engaged in volun-
engage in joint provision of care. teer work at different ages. However, shares
The contribution of older people to child are lower in most Asian countries (with the
care is also underscored in multivariate notable exceptions of Japan, Korea, and
labor supply analysis. In China, presence the Philippines) and sharply lower in China
of a child under age six in the household is and India. The deficit is particularly notable
associated with 17.6 and 3.0 percent reduc- among those ages 70 and above, suggesting
tions in the probability of employment for a broader challenge of social participation
urban and rural women, respectively. The for the oldest elderly. Offsetting low average
presence of children or elderly showed no participation somewhat, in most East Asian
effect on male employment. In Indonesia, and Pacific economies shown, those who vol-
the presence of children ages 6 to 12 is asso- unteer work at or above the global average
ciated with 11.7 and 9.7 percent reductions number of hours (Leeson and Harper 2007).
in the probability of employment for urban Participation in volunteer work is important
and rural women, respectively. In Korea, not only for its direct contribution to societal
by contrast, the presence of children ages 6 welfare, but also as a channel for sustaining
to 12 has a more significant negative effect social participation of older people.
on labor force participation of older urban A strong correlation also exists for the
men. However, this effect could be driven by employment of older spouses, suggesting
already low participation of urban women. that spouses have joint labor supply deci-
In Japan, the probability that a woman will sions and may have a preference for retiring
be employed is 8 percent lower in house- together (table 2.5). The most pronounced
holds with an older resident over age 80 effect is in urban China, where employ-
(Giles, Lei et al. 2015). ment of a spouse is associated with 40 and
Finally, in addition to care for family 58 percent increases in probability that men
members, older people in East and Southeast and women are working, respectively, and
Asia may contribute to lower labor force in Japan, especially for women. In rural
participation because of volunteer work, China, a working spouse is associated with
although the scale of involvement is gener- 23 and 33 percent increases in probabilities
ally lower than global averages. Figure 2.19 that older men and women, respectively,
shows that around 20 percent of people ages are working.14 In Indonesia and Korea, the
T H E W E L FA R E A N D L A B O R MA R K E T B E H AV I O R O F O L D E R P E O P L E I N E A S T A S I A A N D PAC I F I C 73
FIGURE 2.19 Older people in East and Southeast Asia are a significant source of volunteer work
Percentage of population by age participating in volunteer work, selected economies
40
35
30
Share of population (%)
25
20
15
10
0
40s 50s 60s 70s
Age cohort
World India Philippines Malaysia China
Taiwan, China Singapore Korea, Rep. Hong Kong SAR, China Japan
TABLE 2.5 Influence of employment of a spouse on associated with longer working hours for
probability of own employment rural men and women in China, Indonesia,
Percentage change in probability of own employment and Korea, and for all men in Japan.
with working spouse Joint decisions by spouses on labor supply,
Country Men Women or retirement, are of considerable policy
relevance when one considers incentives
China
to increase working ages of men and women,
Urban 48 58
Rural 23 33
particularly in countries where male and
Indonesia
female retirement ages differ significantly
Urban 11 19 (as in China and Vietnam).
Rural 9 12
Japan 24 56
Conclusion
Korea, Rep.
Urban 13 21 The diverse patterns of well-being, labor mar-
Rural 14 28 ket behavior, and sources of support across and
Sources: Giles, Hu, and Huang 2015 based on data from CHARLS 2011. within East Asian and Pacific countries mean
that policies will need to take into account dif-
association between own work status and a fering initial conditions. Policy makers will
working spouse is also significant, but lacks need to pay close attention to various aspects
the same effect on probability of employ- of elderly welfare, a task made more complex
ment. In terms of hours worked for those by the diversity of situations across countries
who continue to work, a working spouse is and between different groups of elderly within
74 LIVE LONG AND PROSPER
countries—rich and poor, rural and urban, measures of ability to perform activities
and healthy and frail or depressed, as well important for independent living, such as
as men and women and formal and informal shopping, preparing meals, and managing the
workers. For example, improving the well- household budget.
4. These percentages exist even though median
being of older workers who have spent their
self-rated happiness tends to be higher for
lives working in agriculture or the informal
those above age 50 than for adults between
sector and facilitating an ability to retire could ages 19 and 49 (Deaton 2007).
be addressed by broadening access to pensions 5. Subjective well-being was calculated using the
and other public support in old age. However, Center for Epidemiologic Studies Depression
these interventions will need to be balanced Scale (CES-D 10), which is constructed from
against potential disincentives to work at 10 questions commonly used to study the
older ages, especially for the urban elderly. incidence of depression.
Thus, a key challenge for policy makers lies 6. One should be careful in assigning causality
in extending support for the elderly at risk of to this relationship because having a happier
poverty while also incentivizing later retire- disposition throughout one’s life may have
contributed to raising successful (and
ment among older urban and formal sector
educated) children.
workers ages 50 to 69.
7. Results for women in urban and rural
Building on the microlevel discussion from areas are quite similar in all countries (2–5
this chapter, the following chapters examine percentage points different and lower in most
the policy challenges with respect to the mac- cases) except Korea, where women’s labor
roeconomy, labor policies, pensions, health income share is 11–15 percentage points
care, and long-term care. Chapter 3 focuses lower than that of men.
on the macroeconomic and fiscal implica- 8. Taking account of nonincome support is
tions of rapid aging in East Asia and Pacific. important because older people living with
As noted, behavioral and policy responses to adult children in China, Indonesia, and
aging will be crucial in determining the nature Thailand receive lower new transfers from
adult children but have regular nonfinancial
and scale of effects on growth and the fiscal
support (see Giles, Lei et al. 2015).
positions of East Asian and Pacific countries.
9. Private transfers include all cash and in-kind
The microlevel insights from this chapter are transfers provided to households with elderly
useful for anticipating potential responses. members, and public transfers include all
forms of poverty assistance and pensions.
The approach assumes no labor behavioral
Notes responses.
1. Because of data limitations, old age poverty 10. See Mason et al. (2006) for a description of
is measured according to households with methodology and for subsequent publications
elderly members, households with only elderly edited by Andrew Mason and Ronald Lee
members, and others with mixed composition. for cross-country applications. NTA also
To the extent that intrahousehold resource provided an important underpinning for the
allocation is biased toward younger members, Asian Development Bank’s 2012 volume on
the welfare of individual elderly persons may aging and growth in Asia (Park, Lee, and
be lower. Mason 2012).
2. ADLs are basic self-care tasks, and IADLs 11. The 13 new members of the European Union
are tasks that allow an individual to function (EU13) are Bulgaria, Croatia, Cyprus, the
independently. Unlike ADLs, IADLs can be Czech Republic, Estonia, Hungary, Latvia,
delegated to someone else and thus are not Lithuania, Malta, Poland, Romania, the
uniformly carried out by everyone themselves. Slovak Republic, and Slovenia.
In addition, some IADLs relate to functioning 12. Research conducted in Korea showed that
within a community rather than only within over 92 percent of respondents expressed
the home. a desire to take their pension after the
3. ADLs are measures of physical functioning, eligible age, in part to improve adequacy
ranging from the ability to walk 1 kilometer (Korea National Pension Research Institute
to the ability to get out of bed. IADLs are 2012).
T H E W E L FA R E A N D L A B O R MA R K E T B E H AV I O R O F O L D E R P E O P L E I N E A S T A S I A A N D PAC I F I C 75
13. If one looks at spouse health status using an the Developing World.” Social Protection
index of spouse ADLs and IADLs, there is and Labor Policy Note 17, World Bank,
evidence of a relatively small added worker Washington, DC.
effect of poor spouse health status on Eyres, Harry. 2014. “The Slow Lane: The
employment of men and women in urban and Cruel Cost of the Crisis.” Financial Times,
rural areas of China, but no apparent effect in July 25. http://www.ft.com/intl/cms/s/2
other countries. / b4cbef16-10c9-11e4-b116-00144feabdc0
14. This effect is seen in OECD countries also. .html#axzz3iMcpEjMf.
Schirle (2008) finds that increases in wives’ FIES (Family Income and Expenditure Survey).
participation in the labor force can explain 2009. Philippine Statistics Authority, Manila.
one-fourth, one-half, and one-third of the http://psa.gov.ph/content /family- income-and
increase in older married men’s labor force -expenditure-survey-fies.
participation in the United States, Canada, Giles, John, Yuqing Hu, and Yang Huang. 2015.
and the United Kingdom, respectively. “Understanding the Retirement Decision in
Aging East Asia.” Background paper for the
East Asia and Pacific regional report on aging,
References World Bank, Washington, DC.
Cai, Fang, John Giles, and Xin Meng. 2006. “How Giles, John, and Yang Huang. 2015. “Are the
Well Do Children Insure Parents against Low Elderly Left behind in a Time of Rapid
Retirement Income? An Analysis Using Survey Demographic and Economic Change?
Data from Urban China.” Journal of Public A Comparative Study of the Poverty and Well-
Economics 90 (12): 2229–55. doi:10.1016/j Being of East Asia’s Elderly.” Background
.jpubeco.2006.03.004. paper for the East Asia and Pacific regional
CHARLS (China Health and Retirement report on aging, World Bank, Washington, DC.
Longitudinal Study). 2011. National School of Giles, John, Xiaoyan Lei, Yafeng Wang, and
Development, Peking University. http://charls Yaohui Zhao. 2015. “One Country, Two
.ccer.edu.cn/en. Systems: Evidence on Retirement Patterns in
China Census. 1982. Integrated Public Use China.” World Bank, Washington, DC, and
Microdata Series, International, series 6.3, Peking University, Beijing.
Minnesota Population Center University of Giles, John, and Ren Mu. 2007. “Elderly Parent
Minnesota, Minneapolis. Health and the Migration Decisions of Adult
———. 1990. China Population and Information Children: Evidence from Rural China.”
Research Center 1% Sample of the 1990 China Demography 44 (2): 265–88.
Population Census Data, China Archive, Texas IFLS (Indonesia Family Life Survey). 2007. RAND,
A&M University, College Station. Santa Monica, CA. http://www.rand.org/labor
———. 2000. National Bureau of Statistics 2000 /FLS/IFLS.html.
Fifth Population Census Data, China Statistics Jackson, Richard, and Tobias Peter. 2015. “From
Press, Beijing. Challenge to Opportunity: Wave 2 of the
CHNS (China Health and Nutrition Survey). East Asia Retirement Survey.” Global Aging
Various years. http://www.cpc.unc.edu Institute, Alexandria, VA.
/projects/china/data. Jacobs, Josephine, Audrey Laporte, Courtney Van
Christensen, Kaare, Gabriele Doblhammer, Houtven, and Peter Coyte. 2014. “Caregiving
Roland Rau, and James Vaupel. 2009. “Aging Intensity and Retirement Status in Canada.”
Populations: The Challenges Ahead.” Lancet Social Science and Medicine 102: 74–82.
374 (9696): 1196–208. JSTAR (Japanese Study of Aging and Retirement).
Costa, Dora L. 1998. The Evolution of Retirement: Various years. Research Institute of Economy,
An American Economic History, 1880–1990. Trade, and Industry, Hitotsubashi University
Chicago: University of Chicago. and University of Tokyo, Tokyo. http://www
Deaton, Angus. 2007. “Income, Aging, Health, .rieti.go.jp/en/projects/jstar/.
and Wellbeing around the World: Evidence KLoSA (Korean Longitudinal Study of Aging).
from the Gallup World Poll.” NBER Working 2010. Korea Labor Institute, Seoul. http://
Paper 13317, National Bureau of Economic www.kli.re.kr/klosa/en/about/introduce.jsp.
Research, Cambridge, MA. Knodel, John. 2014a. “Is Intergenerational
Evans, Brooks, and Robert Palacios. 2015. “An Solidarity Really on the Decline? Cautionary
Examination of Elderly Co-residence in Evidence from Thailand.” Asian Population
76 LIVE LONG AND PROSPER
Studies 10 (2): 176–94. doi:10.1080/1744173 McDougall, Fiona A., Fiona E. Matthews, Kari
0.2014.902160. Kvaal, Michael E. Dewey, and Carol Brayne.
———. 2014b. “The Situation of Older Persons in 2007. “Prevalence and Symptomatology
Myanmar: An Overview.” BOLD 24 (3): 9–16. of Depression in Older People Living in
Knodel, John, Vipan Prachuabmoh, and Napaporn Institutions in England and Wales.” Age and
Chayovan. 2013. The Changing Well-Being Ageing 36 (5): 562–68. doi:10.1093/ageing
of Thai Elderly: An Update from the 2011 /afm111.
Survey of Older Persons in Thailand. London: Meng, Annika. 2011. “Informal Caregiving and
HelpAge International. the Retirement Decision.” German Economic
Korea National Pension Research Institute. 2012. Review 13 (3): 307–30.
“Middle- and Old-Aged Koreans’ Preparation Pang, Lihua, Alan de Brauw, and Scott Rozelle.
Realities for and Expectations about Their Old 2004. “Working until You Drop: The Elderly
Age.” Seoul. http://english.nps.or.kr/jsppage of Rural China.” China Journal 52: 73–96.
/english/research/research_01.jsp. Park, Donhhyuan, Sang-Hyop Lee, and Andrew
Lee, Chulhee. 2009. “Labor Force Participation Mason, eds. 2012. Aging, Economic Growth,
of Older Males in Korea: 1955–2005.” NBER and Old-Age Security in Asia. Manila: Asian
Working Paper 14800, National Bureau of Development Bank.
Economic Research, Cambridge, MA. Qin, Huali. 2007. “Adding Life to Years: Predicting
Leeson, George, and Sarah Harper. 2007. “The Subjective Quality of Life among Chinese
Global Ageing Survey (GLAS): Ageing and Oldest-Old.” Gerontology thesis, Georgia
Later Life—Hong Kong and Asia.” Research State University, Atlanta, May 4. http://
Paper 307, Oxford Institute of Ageing, scholarworks.gsu.edu/gerontology_theses/4.
University of Oxford, Oxford, U.K. Reynolds, Charles F., and David J. Kupfer. 1999.
Lei, Xiaoyan, John Strauss, Meng Tian, and “Depression and Aging: A Look to the
Yaohui Zhao. 2012. “Living Arrangements Future.” Psychiatric Services 50 (9): 1167–72.
of the Elderly in China: Evidence from doi:10.1176/ps.50.9.1167.
CHARLS.” IZA Discussion Paper 6249, Schirle, Tammy. 2008. “Why Have the Labor
Institute for the Study of Labor, Bonn, Force Participation Rates of Older Men
Germany. http://papers.ssrn.com/sol3/papers Increased since the Mid-1990s?” Journal of
.cfm?abstract_id=1981205. Labor Economics 26 (4): 549–94.
Liang, Jersey, Xian Liu, and Shengzu Gu. 2001. Sunwoo, Duk. 2013. “Economic Costs and
“Transitions in Functional Status among Opportunities for Ageing and Healthcare.”
Older People in Wuhan, China: Socioeconomic Paper presented at the Global Forum on
Differentials.” Journal of Clinical Epidemiology Innovation for Ageing Populations, Kobe,
54 (11): 1126–38. Japan, December 11.
Lim, Wei-Yen, Stefan Ma, Derrick Heng, Vineta ThaiSES (Thai Household Social and Economic
Bhalla, and Suok K. Chew. 2007. “Gender, Surveys). Various years. National Statistical
Ethnicity, Health Behaviour, and Self-Rated Office, Bangkok. http://web.nso.go.th/survey
Health in Singapore.” BMC Public Health /house_seco/socio.htm.
7 (1): 184. doi:10.1186/1471-2458-7-184. Van Houtven, Courtney, Norma Coe, and Meghan
Liu, Xian, Jersey Liang, Naoko Muramatsu, and Skira. 2013. “The Effect of Informal Care
Hidehiro Sugisawa. 1995. “Transitions in on Work and Wages.” Journal of Health
Functional Status and Active Life Expectancy Economics 32 (2013): 240–52.
among Older People in Japan.” Journals of Van Minh, Hoang, Peter Byass, and Stig Wall.
Gerontology Series B: Psychological Sciences 2010. “Multilevel Analysis of Effects of
and Social Sciences 50B (6): S383–94. Individual Characteristics and Household
doi:10.1093/geronb/50B.6.S383. Factors on Self-Rated Health among Older
Mason, Andrew, Ronald Lee, An-Chi Tung, Mun- Adults in Rural Vietnam.” Geriatrics and
Sim Lai, and Tim Miller. 2006. “Population Gerontology International 10 (2): 209–15.
Aging and Intergenerational Transfers: VHLSS (Vietnam Household Living Standards
Introducing Age into National Accounts.” Survey). 2012. General Statistics Office of
NBER Working Paper 12770, National Bureau Vietnam, Hanoi. http://www.gso.gov.vn
of Economic Research, Cambridge, MA. /Default_en.aspx?tabid=491.
T H E W E L FA R E A N D L A B O R MA R K E T B E H AV I O R O F O L D E R P E O P L E I N E A S T A S I A A N D PAC I F I C 77
WHO (World Health Organization). 2012. Pension Program: Income, Poverty, Expenditure,
“Global Health and Aging.” WHO, Geneva. Subjective Wellbeing, and Labor Supply.” China
Witoelar, Firman, John Strauss, and Bondan Sikoki. Economic Quarterly 14 (1): 203–30.
2012. “Socioeconomic Success and Health in Zimmer, Zachary, and Pattama Amornsirisomboon.
Later Life: Evidence from the Indonesia Family 2001. “Socioeconomic Status and Health
Life Survey.” In Aging in Asia: Findings from among Older Adults in Thailand: An Examina-
New and Emerging Data Initiatives, edited tion Using Multiple Indicators.” Social Science
by James P. Smith and Malay Majmundar, and Medicine, Comparative Studies of Com-
309–41. Washington, DC: National Academies petition Policy 52 (8): 1297–311. doi:10.1016
Press. /S0277-9536(00)00232-X.
World Bank East Asia and Pacific Standardized Zimmer, Zachary, Xian Liu, Albert Hermalin, and
Household Surveys. Various years. Internal Yi-Li Chuang. 1998. “Educational Attainment
data set of World Bank, Washington, DC. and Transitions in Functional Status among
Zhang, Chuanchuan, John Giles, and Yaohui Zhao. Older Taiwanese.” Demography 35 (3):
2014. “Policy Evaluation of China’s New Rural 361–75. doi:10.2307/3004043.
Aging and the Economy
II
Aging and the Macroeconomy in
East Asia and Pacific 3
This chapter is based on a background paper by Thomas Flochel, Yuki Ikeda, Harry Moroz, and Nithin
Umapathi (2014).
81
82 LIVE LONG AND PROSPER
FIGURE 3.1 East Asia and Pacific’s developing economies are getting old before getting rich
Elderly dependency ratio and GDP per capita (2005 PPP), various economies, 1980–2010
60,000
50,000
GDP per capita (US$2005 PPP)
40,000
30,000
20,000
10,000
Sources: GDP data in 2005 US$PPP for 1980 to 2010 from the World Bank’s World Development Indicators database; elderly dependency ratios calculated as
ratio of population over age 65 to population ages 15–64, using data from UN 2013.
Note: PPP = purchasing power parity.
much lower levels of gross domestic product generally migrated upward and to the right,
(GDP) per capita than the red economies and but they have followed very different paths.1
those of countries in the Organisation for The orange group, where the turning point
Economic Co-operation and Development for dependency is either just past or not far
(OECD) at similar stages of old-age depen- away, is particularly heterogeneous. Malaysia
dency. The green economies, which are just experienced high growth, nearly tripling its
beginning to undergo a demographic transi- GDP per capita with a still reasonably low
tion, have lingered in the bottom left-hand ratio of elderly dependency, while Thailand’s
corner at low levels of per capita income grow th was accompanied by particu-
and low elderly dependency. Early declines larly rapid aging, with demographic ratios
in youth dependency have even shifted some approaching one elderly for seven working-
of them toward the left as the working- age adults. China has grown fast but from
age share of the population began to rise. an even lower base; its rate of elderly depen-
Economies in the orange and red groups have dency is not far behind Thailand’s. What is
AG I N G A N D T H E MAC R O E CO N O M Y I N E A S T A S I A A N D PAC I F I C 83
clear for this group is that they will not be affect income per capita. Labor productiv-
able to reach the income levels of red econo- ity interacts with aging in several ways. As
mies at similar points of demographic transi- the population ages, the net effect on GDP
tion. For example, Indonesia in 2010 had an per capita is determined by labor productiv-
elderly dependency ratio similar to that of the ity at each age. Evidence suggests that age
Republic of Korea in 1980, but with purchas- and individual productivity are correlated,
ing power parity per capita income around but the relationship is complex and depends
that of Korea in 1976 and output per worker on the relative endowments of human and
equivalent to Korea’s in 1969 (or around physical capital workers are equipped with.
15 percent of Korea’s 2010 level of output The current cohort of elderly in East Asia and
per worker). At the high end of the spectrum, Pacific is far more productive than the previ-
red economies have continued solid growth ous cohort at any given age, particularly in
(though flattening in Japan) while doubling countries where education and health have
their rates of elderly dependency. improved rapidly. More healthy and better
The contribution of favorable demo- educated, the elderly are increasingly able to
graphics to the region’s growth performance develop skills as they age. Labor participation
in recent decades has been significant. rates and productivity at each age are also
Literature on East Asia and Pacific finds that affected by the two drivers of aging—lower
up to one-third of the high per capita income fertility and extended longevity—through
growth in the region from 1960 to the 1990s effects on economic behavior and on physi-
can be attributed to the demographic divi- cal and human capital accumulation. From a
dend (Bloom, Canning, and Malaney 2000; public policy perspective, demographics may
Bloom, Canning, and Sevilla 2003; Bloom be a given, but labor participation and pro-
and Williamson 1998). 2 Other authors ductivity are not. The size of the economic
have estimated that demographic transi- bonus or burden that results from population
tion accounted for as much as 44 percent aging depends on how policy influences labor
of growth in per capita incomes from 1960 force participation, saving, human capital
to 1990 (Kelley and Schmidt 1995, 2005). accumulation, and total factor productivity.
Part of this effect is driven by rapid falls in
fertility, which result in a large working-age
share of population relative to youth share,
Channels to augment the labor
thus raising per capita incomes. In the case of
force
many East Asian countries, the pure demo- Although the share of working-age popula-
graphic effect was positively compounded tion will experience significant declines in
by improvements in educational attainment, red and some orange economies over the
capital investment, and other factors. coming decades, significant potential exists
Although demographics have provided a to augment the labor force in these countries
clear tailwind for East Asian growth in recent through different channels. The potential
decades (and will continue to do so in lower- channels through which aging may affect
income countries for some time), demograph- growth include labor force effects, impacts
ics are only part of the story. The effect of on saving and investment, indirect impacts
population aging on GDP per capita also of rising fiscal obligations, and productivity.
depends on labor force participation and One body of literature seeks to unpick these
productivity, which in turn are affected by disparate effects for East Asia and Pacific
demographic forces. Figure 3.2 decomposes and other parts of the world. The most pes-
the channels through which aging influences simistic estimates of the impact of aging
per capita income and provides the frame- on growth in East Asia and Pacific (typi-
work for this chapter: the confluence of cally using basic growth accounting models)
demographic, employment, and productivity assume unchanged labor force participation
factors determines how aging has and will rates, retirement behavior, population health,
84 LIVE LONG AND PROSPER
FIGURE 3.2 Aging’s influence on GDP per capita is the product of the employment rate, the working-age share of the
population, and GDP per employed worker
Average
Hong Kong SAR, China 42,556 67% 75% 85,129
Japan 31,409 78% 64% 63,435
Korea, Rep. 26,830 67% 73% 55,133
Malaysia 13,793 61% 67% 33,402
Thailand 7,982 81% 72% 13,800
China 6,692 76% 74% 12,058
Indonesia 3,864 68% 65% 8,741
Mongolia 3,686 60% 69% 8,823
Philippines 3,532 64% 61% 9,095
Vietnam 2,920 82% 70% 5,095
Lao PDR 2,189 81% 59% 4,525
Papua New Guinea 2,187 74% 58% 5,076
Cambodia 1,949 88% 63% 3,526
Timor-Leste 1,482 39% 50% 7,585
Myanmar (missing data) 81% 69%
Sources: World Bank estimates using population data from UN 2013, GDP per capita in 2005 US$PPP from the World Bank’s World Development Indicators database, and employment
data from ILO 2011.
Note: PPP = purchasing power parity.
rise as much. The evidence suggests, however, because of the income effect. According to
that medium-term convergence toward fuller figure 3.3, relative trends in labor force par-
use of a given working-age population is pos- ticipation rates and gains in health-adjusted
sible. Policies to increase female labor force life expectancy years suggest the increas-
participation and outcomes to date in East ing potential for extending working lives in
Asia and Pacific are discussed in chapter 4 of East Asia and Pacific, particularly among
this report. women. For example, on average, Japanese
A second channel for augmenting labor men and women at ages 55–59 in 2010
forces is extending productive working lives could expect to live 1.4 and 1.2 extra years,
of older workers. As shown in chapter 2 of respectively, compared with 1990. This
this report, people in East Asia and Pacific “longevity dividend” is not gender neutral:
work until fairly advanced ages already, since 1990, improvements in health-adjusted
with notable exceptions in urban areas of life expectancy in the region have increased
some middle-income countries. Many rural participation by women in China, Japan ,
people continue to work well into older age and Indonesia.
out of necessity, not infrequently even when Apart from effects on quality of the labor
in poor health. Among richer economies of force (see below), rising educational attain-
the region, Japan and Korea have the high- ment across East Asia and Pacific can be
est effective retirement age among OECD expected to play a role in promoting higher
countries where, on average, people retire labor force participation across ages and
later than pensionable age (D’Addio, Keese, genders through two channels. First, higher
and Whitehouse 2010). Presumably, there- educational attainment increases incentives
fore, this channel for enhancing labor force to work in better-paid employment. Second,
participation would seem more constrained studies find that more education is associated
than in other regions. However, certain with better health outcomes and lower mor-
countries—crucially China and Vietnam— tality, thereby improving the ability and will-
have significant potential to increase pro- ingness of older people to participate in the
ductive working lives of urban people and labor force (Burtless 2013; Cutler and Lleras-
mitigate the impacts of aging. The risk also Muney 2008). According to this logic, the
exists that growth, urbanization, and expan- regional education projections to 2040 imply
sion of social security systems in middle- and that labor force participation, especially at
lower-income countries of East Asia and middle and older ages, should be expected
Pacific will see current rural workers increas- to expand. Overall, as long as an increasing
ingly adopt the retirement behavior of urban share of the population is better educated and
workers, which would result in significant successive cohorts supplant the less-educated
downward shifts in average participation generations, labor participation rates are
rates at older ages in most countries of the expected to grow—especially among the
region. That downside risk is considerable older population for whom the participation
in the absence of further reform of pension rate tends to be lower.
systems. Chapter 4 of this report discusses In East Asian countries with aging pop-
regional and global experience with policies ulations, immigration of younger work-
to extend active working lives. ers from other countries can help lift labor
The potential for higher employment at force participation rates. Migration can
older ages is influenced by gains in life expec- provide a certain level of relief, especially
tancy and, especially, by gains in healthy in the short term, in the labor markets of
years (Eggleston and Fuchs 2012). Bloom, countries with aging populations suffering
Canning, and Moore (2004) show that from labor shortages. Migrants, on average,
theoretically healthier life expectancies will tend to be younger and exhibit higher rates
extend working lives but that improvements of participation relative to natives. Two
tend to increase less than proportionately features distinguish East Asia and Pacific.
86 LIVE LONG AND PROSPER
FIGURE 3.3 Older women in East Asia and Pacific will be able to work later in life by virtue of rising
health-adjusted longevity
Percentage of women in the labor force and healthy years of life expectancy by age between 1990 and 2010, various
countries
a. China b. Indonesia
70 25 65 25
65 60
60 20 55 20
55
Percent
Percent
Years
50
Years
50 15 15
45
45
40 10 40 10
35 35
30 5 30 5
50–54 55–59 60–64 65+ (65–69 50–54 55–59 60–64 65+ (65–69
for HALE) for HALE)
75 25 65 25
70 60
65 20 20
55
60
Percent
Percent
50
Years
55
Years
15 15
50 45
45
40 10
40 10
35 35
30 5 30 5
50–54 55–59 60–64 65+ (65–69 50–54 55–59 60–64 65+ (65–69
for HALE) for HALE)
Health-adjusted life expectancy, 1990 (right axis) Health-adjusted life expectancy, 2010 (right axis)
Female labor force participation rate, 1990 (left axis) Female labor force participation rate, 2010 (left axis)
Sources: World Bank calculations based on IHME and World Bank 2013 and ILO 2011.
Note: HALE = health-adjusted life expectancy.
First, significant variation exists in terms about the long-run levels of the United
of demographic transitions. Thus, migra- States and a number of western European
tion of young workers from countries with countries). However, the low base effect
a younger workforce would benefit aging also indicates sociocultural, linguistic, and
host countries now and the lower-income political barriers to immigration that may
sending countries in the future by easing be more difficult to overcome. In contrast,
the aging challenge in both (“chronological Hong Kong SAR, China, and Singapore
arbitrage”). Second, some of the intensely stand out with high international migrant
aging countries in the region are marked by stocks that make up about 40 percent of the
very low rates of immigration. For example, total population.
in Japan and Korea only 1 to 2 percent of The policy changes required will be
the population is international migrant politically challenging and require signifi-
stock (Özden and Testaverde 2014; World cant behavioral change on the part of work-
Bank 2013d). This low base effect pres- ers, employers, and society more broadly.
ents a significant opportunity if inflows of Migration is an unusually sensitive issue in
new migrants ages 25–35 can be raised to potential host countries in East Asia and
10 percent of the labor force that age (that is, Pacific, and the historical record is modest
AG I N G A N D T H E MAC R O E CO N O M Y I N E A S T A S I A A N D PAC I F I C 87
in some regional countries most affected by T ABLE 3.1 The educational attainment of
aging.4 Increasing participation of older future generations in East Asia and Pacific will be
workers will require reforms of social secu- considerably higher than for the current working-
rity systems and labor policies, which have age population (secondary and above)
Percentage of population ages 20–64 with secondary or
proven politically difficult in many coun-
higher education, 1990, 2030, and 2040
tries, though recent reforms in some East
Asian countries provide encouragement. And Economy 1990 2030 2040
increasing female labor force participation Hong Kong SAR, China 72.5 89.4 88.9
will require incremental investments in child Japan 99.9 100.0 100.0
care and other programs as well as overcom- Korea, Rep. 74.6 99.5 99.8
ing inertia and long-held social attitudes to Singapore 63.3 90.3 92.5
the role of women in society once they have China 48.5 83.7 89.7
had children. These issues are discussed in Indonesia 31.6 72.3 79.2
chapters 4 and 5 of this report. Malaysia 50.4 89.9 93.0
Mongolia 75.8 91.4 92.3
Thailand 20.8 65.6 75.7
Labor force quality and Vietnam 24.2 49.1 56.9
productivity Cambodia 16.2 45.2 52.7
Lao PDR 17.8 48.5 56.3
Just as important as mitigating quantity of Myanmar 27.6 58.2 65.9
workforce declines will be efforts to enhance Philippines 52.7 84.0 88.3
worker quality. Better-educated people are
Source: World Bank calculations based on data from World Population
more prepared for lifelong learning, health- Program, International Institute for Applied Systems Analysis, http://www
ier, and more productive. On this front, .iiasa.ac.at/web/home/research/researchPrograms/WorldPopulation
/Introduction.html.
as shown in tables 3.1 and 3.2, East and
Southeast Asia have a positive starting point
in terms of the massive expansion in second-
ary and higher education in recent decades, TABLE 3.2 The educational attainment of
which will flow through into the labor force future generations in East Asia and Pacific will be
in coming decades. At all income levels, considerably higher than for the current working-
each generation of future East and Southeast age population (tertiary)
Asian workers will be significantly more Share of population ages 20–64 years old with tertiary
educated than the last. The second trend is education in 1990, 2030, and 2040
that the educational gap between older and
Economy 1990 2030 2040
younger workers is expected to diminish con-
siderably in China, Indonesia, Malaysia, the Hong Kong SAR, China 15.9 34.0 35.7
Philippines, and Thailand—and even disap- Japan 23.5 55.2 60.5
Korea, Rep. 15.6 49.7 55.6
pear in countries such as Korea. Although
Singapore 15.1 50.2 55.2
productivity differences may persist because
China 3.1 11.5 14.7
of age-specific factors, intergenerational Indonesia 4.2 15.3 18.7
inequality of educational attainment is going Malaysia 5.7 26.9 32.4
to diminish significantly. This outcome Mongolia 20.1 30.5 35.1
implies improving prospects for employment Thailand 6.8 27.6 33.9
at older ages. Vietnam 3.0 9.3 12.1
Several studies find that significant Cambodia 0.5 3.3 4.3
increases in human capital such as those Lao PDR 5.0 19.0 22.3
experienced in East Asia and Pacific may Myanmar 3.4 12.7 15.6
be sufficient to offset the effects of popula- Philippines 15.4 37.4 42.9
tion aging on aggregate productivity. Bloom, Source: World Bank calculations based on data from World Population
Program, International Institute for Applied Systems Analysis, http://www
Prettner, and Strulik (2013) show that under .iiasa.ac.at/web/home/research/researchPrograms/WorldPopulation
plausible production function specifications, /Introduction.html.
88 LIVE LONG AND PROSPER
the rise in education and health investments the share of students achieving basic literacy
that is in part spurred by declines in fertil- has been estimated to translate into an annual
ity rates raises workers’ productivity enough growth rate that is 0.3 percentage points
to compensate for declines in labor supply. higher than it would otherwise be for that
And according to Lee and Mason (2010), country (Hanushek and Woessmann 2009).
the effect of spending on education is strong Although red economies are in a strong posi-
enough to offset the adverse effects of popu- tion in terms of educational attainment and
lation aging. Significant increases in educa- skills, some orange economies and many
tion levels will be particularly important green economies are still far from par. OECD
sources of growth in aging orange econo- (2012) PISA (Programme for International
mies where the levels in 1990 were low. For Student Assessment) results show that parts
red economies, where the educational gap of the East Asian region outperform the
across cohorts of workers is narrower, other world. Hong Kong SAR and Shanghai,
forms of lifelong education and retraining China; Japan; Korea; and Singapore are
of older workers that are not reflected in above the OECD average, and Vietnam’s per-
tables 3.1 and 3.2 deserve further attention formance—with 15-year-olds performing on
(see box 3.1). a par with those in Germany and Austria—is
Falling fertility rates are usually associated exceptional. But other orange economies, such
with greater parental investments in educa- as Indonesia, Malaysia, and Thailand, lag the
tion and health of children, but translating OECD average. Furthermore, according to
these investments into productivity growth the Program for International Assessment of
requires investments in the quality of edu- Adult Competencies, which measures capac-
cation. Becker and Lewis (1973) and Willis ity for retraining and adaptation among
(1973) first described the quality-quantity adults, Japan leads globally, with Korea
trade-off that occurs as parents have fewer close to the OECD average. More alarming
children and they invest more in the human is the state of adult skills and competencies
capital of each child. A 10 percent increase in in green economies, such as the Lao People’s
A commonly held belief is that the ability to learn older workers should not be judged by the learning
new skills declines with age. In particular, Baltes, abilities of the current elderly population. Second,
Lindenberger, and Staudinger (2006) find that the because healthy life expectancy is increasing, cogni-
speed at which an individual processes new infor- tive decline is found to start later in life. Skirbekk
mation declines from 25 years of age onward. et al. (2013) conclude that the cognitive plasticity
Cognitive plasticity, the ability to learn, appears curve is shifted not only up, but also outward: indi-
to follow an inverted U shape over an individual’s viduals are more able to learn at each age, and decline
lifetime. Should we then conclude that older work- occurs increasingly late in life. Third, scientific evi-
ers cannot be retrained to renew their skills if theirs dence shows that the decline is not deterministic and
become obsolete? the elderly’s ability to learn improves given the right
Recent research nuances the conventional wisdom. circumstances. Physical exercise has been found to
First, because successive cohorts are healthier and revitalize key parts of the brain that improve learning
more educated, their brains function better. This phe- abilities (Christensen et al. 1997; Voelcker-Rehage,
nomenon, known as the Flynn effect, suggests that Godde, and Staudinger 2011), and learning outcomes
testing a cross-section of elderly people at one point among the elderly are particularly responsive to moti-
in time ignores improvements across cohorts. Because vational factors and supportive attitudes in the work-
cognitive plasticity rises with each cohort, future place (Kessler and Staudinger 2007).
AG I N G A N D T H E MAC R O E CO N O M Y I N E A S T A S I A A N D PAC I F I C 89
Democratic Republic (World Bank 2013c) profiles that emerge from industry or firm-level
and Timor-Leste (World Bank 2013b), which empirical research in Japan have consistently
have scored very low on core literacy tests found productivity to follow an inverted U
as part of the World Bank’s STEP (Skills shape with respect to age, peaking for employ-
Toward Employability and Productivity) ees at about 20 years of experience, which cor-
skills measurement program. responds to an age of 40 to 46 years (Ochiai
Severe child health and nutrition problems 2008; Shirakawa 2010). But productivity
in some green economies could seriously cur- does not seem to fall rapidly after that. In
tail the demographic dividend. Several green fact, research from Japan finds that the labor
economies have some of the highest rates of productivity of the 55–59 age group is virtu-
stunting in the world, with up to 44 percent ally identical to that of the 40–44 age group
of children below age 5 being stunted (Murphy and Welch 1990). Evidence from
(figure 3.4). Children who are stunted are Europe points to similar findings. In Austria
more likely to have lower cognitive abilities. and Sweden, for example, at worst, no evi-
Key parts of the brain develop less in children dence of age-related productivity decline is
who are severely malnourished, making these found (Göbel and Zwick 2013; Malmberg,
children much less likely to attend school or Lindh, and Halvarsson 2008).
more likely to drop out early. Compared with In red economies, past trends in the age
nonstunted children, stunted children score distribution of the population have been par-
7 percent lower on math tests, are 19 percent ticularly propitious to workforce productiv-
less likely to be able to read a simple sentence ity growth, but their reversing trends do not
at age 8, are 12 percent less likely to be able necessarily imply dramatic reductions in pro-
to write a simple sentence, and are 13 percent ductivity. A higher share of prime-age indi-
less likely to be in the appropriate grade for viduals in the population, sometimes defined
age at school. as the 30–54 age groups, has been found to be
A second important question on which significantly associated with high productiv-
evidence is encouraging but inconclusive is ity growth (Gómez and de Cos 2008).5 From
how a falling share of “prime working-age” the mid-1960s to the mid-2000s, the share of
population in red and a number of orange those ages 35–54 rose and peaked at a ratio
economies could affect productivity. How
individual productivity varies with age is
a challenging empirical question that has FIGURE 3.4 Prevalence of stunting among children under age five
important implications for aggregate pro- could undermine investments in education and affect cognitive
skills in East Asian and Pacific economies of the green group
ductivity, particularly in advanced econo-
Percentage of children under five who are stunted and GNI per capita
mies where the workforce is aging and (2005 dollars), selected countries
shrinking. Population aging has implica-
tions for growth not only through its direct 50
effect on workers’ labor productivity as they Ethiopia
Share of children under age 5 (%)
45 Lao PDR
age, but also through its effect on capital 40 Cambodia
Nepal Indonesia
accumulation, total factor productivity, 35 Uganda Philippines
30 Togo
and innovation. Empirical evidence on this
25
question is mixed and limited in East Asia 20
Vietnam
Sri Lanka
and Pacific but, in general, provides some 15 Thailand
grounds for encouragement. 10 China
Empirical evidence from Japan and 5
other aging countries suggests that the 0
0 500 1,000 1,500 2,000 2,500 3,000 3,500
age-productivity profile of workers follows an
GNI per capita (2005 dollars)
inverted U shape, with a mild decline after the
mid-40s and a significant dip only after age Source: World Health Organization, Global Database on Child Growth and Malnutrition.
65 (box 3.2). Estimates of age-productivity Note: GNI = gross national income.
90 LIVE LONG AND PROSPER
Evidence on the relative productivity of older and verbal abilities matter more (“crystallized”
workers that emerges from empirical research in abilities), less or no reduction in productivity occurs
advanced economies is somewhat contradictory. among elderly workers (Skirbekk 2008). Findings
While the productivity-age profile is robustly found also appear to differ depending on the measure of
to follow an inverted U shape, with the peak age productivity used and even using same approach.
somewhere between 35 and 54, macro estimates Thus, supervisor assessment studies have found no
of the effect of age on aggregate productivity find clear effect of age on productivity (McEvoy and
that older workforces are at worst as productive as Cascio’s [1989] review of 96 studies) while others
younger workforces. using quantity and quality of output measures have
Several possible explanations exist. First, indi- found lower productivity of older workers. Matched
vidual worker productivity does not dramatically employer-employee surveys show mixed results
decline after it peaks around age 40. The decline across studies and specific measures of productivity.
is found to be very mild relative to the rise in pro- But some surprising results have emerged from
ductivity in the first 20 years of work experience in the research. Older workers who remain in the
various studies in Japan (Fukao et al. 2006; Shinada labor force despite technological changes that affect
2011). In a famous study of BMW, Börsch-Supan their job content are found to have adapted well,
and Weiss (2008) also find that productivity does experiencing greater growth in tasks with intense
not begin to decline until age 60. Among Organisa- use of cognitive abilities (Romeu Gordo and Skir-
tion for Economic Co-operation and Development bekk 2013, following a line of research initiated by
countries, the share of those ages 50–64 was found Autor 2003). Focusing on the distinction between
to have a positive influence on growth, while a mental and physical productivity, Van Ours (2009)
higher share of those 65 and older contributed nega- finds that although the latter declines after age 40,
tively (Lindh and Malmberg 1999). Second, as noted the former does not. Similarly, perhaps counter to
previously, some issues with the research method- popular opinion, studies have found lower absen-
ology may bias the results. The decision to retire teeism (Martocchio 1989) and fewer job shifts
or retain a worker as he or she ages likely depends among older workers than younger ones, perhaps
on the worker’s productivity. It follows that when saving on training costs (AARP 2002). The most
measuring the productivity of older workers, only recent studies provide a generally more upbeat
those elderly workers with “high enough” produc- assessment of older-worker productivity (Bloom
tivity can be surveyed, whereas this selection effect and Sousa-Poza 2013).
is much less present among younger workers. Finally, productivity is not only an individual but
Third, the link between age and productivity is also a team concept, and evidence indicates that skills
also highly dependent on the context and task under of the old and the young may be complementary in
consideration. In general, productivity reductions at some contexts. Grund and Westergaard-Nielsen
older ages are strongest in job tasks where problem (2008) find that those companies with workers of
solving, learning, and speed are important (“fluid” mixed ages are actually more profitable than those
abilities), whereas for work tasks where experience with exclusively young or old workers.
of one in three people on average in the red cross-sectional evidence ignores cohort effects
economies before declining again (figure 3.5). that drive a worker age 60 today to be both
Other things being equal, the workforce in more productive and more able to acquire
these countries could be expected to experi- new skills than a worker of the same age three
ence a steady decline in labor productivity as decades earlier.
it ages rapidly in the next 30 years and prime In orange and green economies, the
age groups are less represented. However, maturing of the population offers increas-
all things will not be equal. To begin with, the ing opportunities for productivity growth,
AG I N G A N D T H E MAC R O E CO N O M Y I N E A S T A S I A A N D PAC I F I C 91
FIGURE 3.5 Productivity depends on the age distribution of the population, but mitigation channels
exist
Share of population ages 35–54, 1950–2050
32%
32
Share of population ages 35–54 (%)
30 29%
28
26
24 25%
22
20
20%
18 17%
16
16%
1950 1960 1970 1980 1990 2000 2010 2020 2030 2040 2050
Years
Hong Kong SAR, China; Japan; Republic of Korea; Singapore
China; Indonesia; Malaysia; Mongolia; Thailand; Vietnam
Cambodia; Lao PDR; Myanmar; Papua New Guinea; Philippines; Timor-Leste
yet the pace of growth will depend on physi- entrepreneurship, which is of particular con-
cal and human capital investments and how cern in countries seeking to promote innova-
efficiently they are put to use. The next 15–20 tion. Figure 3.6 shows the share of workers
years will see a rise in the average share of the by age who manage and own a business and
prime-age population of orange economies, those who expect to launch a start-up busi-
although a flatter rise than that experienced ness in the coming three years. Nonetheless,
in the 30 years since 1985. Green economies China, Japan, and Korea have relatively
should experience some positive growth in high ownership of businesses among older
labor productivity from the maturing of workers, and Japan is distinctive in that this
their workforce, peaking around 2045. Yet increases as a share with age.
although the age-productivity profile is the
most obvious link between population aging
and labor productivity, aggregate productiv-
Saving in East Asia and Pacific:
ity growth in the future will be determined by
Will aging play an influential role?
investments in human capital and the quality The net effect of aging on labor produc-
of education. Saving behavior and the types tivity depends in part on the capital stock
of saving mechanisms available to households and how capital accumulation evolves with
will also have a role, as will the capacity of population aging.6 The concerns over saving
the economy to make efficient use of these in East Asian countries with aging popu-
human and physical inputs. The next section lations reflect the important role played
focuses on the saving channel and aging. by domestic savings in the high growth of
One feature that appears to decline the region to date. Savings not only fund
with age of the workforce globally is investment but also allow lower-income
92 LIVE LONG AND PROSPER
FIGURE 3.6 Although in many countries, entrepreneurship seems to decline with worker age, business
ownership in East Asian countries is significant among older workers
Adults who own a business and engaged in start-up activities, by age, selected areas
15
10
0
Japan China Malaysia Korea, Rep. United States ECA LAC
40
20
0
Japan China Malaysia Korea, Rep. United States ECA LAC
Ages 18–34 Ages 35–44 Ages 45–54 Age 55 and over
countries to adopt more advanced technolo- 2011). It could possibly even accelerate
gies, which matters enormously for growth. productivity growth and thus mitigate the
Aging may affect supply and demand for effects of a shrinking labor force. However,
household savings, potentially corporate as dependency ratios stabilize, the supply
savings, and almost certainly government channels discussed below ultimately deter-
savings because of the effects of age-related mine the magnitude of long-term productiv-
public spending. On the demand side, one ity and growth rates.
important channel through which chang- Household saving in much of East Asia is
ing demographics affect growth is increases high and provides a robust base as the region
in labor productivity. A decrease in the size ages. In East Asian economies, the average pri-
of the labor force in aging economies leads vate savings rate between 1960 and 2000 was
to an increase in capital intensity (or the 25 percent—exceeding those of most other
capital-to-labor ratio), thus increasing labor regions. China has accounted for the predom-
productivity. Such capital deepening and inant share of this impressive increase since
consequent increase in productivity leads the 1960s (figure 3.7). That said, domestic
to higher standards of living, even in the savings across the region vary substantially:
absence of higher savings (Lee and Mason from 1966 to 2007, the savings rate ranged
AG I N G A N D T H E MAC R O E CO N O M Y I N E A S T A S I A A N D PAC I F I C 93
FIGURE 3.7 Compared with household savings rates in other regions, the rate in East Asia and Pacific is relatively
high and stable
Private savings as a percentage of GDP by region, 1980–2010
a. East Asia and Pacific b. Europe and Central Asia c. High-income OECD countries
40 40 40
Private savings as a
Private savings as a
Private savings as a
20 20 20
10 10 10
1980 1990 2000 2010 1980 1990 2000 2010 1980 1990 2000 2010
Year Year Year
d. Latin America and the Caribbean e. Middle East and North Africa f. South Asia
40 40 40
Private savings as a
Private savings as a
Private savings as a
30 30 30
20 20 20
10 10 10
1980 1990 2000 2010 1980 1990 2000 2010 1980 1990 2000 2010
Year Year Year
from well over 30 percent in China; Hong expected duration of consumption in retire-
Kong SAR, China; Malaysia; and Singapore ment. Predictions on which of these two
to below 20 percent in the Philippines and effects will dominate in East Asia and Pacific
Vietnam (Horioka and Terada-Hagiwara are mixed. Some authors suggest the compo-
2012). Although household saving has fluctu- sitional effect will dominate the behavioral
ated considerably, corporate saving has been effect; others maintain that extensions in
on a steady upward trajectory since 2001. longevity will positively affect savings rates
The demographic effect on the savings in the future, while others find mixed effects
rate is difficult to predict because the dif- (see below). On balance, the studies suggest
ferent demographic transformations driv- less cause for concern in East Asia and Pacific
ing population aging will have ambiguous than in other parts of the world.
effects on the savings rate.7 Two offsetting Whether aging in East Asia and Pacific
effects determine the net effect of longev- will adversely affect savings growth prospects
ity on aggregate saving: a compositional depends on three regional characteristics.
effect that leads to a reduced aggregate sav- First are the patterns of life-cycle profiles of
ings rate because of an increase in the share savings, which affect the scale of the compo-
of older people in the adult population; and sitional effect as the population age distribu-
behavioral effects, such as the rise in the tion changes. Second are the adjustments to
individual savings rate to finance a longer saving behaviors in response to lower fertility
94 LIVE LONG AND PROSPER
accounting for three-fourths of the increase be too negative because the precautionary
in Asia. Moreover, their analysis predicts that motive suggests efficiency gains can be made
the behavioral effect of longevity increases by shifting to lower levels but higher quality
outweighs the compositional effect, implying of saving: for example, by expanding social
that population aging will not lower saving insurance. A precautionary motive also
in Asia. Using cross-country panel data, Li, appears to explain the high corporate savings
Zhang, and Zhang (2007) find that the effect rate in China. In the future, policies decreas-
of increased longevity on aggregate sav- ing income volatility can also lessen the need
ings is positive whereas that of an increased for inefficient precautionary saving by firms.
old-age dependency rate is negative. World An increase in the variance of permanent
Bank (2013a) also argues that the drag on income shocks implies not only a higher sav-
saving from the compositional effect will be ings rate but also a change in the portfolio
offset by the behavioral effect in East Asia allocation of savings toward safe assets and a
and Pacific. decrease in investment (Cherif and Hasanov
The concern about reduced household sav- 2012). Conversely, International Monetary
ing induced by aging in East Asia and Pacific Fund (IMF) analysis of the impact of finan-
may thus be overstated because broader cial liberalization on corporate saving glob-
macroeconomic channels are likely to play ally reveals that financial reforms may have
a dominating role. First, a precautionary a considerably larger effect in Asia (IMF
motive appears to be driving high saving in 2009). This evidence suggests that financial
East Asia and Pacific, suggesting important sector reforms are likely to reduce precau-
inefficiencies despite high national rates of tionary saving by the corporate sector in
saving. Second, further financial liberaliza- Asia. In fact, recent policy changes indicate
tion may contribute to a decline in the sav- that many middle-income countries in East
ings rate, which will be a sign of improved Asia and Pacific are already moving in this
efficiency. Finally, improvements in financial direction.
intermediation, inclusion, and social security Improvements in financial inclusion will
coverage will play a crucial role in determin- play an important role for channeling house-
ing levels of saving and in channeling savings hold savings into productive investment.
into productive investment. Greater financial inclusion, either through
A precautionary motive appears to be bank or nonbank financial institutions,
driving high household and corporate sav- allows for efficient flow of savings and invest-
ing in East Asia and Pacific, suggesting ments in the economy (Levine 2003; Levine,
inefficiencies that, if reduced, would allow Loayza, and Beck 2000). At the household
stronger economic effects from a given rate level, bank account penetration differs across
of saving. Blanchard and Giavazzi (2005) countries in East Asia and Pacific. With a rate
stress the importance of reducing precau- of 55 percent, East Asia and Pacific is ahead
tionary saving to boost consumption and of most regions in bank account penetration
rebalance growth. In China, market-oriented and only behind high-income economies, yet
reforms led to large-scale corporate restruc- access to personal saving instruments and
turing and downsizing of the public sec- credit varies significantly across countries.
tor between 1995 and 2005, consequently According to Demirgüç-Kunt and Klapper
weakening the public social safety net. (2012), in countries such as Cambodia more
Households responded by building up sav- than 95 percent of adults do not have an
ings to self-insure against future health and account at a formal financial institution. In
old-age income needs (Chamon and Prasad sum, these regional stylized facts reflect that,
2010; Ma and Yi 2010). Therefore, even aging pressures notwithstanding, East Asian
if savings rates decline in future, the aging and Pacific countries have varying abilities to
effect means the impact on growth may not mobilize savings, allocate capital efficiently,
96 LIVE LONG AND PROSPER
and diversify risks, which suggests that sig- cases and that will require strong policy
nificant gains remain to be made. leadership to mitigate. As noted, the most
obvious fiscal risks are posed by pension,
health, and long-term care systems, all of
Fiscal effects of aging which are affected by aging (see chapters 5
The biggest risk to aggregate savings rates to 7 of this report for detailed discussion
in East Asia and Pacific will be through the and summaries of country-specific expendi-
impact on government saving caused by ture projections). The fiscal risks are com-
rising pension, health, and long-term care pounded in many East Asian and Pacific
spending. This risk is of particular concern countries because governments are pur-
in developing East Asian and Pacific coun- suing rapid coverage expansion of health
tries where projected fiscal space is very insurance and pensions and exploring more
limited (see Pradelli and Van Doorn 2015). active public financing for long-term care.
Although reasons exist to be sanguine Various projections of pension spending
about the potential effects on household in East Asia and Pacific reveal significant
and corporate savings from aging, the same fiscal pressure from rising pension spending
is not the case for government saving. Even over coming decades. Various cross-country
assuming relatively benign effects of aging projections are available that include East
on economic growth in the region, major Asia and Pacific, some country-specific and
fiscal risks from aging remain that are some more stylized around regional coun-
already manifesting themselves in several try groupings. Although estimated impacts
cover a considerable range (and are impor-
tant to supplement with more detailed
FIGURE 3.8 Projected increases in pension spending to 2030 point country projections presented in chapter 5
to a significant increase in China and Malaysia, with more modest of this report), they all point to the fiscal
increases for Indonesia, the Philippines, and Thailand risk presented by pension systems in the
Projected increase in pension spending as a percentage of GDP, 2010–30 region. In the following, we examine a few
estimates.
Turkey IMF projections with simplified but
Egypt, Arab Rep. standard assumptions across countries
China
Jordan
(figure 3.8) point to a significant increase in
Russian Federation pension spending in China and Malaysia,
Saudi Arabia with more modest increases for Indonesia,
Malaysia
Ukraine the Philippines, and Thailand (although
Romania the projection for Indonesia does not take
Argentina into account the new pension scheme). The
Brazil
Mexico increase in China of over 3 percent of GDP
Latvia is particularly notable. Not shown in the fig-
Philippines ure, but available from the same source, are
South Africa
Lithuania estimates for Japan and Korea. Japan exhibits
Thailand a slight fall in pension spending as a share of
Indonesia
Pakistan
GDP, reflecting its already aged population
India and cost-containment measures in its pension
Bulgaria scheme (see chapter 5 in this report), whereas
Average = 1.0
Chile
Poland Korea experiences a significant increase in
Estonia pension spending approaching 4.5 percent
Hungary in the 2010–30 period, reflecting the matu-
–4 –2 0 2 4 6 8 ration of its relatively young pension system
% of GDP and the continuation of rapid aging.
Simplified estimates of pension spending in
Source: IMF Fiscal Affairs Department 2010. 2050 done for this report (which incorporate
AG I N G A N D T H E MAC R O E CO N O M Y I N E A S T A S I A A N D PAC I F I C 97
FIGURE 3.9 Estimated increases in East and Southeast Asian pension spending for 2014 to 2050 find significant fiscal risks
from pensions
Projected increase in pension spending as a percentage of GDP, 2014–50, various economies
4
% of GDP
0
a
DR
ar
ea
sia
nd
m
ia
in ,
.
Ch AR
ep
ne
di
in
pa
ys
nm
na
in
ila
ne
oP
Ch
bo
,R
a
S
pi
ala
Ja
Gu
et
a
ng
do
ya
ilip
a
m
La
Th
Vi
re
M
M
Ko
Ca
In
Ph
Ko
Ne
ng
ua
Ho
p
Pa
debt dynamics) also find significant fiscal FIGURE 3.10 Pension spending projections by stylized groupings
risks from pensions. China continues to have for APEC economies, 2010–70, illustrate rapid increase in pension
a substantial fiscal burden from pensions spending in a no-reform scenario
(similar in magnitude to the IMF estimate), Projected pension spending as a percentage of GDP in APEC economies (no-
while the incremental costs in Malaysia, the reform scenario), 2009–75
Philippines, and Vietnam are also substantial
(figure 3.9). 18
16
A third estimation exercise uses a stylized
14
approach to capture broad groupings of the
12
Asia-Pacific Economic Cooperation (APEC)
% of GDP
10
economies and potential fiscal costs of future 8
pensions (figures 3.10 and 3.11) (Hinz and 6
Zviniene 2011).12 Notable about this projec- 4
tion, which uses the World Bank’s PROST 2
(Pension Reform Options Simulation Toolkit) 0
model, is that similar long-term outcomes in 2009 2019 2029 2039 2049 2059 2069 2079
overall pension spending are reached at dif- High-income economies, mature pension system
ferent rates and through the interaction of Middle-income economies, maturing pension system
different influences. The high-income mature High-income economies, maturing pension system
systems experience increases earlier, add- Low-income economies, developing pension system
ing more than 5 percent of GDP to pension
spending costs in the first 20 years and nearly Source: Hinz and Zviniene 2011.
Note: APEC = Asia-Pacific Economic Cooperation.
10 percent of GDP after 40 years. These addi-
tions are primarily owing to the advanced
stage of population aging, but assume and Pacific. The lower- and middle-income
modest accrual rates that are characteristic cases experience lower spending increases in
of mature high-income systems in East Asia the initial decades of only 2 to 3 percent of
98 LIVE LONG AND PROSPER
FIGURE 3.11 Pension spending increases can be substantially GDP but then rapidly increase an additional
reduced with key reforms 10 percent of GDP in spending between 2030
Projections by stylized groupings for APEC economies, 2010–70, assuming a and 2070. The rapid increase in the later
gradual five-year increase in pension age and price indexation years is caused by the accelerating process of
coverage expansion. The baseline is driven
20
18 by the assumption that these economies
16 continue to retain high annual accrual rates
14
12 that are characteristic of low- and middle-
Percent
available only for China, Indonesia, Japan, FIGURE 3.13 Projected increases in public health spending
and Korea in East Asia and Pacific (De la reflect the importance of policy choices in sustainability of health
Maisonneuve and Oliveira Martins 2013). All care systems
East Asian and Pacific and OECD economies Percentage of GDP increase in public health care spending under cost
pressure and cost containment scenarios, 2010–30
will experience significant increases in system
costs over time. A few other pertinent points 3.0
emerge from figure 3.13. First, although the
absolute projected increase in public health 2.5
spending is higher in red economies, the
p.
sia
il
ico
ey
demographic effect than previous IMF esti-
az
in
di
pa
tio
Re
rk
ne
ex
Ch
In
Br
Ja
ra
Tu
a,
do
de
mates. Second, reflecting this assumption,
re
In
Fe
Ko
ian
the gap between cost containment (benign
ss
Ru
policy) and cost pressure (pessimistic policy)
Cost pressure Cost containment o/w demographic
is significant and in Indonesia, for example, effect
almost as large as the entire effect of aging.
In short, aging matters and should be taken Source: De la Maisonneuve and Oliveira Martins 2013.
Note: o/w = out of which.
into account, but policy choices also matter a
lot for the sustainability of health and long-
term care systems.
A final source to consider is country- FIGURE 3.14 Projected increases in public long-term care spending
specific projections of health spending and indicate additional fiscal pressure in countries where the aging
the impact of aging. Unfortunately, these effect is strong
projections are limited for East Asian and Percentage of GDP increase in public spending on long-term care,
Pacific economies, but the two available 2010–30, various countries
actuarial studies that look at this question
suggest effects from aging more aligned 0.9
with IMF estimates. For Malaysia, about 0.8
one-third of the total projected increase in 0.7
GDP increase (%)
sia
zil
ico
y
ep
e
in
di
pa
io
a
rk
ne
ex
Ch
,R
In
at
Br
Ja
Tu
M
ea
d
r
In
Fe
Ko
is public funding for long-term care. As chap- Cost pressure Cost containment o/w demographic
ter 7 of this report shows, publicly financed effect
long-term care is present in richer economies of Source: De la Maisonneuve and Oliveira Martins 2013.
the region such as Japan, Korea, and Taiwan, Note: o/w = of which.
100 LIVE LONG AND PROSPER
China, but remains in its infancy in the region’s the World Bank do not incorporate such
developing economies. However, long-term interactions between demographics and fis-
care is an area in which governments in rapidly cal accounts. This decision is understandable
aging countries such as China, Thailand, and in settings where the focus is on short-run
Vietnam are showing increasing interest and macrofiscal situations, but such interactions
already making policy commitments. OECD become relevant when looking at long-term
projections for long-term care in the region debt sustainability. A highly stylized effort
point to a further (though for the foreseeable at incorporating demographics into debt sus-
future relatively modest) source of fiscal pres- tainability analysis follows here.
sure where—as for pensions—the aging effect The model underlying the following
is strong (figure 3.14). simulations is a standard fiscal sustainability
Each estimation approach has its strengths model with additional links to demograph-
and limitations, but all point toward ics and age-related expenditure. The analysis
significant fiscal risks from increased age- uses a standard fiscal sustainability frame-
related spending in East Asia and Pacific, work linking GDP, fiscal accounts, and debt
particularly in the aging middle-income with long-term demographic projections
countries (largely orange economies in the and simple models that link demographics
report typology). Chapters 5, 6, and 7 of to production and to age-related expendi-
this report look at current pension, health, ture. The model is highly stylized to make
and long-term care systems in the region and sure that interactions and results are well
discuss reforms that aim to strike a balance understood, so only the channels from demo-
between managing cost escalation and fiscal graphics to GDP and from demographics to
risk, ensuring basic adequacy of benefits and age-related expenditure have been modeled.
services, and expanding coverage where it is Figure 3.15 shows a schematic overview of
currently less than complete. the links between demographics and fis-
A more complex (and uncertain) picture cal accounts that underpin the simulation
of fiscal pressures from aging emerges if we results.
try to incorporate debt dynamics into fiscal Using the characteristics of 16 developing
projections. Although the preceding fiscal East Asian and Pacific countries, a number of
projections are very useful for understanding stylized cases are presented on debt dynamics
orders of magnitude of age-related spend- that reflect different combinations of three
ing pressures, they fail to take into account dimensions:
that countries may be running fiscal deficits
and hence financing public spending with • Aging pattern. Broadly, two types of
debt that incurs further interest obliga- countries exist: those with young and
tions. To develop a more complete analysis slowly aging populations (green econo-
of whether an economy will have the fiscal mies) and those with older and faster-
resources to deal with the effects of aging, aging populations (orange economies such
a few more factors need to be taken into as China, Indonesia, Malaysia, Thailand,
account: national investment, external cur- and Vietnam).
rent account, foreign direct investment ver- • Age-related expenditure levels. For each
sus debt financing, and total external debt. group, the simulations analyze the impact
Countries that already have a structural cur- of current per capita public spending on
rent account deficit and high external debt health care, education, and pensions in
will have a smaller envelope to work with the long run. Thus, two levels of current
than those with the same demographics and aging spending are seen: low spending
saving behavior but with a current account (spending on health care, education, and
surplus and low external debt. However, the pensions at 1.5 percent of GDP each at
standard debt sustainability analysis mod- end-2013) and high spending (spending at
els used by agencies such as the IMF and 2 percent of GDP each).
AG I N G A N D T H E MAC R O E CO N O M Y I N E A S T A S I A A N D PAC I F I C 101
FIGURE 3.15 Schematic overview of the links between demographics and fiscal accounts
Y=A×L
Labor force A: productivity
(exogenous) GDP
Participation rate L: labor force
(exogenous)
Revenue (fixed as % of
Demographics
GDP)
(exogenous
projections)/working-age Expenditure:
population • Nonaging, noninterest
(exogenous)
• Pensions
• Health
• Education
• Interest
Interest
Vector of age-related Fiscal balance payments
expenditure parameters
× Vector of
demographics = Age- Financing
related expenditure
Debt accumulation
Financing
assumptions Debt at t = 0
(interest rate)
• Initial debt levels. The initial (end-2013) populations and low age-related spending
public debt burden affects long-term sus- have the best fiscal sustainability outlook,
tainability because it amplifies deficits as we might expect. Maintaining a primary
driven by age-related expenditure: low fiscal surplus and a healthy growth rate
debt means public debt at 20 percent of (higher than the interest rate paid on pub-
GDP, and high debt means 60 percent of lic debt) are the key determinants of the fis-
GDP.13 cal sustainability of aging. In cases 1 and 2
(aging countries), both high primary fiscal
The simulations demonstrate for a stylized deficits and adverse interest rate growth
set of East Asian and Pacific countries the dynamics lead to continuous debt accumula-
major variability in public debt over time tion. Cases 3 and 4 have the same spending
from different interactions of initial debt lev- in the beginning but more favorable demo-
els, spending on age-related programs, and graphics, which leads to both a narrower
demographic trends (figure 3.16). The initial primary deficit and higher growth. The last
level of debt is almost a pure level effect. four cases start with low age-related spend-
The results show that the case of an aging ing, which allows them to run primary
country with a high level of age-related surpluses, at least for the first 15 years.
spending and a high level of initial debt is The countries with younger populations still
clearly unsustainable, followed closely by grow relatively fast and are able to reduce
the equally unsustainable case with a lower debt. Although these are only stylized cases
level of initial debt. Countries with younger (built on composites of real country initial
102 LIVE LONG AND PROSPER
FIGURE 3.16 Initial debt and spending levels and future debt Addressing concerns about saving in the
dynamics have important interactions with demographics that can region will require improved social secu-
sharply increase the fiscal impacts of aging rity and the deepening of financial mar-
Stylized debt dynamics and composition with East Asian and Pacific kets and the role of banks. Ultimately, the
demographics, 2013–60, under different starting combinations of spending
impact of aging on capital formation in
levels, debt, and demographics
East Asia and Pacific will depend on the
interaction of a number of factors: the pace,
a. Public debt evolution for each stylized country case magnitude, and intensity of demographic
300 changes; the consequent behavioral adjust-
250 ments in response to fertility and longevity
changes; and the macroeconomic environ-
Share of GDP (%)
0
–1
–2
–2
–3
–3
–4
–4
–5
–5
–6
19
24
29
34
39
44
49
54
59
20
20
20
20
20
20
20
20
20
20
150 Conclusion
100 The preceding analysis suggests that even
Percent
The bigger risks from aging in the region account. The implication is that they will be
are fiscal and require urgent policy attention aging faster.
in rapidly aging countries. Distinguishing the 2. This literature generally excludes the Pacific.
demographic and economic challenges fac- 3. See for example Park, Lee, and Mason (2012),
ing different groups of East Asian and Pacific which provides useful growth accounting
countries is important as they think about results, finding for a selection of East and
South Asian economies that only in Korea and
aging populations:
Hong Kong SAR, China, does the negative
• For red economies, reducing the structural growth effect of rising old-age dependency
outweigh the positive contribution of falling
labor force decline from aging will be the
youth dependency from 2011 to 2020.
most critical challenge. Although the fis-
4. Interestingly, however, surveys of global
cal risks are significant, a number of red
attitudes to migration find more positive
economies have already adopted forward- attitudes to immigration in some potential
looking policies in areas such as pensions host countries in East Asia and Pacific.
and health that will help contain the fis- Korea, for example, had the lowest share
cal impact of further aging, although areas of respondents globally agreeing that
such as long-term care will continue to immigration should be further restricted (at
pose challenges (see chapters 5–7 of this 25 percent), and Japanese respondents were
report for details). almost evenly split. These findings contrast to
• For orange economies, the challenges of those in Malaysia, where almost 90 percent
the demographic transition require sus- of respondents wanted more restrictions on
immigration (Pew Research Center 2007).
taining high productivity growth. At the
5. For literature on the link between age
same time, they will need to mitigate the
composition and growth, see Bloom and
labor supply and fiscal impacts of rapid Williamson (1998); Feyrer (2007, 2008);
aging through ongoing pension and health Lindh and Malmberg (1999); Modigliani
reforms and labor policies to extend the (1986); Modigliani and Cao (2004); Persson
working lives of urban and formal sector (1999); and Sarel (1997).
workers. 6. This section draws on a background paper by
• For the younger green economies, the pri- Malathi Velamuri (2015).
orities are to establish the conditions to 7. Studies that have shown strong links between
realize maximum GDP growth from the national savings rates and age structure
demographic dividend (that is, invest- include those by Bloom et al. (2007); Deaton
ments to raise productivity and maximize and Paxson (1997); Fry and Mason (1982);
youth employment) and to avoid adopt- Higgins (1998); Kelley and Schmidt (1995);
ing policies in areas such as pensions and Lee, Mason, and Miller (2000); and Mason
(1988).
health that are affordable now but would
8. In theory, a stationary population has no
rapidly become unsustainable once aging
aggregate savings because in the absence of
accelerates. a bequest motive, the dissaving of the old
The main areas of services and entitlement and the saving of the young cancel out (Ando
reform—pensions, health, and long-term and Modigliani 1963; Bloom, Canning, and
Sevilla 2003). However, when the population
care—are discussed in chapters 5–7 of this
grows or the economy is growing rapidly and
report. incomes of wage earners are high relative to
the incomes of the retired (as in East Asia and
Pacific), aggregate savings or dissavings may
Notes occur.
1. Because only elderly dependency ratios are 9. See Horioka (2010) and the references therein.
shown here, the patterns may not perfectly 10. See Deaton (1992) for Thailand, and Deaton
match the patterns of aging across groups. and Paxson (1993) for Taiwan, China.
Some countries, such as Singapore, appear 11. See the National Transfer Accounts website,
much younger than they are when the rest http://www.ntaccounts.org/web/nta/show
of the population age structure is taken into / NTA%20Data.
104 LIVE LONG AND PROSPER
Cutler, David, and Adriana Lleras-Muney. Fukao, Kyoji, Ryo Kambayashi, Daiji Kawaguchi,
2008. “Education and Health: Hyeog Ug Kwon, Young Gak Kim, and Izumi
Evaluating Theories and Evidence.” In Yokoyama. 2006. “Deferred Compensation:
Making Americans Healthier: Social and Evidence from Employer-Employee Matched
Economic Policy as Health Policy, edited by Data from Japan.” Discussion Paper 187,
Robert F. Schoeni, James S. House, George Institute of Economic Research, Hitotsubashi
A. Kaplan, and Harold Pollack, 29–60. New University, Tokyo.
York: Russell Sage Foundation. Global Entrepreneurship Monitor (database).
D’Addio, Anna, Mark Keese, and Edward 2011. Global Entrepreneurship Research
W hitehouse. 2010. “Population Ageing A ssociation, L ondon Business S chool,
and Labour Markets.” Oxford Review of London, http://www.gemconsortium.org
Economic Policy 26 (4): 613–35. / Data.
De la Maisonneuve, Christine, and Joaquim Göbel, Christian, and Thomas Zwick. 2013. “Are
Oliveira Martins. 2013. Public Spending on Personnel Measures Effective in Increasing
Health and Long-Term Care: A New Set of Productivit y of Old Workers?” L abor
Projections. Paris: Organisation for Economic Economics 22: 80–93.
Co-operation and Development. Gómez, Rafael, and Pablo Hernández de Cos.
Deaton, Angus. 1992. “Household Saving 2008. “Does Population Ageing Promote
in LDCs: Credit Markets, Insurance and Faster Economic Growth?” Review of Income
Welfare.” Scandinavian Journal of Economics and Wealth 54 (3): 350–72.
94 (2): 253–73. Grund, Christian, and Niels Westergaard-Nielsen.
Deaton, Angus, and Christina H. Paxson. 1993. 2008. “Age Structure of the Workforce and
“Intertemporal Choice and Inequality.” Firm Performance.” International Journal of
NBER Working Paper 4328, National Bureau Manpower 29 (5): 410–22.
of Economic Research, Cambridge, MA. Hanushek, Eric, and Ludger Woessmann. 2009.
———. 1997. “The Effects of Economic and “Do Better Schools Lead to More Growth?
Population Growth on National Saving and Cognitive Skills, Economic Outcomes, and
Inequality.” Demography 34 (1): 97–114. Causation.” NBER Working Paper 14633,
Demirgüç-Kunt, Aslı, and Leora Klapper. 2012. National Bureau of Economic Research,
“Measuring Financial Inclusion: The Global Cambridge, MA.
Findex Database.” Policy Research Working Higgins, Matthew. 1998. “Demography, National
Paper 6025, World Bank, Washington, DC. Savings, and International Capital Flows.”
Eggleston, Karen, and Victor Fuchs. 2012. “The International Economic Review 39: 343–69.
New Demographic Transition: Most Gains in Hinz, Richard, and Asta Zviniene. 2011. “The
Life Expectancy Now Realized Late in Life.” Potential Fiscal Effects of Population Aging:
Journal of Economic Perspectives 26 (3): Simulation of Aggregate Pension and Health
137–56. Expenditure for Four Stylized Countries at
Feyrer, James. 20 07. “Demographics and Various Levels of Economic Development.”
Productivity.” Review of Economics and World Bank, Social Protection and Labor
Statistics 89 (1): 100–109. Practice, Washington, DC.
———. 2008. “Aggregate Evidence on the Link Horioka, Charles Yuji. 2010. “The (Dis)saving
between Age Structure and Productivity.” Behavior of the Aged in Japan.” Japan and the
Population and Development Review 34 World Economy 22 (3): 151–58.
(suppl.): 78–99. Horioka, Charles Yuji, and Akiko Terada-
Flochel, Thomas, Yuki Ikeda, Harry Moroz, and Hagiwara. 2012. “The Determinants and
Nithin Umapathi. 2014. “Macroeconomic Long-Term Projections of Saving Rates in
Implications of Aging: Demographics, Labor Developing Asia.” Japan and the World
Markets, and Productivity in East Asia Economy 24 (2): 128–37.
Pacific.” Background paper for the East Asia IHME (Institute for Health Metrics) and World
and Pacific regional report on aging, World Bank. 2013. The Global Burden of Disease:
Bank, Washington, DC. Generating Evidence, Guiding Policy—East
Fry, Maxwell J., and Andrew Mason. 1982. Asia and Pacific Regional Edition. Seattle,
“The Variable Rate-of-Growth Effect in the WA: IHME.
Life-Cycle Saving Model.” Economic Inquiry ILO (International Labour Organization). 2011.
20 (3): 426–42. LABORSTA (database), 6th ed., ILO, Geneva.
106 LIVE LONG AND PROSPER
[in Japanese], 29–47. Tokyo: Japanese Center Paper 97/97, International Monetary Fund,
for Economic Research. Washington, DC.
OECD (Organisation for Economic Co-operation S c hu l t z , T. Pau l . 2 0 0 4 . “ D e m o g r ap h i c
and Development). 2012. PISA 2012 Results: Determinants of Savings: Estimating and
What Students Know and Can Do: Student Interpreting the Aggregate Association in
Performance in Mathematics, Reading and Asia.” Working Paper 901, Economic Growth
Science. Vol. 1. Paris: OECD. Center, Yale University, New Haven, CT.
Özden, Çaglar, and Mauro Testaverde. 2014. Shinada, Naoki. 2011. “Quality of Labor, Capital,
“International Migration in Aging Societies: and Productivity Growth in Japan: Effects
Impacts and Benefits in East Asia and Pacific of Employee Age, Seniority, and Capital
Countries.” Background paper for the East Vintage.” RIETI Discussion Paper 11-E-036,
Asia and Pacific regional report on aging, Research Institute of Economy, Trade, and
World Bank, Washington, DC. Industry, Tokyo.
Park, Donghyun, Sang-Hyop Lee, and Andrew Shirakawa, H i romichi. 2010. “E mpirical
Mason, eds. 2012. Aging, Economic Growth, Analysis of Relationship between Worker
and Old-Age Security in Asia. Manila: Asian Age, Productivity, and Real Wages.” In Is the
Development Bank; Cheltenham, U.K.: Aging of Society a Threat to Japan?, 48–71.
Edward Elgar. Tokyo: National Institute for Research
Persson, Joakim. 1999. “Demographic and Per Advancement.
Capita Income Dynamics: A Convergence Skirbekk, Vegard. 2008. “Age and Productivity
Study on Demographics, Human Capital, Capacity: Descriptions, Causes, and Policy
a n d Pe r C ap i t a I n c o m e f o r t h e U S Options.” Ageing Horizons 8: 4–12.
States.” FIEF Working Paper 1999/156, Skirbekk, Vegard, Marcin Stonawski, Eric
Fackföreningsrörelsens Institut för Ekonomisk Bonsang, and Ursula M. Staudinger. 2013.
Forskning, Stockholm. “The Flynn Effect and Population Aging.”
Pew Research Center. 2007. Global Unease Intelligence 41 (3): 169–77.
with Major World Powers: 47-Nation Pew UN (United Nations). 2013. “World Population
Global Attitudes Survey. Washington, DC: Prospects: The 2012 Revision—Highlights
Pew Research Center. http://www.pewglobal and Advance Tables.” Working Paper ESA/P/
.org/files/pdf/2007%20Pew%20Global%20 WP.228, Population Division, Department of
Attitudes%20Report%20-%20June%2027 Economic and Social Affairs, United Nations,
.pdf. New York.
Pradelli, Juan, and Ralph Van Doorn. 2015. Van Ours, Jan C. 2009. “Will You Still Need
“Fiscal Impact of Aging.” Background paper Me: When I’m 64?” De Economist 157 (4):
for the East Asia and Pacific regional report 441–60.
on aging, World Bank, Washington, DC. Velamuri, Malathi. 2015. “Aging and Savings in
Rannan-Eliya, Ravi P., Chamara Anuranga, Adilius Asia: Current Trends and Future Prospects.”
Manual, Ruwani Wickramasinghe, Prasadini Background paper for the East Asia and
Perera, Shanaz Saleem, Jayalal Chandrasiri, Pacific regional report on aging, World Bank,
Indra Pathmanathan, and Institute for Health Washington, DC.
Systems Research Core Team. 2013. Malaysia Voelcker-Rehage, Claudia, Ben Godde, and
Healthcare Demand Analysis: Inequalities in Ursula M. Staudinger. 2011. “Effects of
Healthcare Demand and Simulation of Trends Different Exercise Interventions on Cognitive
and Impact of Potential Changes in Healthcare Functioning in Older Adults.” Journal of
Spending. Kuala Lumpur: Institute for Health Psychophysiology 25: 29.
Systems Research. World Development Indicators (database).
Romeu Gordo, Laura, and Vegard Skirbekk. Various years. World Bank, Washington, DC.
2013. “Skill Demand and the Comparative http://data.worldbank.org/data-catalog/world
Advantage of Age: Jobs Tasks and Earnings -development-indicators.
from the 1980s to the 2000s in Germany.” Willis, Robert. 1973. “A New Approach to the
Labor Economics 22 (suppl.): 61–69. Economic Theory of Fertility Behaviour.”
Sarel, Michael. 1997. “Growth and Productivity Journal of Political Economy 81 (2, pt. 2):
i n A SE A N C ou nt ries.” I M F Work i ng S14–64.
108 LIVE LONG AND PROSPER
World Bank. 2013a. “Global Saving in 2030.” In ———. 2013c. Skills for Quality Jobs and
Capital for the Future: Saving and Investment Development in Lao PDR: A Technical
in an Inte rde pe n de nt World, 61–112 . A s s e s s m e n t of t he C u r re n t C o n t e x t.
Washington, DC: World Bank. Washington, DC: World Bank.
———. 2013b. “Labor Market Issues in Timor-Leste: ———. 2 0 1 3 d . W o r l d D e v e l o p m e n t
Current State, Prospects, and Challenges.” I n d i c a t o r s 2 0 1 3 . Wa s h i n g t o n , D C :
World Bank, Washington, DC. World Bank.
Labor Market Policies to Address
Aging in East Asia and Pacific 4
This chapter draws from a background literature review for this report by Shonali Sen (2014), a back-
ground paper on migration in East Asia and Pacific by Çaglar Özden and Mauro Testaverde (2014),
inputs on labor and aging policies in the Republic of Korea by Dewen Wang, and a study visit to Japan
by the task team.
109
110 LIVE LONG AND PROSPER
and OECD countries, because they are more breaks and rebates for families with young
advanced in trying to address the labor mar- children or large families; and generalized
ket consequences of rapid aging. child allowances. In some cases, national
initiatives have been supplemented by sub-
national programs, as in Seoul. For several
Fertility incentives East Asian and Pacific countries, such mea-
In recent decades, multiple initiatives have sures have lagged the decline in fertility below
been undertaken in East Asian and Pacific replacement, reflecting the legacy of earlier
countries with aged populations to reverse very family planning policies that were successful
low fertility rates and to increase the participa- in rapidly reducing previously high fertility
tion of women in the labor force. As noted in rates. Korea, for example, introduced its first
chapter 1, the richer red economies have expe- pronatalist strategy (First Basic Plan for Low
rienced very low fertility rates for some time, Fertility and Aged Society) only in 2006, well
and their earlier trajectories of fertility decline after fertility had declined to very low levels.
are seemingly being followed by a number of The balance of evidence from East Asia
middle-income countries in the region, includ- and Pacific on direct incentives to have more
ing China, Thailand, and Vietnam. In addi- children suggests they have not been particu-
tion, although East Asia and Pacific generally larly effective or cost-effective.1 Evaluation of
has relatively high female labor force par- Japan’s history of pronatalist initiatives since
ticipation by global standards, a number of the early 1990s found almost no effect on
countries, including Fiji, Japan, Malaysia, and the likelihood of second and third children,
the Republic of Korea, face major challenges although a slight recovery has occurred in
getting women to reenter the workforce after women having at least one child (Ma 2010).2
childbirth (World Bank 2012). This difficulty The consistently low fertility rates across the
has led to many initiatives to stimulate fertil- region’s red economies suggest that similar
ity directly and to promote family-friendly conclusions are likely to be reached and that
employment policies in countries such as a wider array of policies and changes in social
Japan and Korea that are attempting to attitudes are likely to be needed. In a different
encourage women to return to the labor force context, for example, surveys in urban China
after childbirth. Coherence across fertility and found that only a minority of families who
labor policies has not always been a strong are eligible to have second children wish to do
point, however; for example, evidence from so (about 18 percent to 24 percent of eligible
countries of the Organisation for Economic families in Beijing and less than 40 percent of
Co-operation and Development (OECD) indi- eligible families in Jiangsu) (Jones 2009).
cates that high child allowances reduce incen-
tives for women to return to work (IMF 2012).
Economies including Japan; Korea;
Enhancing female labor force
Singapore; Taiwan, China; and Vietnam have
participation
introduced different combinations of mea- Paid parental leave has been spreading
sures to help stimulate fertility. Measures used across the region, although for developing
in the region have included direct payments East Asian and Pacific countries, it is mostly
to families for second, third, and fourth chil- restricted to the formal sector and thus
dren (the “baby bonus,” for example, paid excludes the majority of families. A study
for six years after birth in Singapore since of 13 East Asian and Pacific countries finds
2000, and the “mother hero” lifetime ben- maternity leave in all of them, most often
efits in Mongolia for mothers having four or employer financed (World Bank 2012). For
more children) and for any childbirth (as in example, maternity leave in Japan is for up
Vietnam, where a lump sum of two months’ to one year and paid at 50 percent of salary,
pay is provided within the social insurance which is around the OECD average. Over
system); priority access to public housing; tax the years, take-up among eligible women
L A B O R MA R K E T P O L I C I E S TO A D D R E S S AG I N G I N E A S T A S I A A N D PAC I F I C 111
has increased significantly, from just under Van Doorne-Huiskes 2007). Positive impacts
50 percent in 1996 to 84 percent by 2012 on female employment have been found also
(MHLW 2013). In developing East Asian in middle-income countries such as urban
and Pacific countries, however, maternity Brazil and Colombia (under its Hogares
leave is typically only for those within for- Comunitarios program). In East Asia and
mal social security schemes (a minority in Pacific, however, evidence from countries at
nearly all cases) and tends to be shorter: 98 different income levels (including Fiji, Japan,
days in China, three months in Thailand at Korea, and Malaysia) shows sustained drops
50 percent of salary, and up to six months in female labor force participation after
in Vietnam.3 Leave for fathers is limited but giving birth, pointing to the challenges of
also increasing across the region, although it balancing child-rearing responsibilities and
remains modest, where available, in devel- work (World Bank 2012).
oping East Asian and Pacific countries (for In response to the challenges of reentry
example, five to seven days in Vietnam), and into the workforce after childbirth, several
take-up tends to be low. In Japan, fewer than East Asian and Pacific countries have subsi-
2 percent of men took their parental leave in dized child care in different ways. Overall,
2012, similar to the 2 percent who avail them- East Asian and Pacific countries that have
selves of such leave in France and Germany promoted child care tend to use subsidies
but much lower than the 70 percent take- and incentives rather than mandates on
up rate in Sweden (Steinberg and Nakane employers, an alternative approach taken in
2012).4 For OECD countries, parental leave countries such as Brazil, Chile, and India,
after childbirth is positively, although mod- which mandate crèche provision by employ-
estly, associated with female employment and ers over a certain size. Within East Asia and
hours worked, although the positive impact Pacific, Korea has introduced subsidized day
reverses when paid leave is longer than two care for children up to four years of age, with
years (Thévenon and Solaz 2013). Evaluation variable subsidies according to family income.
of the effects of paid parental leave in East Singapore provides a subsidy for children
Asian and Pacific countries is limited, but evi- under seven years old in child care centers
dence from Japan (Waldfogel, Higuchi, and when their mothers are working, with higher
Abe 1999) indicates a much stronger impact subsidies for infants and low-income families.
on likelihood of women’s return to work after Malaysia, in contrast, works through the tax
childbirth than in the United Kingdom or the system, making child care facilities an allow-
United States (Suruga and Cho 2003), and able expense for employers and making pay-
employers providing parental leave report ments by workers tax exempt. However, only
more women return to work. 166 centers—of which only 26 were in pri-
Across the world, public subsidies for vate organizations—were established nation-
child care affect female labor force par- ally in response to this policy. In Thailand,
ticipation positively, although the degree the trade unions are involved in provision of
of impact depends on country context and crèches and receive only very modest support
alternative options for informal care (see from municipalities (for example, through
chapter 2 of this report on family-provided milk subsidies). Vietnam has taken a differ-
child care). A study of 18 OECD countries ent approach for informal sector women by
(Thévenon 2013) found that the level of lengthening the school day, especially where
public spending on child care services for attendance is usually a half day. China has
parents of children under three years old taken a somewhat similar approach in its
significantly affects employment of women efforts to promote mass coverage of early
ages 25–54. Similarly, analysis from Europe child development programs.
finds that public subsidization of child care In East Asian and Pacific countries where
is associated with higher rates of women’s public subsidies for child care are significant,
employment and fertility (Den Dulk and they have had some success in stimulating
112 LIVE LONG AND PROSPER
women’s employment. However, disentan- FIGURE 4.1 The Kurumin mark is given to firms in
gling the effect of child care alone is not Japan that pursue child-friendly policies
always easy when such initiatives are bundled
with other measures, as discussed below.
Available assessments come from richer East
Asian and Pacific economies and show posi-
tive effects of child care provision in Japan
(higher than cash allowances) and Korea
(Kim 2007; Steinberg and Nakane 2012).
When child care provision and subsidies were
sharply reduced in Mongolia (World Bank
2013b; World Bank and ADB 2005) and
China (Chi and Li 2008; Du and Dong 2010;
Maurer-Fazio et al. 2011) in the context of
state enterprise restructuring, female labor
force participation fell in urban areas, which
researchers attribute in part to the rising costs
of child care.
The emerging practice in richer East
Asian and Pacific economies is to bundle
packages of measures that aim to stimulate Korea’s Saeromaji Plan 2010 provided a
female labor force participation, especially similar package of benefits, including lower
after childbirth. In 2007, Singapore intro- taxes for families.
duced a package of training programs, flex- Although OECD countries historically
ible working arrangements, targeted active saw a trade-off between childbearing and
labor programs and work fairs, and other increased labor force participation of women,
measures under the auspices of its Tripartite recent evidence indicates this is mostly no
Workgroup on Enhancing Employment longer the case. As shown in figure 4.2, the
Choice for women. These programs com- relationship between women’s work and
plemented a range of measures to stimulate fertility behavior has reversed over the past
fertility. Taiwan, China, bundled paren- few decades in the OECD, and having more
tal leave benefits, child care subsidies, and children is associated with higher rates of
early childhood development programs employment among women ages 25–54.
under its White Book of Population Policy However, while both Korea and Japan have
(Ministry of Interior 2008). Japan has also seen rising female labor force participation,
had a series of initiatives (including the they have not managed to combine this with
Angel Plans in the 1990s and the Support increased fertility in the way that a number
Plan of 2004) to help combine work and of other OECD countries have. The factors
child rearing, including child care subsidies, behind such divergent patterns are discussed
child allowances, and employer incentives to in box 4.1.
become more family-friendly (Ma 2010).5 Notably, some East Asian and Pacific
Japan’s target is to increase the percentage countries actually provide disincentives
of women resuming their careers after child- to women’s work. A previously common
birth from 38 percent in 2010 to 55 percent (although declining) feature of labor laws in
by 2020. Firms judged to be taking proac- East Asia and Pacific and other parts of the
tive measures to promote family-friendly world has been gender-specific restrictions
workplaces are also given a Kurumin mark on the sectors, types of work, and hours of
(figure 4.1), and women’s reemployment is work for women. Within the International
supported through various public programs, Labour Organization, such policies are no
such as the Mothers Hello Work program. longer considered to promote the interests of
L A B O R MA R K E T P O L I C I E S TO A D D R E S S AG I N G I N E A S T A S I A A N D PAC I F I C 113
FIGURE 4.2 The relationship between women’s work and fertility has reversed sharply in recent decades
in most member countries of the OECD
Employment rates of women ages 25–54 and total fertility rates, 1980 and 2009, selected countries
a. 1980 b. 2009
3.5 2.2
IRL NZL ISL
MEX IRL USA
2.0 AUS FRA NOR
3.0 SWE
KOR GBR FINDNK
1.8 BEL
NLD
Total fertility rate
1.0 1.0
30 40 50 60 70 80 50 60 70 80 90
Employment rate of women ages 25–54 Employment rate of women ages 25–54
BOX 4.1 Fertility and public policy in East Asia and Pacific and the OECD
Much of East Asia and Pacific already has below- and care commitments. The drop happened at a time
replacement levels of fertility. A number of national of increased female participation in higher educa-
governments in the region (Japan, the Republic of tion and the labor market. Increasing participation
Korea, and Singapore) are especially concerned by in higher education, coupled with evolving societal
the ultra-low total fertility rate (TFR) at or below attitudes toward gender roles, influenced career aspi-
1.3 children per woman, which is comparable to rations as well as partnership preferences, such as
some of the lowest TFR countries, such as Germany, lower rates of marriage. These choices resulted in not
Hungary, and Portugal. In comparison, the Organ- only postponement of childbearing but also increased
isation for Economic Co-operation and Development propensity for childlessness. However, the relation-
(OECD) average TFR is 1.74 children per woman. ship with employment has reversed in Europe since
Although China is not considered to be one of the the 2000s (OECD 2011), as seen in many OECD
“lowest low,” its official TFR is well below replace- countries, such as Greece, Hungary, Italy, Japan,
ment level at around 1.6 children per woman (and Korea, Poland, and Spain, where low female employ-
could be considerably lower, according to estimates ment rates coexist with low fertility.
from a number of well-respected researchers). Since 2002, a small rebound has occurred in
These fertility levels raise two important questions fertility in about half the OECD countries (OECD
for East Asian and Pacific governments: What are 2011), though whether this trend will sustain over
the main drivers of fertility decline and recovery, and time and can be replicated in low-fertility countries
what public support is effective in enabling people of East Asia and Pacific is unclear. One part of the
to realize their desired fertility plans? In Europe, fer- recovery is explained by a fertility catch-up, because
tility rates declining from their 1970 levels generally births that occurred at young ages in the past are
were a result of women’s inability to match career no longer taking place but recover as births take
BOX 4.1 Fertility and public policy in East Asia and Pacific and the OECD (continued)
place at older ages (Bongaarts and Sobotka 2012). countries, suggesting that fertility recovery may not be
Another major influence behind the recovery is easy in the region. These features include less flexible
improved public support. An OECD (2011) report labor markets, lack of public support, and traditional
observes that although direct financial support had a attitudes toward gender roles related to care commit-
temporary and limited impact on birth rates, invest- ments. High financial costs of child care and housing
ments in formal child care services and flexible labor in Japan and Korea have also constrained people in
markets that also involve a greater degree of per- realizing their fertility plans. Lack of stable employ-
manent employment have been most effective. For ment for both parents seems to have a particularly
instance, Nordic countries have successfully shifted adverse effect on the decision to have a second child
perceptions about combining parenthood and career (Rendall et al. 2014) and is likely to be another fac-
aspirations through universally accessible parental tor that challenges fertility recovery in East Asia and
support, including paid and job-protected parental Pacific. In China and Vietnam, explicit fertility poli-
leave, flexible work arrangements, and subsidized cies have played a more specific role. Although these
child care and schooling. As a result, Nordic coun- policies clearly affected past fertility trends, to what
tries have the highest fertility rates in the OECD. extent further relaxation of policies would result in
Several structural features distinguish low-fertility a sustained increase in fertility is less clear, given the
East Asian and Pacific countries from the OECD range of other factors at play in fertility decisions.
women, and they seem even less appropriate men in child rearing. This issue is a chal-
in countries experiencing rapid aging. More lenge in much of the world. Countries such
specifically, incentives may exist in tax sys- as Singapore, through public information
tems that encourage women to stay out of efforts, and Japan, through its 2005 Gender
the workforce. Although promoted as part Equality Law, are aiming to address social
of family-friendly policies, such measures attitudes toward the roles of men and women
increasingly run counter to other policies to in family life. Japan is also targeting greater
encourage return to work after childbirth. efforts at men with its Ikumen project, which
The Japanese tax and pension systems are one recognizes individual men and groups of men
example: dependent exemptions for heads of who take an active role in parenting.6
household (usually men) are provided as long
as the dependent spouse’s income is under a
fairly modest threshold, a measure considered
Extending productive working
to discourage full-time work among women.
lives
Similarly, the pension system has incentives As societies in East Asia and Pacific age,
for nonworking women. These policies are urbanize, and formalize, the challenge of
under review. More generally, analysis points ensuring long and productive working lives
to the positive association between neutral becomes ever more pressing. As East Asia
tax treatment of second earners, child care and Pacific continues its rapid urbanization,
subsidies, and paid maternity leave (Steinberg coverage of pension systems grows, and ris-
and Nakane 2012). ing incomes provide new opportunities for
Although policy measures to stimulate people to retire before they are no longer
female employment are welcome, realiz- physically capable of work, thereby creating a
ing their full potential will require longer- significant risk that average lengths of work-
term shifts in social attitudes in many East ing lives will fall. Although the potential trend
Asian and Pacific countries toward more of falling retirement ages for rural and infor-
shared responsibility between women and mal sector workers is in one sense a boon
L A B O R MA R K E T P O L I C I E S TO A D D R E S S AG I N G I N E A S T A S I A A N D PAC I F I C 115
and an indicator of rising prosperity and wel- FIGURE 4.3 Japanese workers show a desire to work in old age
fare, the experience of richer East Asian and well above OECD counterparts
Pacific economies already points to a need for Percentage of Japanese workers desiring to work until indicated
strengthening policies to increase the work- age, 2007
ing lives of existing urban and formal sector
Until when do you want to work?
workers and to ensure that any future fall
in rural and informal working lives is not 2.4%
The idea that increased participation of older work- Economic Co-operation and Development countries
ers will negatively affect employment for younger on the relationship between employment of people
people is known as the lump of labor fallacy. This ages 55–64 and those ages 20–24. For all but one
fear of displacement is grounded in the assumption country (Japan, although the negative impact is mod-
of a zero-sum labor market in which every job occu- est), increases in employment of older workers had
pied by an older worker is one less potential job for at worst a neutral association and in most cases a
a younger person. positive association with employment rates of young
In fact, higher participation of older workers tends workers. This relationship is confirmed in regression
to stimulate the economy and in doing so creates analysis. Although more analysis is needed on the
higher aggregate demand for labor, thereby benefiting question in developing East Asian and Pacific coun-
young people as well. In addition, the types of jobs tries, studies to date for China confirms no impact
among older and younger workers differ, so they can- on youth employment from higher employment rates
not be considered substitutes in general. Figure B4.2.1 of older workers (Munnell and Wu 2013; Zhang and
shows results of analysis for 12 Organisation for Zhao 2012).
0.20
y = 0.257x + 0.023
0.15 R2 = 0.034
Change in employment,
0.10
ages 20–24
0.05
–0.05
–0.10
0 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16
Change in employment, ages 55–64
Employed, ages 20–24 Linear (employed, ages 20–24)
Germany, Hungary, Italy, and Romania), concludes that wage subsidies granted solely
while the rest provide wage subsidies for on the basis of age may not be very effective
employers recruiting older people as part compared to a subsidy that is targeted more
of vulnerable groups such as the long-term narrowly at the older long-term unemployed
unemployed, people threatened by redundan- or at low-income older workers.
cies, and people with disabilities (for example, Wage subsidy schemes that are predomi-
Austria, Denmark, Latvia, Lithuania, Malta, nantly targeted at older workers raise a num-
Portugal, and Sweden) (Feifs et al. 2013). The ber of issues in terms of cost-effectiveness.
only available evidence (from 2006) on impact First, because older workers are a very diverse
L A B O R MA R K E T P O L I C I E S TO A D D R E S S AG I N G I N E A S T A S I A A N D PAC I F I C 117
group, wage subsidies targeted on age alone approach seems defensible in a period when
risk being quite blunt instruments and may governments are trying to change the societal
result in small net employment effects at con- mindset on the productive value of older
siderable cost. They may involve substantial people, but close evaluation will be needed
deadweight loss (that is, a large proportion of to assess the cost-effectiveness of different
subsidized workers would have been employed approaches as the stock of older workers
even without the subsidy) and substitution or grows. Incentives may need to become more
displacement effects (that is, subsidized jobs targeted over time to those at highest risk of
for eligible workers lower job opportunities early exit from the labor force or at highest
for workers ineligible for the subsidy). Korea’s risk of failure to reenter the labor force.
2011 reforms of financial incentives for hir- Richer East Asian and Pacific economies
ing older workers were driven by evaluations have increasingly sought to reinforce finan-
that found substantial deadweight losses from cial incentives to extend working lives with
previous efforts. Second, subsidies for older employment services targeted at older work-
workers as a group may lead to stigmatiza- ers. As with financial incentives, a number of
tion and reinforce negative attitudes among countries in East Asia and Pacific and beyond
employers regarding hiring and retention of have instituted employment service programs
older workers (OECD 2006). specifically targeted to older workers. In
For East Asia and Pacific, the approach to the European Union, Germany provides an
date has more commonly been to adopt poli- example with its Perspective 50 Plus program
cies targeted at all people above a certain age of active labor policies and campaigns to pro-
and to supplement them with targeted sup- mote attitudinal change among employers
plements for particular groups among older and society (OECD 2012). In East Asia and
people (for example, Singapore’s Employment Pacific, Japan has a set of measures from mid-
Assistance Payments, which are targeted at dle age onward, as do Korea and Singapore
workers otherwise unable to be reemployed, (see box 4.3). Evaluations of impact and cost-
in addition to general programs to sup- effectiveness of these programs are generally
port employment of older workers). This lacking. Evidence from Japan suggests small
BOX 4.3 Employment services targeted to older workers in Japan and the Republic of Korea
Japan has strengthened its public employment ser- society” philosophy and are complemented by the Act
vices by establishing special offices that support older on Stabilization of Employment of Older Persons and
job seekers and incumbent workers, as well as sup- Silver Human Resources Centers.
porting the reemployment of older workers through Korea has established a talent bank for older job-
the following programs: Career Exchange Plazas to seekers that has been connected to job centers since
help middle-aged and older white-collar job seekers 2012. Moreover, some private agencies have been
work in close cooperation with Talent Banks; Older designated as employment service centers to support
Persons Vocational Experience Utilization Centers to middle-aged professional workers with a certain level
provide free placement services and dispatching ser- of qualifications in their job search. Korea has also
vices for persons ages 60 and older; Industry Employ- established special offices that support older job seek-
ment Security Centers to provide matching services to ers as well as older incumbent workers through the
middle-aged or older workers; Elderly Employment Aged Service Centers and Manpower Banks for Older
Support Centers/Corners to establish an office in Workers.
every prefecture to provide assistance to middle-aged
and older workers to develop career life plans. These Sources: OECD 2006 and Ministry of Health, Labor, and Welfare presentations to
services all fall under the umbrella of the “ageless World Bank staff for Japan; OECD 2012 for Korea.
118 LIVE LONG AND PROSPER
improvements in placement of workers ages wage rates (IMF 2012). Various mechanisms
60–64 after the introduction of age-targeted are used in OECD and richer East Asian and
employment services, but how much of the Pacific economies to take advantage of this
effect was attributable to the programs can- phenomenon, including the following:
not be determined (OECD 2006).
Supplementing the preceding measures are • Providing incentives in tax and social
programs to incentivize workplace adjust- contribution systems. These include spe-
ments to accommodate older workers and cific income tax credits for older work-
enhance their productivity. Given the onset ers (for example, Australia, Denmark,
of physical limitations for older workers, the Netherlands, Sweden, and the United
many countries have sought to induce adjust- Kingdom, with the tax credit increasing
ments in the physical work environment in with the age of the worker in some cases);
firms through grants, loans, and other tools. reductions in or exemptions from social
This approach is also happening increasingly contributions for workers over a certain
on a purely private basis, as firms realize age (for example, France, the Netherlands,
the financial payoffs of often modest invest- and Spain); and tax credits that are gen-
ments. The most famous example in private eral but for which older people are dis-
enterprise is Germany’s BMW: in response to proportionately represented in the target
concerns about productivity on an assembly groups (for example, in Korea, the earned
line staffed by older workers, BMW spent income tax credit introduced in 2008 for
€40,000 on 70 small changes to workplace households with no children and some
design and equipment to accommodate the self-employed individuals).
particular needs of older workers. Within • Reducing disincentives to work in social
a year, the line’s productivity had improved security systems. Social security systems
by 7 percent (matching that of others staffed may give disincentives to work at older
by younger workers), and within two years, ages in a variety of ways. The first is treat-
absenteeism caused by poor health had fallen ment of work income while in receipt of
to 2 percent below the firm average—payoffs a pension. Historically, countries such as
that much more than offset the initial invest- the former Soviet Union republics pro-
ment (Loch et al. 2010). The positive outcome hibited receipt of both pension and work
is mirrored in a larger study of productivity income, although these systems were
and age profiles in response to different inter- reformed in most cases during transition
ventions (Bussolo, Koettl, and Sinnott 2015; as the state ceased to be the dominant pro-
Göbel and Zwick 2010). In East Asian and vider of both pensions and wages. Within
Pacific red economies, governments provide East Asia and Pacific, Japan provides an
grants and subsidized loans to firms to under- interesting case: it retains a social security
take such adjustments. For example, Korea earnings test whereby pension benefits
since 2011 has provided subsidized loans beyond the basic pension are withdrawn
for installation of age-friendly facilities, and in proportion to labor income at vary-
Singapore provides grants that may be used ing rates, according to age and income
by firms for upgrading. Increasingly, firms in level. For those ages 60–64, benefits are
Japan are also making such investments on a first withdrawn at a 50 percent rate and
purely market basis. then suspended entirely for incomes
A second broad set of policies com- over ¥460,000 (in 2012). For those ages
monly used to stimulate work at older ages 65 and older, a 50 percent withdrawal
is within tax and social security systems, and applies over the same threshold. Analysis,
these methods are spreading in East Asia and not surprisingly, finds negative effects on
Pacific. Overall, older workers globally have labor supply among workers ages 60–64
been found to be more sensitive than younger in Japan (Shimizutami, Fujii, and Oshio
workers to financial incentives and after-tax 2012).7 Such an approach seems to run
L A B O R MA R K E T P O L I C I E S TO A D D R E S S AG I N G I N E A S T A S I A A N D PAC I F I C 119
counter to other policies designed to stim- FIGURE 4.4 Tax force to retire is strongly correlated with labor
ulate employment at older ages in Japan, force participation of older men in OECD countries
and many OECD and transition countries
have moved away from it. 0.7
United R2 = 0.81 France Belgium
0.6 Kingdom
For developing East Asian and Pacific Netherlands Italy
0.5 Spain
countries, the bigger work disincentive in Canada Germany
Proportion
pension systems lies in early retirement pro- 0.4 United
visions, which in a number of countries States Sweden
0.3
(for example, China, the Philippines, and
0.2 Japan
Vietnam) provide strong financial incentives
to retire earlier than official pension ages, as 0.1
discussed in chapter 5 of this report. In con- 0
trast, Japan and Korea (along with a num- 0 1 2 3 4 5 6 7 8 9 10
ber of OECD countries, including Canada, Tax force to retire
the United Kingdom, and the United States) Ages 55–65 not in labor force Log of ages 55–65 not in labor force
structure accrual and other benefit rules to
incentivize retirement after the official age Source: Gruber, Milligan, and Wise 2010.
Note: OECD = Organisation for Economic Co-operation and Development.
(Feifs et al. 2013). A further factor at play is
retirement age, which is discussed below.
East Asian and Pacific countries at all higher productivity of older workers, this
income levels can take lessons from OECD system will make older workers less sought
experience (including in East Asia and Pacific) after in the labor market and will constrain
on the importance of minimizing “tax force efforts to extend their working lives. Evidence
to retire” (that is, the implicit tax on work from OECD countries finds that seniority
over a certain age). As shown in figure 4.4, wage systems have a strongly negative impact
in OECD countries with the lowest tax force on hiring of workers ages 50–64, a find-
to retire, fewer than 40 percent of men ages ing confirmed in firm-level data (Daniel and
55–65 are out of the labor force, whereas that Heywood 2007; OECD 2011). On the basis
share rises to as high as two-thirds in coun- of an early 2000 survey of employers in 10
tries with the highest tax force to retire (for East Asian and Pacific and OECD economies,
example, Belgium). Japan does particularly East Asia and Pacific showed higher-than-
well on this measure, with very low tax force average reliance on seniority wage setting,
to retire and only about 20 percent of men particularly in Japan and Taiwan, China
ages 55–65 out of the labor force. The nega- (Lowe et al. 2002). Although the differences
tive impacts of high tax force to retire can are not dramatic compared with other OECD
already be seen in the formal sectors of some countries, employers in China and Korea,
countries in developing East Asia and Pacific for example, were notable for thinking that
(for example, China and Vietnam), so avoid- seniority should be more important in wage
ing high tax force to retire will be important setting in the future.
for all developing countries in the region as Richer East Asian and Pacific economies
they reform their pension and tax systems. are trying to reduce the influence of seniority
In contrast to various incentives in East in wage setting and increase that of perfor-
Asia and Pacific to extend working lives, some mance in a variety of ways that offer lessons
policies and practices hinder demand for older for other countries in the region. Box 4.4 dis-
workers, such as seniority wage systems. As in cusses recent reforms of seniority wage systems
many countries, East Asian and Pacific coun- in Korea and Singapore and the contrasting
tries have inherited seniority-based wage sys- example of Japan. In the first two cases, sensi-
tems that automatically increase pay with age, bly designed reforms are being adopted slowly
independent of performance. Absent much by firms but show potential. Developing
120 LIVE LONG AND PROSPER
BOX 4.4 Reforming seniority wage systems in the Republic of Korea, Singapore, and Japan
The Republic of Korea, Singapore, and Japan offer of a broad set of incentives to firms to hire or retain
examples of efforts to reduce the role of seniority in older workers.
wage determination. Japan has taken a different route to the challenge
Korea introduced a so-called wage-peak sys- of seniority wage systems. Japan has to date seen most
tem from 2005, which incentivized firms to adopt experimentation in reduction of seniority wages hap-
more performance-based wage systems and con- pen at firm level, though it is also making efforts in
duct productivity-based collective bargaining. The the public sector to flatten wage-to-age schedules.
government has disseminated industry-specific Since the 1990s, the importance of age-based pay
performance-based compensation models and within base-pay setting has seen a gradual reduc-
actively promoted the new system through media, tion, especially among managers, where emphasis
awards, and the like. This strategy has been sup- on role-related pay (which within it has an element
ported since 2006 by allowances for firms adopt- of performance-related pay) has increased (Conrad
ing such measures, which were extended in 2011 to 2009). By 2006, about one-third of firms reported
employers who extended their firm-level retirement relying entirely on role-related pay for their managers,
age and adopted measures such as flexible work and a further third relied on it for setting 50 percent or
and reemployment after retirement (OECD 2012). more of wages, while around three-quarters of firms
Although adoption has been relatively slow (only reported role-related pay not being the primary driver
about 12 percent of firms had adopted the wage-peak of nonmanagerial pay (Shakai Keizai Seisansei Honbu
system by 2011), the pace of demographic transi- 2006). Despite this gradual dilution of seniority pay,
tion and market forces are strengthening incentives the most common approach in response to amend-
to adopt it. Several large employers have already ments to the Job Security Law has been for firms to
adopted the system: Samsung, for example, reduced mandate retirement of workers, typically at age 60,
worker pay by 10 percent annually after age 55 to and then to rehire them at usually half or less of prere-
increase the relative attraction of older workers. The tirement wages and on a term contract as a nonregular
policy is not uncontroversial, however. The system employee. About 85 percent of companies had estab-
has been a subject of dispute between government, lished such rehiring systems (Phang 2011). However,
organized labor, and employer representatives during initiatives exist at the firm level to reduce the wage-
2015. age link. One large Japanese firm interviewed for this
Singapore has for many years aimed to increase report has changed the entire wage-age curve within
the flexibility of wage setting and its links to produc- the firm, creating budget space for rehiring and ade-
tivity growth and economic conditions, starting with quately compensating older workers by flattening the
its “flexi-wage policy” from 1986. More recently, it age-wage trajectory for workers in their 30s and 40s.
has introduced financial grants to firms that move
away from seniority-based wage setting. This is part Source: Sen 2014.
countries are also trying to increase the role availability and retirement. Evidence from
of performance-based pay in many cases, for around the world, including developed and
example in China, where teachers have had developing East Asia and Pacific, finds evi-
an increasing element of performance pay in dence of significant impact of retirement age
recent years. on withdrawal from work, effects which in
A second significant policy factor in the East Asia and Pacific are exacerbated by gen-
formal sectors of East Asian and Pacific coun- erous early retirement provisions in countries
tries is retirement age policy. This is discussed such as China, the Philippines, and Vietnam.
in chapter 5 of this report in terms of fiscal Richer East Asian and Pacific economies
sustainability and pension adequacy, but it have pursued a distinctive approach to delay-
also has a clear labor market dimension, as ing withdrawal from the labor force, which
found in chapter 2 with respect to pension involves a nuanced interaction of mandates
L A B O R MA R K E T P O L I C I E S TO A D D R E S S AG I N G I N E A S T A S I A A N D PAC I F I C 121
and incentives. Compared with many other and Singapore. This is also true in Indonesia,
parts of the world, richer red economies in which has not had a mandated retirement age
East Asia and Pacific have taken a creative (Giles, Wang, and Cai 2012). The second ele-
approach to the issue of retirement age and ment of the approach is that firms are given
stimulating longer working lives. The first flexibility in how they extend working lives
element of this approach is the role of the over specified minimum ages, and in gen-
state and firms. Although the state mandates eral they seem to prefer programs of con-
minimum retirement ages (and imposes man- tinued employment rather than raising the
datory retirement in the public sector), firms firm-level mandatory retirement age itself.
have a greater role in setting firm-specific The approach, most elaborated to date in
retirement in countries such as Japan, Korea, Japan, is outlined in box 4.5. It also provides
BOX 4.5 East Asia and Pacific’s distinctive approach to retirement age and continued work
Richer East Asian and Pacific economies have exhib- retirement age. The adoption of reemployment as a
ited a distinctive but still evolving approach to the measure is inspired by labor practices in Japan and
issues of mandatory retirement ages and promotion is believed to be more useful in raising the effec-
of longer working lives. Japan is a key example of tive retirement age while enhancing mutual ben-
the mixture of mandates with flexibility. The Act efits because it allows for more flexibility in work
on Stabilization of Employment of Elderly Persons arrangements as well as wage and benefit adjust-
2012 obliges employers to do one of the following: ments (Phang 2011).
(a) raise the mandatory retirement age from 60 to The Republic of Korea is an interesting example
65; (b) introduce a continued employment system of mandate and incentives shifting over time with
to ensure that all employees who desire to work respect to retirement age. In Korea, retirement ages
remain in the workforce until age 65; or (c) abol- historically have not been mandated by law, but
ish the mandatory retirement age at firm level. The employment laws have allowed firms to set man-
aim is to improve the opportunities for older work- datory ages. Government employees face manda-
ers to continue their career until at least pensionable tory retirement. One feature of pension eligibility
age (65). The contract for this work differs from traditionally has been an imperfect correspondence
that covering work prior to retirement age. In addi- between mandatory retirement age and the age at
tion, when a system for continuous employment is which pension-eligible retirees may start receiving
introduced in an enterprise, the standard for treat- pension benefits. The national pension scheme does
ment of older persons can be decided through a for- not begin paying benefits until age 60, yet a signifi-
mal labor-management agreement. As of mid-2012, cant share of employees until recently faced manda-
82.5 percent of companies had adopted the contin- tory retirement at age 55 (Cho and Kim 2005). The
ued employment system, and few have taken mea- government over time has introduced various mea-
sures to raise or abolish the mandatory retirement sures to encourage firms to voluntarily extend their
age system (MHLW 2013). mandatory retirement age. The Act on Prohibition
The statutory retirement age in Singapore was of Age Discrimination in Employment and Aged
age 60 at the introduction of the Retirement Age Employment Promotion was adjusted in 2008, so
Act in 1993. In 1999, it was raised to 62, with the that the retirement age an employer sets should be
long-term objective of increasing the retirement 60 or over. Firms with 300 or more workers must
age progressively to 67. In an effort to encour- report their firm-specific mandatory retirement
age employment of more older workers, however, age each year to the Ministry of Employment and
in 2007 the Tripartite Committee recommended Labor, and the ministry may request employers to
reemployment instead of urging a higher statutory raise the minimum if it is below 60. However, this
BOX 4.5 East Asia and Pacific’s distinctive approach to retirement age and continued
work (continued)
monitoring-based approach was considered not to be to employ older people above the industry-specific
achieving sufficiently rapid results, and in 2013 the employment quota for elderly workers and to sub-
act was amended to require firms to set minimum mit a status report each year. The employment
retirement ages of at least 60. The amendments will quota is set by the standard employment rate of
come into effect from 2016 for large firms and from older workers, which varies between 4 percent and
2017 for firms with fewer than 300 employees and 42 percent by industry and occupation. The Ministry
central and local governments. of Employment and Labor may request firms to for-
Korea also uses age-specific quotas to induce mulate and submit plans for implementation of the
longer working lives in firms having 300 perma- standard employment rate when they fall below that
nent employees or more. Those firms are required standard.
an interesting case of industrial relations in reduced average working hours and a subsi-
parts of East Asia and Pacific, whereby trades dized program that allows older workers to
unions, employer associations, firm man- combine paid work and voluntary activities
agement, and government aim to develop at social enterprises and nongovernmental
nuanced policies that reduce the social con- organizations (though take-up has remained
flict often engendered by economywide low) (OECD 2006, 2012).
retirement age increases in other parts of the
world.
Complementing retirement age policy
Promoting immigration into
have been initiatives to promote flexible
aged East Asian and Pacific
work and a “glide” into retirement, although
economies
these have had limited impact in East Asia The potential for migration to mitigate the
and Pacific to date. As shown in chapter 2 of labor force impacts of aging across East Asia
this report, when workers in East Asia and and Pacific is significant and would pres-
Pacific (at all country income levels) continue ent a “win-win” for both sending and host
work into older age, they often tend not to countries.8 As in other parts of the world,
reduce hours significantly until well into their immigration policy is a sensitive issue and
70s. This finding contrasts with patterns one that requires careful brokering of con-
in some OECD countries, where flexible sensus in countries potentially involved to
work arrangements have been promoted to ensure its sustainability and good outcomes
encourage greater participation of both older for both migrants and their hosts. Rapid
workers and women. The United Kingdom, aging, however, may provide increased
for example, has achieved some success in impetus for migration policies that can be
extending working lives through this route, beneficial for both sending and host coun-
but in Germany the impacts have been more tries. Opinion poll evidence from Japan and
limited because of negative perceptions about Korea presented in chapter 3 of this report
part-time work (EC DG-ESAEO 2012). suggests that populations in rapidly aging
Japan and Korea have made efforts to reduce East Asian and Pacific societies increasingly
the average working week, which remains realize the desirability of raising historically
very high by OECD standards, though low immigration rates. China has benefited
progress is challenging. In response, Korea from this economic phenomenon in a differ-
in 2011 introduced subsidies for firms that ent way with the mass internal migration of
L A B O R MA R K E T P O L I C I E S TO A D D R E S S AG I N G I N E A S T A S I A A N D PAC I F I C 123
young rural workers to higher-productivity TABLE 4.1 There is enormous variation in the importance of
work in urban areas. The necessity of immi- emigration in East Asian and Pacific countries
gration for East Asian and Pacific econo- Number and percent of total population emigrating in selected East Asian
mies with older populations is reflected in and Pacific countries as of 2010
the words of the late Singaporean Prime Emigration rate
Minister Lee Kuan Yew (2012), accord- Origin Emigrants (thousands) (% of population)
ing to whom “in the future we will have to Indonesia 2,502 1.1
depend on immigrants to make up our num- Philippines 4,275 4.6
bers, for without them Singapore will face Samoa 120 67.3
the prospect of a shrinking workforce and a Tonga 47 45.4
stagnant economy.” Vietnam 2,226 2.5
In both host and sending countries in East Source: World Bank 2013a.
Asia and Pacific, enormous variation exists
in the scale of migration. In terms of emigra-
tion, several countries are among the most
important points of origin across the globe. TABLE 4.2 The share of immigrants ranges from very low to very high
For example, more than 4.3 million Filipinos by global standards in East Asia and Pacific
are working abroad, followed by 2.5 mil- Immigration to selected East Asian and Pacific economies
lion Indonesians and 2.2 million Vietnamese. Immigrants in Immigration rate in 2010
Yet because of the large populations of those Destination 2010 (thousands) (% of population born abroad)
countries, emigration accounts for only Australia 5,522 25.7
between 1.1 percent and 4.6 percent of their Hong Kong SAR, China 2,742 38.8
populations (table 4.1). In contrast, emigra- Japan 2,176 1.7
tion is critical for several Pacific Island coun- Korea, Rep. 535 1.1
tries, such Samoa and Tonga, where emigrants Malaysia 2,358 8.4
make up 67 percent and 45 percent of total New Zealand 962 22.4
population, respectively. Although a large por- Singapore 1,967 40.7
tion of these emigrants stay within the region, Thailand 1,157 1.7
a significant number go to other parts of the Source: World Bank 2013a.
world, such as the Persian Gulf, Australia and
New Zealand, and North America.
In terms of immigration, several economies
in East Asia and Pacific display high levels, labor market liberalization in the context of
whereas others are among the lowest receiving wider economic integration among members
countries globally (table 4.2). In Australia and of the Association of Southeast Asian Nations.
New Zealand, historically important migrant- The fertility and household formation
receiving economies (mostly permanent immi- behavior of migrants is a critically important
grants), around a quarter of the population and rapidly growing research area.9 Evidence
was born abroad. Singapore and Hong Kong shows that the speed with which migrants
SAR, China, (often temporary immigrants) adapt to the social norms of the destina-
have even higher levels at about 40 percent. tion society depends on personal charac-
Another important case is Malaysia with teristics of the migrants as well as cultural
significant temporary low-skilled migration proximity between origin and destination
from countries such as Bangladesh, Indonesia, countries, such as attitudes on gender pref-
and the Philippines. In contrast, Japan and erences, sexuality, and marriage. In aging
Korea have minimal immigration, at less than societies, the critical issue is the number of
2 percent of their total populations. Overall, children migrants will have and its deter-
large portions of migrants in East Asia and minants. Widespread evidence exists that
Pacific originate from within the region, a migrants who arrive before adulthood have
situation that is likely to be reinforced with fertility behavior that is indistinguishable
124 LIVE LONG AND PROSPER
from natives. For those who arrive at a later to fill the vacancy in question. Admission
age, Kahn (1988) shows that highly edu- is granted on both permanent and tempo-
cated migrants have almost identical fertil- rary basis, but the conditions to “import”
ity behavior as natives of the destination. a low-skilled worker are normally more
Migrants from neighboring countries or with stringent and residency privileges are less
similar cultural backgrounds, regardless of permanent than those for skilled labor. The
age at migration or education, have almost overall intake is again controlled by quo-
identical fertility patterns. All of the evidence tas or other instruments. Korea, Malaysia,
shows that assimilation is completed, at the Singapore, Thailand, and Taiwan, China,
latest, with the children born in the destina- are examples of such systems in East Asia
tion, if not with the migrants themselves. and Pacific.
The impact of migrants on the overall fer- The most common policy tool across the
tility level of the destination seems to be, at region and the world in managing immigra-
best, a medium-term phenomenon. tion flows is quotas, and less common but
Although the impact of migration on rapidly expanding tools are levies or taxes.
migrants’ own fertility behavior is a criti- Many key destination countries implement
cal issue, migration may also affect the quantitative limits on the number of visas
fertility behavior of natives when large issued for migrants. These quotas can be
numbers of migrants are involved. Among specific to education levels, occupations,
the main determinants of fertility behavior or origin countries. They can be modi-
are the opportunity costs of having chil- fied frequently or be set at a certain level
dren for women and how migration affects for years, regardless of changing needs and
these costs. More specifically, availability conditions. Under the levy approach, either
of migrant workers for household services the employees or the migrants are required
as nannies and maids is known to affect to pay a specific per worker tax to the gov-
the labor force participation and fertility ernment. This tax can be imposed annually
decisions of women in destination coun- or collected at the issuance of the employ-
tries (Cortés and Tessada 2011). This issue ment visa or upon entry. Again, the levies
is especially relevant for Singapore and can vary depending on the education level,
Taiwan, China, which host significant num- the occupation, or the sector of employ-
bers of female migrant workers. ment of the migrant. In recent years, coun-
Host economies in East Asia and Pacific tries such as Singapore have implemented
exhibit a mix of supply- and demand- levies that are employee specific and depend
driven migration systems. A supply-driven on the share of foreigners in the firm’s
immigration system selects workers based total labor force. To discourage employ-
on their human capital credentials (educa- ment of migrants, firms with higher shares
tion and experience), and migrants do not pay higher per worker levies. The quota
need prearranged employment contracts. and levy distinction closely resembles the
Admission is normally granted on a tempo- quotas-versus-tariffs distinction in interna-
rary basis but with high likelihood of even- tional trade. Over time, quotas have been
tual permanent residency and citizenship. removed or replaced by tariffs as the latter
Quotas set by central authorities limit the are observed to be more flexible and eco-
total annual intake. Australia is an example nomically efficient, since they lead to col-
of such a system in East Asia and Pacific. lection of taxes rather than to the surplus
In a demand-driven system, migrants nor- being captured by the quota permit holder.
mally enter the host country with prear- The same economic benefits would be real-
ranged employment contracts. Employers ized in the migration context, but political
are generally required to obtain permission and public opposition to migration usu-
from the government beforehand and to ally prevents full removal of quantitative
show that no native worker can or is willing restrictions.
L A B O R MA R K E T P O L I C I E S TO A D D R E S S AG I N G I N E A S T A S I A A N D PAC I F I C 125
Origin countries can also play an protection of workers while they are abroad,
important role in enhancing the benefits of helping them maximize the long-term eco-
migration for all parties involved by imple- nomic and social benefits for both the
menting effective and efficient policies. migrants and their families. Other countries
Among the roles that these governments can implement similar programs with varying
play are improving informational efficien- degrees of success (see Ahsan et al. 2014).
cies in the system, enforcing contracts, and Encouraging migration that maximizes
protecting migrants’ rights. Informational benefits for both host and sending coun-
problems are prevalent in migration. Among tries requires policy action in a range of
the key predeparture policies is supporting areas, and practice from within East Asia
and overseeing the recruitment process. The and Pacific already offers lessons. A recent
first step is providing accurate information World Bank report (Ahsan et al. 2014) has
to potential migrant workers. Creating infor- analyzed challenges, market failures in need
mation centers and having nongovernmental of public policy focus, and policy implica-
organizations involved in the process can tions and options for stimulating migration
generate significant benefits. Regulation of in East Asia and Pacific in greater detail (see
recruitment agencies through effective polic- also Adams and Ahsan 2014). The summary
ing and enforcement and increased competi- findings of Ahsan et al. (2014) are therefore
tion are critical. reiterated in table 4.3 and provide a broad
The main concern of origin govern- framework for action. Details on specific
ments should be to protect the rights of recommendations and good practices are
their citizens abroad while finding appro- provided in that report and in Özden and
priate employment opportunities for them. Testaverde (2014).
These dual aims require close coordina-
tion of different government agencies, such
as departments of education, labor, and
Policy effects on the labor force
foreign affairs. The Philippines Overseas Even if sensible policy responses raise par-
Employment Administration (POEA) has ticipation rates, the question remains whether
been relatively successful in designing poli- they will be sufficient to compensate for the
cies that generate long-term benefits. In declining share of working-age population.
2010, around 1.5 million workers registered The following section examines this issue
through POEA before leaving to work in using a simple simulation framework looking
more than 150 different countries. One of at red and orange economies, for which the
the key features of POEA is the licensing shrinking working-age population is a press-
of more than 1,000 recruitment agencies, ing challenge (see box 4.6 for assumptions
which enables the government to prevent underlying these simulations).
abusive practices by both these agencies and For both the upper- and middle-income
potential employers. POEA is able to pro- countries shown, the biggest impact on
tect the rights of the workers from the first labor force participation is from increasing
stages of the migration process until they female labor force supply; for China and
return home. The sister organization, the Thailand, increasing the participation of
Overseas Workers Welfare Administration, older people also has significant potential
collects contributions from the migrants to (figure 4.5). For red economies, significant
fund predeparture training programs and life scope exists to increase female labor force
insurance schemes as well as to operate wel- supply through a gradual convergence
fare centers abroad. Finally, the Department toward male participation rates, whereas
of Foreign Affairs coordinates the activi- the potential for labor force growth from
ties of migration-related agencies through a older workers is more modest because of
network of labor offices attached to consul- their already high participation rates. For
ates. Such coordinated action leads to better China, in particular, increasing both elderly
126 LIVE LONG AND PROSPER
TABLE 4.3 Policies to maximize the benefits of migration in East Asia and Pacific for sending and receiving countries
Objective: Address possible adverse labor market impacts of migrant labor on native unskilled workers
Social returns of migration are Higher profits made by firms are not Ensure that foreign-worker levy is charged to and borne by
less than private return to firms. shared with the economy although employing firms, not by workers.
migrant workers have a social presence. Ensure that minimum wage or minimum-wage-indexed
Lower any possible social costs by adding thresholds and benefits for migrant workers are used
a cushion against lower wages and to prevent excessive lowering of wages or a general
increasing unemployment. deterioration in working conditions.
Although firms profit, the economy can Provide job training, retraining, and education for native
suffer from the use of less productive, workers to upgrade their skills. This will probably happen
labor-intensive technologies. normally, but government assistance can accelerate
Migrant workers use public facilities the process. Ensure labor market flexibility to facilitate
and social benefits, crowding out mobility of native workers from less productive to more
local citizens. productive occupations. Wage subsidies and hiring
bonuses can be provided to that end.
Improve climate for investment and technology innovation
and retention of professional and skilled labor.
Provide job training and technical upgrading for workers to
match demand in more skilled occupations.
Provide normal fiscal incentives to firms for innovation and
investment in new technology.
Enable migrants to join and contribute to social security;
make benefits and public facilities accessible to
documented workers who also pay taxes.
Social and private returns are Poorly regulated recruitment agencies Ensure workers receive orientation about rights and
lower because of regulatory and rent-seeking lead to defrauding obligations, certified by recruitment firms and employers.
and information failures. of migrant workers and excessive Ensure minimum capital experience thresholds for
importation of foreign workers. Workers recruitment firms.
are not adequately informed about their Standardize, monitor, and enforce contracts between
rights and working conditions. employers, recruitment agencies, and workers.
Take advance deposits from recruitment agencies and
employing firms to use as penalties against abuse of
migration laws through defrauding workers or violating
contracts.
Source: Ahsan et al. 2014.
and female labor force participation has working-age population in richer East Asian
significant potential to stave off what is and Pacific economies. Figure 4.6 shows the
otherwise a significant decline in labor impact of permanent and temporary migra-
force size. Thailand also has potential from tion scenarios on the size of the labor force
both channels, although less so for elderly in destination economies relative to the base-
participation. line case and the enhanced female and older-
Permanent or temporary migration also worker participation scenarios. The impact
has potential to mitigate the decline in of increased female participation continues
L A B O R MA R K E T P O L I C I E S TO A D D R E S S AG I N G I N E A S T A S I A A N D PAC I F I C 127
The scenarios span the period 2010–40 (2010–50 to the maximum rate in 2050 is determined dif-
for migration) and combine population projections ferently for the red and orange groups. For the
from the 2012 revision of the United Nations World red group, the participation rate for older men
Population Prospects (UN 2013) with labor force and women is assumed to gradually converge to
participation data from the International Labour the current Japanese rates for workers ages 60
Organization’s LABORSTA database (ILO 2011). and older. For the second group of aging coun-
The assumptions for the base case and alternative tries, the elderly participation rate will move to
scenarios are as follows: current Indonesian rates for workers ages 60 and
older. No scenarios were estimated for the group
• Base-case scenario. The International Labour with the lowest share of older workers: the Lao
Organization publishes male and female partici- People’s Democratic Republic, Myanmar, Papua
pation rates for each five-year age group from age New Guinea, the Philippines, and Timor-Leste.
15 and average participation for individuals ages • Migration scenarios— permanent and tempo-
65 and older. Taking these age and sex-specific rary. With permanent migration, newly arriving
participation rates as given at 2010 levels and immigrants are assumed to be ages 25–35 and to
combining them with official population projec- constitute 10 percent of the labor force of destina-
tions implies that the projected changes in the tion countries in that age group. The assumption
national labor forces are entirely determined by is based on the current levels observed in many
the population age structure projections for 2010 countries that accept mostly permanent migrants,
to 2040. including the European Union and the United
• Convergence for women. The convergence sce- States. Migrants are assumed only to arrive in
nario assumes that female labor force participa- that age group and never to leave. In addition,
tion grows at an annual rate of 1/40th of the 2010 migrants adopt the fertility, mortality, and labor
difference between men and women. Although force participation patterns of natives. The tem-
women’s rates are assumed to converge toward porary migration scenario assumes that new
male levels in 2050, the gender gap in partici- migrants constitute 20 percent of the labor force
pation will not be eliminated by the end of the between ages 25 and 35 but that migrants stay
period in 2040. for only 10 years (with a 100 percent labor force
• Increasing participation by older workers. These participation rate) and then return to home coun-
projections assume increased participation for tries. After 10 years, a new cohort of the same
populations ages 60 and older, where convergence ages 25–35 arrives.
to be the most significant in the long run for However, within 20 years, the permanent
the overall size of the labor force in destina- migration scenario dominates as migrants
tion countries. At the same time, both types of settle in host countries. The cumulative nature
migration increase the labor force size in host of permanent migration has a larger impact in
countries considerably in the short run and the long term, almost twice that from tempo-
are superior in this time frame to effects from rary migration. Over time, as the migrants age
higher female and elderly labor force partici- and retire, the total labor force again declines,
pation rates. Temporary migration leads to the even with migration.
largest immediate increase in the labor force The impact of migration on the origin coun-
in destination countries because the arriv- tries is quite different because those countries
ing cohort is bigger and temporary migrants have younger and expanding populations.
have higher labor force participation rates. The initial decline is very similar under both
128 LIVE LONG AND PROSPER
FIGURE 4.5 The potential channels for mitigating labor force decline vary across aging countries in
East and Southeast Asia
Percentage changes in labor force size relative to 2010 baseline from increases in female and elderly labor force
participation
a. China b. Japan
106 105
104
100
% change relative to 2010
94
85
92
90 80
2010 2016 2022 2028 2034 2040 2010 2016 2022 2028 2034 2040
125
110
120
105 115
110
100
105
100
95
95
90 90
2010 2016 2022 2028 2034 2040 2010 2016 2022 2028 2034 2040
e. Thailand
110
108
106
% change relative to 2010
104
102
100
98
96
94
92
90
2010 2016 2022 2028 2034 2040
Baseline Female Elderly
scenarios for the first two decades, but the FIGUR E 4.6 Greater participation by women, older people, and
gap widens over time (figure 4.7). Permanent migrants can mitigate the projected decline in the total labor
migration leads to a decline of about 5 percent force in richer East and Southeast Asian economies
in the labor force by 2050 while the impact Projected changes in labor force size of destination economies with changes
in female, elderly, and migrant labor force participation
of temporary migration is about half that.
Given the growth in origin-country labor
130
forces during that period and the immediate
benefits of remittances to the origin countries,
to OECD averages and contrasts with coun- gradual increase in targeted training schemes
tries such as the United States, which exhibits for older workers. Co-financing or providing
more equal access across the work life cycle government subsidies for training programs is
(figure 4.8). Apart from this transition chal- a popular measure used by Australia and sev-
lenge, a broader challenge exists of building eral countries in Europe to incentivize training
lifelong learning systems for continuous skill and skill building for older workers. Sweden
upgrading of workers across the life cycle. provides an innovative example of a “learning
Globally, education, skill, and labor systems account” to which both employers and
are aiming to evolve from a traditional pat- employees can make tax-exempt deposits for
tern of study, work, and retirement to one training. Such measures are driven by the need
where skill upgrading and work lives blend to improve employability of older workers,
more seamlessly and the work-retirement since schemes to improve training rates across
divide becomes more blurred. the adult population have the lowest take-up
Australia, high-income countries of East among older workers (EC DG-ESAEO 2007,
Asia and Pacific, and Europe have adopted based on the European Commission’s Ageing
a variety of measures that subsidize training and Employment: Identification of Good
and lifelong learning activities for older work- Practice study; OECD 2006).
ers to improve their productivity and employ- Richer East Asian and Pacific economies
ability. Globally, countries face challenges have also instituted training programs for
in uptake of training by older workers rela- older workers, both on a targeted basis and
tive to others, and examples from East Asia within the context of broader programs. The
and Pacific suggest that the challenge may be government of Singapore in Budget 2014
even more acute.10 This situation has led to a announced a S$500 million top-up to its
FIGURE 4.8 Japan and the Republic of Korea show low rates of on-the-job training for middle-aged and
older workers relative to countries such as the United States
Index of learning at work by age and country, 2012, selected OECD countries
2.6
2.4
2.2
2.0
Index
1.8
1.6
1.4
1.2
1.0
24 or less 25–34 35–44 45–54 55 plus
Age group
Australia Canada Finland France Germany
Japan Poland Korea, Rep. United States Russian Federation
Lifelong Learning Endowment Fund, taking Smyth 2012). In Korea, less than 10 percent
the total size to S$4.6 billion with a lifelong of workers ages 50–64 participated in train-
commitment to continuing education and ing of any form (and fewer older women than
training to enhance employability, though men), less than half the rate for other work-
to what extent older workers are accessing ers, and well below age-specific rates of about
the scheme remains unclear (Government of 40 percent in the United States and Nordic
Singapore 2014). In addition, ADVANTAGE countries (Phang 2011).12 Part of the issue is
and Flexi-Works Policy grant financial assis- simple cost-benefit, with a justified percep-
tance to employers who train older workers. tion that the shorter the remaining working
To better reach out to low-wage workers, life the more returns to training fall, a factor
the government introduced a three-year that has led James Heckman, for example, to
Workforce Training Scheme that subsidizes suggest that wage subsidies may be a more
90 percent to 95 percent of absentee payroll efficient intervention as workers get older.
and course fee outlay for employers who However, another strain of literature finds
send their older low-wage workers for train- significant discrimination against older work-
ing. Low-wage workers are encouraged to ers by employers in providing training and a
go for training to upgrade their skills and feedback loop whereby some older workers
improve their employability, upgrade to bet- internalize the effect and seek less training
ter jobs, and earn more. Since the Workforce (OECD 2006).
Training Scheme started in July 2010, more “Universities of the third age”—in which
than 90,000 Singaporeans have benefited, fol- skilled older people provide education and
lowing the guidance of an old Chinese phrase: skill training for other older people—have
㋀ܦ冾븽ᘕܦ冾 (“live till old, learn till old”) proliferated across East Asia and Pacific.
(Government of Singapore 2013; Phang 2011; Initiated in France, this type of organization
UNFPA and HelpAge International 2012). has spread to transition and OECD coun-
In Korea, job training subsidies are paid to tries and has been taken up in East Asian
employers who provide training opportuni- and Pacific countries such as Australia,
ties to workers under the Occupational Skill China, Malaysia, and Singapore. These
Development Scheme (Phang 2011). More mixed programs are oriented toward work-
specifically, the scheme focuses on providing related skills as well as general intellectual
training for older people who want to start enrichment and social interaction, with
their own businesses (given the prevalence of an aim to provide stimulation and slow
self-employment among older workers) or cognitive decline (UNFPA and HelpAge
return to rural areas. International 2012).
As in other parts of the world, participa- Although limited, the global evidence on
tion in training among older workers tends effects of targeted training schemes on the
to be relatively low, though how much this productivity of older workers is mixed, and
is driven by weak demand, employer atti- more assessments of cost-effectiveness are
tudes, or inflexible supply is difficult to dis- needed. Assessments from Germany find that
entangle.11 This low participation rate can training of older workers did not appear to
be seen throughout East Asia and Pacific. significantly affect productivity among those
In Singapore, about 37 percent of workers already working (Göbel and Zwick 2010,
ages 20–29 undertook training with support 2013). Similarly, in Sweden training for
of the Skills Development Fund, whereas unemployed workers ages 50–64 resulted in
the share was only 19 percent in the 50–64 regular job placement for only one in eight
age group (Phang 2011). In China, the trainees (OECD 2006). In contrast, placement
effect was found to be more pronounced, rates of older trainees in the Czech Republic
with a 1 percent increase in age resulting in and Spain were encouraging, with 70 and
a 27 percent to 35 percent decline in prob- 50 percent of trainees, respectively, in work
ability of undertaking training (Mishra and 6–12 months after training (OECD 2006).
132 LIVE LONG AND PROSPER
In East Asia and Pacific, evaluations of initia- under its Advantage! Scheme; the WorkPro
tives in red economies would be beneficial. scheme of funding to help employers rede-
The most notable feature of policies to sign jobs and work environments to adapt
extend working lives in East Asia and Pacific to an older workforce; and the Special
in recent years has been the adoption of mul- Employment Credit and Workfare Income
tipronged strategies to promote productive Supplement, which aims to support employ-
aging. Richer East Asian and Pacific econo- ers and employees to raise the employability
mies have heightened their policy focus on of older, low-wage Singaporeans. Japan has
employment of older workers through com- had several packages of measures over the
prehensive programs combining different years under its Basic Policy of Employment
types of interventions. Box 4.7 discusses the Measures for the Elderly and the 2012
example of Korea, which mixes incentives revised Act on Stabilization of Employment
for enhancing productivity of older work- of Older Persons, including wage subsidies,
ers, subsidies to employers to retain older financial support to employers with age-
workers, and efforts to activate unemployed friendly workplaces, dedicated training cen-
older workers or those approaching retire- ters (Silver Human Resources Centers), and
ment or dismissal. Singapore has adopted employment services (sections for the elderly
a similar range of programs, including in Hello Work job centers). These are part of
BOX 4.7 Second Basic Employment Promotion Plan (2012–16) for the aged in
the Republic of Korea
1. Strengthening support for intergenerational job 3. Strengthening support for retirement preparation
sharing and skills development
• Skilled middle-aged and old retirees will serve • Large firms will be required to provide a cer-
as mentors and instructors for young employ- tain period of preretirement training to their
ees at 300 small and medium enterprises, and middle-aged and older workers who are forced
1,600 retired experts will be used as “on-site to leave their jobs involuntarily.
professors.” • Long-serving middle-aged and older workers
• The job-sharing program, which provides a will be granted the right to ask for unpaid
subsidy of W7.2 million per year to firms if educational leave of one year or less.
they hire youth to fill vacancies that arise as 4. Expanding support for early reemployment and
middle-aged and older employees work shorter jobs for the aged
hours or receive training, will be promoted. • Opportunities to participate in the Employment
2. Expanding support for older workers to stay Success Package that links counseling,
longer in their principal jobs vocational training, and job placement and to
• Subsidies for the wage-peak system will be receive on-the-job training at small and
increased. medium enterprises will be expanded.
• Subsidies for extending the retirement age • Relevant training and funding will be provided
or reemploying older workers will be differ- to older people who want to start up their
entiated based on the length of the extended own business or return to farming or rural
employment period to encourage firms to keep areas.
older workers for a longer period. 5. Promoting social contribution and talent-sharing
• T he coverage of workplace su r veys of activities
retirement-age schemes will be extended from 6. Improving systems and infrastructure to cope
those with 300 workers or more to those with with an aged society
100 workers or more, to prepare for discussions • The retirement pension system will be further
on retirement-age reform. promoted to ensure old-age income security.
L A B O R MA R K E T P O L I C I E S TO A D D R E S S AG I N G I N E A S T A S I A A N D PAC I F I C 133
a broader “ageless society” theme of public programs is needed to inform policy deci-
policy in Japan. Such active labor programs sions. For many programs aiming to stim-
along with reforms of retirement age and ulate enhanced labor force participation,
continuous employment effectively combine very little is known about their employ-
incentives with mandates and give firms ment impacts and cost-effectiveness. As
considerable latitude for how they comply. discussed earlier in the chapter, one spe-
The lead from public policy is increasingly cific issue is the extent to which policies
being taken by large employers, with com- promoting higher labor force participa-
panies such as JFE Steel introducing the tion are more cost-effective when targeted
Senior Expert Program to retain and maxi- purely by demographic characteristics
mize the experience of workers over age 60 such as age or gender or by more general
(Shimbun 2015). economic characteristics such as unem-
ployment or income level.
• An appropriate balance must be struck
Conclusion between mandates (on employers or work-
ers) and incentives. Emerging experience
Fortunately, the middle-income countries
from East Asia and Pacific suggests that
of East Asia and Pacific can benefit from
neither approach in isolation is likely to
the lessons of various labor market policies
achieve the desired results. A related con-
implemented in richer East Asian and Pacific
sideration is the role of public policy and
economies In responding to the challenges of
market forces.
shrinking labor forces and outputs, the expe-
rience from East Asian and Pacific countries
suggests a set of overarching issues in devel- Notes
oping labor market policy responses:
1. Jones (2009) provides an excellent summary
• As supported by the recommendations pre- of pronatalist policies in richer East Asian and
sented in this chapter, an effective labor Pacific economies. See also Tsuya, Choe, and
market policy response to aging in East Feng (2009).
Asia and Pacific requires policy action 2. This finding is consistent with other analysis
across the life cycle, ranging from efforts for Japan (Tsuya 2005, 2008; Tsuya, Choe,
to stimulate fertility in countries with older and Feng 2009).
3. U.S. Social Security Administration (2015)
populations to initiatives aimed at extend-
provides a comprehensive summary of
ing the productive lives of older workers.
maternity and other insurance benefits in Asia.
• Attention is needed to policy coherence 4. See Hegewisch and Gornick (2011) on slow
and trade-offs in trying to mitigate the take-up of paternity leave in OECD countries.
adverse labor market impacts of aging. For Japan, see Basic Survey of Gender Equality
Although many interesting and potentially in Employment Management, MHLW (2013).
useful initiatives are available to augment 5. Employer initiatives to promote family-
the labor forces in aging East Asian and friendly workplaces are based on the 2003
Pacific countries, ensuring that labor mar- Act on Advancement of Measures to Support
ket, social security, tax, and other policies Raising the Next Generation of Children
all push in the same direction remains a and the Basic Act for Measures to Cope with
a Society with a Declining Birthrate, the
challenge in some cases. Thinking about
former amended and strengthened in 2014.
policy coherence raises potential trade-offs,
6. Presentation of Japanese MHLW to World
for example between the economic impera- Bank team, November 2013.
tives of sustaining labor forces and socio- 7. The earnings test on pensions in Japan was
cultural norms around the role of women repealed in 1985 for workers ages 65–69 but
in society. reinstated in 2002.
• A better understanding of the cost- 8. This section draws from a background paper
effectiveness of labor market policies and by Özden and Testaverde (2014) and the
134 LIVE LONG AND PROSPER
World Bank regional migration report for Community Society of Elderly People” [in
East Asia and Pacific (Ahsan et al. 2014). Japanese]. Government of Japan, Tokyo. http://
9. See Adsera and Ferrer (2014) for an excellent www8.cao.go.jp/kourei/ishiki/h20 / sougou
review. /zentai/.
10. Available evidence suggests that older Chi, Wei, and Bo Li. 2008. “Glass Ceiling or
workers are much less likely to participate Sticky Floor? Examining the Gender Earnings
in job training programs relative to younger Differential across the Earnings Distributions
workers. An OECD (2006) report suggests in Urban China.” Journal of Comparative
on-the-job training participation rates of Economics 36 (2): 243–63.
workers ages 50–64 to be under 10 percent Cho, Joonmo, and Sunwoong Kim. 2005. “On
in Korea and over 40 percent in Demark, Using Mandatory Retirement to Reduce
Norway, and the United States. In terms Workforce in Korea.” International Economic
of lifelong learning opportunities, whereas Journal 19 (2): 283–303.
just over 10 percent of Koreans ages 50–64 Conrad, Harald. 2009. “From Seniority to
participated in some form of lifelong learning Performance Principle: The Evolution of Pay
activity during 2000, the corresponding Practices in Japanese Firms since the 1990s.”
proportion was about 36 percent in the Social Science Japan Journal 13 (1): 115–35.
United States and about 46 percent in Sweden Cortés, Patricia, and José Tessada. 2011. “Low-
(Phang 2011). Skilled Immigration and the Labor Supply of
11. See Mayhew, Elliott, and Rijkers (2008) for a Highly Skilled Women.” American Economic
useful discussion of training and employment Journal: Applied Economics 3 (3): 88–123.
of older workers. Daniel, Kirsten, and John Heywood. 2007. “The
12. Japanese MHLW survey findings suggest Determinants of Hiring Older Workers:
higher training among older men (36 percent U.K. Evidence.” Labour Economics 14 (1):
in 2005), though the rate is less than half that 35–51.
among older women. Den Dulk, Laura, and Anneke Van Doorne-
Huiskes. 2007. “Social Policy in Europe: Its
Impact on Families and Work.” In Women,
References Men, Work, and Family in Europe, edited by
Adams, Richard H., Jr., and Ahmad Ahsan, eds. Rosemary Crompton, Suzan Lewis, and Clare
2014. Managing International Migration for Lyonette, 51–70. Basingstoke, U.K.: Palgrave
Development in East Asia. Washington, DC: Macmillan.
World Bank. Du, Fenglian, and Xiao-yuan Dong. 2010.
Adsera, Alicia, and Ana Ferrer. 2014. “Immigrants “Women’s Labor Force Participation and
and Demography: Marriage, Divorce, and Childcare Choices in Urban China during
Fertility.” IZA Discussion Paper 7982, Institute Economic Transition.” In Gender Equality and
for the Study of Labor, Bonn, Germany. China’s Economic Transformation: Informal
Ahsan, Ahmad, Manolo Abella, Andrew Beath, Employment and Care Provision, edited by
Yukon Huang, Manjula Luthria, and Trang Xiao-yuan Dong and Sarah Cook, 173–91.
Van Nguyen. 2014. International Migration Beijing: Economic Science Press.
and Development in East Asia and the Pacific. EC DG-ESAEO (European Commission,
Washington, DC: World Bank. Directorate-General for Employment, Social
Bongaarts, John, and Tomáš Sobotka. 2012. Affairs, and Equal Opportunities). 2007.
“A Demographic Explanation for the Recent Employment in Europe 2007. Brussels:
Rise in European Fertility.” Population European Commission.
and Development Review 38 (1): 83–120. ———. 2012. European Employment Observatory
doi:10.1111/j.1728-4457.2012.00473.x. Review: Employment Policies to Promote
Bussolo, Maurizio, Johannes Koettl, and Emily Active Ageing 2012. Luxembourg: Publications
Sinnott. 2015. Golden Aging: Prospects for Office of the European Union.
Healthy, Active, and Prosperous Aging in Feifs, Toms, Tina Weber, Oscar Vargas, Karel Fric,
Europe and Central Asia. Washington, DC: Maurizio Curtarelli, and Irene Mandl. 2013.
World Bank. Role of Governments and Social Partners in
Cabinet Office, Government of Japan. 2008. Keeping Older Workers in the Labour Market.
“Survey Concerning the Participation in the Dublin: European Foundation for the
L A B O R MA R K E T P O L I C I E S TO A D D R E S S AG I N G I N E A S T A S I A A N D PAC I F I C 135
Improvement of Living and Working Kahn, Joan R. 1988. “Immigrant Selectivity and
Conditions. Fertility Adaptation in the United States.”
Giles, John, Dewen Wang, and Wei Cai. 2012. Social Forces 67 (1): 108–28.
“The Labor Supply and Retirement Behavior Kim, Jeong-hee. 2007. “Child Care Support
of China’s Older Workers and Elderly in Programs for Double Income Families.”
Comparative Perspective.” In Aging in Asia: In Generational Change and Social Policy
Findings from New and Emerging Data Challenges: Australia and South Korea, edited
Initiatives, edited by James P. Smith and by Ruth Phillips, 140–163. Sydney, Australia:
Malay Majmundar, 116–47. Washington, DC: Sydney University Press.
National Academies Press. Loch, Christoph, Fabian J. Sting, Nikolaus
Göbel, Christian, and Thomas Zwick. 2010. B a u e r, a n d H e l m u t M a u e r m a n n .
“Which Personnel Measures Are Effective 2010. “How BMW Is Defusing the
in Increasing Productivity of Old Workers?” D e m o g r a p h i c Ti m e B o m b . ” H a r v a r d
ZEW Discussion Paper 10-069, Zentrum für Business Review 88 (3): 99–102. https://
Europäische Wirtschaftsforschung, Mannheim, h b r. o r g / 2 0 1 0 / 0 3 / t h e - g l o b e - h o w - b m w
Germany. -is-defusing-the-demographic-time-bomb.
———. 2013. “Are Personnel Measures Effective Lowe, Kevin B., John Milliman, Helen De Cieri,
in Increasing Productivity of Old Workers?” and Peter Dowling. 2002. “International
Labour Economics 22: 80–93. Compensation Practices: A Ten-Country
Government of Singapore. 2013. A Sustainable Comparative Analysis.” Human Resource
Population for a Dynamic Singapore: Population Management 41 (1): 45–66.
White Paper. Singapore: National Population Ma, Li. 2010. “Social Policy and Childbearing
and Talent Division, Prime Minister’s Office. Behavior in Japan since the 1960s: An
———. 2014. “Budget 2014: Opportunities for Individual Level Perspective.” Paper presented
the Future—Assurance for Our Seniors.” in Poster Session 1 at the European Population
http://www.singaporebudget.gov.sg/data Conference 2010, Vienna, September 1–4.
/budget_2014/download/FY2014_Budget_in Maurer-Fazio, Margaret, Rachel Connelly, Lan
_Brief.pdf. Chen, and Lixin Tang. 2011. “Childcare,
Gruber, Jonathan, Kevin Milligan, and David A. Eldercare, and Labor Force Participation of
Wise. 2010. “Introduction and Summary.” Married Women in Urban China, 1982–2000.”
In Social Security Programs and Retirement Journal of Human Resources 46 (2): 261–94.
around the World: The Relationship to Youth Mayhew, Ken, Matt Elliott, and Bob Rijkers.
Employment, edited by Jonathan Gruber and 2008. “Upskilling Older Workers.” Ageing
David A. Wise, 1–45. Chicago: University of Horizons 8: 13–21.
Chicago Press. MHLW (Ministry of Health, Labor and Welfare,
Hegewisch, Ariane, and Janet C. Gornick. 2011. Japan). 2013. 2012 Basic Survey of Gender
“The Impact of Work-Family Policies on Equality in Employment Management [in
Women’s Employment: A Review of Research Japanese]. Tokyo: MHLW. http://www.mhlw
in OECD Countries.” Community, Work, and .go.jp/toukei/itiran/eiyaku.html.
Family 14 (2): 119–38. Ministry of Interior, Taiwan, China. 2008.
ILO (International Labour Organization). 2011. Population Policy White Book. Taipei:
LABORSTA (database), 6th ed., ILO, Geneva. Ministry of Interior.
IMF (International Monetary Fund, Fiscal Mishra, Vinod, and Russell Smyth. 2012. “High
Affairs Department). 2012. Fiscal Policy and Performance Work Practices and Workplace
Employment in Advanced and Emerging Training in China: Evidence from Matched
Economies. Washington, DC: IMF. Employee-Employer Data.” Monash
Jones, Gavin W. 2009. “Recent Fertility Trends, Economics Working Paper 30-12, Monash
Policy Responses, and Fertility Prospects in University, Department of Economics, Victoria,
Low Fertility Countries of East and Southeast Australia.
Asia.” Report for the United Nations Expert Munnell, Alicia, and April Wu. 2013. “Will
Group Meeting on Recent and Future Trends in Delayed Retirement by the Baby Boomers
Fertility, Population Division, United Nations Lead to Higher Unemployment among
Department of Social and Economic Affairs, Younger Workers?” CRR Working Paper
New York, December 2–4. 2012-22 (expanded version), Center for
136 LIVE LONG AND PROSPER
141
142 LIVE LONG AND PROSPER
pensions, and (d) noncontributory pensions, global pension landscape. Most countries in
or “social pensions.” As shown in table 5.1, East Asia and Pacific have adopted this
all four elements are present in some but not approach, but how different countries have
all East Asian and Pacific economies. done so varies widely.
East Asian and Pacific economies can be
broadly divided into three groups in terms of
Government-mandated contributory
the introduction of nationally mandated pen-
pension schemes
sion systems relative to demographic charac-
The ultimate objectives of public provision of teristics: the globally average adopters, the
pensions are twofold: to reduce poverty in old late adopters, and the global outliers that
age and to smooth consumption over the life have yet to adopt a pension system. Figure 5.1
cycle. Globally, a small number of countries shows the year that different economies intro-
(for example, New Zealand and South Africa) duced pension mandates that applied to most
focus on only the first objective using budget- workers in the formal sector rather than to
financed transfers to the elderly. However, just civil servants, compared with the share of
most countries have mandated retirement the population over age 60 at the time.
savings in an attempt to address both The globally average adopters—China
objectives. Today, government mandates to (1951), Japan (1941), Malaysia (1951),
save for retirement and usually to insure Mongolia (1950s), the Philippines (1954),
against death and disability dominate the and Singapore (1953)—introduced national
pension schemes relatively early. The timing FIGURE 5.1 East Asian and Pacific economies can be divided into
of the introduction is significant because in globally average adopters, late adopters, and global outliers
defined benefit schemes, generally the longer Timing of introduction of national mandated pension schemes
the pension scheme has been operating, the
higher the ratio of pensioners to contributors 16
and the greater the pressure on finances. The Hong Kong
14
Degree of
Country Benefit target redistribution Member risk Risk to government
China High Large Low returns on defined contribution Large unfunded liability
component
Hong Kong SAR, Low Small Low returns on defined contribution; None
China longevity risk
Indonesia Low Small Low returns on defined contribution; None
longevity risk
Japan Moderate Moderate Sponsor risk or default Large unfunded liability
Korea, Rep. Moderate Large Sponsor risk or default Moderate unfunded liability
Malaysia Low Small Low returns on defined contribution; None
longevity risk
Mongolia Moderate Large Inflation risk; sponsor risk or default Large unfunded liability
Philippines High Large Inflation risk; sponsor risk or default Large unfunded liability
Singapore Low None Low returns on defined contribution None
component
Thailand Moderate Moderate Inflation risk; sponsor risk or default Moderate unfunded liability
Vietnam High Small Sponsor risk or default Large unfunded liability
other extreme, the Philippines and Vietnam Civil service schemes in East Asia and
provide generous benefits through a DB Pacific tend to be more generous than those
scheme in which government bears all of the for private sector workers, sometimes signifi-
risk and allows a large unfunded liability or cantly so. The relationship between contribu-
financing gap to grow over time.6 In the table, tions and benefits tends to be more imbalanced
sponsor risk refers to the possibility that, as in from an actuarial perspective than national
the case of the recent reforms in Japan, Korea, schemes in the same country. Benefit parame-
and Mongolia, governments may renege on ters are more generous, and public sector pen-
DB promises when schemes become sioners are likely to have greater life
unsustainable. expectancy at retirement. In some cases, such
as in China (prior to changes announced in
Civil service pensions 2015) and in Myanmar, public sector workers
have made no contributions, and pensions are
As in other parts of the world, civil servants
paid directly from the budget. Aside from fis-
in East Asia and Pacific have usually been
cal pressures caused by these maturing and
covered by pension schemes before the rest of
often generous civil service pensions, the
the workforce (see Palacios and Whitehouse
schemes demonstrate the apparent inequity of
2006). In every economy in the region, civil
providing more generous pensions for public
service pensions preceded mandates for pri-
sector workers. Although a case may be made
vate sector workers, often by decades. Japan’s
for the government providing supplementary
civil service pension scheme was introduced
pensions as an employer, it is difficult to justify
in the late 19th century (see Casey 2004). The
different levels of minimum pensions or
most recent example is Hong Kong SAR,
implied rates of return on contributions or
China, which, after inheriting the British
even different eligibility ages.
colonial pension scheme for civil servants,
Administering separate schemes raises
waited until 2001 to mandate retirement sav-
costs to the extent that economies of scale
ings for private sector workers.
cannot be exploited, especially in countries
Civil service pensions in East Asia and
with a small number of covered workers (see
Pacific fall into three distinct groups:
Sluchynskyy 2015). For example, the admin-
• Countries with separate pensions for civil istrative costs of the Philippine civil service
servants include China,7 Indonesia, Korea, pension scheme were roughly the same as
Lao PDR, Malaysia, the Philippines, and those for the social security system, despite
Thailand.8 the fact that the latter had seven times as
• Economies with integrated pensions for many contributors and beneficiaries.
all formal sector workers include Hong Similarly, the cost per participant was six
Kong SAR, China;9 Japan;10 Mongolia; times higher for Thailand’s civil service pen-
Singapore; and Vietnam following its sion scheme compared with the
2014 reforms. Indonesia also plans to national scheme. Sluchynskyy (2015) pro-
move gradually in this direction, although vides the strongest cross-country evidence of
implementation has not yet started. economies of scale based on a sample of 104
• Countries with coverage only for civil public pension funds from 87 countries:
servants include Cambodia, Myanmar, and administrative costs per beneficiary are found
Timor-Leste, which passed its civil service to fall by half when moving from 100,000 to
pension legislation in 2012 but has not fully 500,000 participants and by another 25
implemented it.11 Except for Timor-Leste, percent with 2 million participants.
these separate civil service pension schemes Finally, parallel pension schemes may
are all DB and are fully mature, with retir- inhibit labor mobility between the public and
ing civil servants having earned full pen- private sectors, because pension rights typi-
sions and with relatively high ratios of cally are not portable.13 Civil service schemes
pensioners to active employees.12 are almost always DB (Hong Kong SAR,
146 LIVE LONG AND PROSPER
China, is the exception), and the typical for- rate targets and contribution rates in their
mula rewards the later years of participation mandated schemes, and the majority of
disproportionately (known as backloading). workers and employers pay income taxes.
Given the lack of portability arrangements in • In China and Thailand, the existing occu-
national schemes that cover private sector pational arrangements need restructuring
workers, a public sector worker who moves to encourage and incentivize additional
to the private sector suffers a large loss in pen- voluntary coverage. In China, members
sion wealth. of the “enterprise annuity” plan (which
For all of these reasons, an increasing num- are DC accounts) have been given limited
ber of countries in all regions have started to tax preferences. Membership is around
integrate civil servants into their national pen- 18.5 million workers, consisting mostly of
sion schemes. Recent reforms in China, public enterprise employees—a small frac-
Japan, Vietnam, and eventually Indonesia are tion of the overall formal sector workforce.
examples in East Asia and Pacific. Integration In Thailand, private pension coverage rose
presents an important policy question for the to around 7 percent of the labor force in
rest of the countries that continue to have 2014.16 Thailand has relatively low ben-
separate civil service schemes,14 but integra- efit targets and contribution rates, but few
tion is particularly urgent in the poorer coun- qualify for favorable tax treatment.
tries that do not yet have national mandates • Malaysia and Singapore have attempted
that apply to private sector workers. to increase voluntary savings through
Relatively small numbers of formal sector their existing provident fund schemes.
workers and low administrative capacity in The programs offered by Malaysia17 and
these countries suggest that they should seri- Singapore draw few participants—less
ously consider avoiding parallel systems. than 5 percent of those covered by their
respective provident funds. This low
participation rate is partly the result of
Supplementary private pensions
already high forced savings rates in those
The role of private, voluntary pension countries, which leave little space for addi-
schemes is relatively limited throughout most tional voluntary pensions.
of the region. Participation in these schemes • In the rest of the region, private voluntary
typically is driven by tax preferences, and cov- pensions currently play a negligible role or
erage tends to overlap with workers already are nonexistent. In Cambodia, Indonesia,
participating in mandated schemes (with Lao PDR, Myanmar, and Timor-Leste, the
some exceptions among the self-employed). enabling conditions for private pensions are
When mandated pension schemes have high not yet in place, and few workers would
target replacement rates and when contribu- benefit from tax preferences. The focus
tion rates are already high, little room remains should be on reforming the existing public
for voluntary provision. Another important systems of these countries. Vietnam offers
factor is the share of the population subject to scope for launching or developing a vol-
the income tax, which is extremely limited untary occupational provision targeted at
outside of the richer countries. small and medium enterprises as the replace-
The situation with regard to voluntary ment rates in the public system decline.
pensions in the region can be summarized as
follows:
Noncontributory (social) pensions
• The only countries with significant private Globally, the role of social pensions—cash
pension coverage are Japan and Korea transfers paid to the elderly, financed from the
with around 23 and 30 percent of the labor general budget and not linked to prior
force covered, respectively.15 These coun- contributions—has expanded dramatically in
tries have relatively modest replacement the past two decades (see Palacios and
PENSION SYSTEMS AND AGING: REFORMS FOR WINNING THE RACE 147
FIGURE 5.2 East Asian and Pacific economies currently spend less on pensions than other economies
after taking into account their demographics
Public pension spending in East Asian and Pacific economies and the world
20
y = 0.0037x2 + 0.2572x – 0.3111
R2 = 0.6907
18
16
Public pension spending (% of GDP)
14
12
10
Japan
8 Vietnam
6 Mongolia
4
Timor-Leste China
2 Philippines
Thailand Hong Kong SAR, China
Korea, Rep.
0
0 5 10 15 20 25 30 35 40
Share of population age 60 and over (%)
Sources: World Bank pensions database, http://www.worldbank.org/pensions; United Nations population data for the latest available year.
Several factors contribute to this difference • The low spending levels in Cambodia and
in pension spending: Myanmar can be explained by the fact
that no mandated scheme exists for the
• DC schemes in Hong Kong SAR, China;
private sector or for social pensions.
Indonesia; Malaysia; and Singapore limit
• Another factor may simply be culture,
their pension spending, at least in terms of
which influences the three factors above.
current budgetary outlays. Indonesia and
Much of East Asia and Pacific tradition-
Malaysia spend only on DB civil service
ally has placed a strong reliance on fam-
pension schemes, whereas Hong Kong SAR,
ily and informal support networks such as
China, spends only on social pensions.
wantoks in the Pacific. Policy makers con-
• Scheme maturation plays a role in Korea,
tinue to assume that traditional support
Lao PDR, and Thailand, where mandates
can substitute or at least complement pub-
for DB schemes have been introduced
lic pension support, although chapter 2
recently. Their expenditures will rise
presented evidence questioning the future
toward the fitted line seen in figure 5.2
validity of this assumption.
in the coming years as a greater share of
workers qualifies for full pensions. Even Despite the strong relationship between
after they mature, however, their low aging and public pension spending across
target benefit levels suggest modest spend- countries, policy choices have room to play a
ing by global standards. role. Independent of the policy approaches
PENSION SYSTEMS AND AGING: REFORMS FOR WINNING THE RACE 149
taken, for decades pension spending in Japan earmarked budgetary resources are used to
has been lower than in other countries at the cover part of pension spending by design.
same stage of demographic transition, and According to the 2014 actuarial report,
projections for Korea suggest that, even after under all but the most pessimistic case,
its scheme matures, it will also spend less than reserves are never exhausted through the
other aging countries. projection period that runs to 2055. Among
Mongolia and Timor-Leste are exceptions the other countries, Mongolia has no
to the general pattern of low pension spending reserves, and the scheme is already running
in the region. Timor-Leste introduced univer- deficits. Spending is expected to exceed con-
sal pensions for individuals age 60 that pay tributions in Vietnam by 2021, according to
around 44 percent of average, non-oil income International Labour Organization projec-
per capita, including some disability benefits. tions (although this projection was prior to
In Mongolia, an inherited Soviet-style system 2014 reforms, which have somewhat
had much higher coverage because of the role improved sustainability), and somewhat
of the state as employer. These two countries later in Korea and the Philippines. Without
spend notably more than other countries at reforms, these cash flow deficits will lead to
the same stage of demographic aging. higher general government deficits.19
Although the regional picture is incom- Two caveats apply in the case of China.
plete, available estimates of pension system First, some provinces still have reserves,
sustainability point to substantial fiscal risks whereas others run significant deficits, and
in a number of national pension systems. there is a separate but relatively small national
The results are not strictly comparable fund, the National Social Security Fund
because they are generated with different (NSSF). Second, the figures refer only to the
models and assumptions and use data of DB scheme. However, as noted, evidence
varying degrees of quality. Nevertheless, gen- shows that most of the contributions into DC
eral patterns are evident in table 5.3, which accounts in China are used for current spend-
shows the projected year of cash flow defi- ing and are therefore not accumulating (that
cits as well as the year in which reserves are is, empty accounts). Nationally, subsidies to
exhausted. Cash flow deficits are expected to the pension system have been estimated to cost
emerge in all East Asian and Pacific DB around 3.8 percent of gross domestic product
schemes in coming decades. In Japan, (GDP) annually (Dorfman et al. 2013).
TABLE 5.3 Defined benefit schemes in East and Southeast Asia face significant sustainability challenges
Indicators of long-run financial sustainability of main public defined benefit schemes, selected East Asian and
Pacific economies
Only Timor-Leste spends a large amount on As noted earlier, no mandated scheme for pri-
social pensions: more than 2.5 percent of GDP vate sector workers exists in Cambodia,
(non-oil), suggesting significant sustainability Myanmar, and Timor-Leste, where only civil
issues. By comparison, no other East Asian servants are covered. In the case of Indonesia,
and Pacific countries spent more than 0.4 poor collection and enforcement of the man-
percent of GDP on social pensions. Based on date are likely to explain some of the short-
current parameters, the present value of spend- fall. In Malaysia, low labor force participation
ing on this program in Timor-Leste through rates of women are part of the explanation, as
2040 as a share of today’s GDP is more than well as the exclusion of most migrant work-
100 percent. The same present value figure for ers. The explanation for the low contributory
Thailand, where coverage is high but benefits coverage found in Indonesia, Lao PDR, and
are much lower, is about 20 percent. Thailand is less obvious. Overall, if current
international patterns are a guide, most of the
Coverage region is likely to experience its demographic
In terms of participation in contributory pen- aging process before coverage rates in man-
sion schemes, East Asia and Pacific is at or dated schemes rise to meet the challenge.
somewhat below par globally, with the excep- The pension schemes in the Philippines and
tion of several low-income countries where Mongolia cover a slightly larger share of their
coverage is minimal—a worrying picture in a labor forces than predicted by their income
region where aging is exceptionally rapid. (figure 5.3). In the Philippines, the main social
FIGURE 5.3 Coverage of contributory pension schemes is around par globally for income levels
and demography
Coverage of working-age population in contributory pension schemes and income per capita, most recent year
120
Singapore
80
Japan
Korea, Rep.
60 China Hong Kong
Mongolia SAR, China
40 Vietnam,
Philippines Malaysia
Cambodia, Lao PDR,
20 Myanmar
Indonesia
0
5 6 7 8 9 10 11 12
Log of PPP adjusted GDP per capita
security scheme is known for its proactive FIGURE 5.4 Elderly pension coverage is low in the bottom income
stance on collecting from the self-employed deciles in China, Indonesia, and the Republic of Korea
and its migrant workforce. The higher-than- Pension coverage among the elderly by income decile
predicted coverage rates in Mongolia are
almost certainly due to the legacy of the cen- 100
trally planned economy. As the country contin- 90
FIGURE 5.5 Increasing contributory pension coverage among the schemes. To compare schemes in terms of
poorer segments of populations is difficult individual outcomes or target benefit levels,
Share of households with elderly in bottom 40 percent of distribution the analysis can simulate the results for a
receiving pensions hypothetical worker who contributes during
his or her entire career.23 Figure 5.7, which
Rwanda
Zambia compares the results in terms of gross and net
Afghanistan
Nigeria replacement rates,24 shows a wide range of
Malawi
Bhutan design choices in the benefit targets for aver-
Uganda
Mozambique age wage workers. China, the Philippines, and
Lao PDR
Indonesia Vietnam have high target replacement rates
Honduras
Maldives compared with the rest of the region and
Cambodia
Pakistan OECD countries.25 Japan, Korea, Mongolia,
Guatemala
Dominican Republic and Thailand have moderate net replacement
Slovenia
Nicaragua rate targets at around 40 to 50 percent,
Bangladesh
Ghana whereas the four countries that rely on DC
Paraguay
Sri Lanka schemes have the lowest target benefit levels.
Vietnam
Bolivia Indonesia stands out—with replacement rate
Ecuador
Philippines targets implied by the design of its schemes—
Peru
Venezuela, RB
at below 20 percent.26
Colombia
Mexico
The degree of redistribution between low-
Panama
Iraq and high-income workers also varies consid-
Costa Rica
Thailand
erably in the region’s pension systems. Many
Chile
Kosovo
pension schemes have minimum pensions
Bosnia and Herzegovina
Argentina
and ceilings on the wages on which contribu-
Turkey
Brazil
tions are made. As a result, target replace-
Uruguay
Montenegro
ment rates are often higher for low-wage
Albania workers and lower for high-wage workers.
Croatia
Serbia Figure 5.8 shows the difference in simulated
Armenia
Tajikstan net replacement rates for workers with half
Bulgaria
Moldova and twice the average wage, respectively.
Romania
Poland Except in Indonesia, Malaysia, Singapore (all
Hungary
Mongolia DC), and Vietnam, target replacement rates
Kyrgyz Republic
Belarus are significantly higher for lower-income
Slovak Republic
Mauritius workers. The biggest differences are seen in
Ukraine
Korea and the Philippines.
0 10 20 30 40 50 60 70 80 90 100
Share of households with elderly receiving pension (%)
In contrast to most high-income countries,
actual replacement rates in DB schemes in
Other countries EAP countries
East Asia and Pacific (except in Vietnam)
Source: World Bank Aspire data.
tend to be lower than would be expected
given the design parameters and are modest
by global standards. Figure 5.9 shows the
housewives and women with low rates of actual average replacement rates for
labor force participation, one of those being earnings-related DB schemes in the region.
provision of a basic pension. The gap is high- Unlike target replacement rates, this indicator
est in Korea, perhaps because of how cover- applies to actual pensions today rather than
age has expanded since the National Pension simulated pension outcomes for those start-
Service (NPS) was set up in 1988. ing to contribute today, and it is an average of
all existing pensions divided by the average
Adequacy wage for all workers covered. Thus, it
The target replacement rate varies signifi- includes early retirement pensions, retirement
cantly among East Asian and Pacific pension pensions for workers across the income
PENSION SYSTEMS AND AGING: REFORMS FOR WINNING THE RACE 153
distribution, survivor and disability pensions, FIGURE 5.6 Elderly pension coverage is higher for men than for women
and pensions granted many years ago. The
low replacement rates in Korea and Thailand 90
reflect the immaturity of their schemes—as
FIGURE 5.8 In most economies, target replacement rates are Singapore, where the nominal return has
significantly higher for lower-income workers essentially been fixed since the scheme began.
Simulated net replacement rates for hypothetical high- and low-income Second, in Hong Kong SAR, China, the invest-
workers ment return is an average, and returns for
individuals vary widely according to choice of
China investment portfolio and fund (MPFA 2014).
The most disappointing case in terms of
Vietnam
DC returns is that of China, where returns are
Korea, Rep. linked to short-term bank deposits. As a result,
Mongolia returns on funds in DC accounts have barely
Philippines
exceeded inflation and have lagged far behind
wage growth since long before individual
Japan accounts were introduced. Figure 5.11 shows
Thailand that they continue to do so. To the extent that
Singapore
this lag continues, the replacement rates will
be much lower than those shown earlier.
Malaysia Critics have noted that replacement rates
Hong Kong SAR, China experienced by cohorts retiring in the past few
Indonesia decades in Singapore have been very low, but
recent trends in wage growth and policy
0 20 40 60 80 100 120 changes may improve the situation (Asher
Share of net wage (%) 2012; Asher and Nandy 2011).29 Its Central
Net replacement rate, low-income male Provident Fund (CPF) payouts are substantial
Net replacement rate, high-income male relative to lifetime income but not relative to
current incomes. Now that Singapore is a
Sources: OECD 2011, 2013; Wiese 2006. high-income country, wage growth is stabiliz-
Note: Simulated results are prospective and refer to new male labor market entrants employed and
making contributions throughout their careers. High and low income here refers to workers with ing at lower rates. If combined with an
twice and half the average wage, respectively. For defined contribution schemes, net rate of return increase in the effective retirement age, this
is assumed to be 1.5 percentage points higher than wage growth.
stability should result in higher replacement
rates in the future. However, cohorts that are
Public awareness that the accumulated bal- currently retiring do not have adequate
ances are not sufficient is increasing, and savings—thanks to their CPF housing invest-
recently a number of remedial measures have ments and appreciation of property values,
been proposed, including delayed withdrawal many are asset rich and cash poor—which has
(see, for example, Mahalingam and Wong led to several recent adjustments to the pro-
2013). Balances are even lower relative to gram. A lease buyback program allows elderly
member incomes in Indonesia largely because Singaporeans to sell their housing back to the
of low contribution rates combined with lib- government, along the lines of a reverse mort-
eral withdrawal rules.28 gage. Most recently, the government has
In some cases, another driver of low ade- announced the Silver Support scheme to sup-
quacy is the actual returns on DC accounts plement the pensions of those without suffi-
relative to the growth of wages. The simula- cient accumulated savings (Rodan 2014).
tions presented assumed that the net rate of For many people in East Asia and Pacific,
return on DC schemes exceeds growth of social pensions are the key determinant of
wages by 1.5 percentage points annually. overall pension system adequacy, and benefit
Figure 5.10 shows that the differential has levels are typically modest. This situation is
been slightly higher than assumed in Hong confirmed by figure 5.12, which compares the
Kong SAR, China, but much lower in China ratio of social pensions to income per capita
and Singapore. Two important caveats should in selected countries with the ratio of social
be pointed out: first, returns in Hong Kong pension recipients to the population ages 65
SAR, China, are much more volatile than in and above.
PENSION SYSTEMS AND AGING: REFORMS FOR WINNING THE RACE 155
The East Asian and Pacific economies with FIGURE 5.9 In most economies, actual replacement rates in DB
social pension programs fall roughly into schemes tend to be lower than would be expected
three groups:30 Actual average replacement rates in selected countries
Percent
Hong Kong SAR, China; Korea; and
Thailand—but the ratio of social pension- 30
ers to the population ages 65 and above is
quite high, ranging from 54 to 94 percent. 20
As noted earlier, Mongolia has increased
its reliance on social pensions over the 10
past decade, with benefits totaling about
20 percent of income per capita (World 0
Bank 2011). Korea, Rep. Mongolia Thailand China Philippines Vietnam
• The only East Asian and Pacific coun- Sources: ILO FACTS 2013; NPS 2012a, 2012b; Mesa-Lago, Viajar, and Castillo 2011; SSO 2009; Wang,
tries shown with relatively high social Béland, and Zhang 2014 (for urban workers only).
Note: DB = defined benefit. Percentages refer to average of old-age, disability, and survivor benefits
pensions and high coverage are Kiribati divided by relevant average covered wage.
and Timor-Leste. Their programs are
significantly more generous than other FIGURE 5.10 Real returns on DC accounts and growth of real wages
programs for which data are available Average annual real returns on defined contribution accounts and growth
and even compared with countries that of real wages
have high coverage levels, such as South
Africa. Timor-Leste’s ratio is actually 16
greater than 100 percent because its eligi-
14
bility age is 60.
12
The role of social pensions should be
% change per annum
FIGURE 5.11 Returns on funds in DC accounts in China have barely social pension is in place. These countries
exceeded inflation and have lagged far behind wage growth include Cambodia, Lao PDR, and Myanmar,
The growth of wages versus interest rates and DC returns which have no income support for the elderly
outside of contributory formal sector
3,000 schemes that cover less than 10 percent of
the population. The Philippines and Vietnam
2,500 have somewhat higher coverage through
their contributory pensions, but their social
2,000 pensions are very small and pay very low
Index (1990 = 100)
Kiribati
45 Timor-Leste personal savings decisions and the ability to
40 work into old age. Korea, for example, is
35 Australia clearly an outlier among OECD countries in
France Brazil
30 Chile terms of its elderly, with high poverty rates
25
Germany
South Africa and reliance on labor income. Rapid wage
Japan
20 Mongolia Canada Georgia growth also affects Malaysia and Singapore
Samoa
United States because of the relationship between wage
15 Bolivia
Malaysia
Hong Kong
growth and returns on account balances.
10 Korea, Rep.
Vietnam SAR, China The DC schemes used in these countries do
5 Philippines Mexico
India China
Thailand not allow for intergenerational transfers, and
0 unless investment returns exceed wage
0 20 40 60 80 100 120
growth, the approach will generate low
Social pensioners as a share of population age 65 and older (%) replacement rates.
several fronts, starting with the effects on the FIGURE 5.13 Older economies have higher pension contribution
demand and supply of labor. Two potential rates than do younger economies
effects are important: higher rates of informal Total contribution rates for pensions in East Asia and Pacific compared with
sector activity and lower labor force partici- selected economies
pation of the elderly. (A third potential effect
Honduras
of the pension system is on individual behav-
Rwanda
ior, caused by the lack of benefit portability.) Indonesia
Thailand
The international evidence for the first effect Congo, Dem. Rep.
is limited, and what evidence does exist is Côte d’lvoire
mixed. Two studies in Latin America found
Ireland
that a reduction in social security contribu-
tions led to an increase in formal Mauritania
Korea, Rep.
Lao PDR
employment—in both cases, a 10 percent Trinidad and Tobago
reduction led to a roughly 5 percent increase Cabo Verde
in formal employment (Heckman and Pagés
Hong Kong SAR, China
2004; Kugler and Kugler 2009). Two other Philippines
empirical studies, however, found no evidence Madagascar
East Asia and Pacific can be divided the population that is covered by the pension
roughly into two groups in terms of pension system. This results in highly divergent
contribution rates.32 The first group—which retirement patterns, particularly between
includes Hong Kong SAR, China; Indonesia; urban and rural workers, as shown in chap-
Korea; Lao PDR; the Philippines; and ter 2. That discussion notes, for example,
Thailand—has modest total contribution that the strongest determinant for exiting
rates that place them in the bottom third of work at older ages among urban Chinese is
the distribution at between 5.7 and 10.3 the receipt of a pension. Receipt of a pension
percent. The second group—China, Japan,33 was also associated with exiting the labor
Malaysia, Mongolia, Singapore, and force in Indonesia and (with a weaker rela-
Vietnam—has contribution rates between 18 tionship) in Japan and Korea, which is con-
and 28 percent. sistent with the design of their pension
Pension contribution rates in the second schemes.34
group of countries are likely to have a signifi- Retirement ages in national pension
cant impact on formality decisions, especially schemes are relatively low in East Asia and
contributions by smaller firms and the self- Pacific compared with life expectancy at age
employed. This impact is a particular concern of retirement, as shown in figure 5.14.
in China and Vietnam, which are aging Indonesia, Malaysia, and Thailand are well
quickly and where the race to extend cover- above the OECD average for life expectancy
age is most urgent. High historical rates of at retirement, and only the Philippines is
coverage in Mongolia are also not guaranteed below average. The situation is more dra-
to continue as the economy evolves. matic for women: their life expectancy is
Ultimately, the factors that lead to formaliza- more than 25 years at the normal retirement
tion, including incentives, economic structure, age, except in Korea and the Philippines.
and enforcement capacity, encompass much Moreover, in countries with lower coverage,
more than the pension system. However, in these figures understate the extent to which
the context of rapid aging, upward pressure life expectancy at retirement exceeds that in
on pension and other social insurance contri- OECD countries, because the mortality rates
bution rates will continue unless financing used for the calculation are based on national
policies change. To some extent this increase averages. The covered population, which
is already happening, as in the case of the tends to be urban and in the upper half of
growing role of earmarked value-added taxes the income distribution, almost certainly
in Japan. The policy options will be discussed lives longer than the national averages.
in the next section, on reforms. For most East Asian countries, LFPRs are
The impact of the pension system on indi- at or above levels found in countries whose
vidual decisions regarding when to retire is elderly have similar health-adjusted life expec-
much more direct and can be attributed to tancy. The clear exception is Mongolia, where
the design of the pension scheme itself. LFPRs are much lower than in comparable
Several studies that focused on OECD coun- countries. However, figure 5.15 shows that
tries have shown that the timing of retire- ability to work at older ages is not closely
ment is influenced by the incentives built linked to LFPRs of the elderly.35
into the pension and tax rules. Gruber and A much stronger relationship is seen
Wise (1999) and Wise (2004) find a strong between pension spending and LFPRs of the
relationship between the incentives built into elderly. Figure 5.16 compares the average
the pension systems and actual retirement share of GDP that countries spent on pub-
patterns across 12 OECD countries. Duval licly mandated pensions in 2010 and LFPRs
(2003) finds similar results for a sample of of people ages 60 and above. The expected
22 OECD countries. For East Asia and negative correlation is evident, but its inter-
Pacific, the effect on labor force participation pretation is not straightforward. Higher pen-
rates (LFPRs) will be restricted to a subset of sion spending is associated with both higher
PENSION SYSTEMS AND AGING: REFORMS FOR WINNING THE RACE 159
FIGURE 5.14 Retirement ages in national pension schemes are relatively low in East Asia and Pacific
compared with life expectancy at age of retirement
Life expectancy at retirement age
a. Male b. Female
Thailand Philippines
Malaysia United States
Indonesia Germany
Singapore Korea, Rep.
Vietnam Singapore
Japan Mongolia
China Sweden
Sweden Indonesia
Korea, Rep. Japan
United States Malaysia
Germany Thailand
Mongolia Vietnam
Philippines China
0 5 10 15 20 25 0 10 20 30 40
Life expectancy after retirement (years) Life expectancy after retirement (years)
average pensions and larger numbers of pen- FIGURE 5.15 Ability to work at older ages is not closely linked to
sioners, both of which should contribute to labor force participation rates
lower labor force participation of the elderly. Health-adjusted life expectancy and labor force participation rates for males
As public pension spending rises, the ages 60 and older
expected decline in LFPRs of older workers
100
is observed, and the effect is generally larger y = 0.2414x2 – 11.302x + 140.21
90
Male labor force participation rate
70
other important factors are not taken into
60
account.36 The major outlier is Japan, which Indonesia
50 Philippines
has much higher LFPRs than would be
40 Myanmar Vietnam
expected, given its pension spending levels. Lao Korea, Rep.
This analysis may provide evidence that the 30 PDR Japan
measures Japan has introduced, such as par- 20 Singapore
10 Mongolia
tial pensions with continued part-time work,
may have helped keep rates higher than they 0
6 8 10 12 14 16 18
would have otherwise been.
Among East Asian countries, Singapore’s Health-adjusted life expectancy at age 60 (years)
LFPR stands out as having LFPRs that have
Sources: For labor force participation rates, World Bank 2013; for health-adjusted life expectancy,
risen more in percentage terms than those of WHO 2010.
any other country in the past two decades.
Although Singapore provides incentives to living standards (see Asher and Bali 2013).
employers to employ older workers, at the In fact, Singapore has the highest LFPR for
same time its CPF balances are generating people ages 60 and above of any country
very low replacement rates, possibly forcing close to its income level. Recent changes that
people to work longer to maintain their restrict withdrawals before a certain age are
160 LIVE LONG AND PROSPER
FIGURE 5.16 Pension spending is closely linked to labor force increase in pension wealth. Only Japan and
participation rates Korea provide a significant reward for delay-
Pension spending and labor force participation rates of persons ages 60 ing retirement.
and above In developing East Asian and Pacific econ-
omies, in many cases incentives for early
20
Pension spending as a share of GDP (%)
TABLE 5.4 Retirement rules for East Asian contributory pension schemes
Normal retirement
age (men/women) Conditions for early retirement Conditions for late retirement
China 60/55 Blue-collar manual workers can retire at Allowed, but pension is not increased.
55/50.
Hong Kong 60/60 Allowed only in cases of terminal illness Allowed, with interest on the balance
SAR, China or permanent migration. accruing for the additional period.
Indonesia 55/55 for formal Withdrawal of balance allowed at any age Not allowed.
sector; 58 for most subject to having contributed for five years
civil servants; 56 and after one month of unemployment.
proposed in SJSN
Japan 65/65 Allowed at 60 with 6 percent reduction per Deferring retirement increases the
year. pension benefit by 0.7 percent per
Early retirement at a reduced benefit is month or 8.4 percent per year.
possible in both basic and earnings-related
schemes. The benefit is reduced by 0.5
percent per month of early retirement, or
6 percent per year.
Korea, Rep. 60/60 Allowed at age 55, but the minimum age At 60, early old-age pension will be
is rising to 60 by 2033. 70 percent of normal old-age pension.
Benefit is increased by 6 percent every
year, so a person who retires at age 64
will be entitled to 94 percent of the full
old-age pension.
People can earn extra pension from
retiring late. The benefit is increased
by 7.2 percent every year, and the
maximum deferral is five years until
age 70.
Lao PDR 60/60 Allowed at 55 with reduction based on n.a.
points system.
Malaysia 60/60 Some withdrawal possible prior to age 55. Allowed; interest on balance accrues
for the additional period.
Mongolia 60/55 Allowed for workers in hazardous jobs. n.a.
Philippines 65/65 Allowed at age 60 subject to minimum Allowed, but pension is not increased.
contribution period. Working while receiving a pension is
not allowed.
Singapore 63/63 (rising to 65 Subject to having the minimum required Allowed; interest on balance accrues
in 2018) balance, a lump sum can be taken for the additional period.
before the standard age of withdrawal.
Annuitized unless retiree opts out.
Thailand 55/55 Not allowed. Allowed with additional 1.5 percent per
year accrual.
Vietnam 60/55 Workers in hazardous jobs and certain Not allowed. Working while receiving a
areas can retire at 55/50; pension reduced pension is allowed.
1 percent per year before normal age.
Source: Palacios 2015.
Note: n.a. = not applicable. SJSN = National Social Security System. Indonesia’s new regulations, effective July 1, 2015, will gradually increase the retirement
age from 56 to 65 by 2043.
Pension systems also can have indirect An important point for fiscal accounting pur-
effects on the economy through their role in poses is that these two ways of accumulating
public and private savings behavior. pension assets are treated very differently.
Prefunding of public pension liabilities gener- Contributions that lead to a surplus in a DB
ally takes two forms: partial funding of DB scheme are considered part of general govern-
schemes and full funding of DC schemes. ment revenues, and the accumulation of
162 LIVE LONG AND PROSPER
TABLE 5.5 Retirement rules and incentives for male workers in negative, the DB schemes contribute directly
selected countries to larger deficits, a situation now common in
Early retirement Late retirement
Europe and already a reality in Mongolia.
Country Normal age penalty (% change) reward (% change) In contrast, evidence indicates that DC
schemes are likely to have a neutral to slightly
Japan 65 −23 23
Korea, Rep. 65 −22 15
positive effect on private savings.39 Most of
Philippines 65 18 −17 the literature examines the changes in private
Thailand 55 Not possible 1 savings when economies move from unfunded
Vietnama 60 15 Not possible DB schemes to privately managed DC
Source: World Bank calculations.
schemes. The amount of savings depends on
Note: Figures in the last two columns refer to the annual change in present value of pension how much of the transition costs are covered
payments from retirement until death for early and late retirement. Figures use national mortality
rates at relevant ages.
by borrowing from the new DC scheme. In
a. Amendments to Vietnam’s social insurance law in 2014 increase the early retirement penalty four Latin American and Caribbean countries,
slightly but not enough to offset the increase in pension wealth.
estimates of the overall increase in savings
ranged from 1 percent to 2.5 percent of
pension reserves in DB schemes can reduce GDP.40 However, no evidence shows that the
explicit government deficits and debt in the Eastern European reforms of this type increase
near term if surpluses are not borrowed for net savings. In East Asia and Pacific, at least
additional consumption. However, a measure two reforms have involved moving from an
of the true impact on public savings should unfunded DB to a funded DC model—namely,
take into account the fact that unreported in China and Thailand41—but no studies have
pension liabilities are growing (see Goebel, been conducted on the impact of these reforms
van de Ven, and Zwijnenburg 2015; on savings. Research on the provident funds
Holzmann, Palacios, and Zviniene 2004). In in Malaysia and Singapore generally shows a
contrast, contributions to individual account positive effect on private savings, although the
schemes are not counted as part of revenues results vary. Husain (1995) found that man-
because they are considered the property of dated savings in the CPF in Singapore were
their owners, the individual workers. Pension fully offset by a reduction in voluntary private
savings in the form of individual accounts are savings. In contrast, Hopf (2009) found that
considered private savings, and here too, the despite some offset through lower private vol-
government may offset any positive effect by untary savings, Singapore still showed an
borrowing more than it would otherwise. important positive net effect. Wickramanayake
However, in this case, no hidden liability is (1998) also found that the CPF contributed to
affecting the government’s long-term finances. national savings during the period 1970–94.
The evidence suggests that mandated DB The International Monetary Fund found no
schemes are likely to reduce savings. In the statistically significant effect of Malaysia’s
case of unfunded or partially funded DB EPF on private saving, suggesting that the
schemes, as individuals accumulate pension reduction in voluntary savings and withdraw-
wealth, they may reduce their own private als offset the impact of the mandate (IMF
savings for old age. Some evidence of this 1994).
effect can be seen in economies of East Asia Prefunding can also affect capital markets.
and Pacific, though mostly it occurs in richer Walker and Lefort (2002) outline the possible
countries.37 Yamada, Yamada, and Liu (1992) channels for such an impact. Their empirical
found a reduction of 40 percent in private sav- results for Latin America and the Caribbean
ings in Japan for every increased unit of public suggest that the presence of these funds can
pension wealth.38 Especially in DB schemes lead to “a reduction in the cost of capital;
that ignore growth in unfunded liabilities dur- lower security price volatility; and higher
ing the immature stage of the system, investing traded volumes” (Walker and Lefort 2002, 3).
heavily in government bonds will tend to Catalán, Impavido, and Musalem (2000) note
reduce public savings. When cash flows turn that the causality could run in the opposite
PENSION SYSTEMS AND AGING: REFORMS FOR WINNING THE RACE 163
direction, namely from improvements in the FIGURE 5.17 Only higher-income economies have pension assets
capital markets to the pension funds. They that are large relative to GDP
found that in most wealthier countries, pen- Pension fund assets as a share of GDP
sion funds led to greater stock market capital-
ization and liquidity, although the results Singapore
varied across countries. In Malaysia and Malaysia
Singapore, little or no evidence demonstrated Korea, Rep.
causality between institutions and markets, Japan
Hong Kong SAR, China
which the authors attribute to centralized
Thailand (total)
public management. In a recent study, Hu
Philippines (total)
(2012) used data from 10 Asian economies, Vietnam
including China; Hong Kong SAR, China; China
Korea; Malaysia; Singapore; and Thailand, to Indonesia
explore the effects of the growth of pension 0 10 20 30 40 50 60 70 80
assets on financial markets. The study found a Public pension assets as share of GDP (%)
significant impact on stock market capitaliza-
tion and, for developing countries, on liquid- Sources: CPF Board 2013; Government of Korea 2012; Mesa-Lago, Viajar, and Castillo 2011; MPFA 2014;
ity. 42 In addition, it found no impact on OECD 2013; SSO 2009; World Bank 2013.
Note: Thailand includes Social Security Office and Government Pension Fund, plus the Social
government bond markets but a positive Security System and the Government Service Insurance System. China refers to the National Social
impact on corporate bond markets. The Security Fund only.
FIGURE 5.18 Pension agencies in East Asia and Pacific generally been seen in the provident funds of some
have conservative investment practices Pacific Island countries, particularly with
Portfolio composition of mandated pension funds in East Asia respect to investments in domestic real estate
ventures.46
Hong Kong SAR, China Pension agencies in East Asia and Pacific
generally have conservative investment
Philippines (SSS) practices, with the exception of Hong Kong
SAR, China, and, more recently, Malaysia.
Philippines (total) Figure 5.18 shows that Hong Kong SAR,
China, has by far the most investments in
Philippines (GSIS)
equity markets among these economies.
Malaysia However, many public pension funds glob-
ally have increased their holdings in private
Korea, Rep. securities and other risky investments, par-
ticularly during the past decade in Korea
Thailand (GPF) and Malaysia. At the other end of the spec-
trum, Vietnam’s and Thailand’s SSOs are
Indonesia
heavily concentrated in cash and govern-
Japan ment- or state-owned enterprise bonds. In
Vietnam, the portfolio consists completely
Thailand (SSO) of government bonds, and returns have
been negative in real terms in recent years
Thailand (total)
(IHME 2013). In the case of Thailand, the
Singapore
fact that the GPF is able to have a signifi-
cantly more diversified portfolio than the
Vietnam SSO suggests that the portfolio is the result
of a policy choice.
0 10 20 30 40 50 60 70 80 90 100 The size of pension or provident funds
Percent relative to domestic markets is a factor even
Bank deposits/fixed income Equities Other
in East Asian and Pacific economies with
deep stock and bond markets. A clear case
Sources: CPF 2013; Government of Korea 2012; Mesa-Lago, Viajar, and Castillo 2011; MPFA 2014; OECD has been made for diversifying pension and
2013; SSO 2009; World Bank 2013. provident fund investments internationally
Note: GPF = Government Pension Fund; GSIS = Government Service Insurance System; SSO = Social
Security Office; SSS = Social Security System. to obtain the best risk-adjusted returns. The
accumulated balances in Malaysia and
instructive for demonstrating the link between Singapore are equivalent to 40 to 45 percent
governance and performance. Mesa-Lago, of the value of their respective stock markets.
Viajar, and Castillo (2011) document how In Korea, the NPS already has shares worth
political objectives have influenced the choice about 5 percent of the domestic stock mar-
for board membership and cite examples in ket, owns 15 percent of domestic bonds
which projects and personal loans were made traded, and is expected to grow rapidly for
under pressure from senior politicians. The the next two decades (Palacios 2015). Earlier
authors also cite politically motivated invest- studies have shown that larger pension funds
ments in airlines, hotels, and commercial with long track records have decent perfor-
banks that had negative repercussions in capi- mance relative to other public pension funds
tal markets and led to losses for the fund. in developing countries, and this record has
Between 2000 and 2010, the real rate of persisted and even improved in the case of
return in the SSS and the Government Service Korea and Malaysia (see Palacios 2002 for a
Insurance System (GSIS) was 3.3 and 5.1 cross-country comparison). Nevertheless,
percent, respectively, even as GSIS returns these figures show room for improvement in
were less volatile. Similar challenges have terms of diversification and its potentially
PENSION SYSTEMS AND AGING: REFORMS FOR WINNING THE RACE 165
TABLE 5.6 Assessment of pension system performance and level of reform needed
In Vietnam, low coverage is a major shaping their future pension model. Because
problem in the context of rapid popu- of Thailand’s long retirement duration and
lation aging. These countries also have low contribution rates, its main pension
high contribution rates and incentives scheme is projected to run deficits in about
to retire early. Prefunding is not likely to two decades. Through the massive expan-
deliver benefits as designed in China and sion of the social pension, coverage has
Mongolia and increasingly so in Vietnam, become close to universal, but benefit lev-
where management of reserves needs to be els are modest and have a relatively modest
improved. The Philippines also requires impact on poverty among the old (although
significant reforms in these areas. a more significant impact on extreme pov-
• The DC schemes in Hong Kong SAR, erty). In contrast, pensions of civil servants
China; Malaysia; and Singapore mainly in Thailand are much higher than for pri-
face problems related to adequacy, albeit vate sector workers. In terms of indirect
for different reasons. Having started its effects, the private sector scheme provides
mandate only recently, Hong Kong SAR, incentives to retire earlier, and the SSO’s
China, is already demographically aged prefunding approach undermines poten-
before its scheme has matured, and even tial saving and capital market impacts. The
when it does, retirement income will be situation in Indonesia is similar in many
low for the lifetime poor and the unlucky ways: the low contribution rate for the DC
(in terms of rate of return). Malaysia and scheme ensures low benefits, and coverage
Singapore have been the victims of their is very low. The major difference is the lack
own success in that growth in wages has of any social pension, which makes both
been a boon to lifetime incomes. However, coverage and adequacy urgent priorities.
modest rates of return on individual The proposed new pension scheme, which
accounts are resulting in low replace- would include a DB plan, fundamentally
ment rates for those retiring recently or changes the picture but only for those who
in the near future. This pattern is likely to will be covered. At the time of writing,
change in the long run and could lead to Timor-Leste was also planning to introduce
higher replacement rates, but it will not a national DB scheme. The current univer-
prevent poverty for the lifetime poor. sal pension raises concerns over longer-
• Japan and Korea have chosen to rely term sustainability. Cambodia, Lao PDR,
on a combination of partial funding and Myanmar have very low coverage rates
and modest benefit targets to ensure even for their income levels and have no
sustainability. Japan appears somewhat social pensions or broad social assistance
more sustainable after its shift to financing schemes to address elderly poverty.
through its consumption tax (the goods
and services tax) and other measures that
will effectively prevent an increase in the
Reforms to date and options
pension bill as a share of GDP in the long
looking forward
run (see Kashiwase, Nozaki, and Tokuoka Although a number of East Asian and Pacific
2012).47 As in Hong Kong SAR, China, countries have undertaken major pension
the immaturity of the scheme in Korea reform efforts since the East Asian financial
means that cohorts retiring soon will have crisis of 1997, many challenges have still not
low pension income. This transitional been addressed. This section first reviews
phenomenon in Korea helps explain why pension reforms in East Asia and Pacific,
it has among the highest elderly poverty some of which were touched upon earlier in
rates in the OECD. the chapter, then outlines key reform recom-
• The picture is mixed among countries mendations. These recommendations deal
of Southeast Asian countries, with low- mainly with changes to existing parameters
income countries facing major choices in and, in some cases, to the integration of the
PENSION SYSTEMS AND AGING: REFORMS FOR WINNING THE RACE 167
separate systems for civil servants and private In both DB and DC systems, most reform
sector workers. The last subsection looks at activity has taken place in the past 15 years,
more fundamental measures to move away in the period following the 1997 East Asian
from the traditional social insurance approach financial crisis. For example:
that is starting to fail in some parts of the
• Indonesia and Lao PDR passed major
region even before countries have grown old.
pension legislation affecting DB systems
during this period. In Lao PDR, a man-
date for private sector workers was intro-
Reforms since 2000
duced in a 1999 law and has subsequently
Over the past 15 years, pension reforms in been implemented. The DB scheme uses
East Asia and Pacific have been diverse in a point system to determine benefits and
terms of the relative focus on coverage, ade- has a number of well-considered param-
quacy, and sustainability. Reforms tended ini- eters. The Indonesian law represented a
tially to focus on sustainability and included fundamental departure from historical
new approaches to funding as well as reduc- pension policy by introducing a DB com-
tion in liabilities through benefit reductions. ponent with much higher benefit targets
In contrast, efforts in the past 10 years have (along with significant modification of the
been much more focused on expanding cover- DC element). However, implementation
age and improving adequacy, at least in terms has been severely delayed, and 10 years
of alleviating poverty in old age. Table 5.7 after the law was passed, key parameters
lists the pension reforms that have taken place have only just been determined. The impli-
in the region since 2000. cations for adequacy and sustainability,
as well as the indirect impact on labor and Service itself and the public debates that
capital markets, remain to be assessed. took place. The result was an increase in the
• China’s shift from a DC system to a hybrid retirement age and reduction in benefit lev-
DB/DC system with individual accounts els, with further reduction in accrual rates
implied a major policy change in terms in the 2007 reform. Very few countries have
of prefunding (at least nominally) and the made such farsighted reforms at a stage
nature of benefits. Mongolia’s law in 1999 when the scheme is immature and running
also fundamentally changed how benefits large surpluses.
would be calculated by moving to an NDC
system. As discussed earlier, in neither case The period 2006–10 was notable for the
have the reforms unfolded as planned. In spread of noncontributory or social pensions
China the funds allocated to the new indi- in East Asia and Pacific (see Rofman, Apella,
vidual accounts have not in practice been and Vezza 2014 on Latin America). This was
segregated in most parts of the country a major element in the 2007 reform in Korea
and are used to pay accrued benefits. In and resulted in a huge increase in coverage. As
Mongolia, spending in 2012 reached close discussed earlier, China and Thailand also fol-
to 4.9 percent of GDP, exceeding projec- lowed a policy of extending social pension
tions made after the reform that spending coverage to the vast majority of those not eli-
in 2012 would be around 3.2 percent of gible for contributory pensions. More recently,
GDP. The introduction of the mandatory Japan significantly expanded social assistance
DC scheme in Hong Kong SAR, China, in for pensioners. The most ambitious social
2001 was another important milestone for pension was introduced in Timor-Leste, which
individual accounts in the region, which joined a substantial number of countries with
contrasts sharply with the developments universal social pensions, including Fiji,
in mainland China. Nowhere is the “one Mauritius, New Zealand, Samoa, and Tonga.
country, two systems” approach more evi- The 2007 Workfare Income Supplement in
dent than in pension policy, with Hong Singapore took a different approach, with the
Kong SAR, China, amassing close to a government supplementing wages and retire-
third of GDP in assets in just over a decade. ment savings for low-income workers, signifi-
• The most comprehensive reforms to an cantly raising replacement rates for these
existing system during this period took workers (see Choon and Tsui 2012).
place in Japan. All major parameters were The most important pension reforms since
adjusted: retirement ages were increased, 2011 have aimed at expanding coverage.
indexation was changed, and benefits were Myanmar’s Social Security Law includes, for
reduced. Most important among those were the first time, a mandated contribution by pri-
(a) the introduction of an automatic stabili- vate sector workers and employers. Thailand
zation mechanism that reduced future ben- introduced two schemes to bring informal
efits to offset increases in dependency ratios sector workers into the pension system, both
and (b) an increase in the consumption tax with significant subsidies to encourage take-
that was earmarked to partly pay the basic up, and Timor-Leste introduced a civil service
pension. Together, these measures appear to pension scheme. The most dramatic develop-
have achieved long-term financial sustain- ment, however, has been the massive increase
ability (Kashiwase, Nozaki, and Tokuoka in coverage in China in the past five years
2012). Reforms in Korea were passed in thanks to the reform that links voluntary con-
the midst of the financial crisis, but only a tributions from informal sector workers to
decade after the scheme started, the gov- provision of a social pension for both work-
ernment acknowledged that the original ers and their parents. By the end of 2014, just
parameters were not sustainable. What was under 500 million rural and urban informal
impressive was the focus on long-term pro- workers contributed, and 140 million elderly
jections produced by the National Pension were receiving basic pensions.
PENSION SYSTEMS AND AGING: REFORMS FOR WINNING THE RACE 169
Recommended reforms FIGURE 5.19 Except in Japan, contribution rates required to pay
for defined benefits are significantly higher than actual rates
Parametric reforms Actual versus required contribution rates for public pension schemes
Across East Asia and Pacific, ongoing reforms
of existing contributory systems are needed to China
make the systems sustainable and to create fis-
Vietnam
cal space for expanding pension system cover-
Thailand
age. Figure 5.19 shows that the contribution
Philippines
rate required to pay for the DB in each coun-
Korea, Rep.
try is significantly higher than the actual con-
tribution rate, with the exception of Japan. Japan
Because contribution rates have risen over 0 10 20 30 40 50 60
time in most of these countries, the gap was Share of wage (%)
even larger during the early years of schemes. Actual contribution rate Breakeven contribution rate
Payroll tax rates at these levels would encour-
age evasion or unemployment, and most Source: OECD 2013.
Note: The Japanese contribution rate refers to 2017.
countries would either have to use other
financing sources or reduce benefits. This was salaries are typically higher relative to lifetime
the conclusion of Korea’s reformers in two salaries for high-income workers, giving them
rounds of parametric reforms that ultimately a better rate of return than is provided to low-
reduced the target benefit from 70 percent to income workers. Price indexation provides
40 percent over a roughly 30-year period. protection against inflation and eliminates the
Some consensus exists over desirable para- arbitrary treatment of different cohorts that
metric reforms for existing pension systems, can occur when adjustments are discretionary,
which are particularly pressing priorities in as is current practice in several East Asian and
the middle-income countries. The following Pacific pension systems. Almost all OECD
seven parametric measures have all been countries have reformed their DB schemes in
implemented in a range of OECD and Eastern this way, although reforms in civil service
European and Central Asian countries over schemes typically lag those of national
the past two decades, but much less widely in schemes.48
East Asia and Pacific: The last three measures are related to
• Moving away from final salary and toward eligibility. Again, most OECD countries are
lifetime, revalued average wages as the base using these kinds of parameters, at least in
wage for calculating replacement rates their main national schemes. Lower retirement
• Using linear accrual schedules ages for women increasingly is recognized as
• Indexing pensions to changes in prices leading to lower pensions for them. It is also
• Having minimal vesting requirements clear that life expectancy has increased signifi-
• Linking retirement-age changes to life cantly since retirement ages were set decades
expectancy ago and that people can work longer than they
• Equalizing retirement ages for men and did earlier. The average retirement age in
women OECD countries has been rising gradually
• Providing actuarially fair (neutral) incen- over the past 15 years, and legislation already
tives for early or late retirement passed will continue to increase normal retire-
ment ages.49 Retirement ages of men and
The first four measures relate to the benefit women also are converging (figure 5.20). One
formula used to calculate the value of the of the perceived advantages of notional
pension and how it is adjusted. Longer aver- defined contributions, the approach followed
aging periods, linear accrual rates, and short in Italy, Latvia, Poland, and Sweden, is that
vesting periods reduce distortions and favors the value of the pension is automatically
toward certain workers. For example, final adjusted according to age of retirement and
170 LIVE LONG AND PROSPER
takes life expectancy changes into account. East Asia and Pacific, but with the exception
However, other mechanisms can achieve the of Japan, Korea, and Singapore, little prog-
same result, as described in box 5.1. ress has been made. The arguments for
East Asian and Pacific countries are at dif- increasing and equalizing retirement ages,
ferent stages in implementing parametric linking them to future changes in life expec-
reforms. These types of changes have been tancy, and rewarding voluntary later retire-
thoroughly implemented in Japan and ment are most relevant for China, Mongolia,
Korea, and Mongolia’s new NDC scheme Thailand, and Vietnam. Such reforms are
addresses many of these parametric issues. In gradual, and in the OECD (including Japan
all other countries, the seven parametric and Korea), they have been phased in over
measures described above are relevant. years, even decades. It will also be an impor-
Retirement age reforms have been recom- tant design choice for new schemes in
mended for many years by many experts in Indonesia, Lao PDR, Myanmar, and Timor-
Leste, where policy makers are in the midst
of making decisions and should avoid the
FIGURE 5.20 Retirement ages for men and women are converging temptation to legislate low retirement ages.
in OECD countries
Retirement ages, 1949–2029 Integration of civil service pension schemes
In terms of labor market, equity, and fiscal
65 effects, deepening the integration of civil ser-
Normal retirement age (years)
BOX 5.1 Automatic adjustments for changes in life expectancy in defined benefit schemes
In the past decade, at least 20 Organisation for Eco- France has linked the minimum contribution years
nomic Co-operation and Development countries have required for retirement to life expectancy.
introduced parametric reforms linking the scheme’s The adjustment in Japan as part of the 2004
parameters to changes in life expectancy. In Finland reform was quite different. A reduction of 0.9 percent
and Portugal, benefits will be reduced by a factor that in the benefit at retirement is being applied each year
is directly related to life expectancy. The link is less through 2023 based on the estimated increase in life
direct in Germany, where benefit adjustment is based expectancy during this period. However, the reduc-
on the ratio of pensioners to workers. Pensioners liv- tion is not adjusted if the estimate turns out to be
ing longer translates into a lower value of initial pen- lower or higher than the actual change.
sions. Italy and Greece will link their retirement age to
life expectancy starting in 2015 and 2020, respectively. Source: Based on OECD 2011.
PENSION SYSTEMS AND AGING: REFORMS FOR WINNING THE RACE 171
of contributors) and ensuring portability of FIGURE 5.21 Expanding coverage has become difficult in recent
benefits between public and private sectors. years
Another argument in favor of integration is Predicted labor force coverage by income level, 1994 versus 2014
based on equity grounds but also has implica-
FIGURE 5.22 The increase in coverage of contributory pension Since 1995, the government has provided
systems in East Asia and Pacific is unlikely to keep pace with matching contributions to farmers and fisher-
population aging man, increasing their participation signifi-
Change in mandated contributory pension coverage from the early 1990s to cantly (Moon 2013). 51 The result of the
around 2010
policy changes is shown in figure 5.23. From
1988 to 2012, labor force coverage rose by
60
about 50 percentage points.
50 Frustration with the pace of expanding
% change in labor force coverage
sia
ia
re
p.
ne
in
Re
po
na
go
ne
Ch
pi
ala
et
ga
a,
on
do
ilip
Vi
re
M
In
Ph
Ko
FIGURE 5.23 In the Republic of Korea, a dramatic change in coverage took place
Number of workers covered by Korea’s National Pension Service, 1988–2012
25,000
20,000 20,329
Number of workers (thousands)
17,740
17,182
15,000 16,262 All businesses
Firms with
Firms with
Urban fewer than
5 or more
residents 5 workers
workers
10,000 Firms with
10 or more
workers
7,497
5,000 Match for
5,021 farmers and
4,433
fishermen
0
1988 1992 1995 1999 2003 2006 2012
that operate in parallel to the social pension. One risk of expanding social assistance or
Another option may be to consolidate pro- social pensions in all but the poorest countries
grams as the overall budget envelope is can be the increased incentive to stay in the
increased and to use a common targeting informal sector.53 In Thailand, for example,
approach and delivery platform. the fact that a person has any pension from a
In lower-income countries such as formal sector scheme negates eligibility for
Cambodia or Myanmar, which have neither the social pension. If the value of the social
social pensions nor significant social assis- pension were to rise significantly, so would
tance programs, the best approach may be to the effective tax on contributions to the pub-
expand targeted social assistance programs. lic pension scheme. A better approach where
Survey data suggest that most poor elderly capacity exists would be to withdraw the
reside with families. Thus, a well-targeted transfer gradually at higher income levels.
transfer will tend to reach the elderly poor, This was done in Chile as part of reforms in
achieving the same objective as a social pen- 2008 that expanded the social pension to
sion while addressing broader antipoverty cover the bottom 60 percent of the distribu-
objectives. In such cases, a social pension is tion (see box 5.2). It reduced incentive prob-
unlikely to reduce poverty more than a pro- lems significantly relative to imposing an
gram that determines eligibility based on effective 100 percent marginal tax on formal
poverty rather than age. Notably, transfers wages and pensions.
made to the bottom segments of the income Another approach to expanding coverage is
distribution are unlikely to reduce formal to subsidize contributions of low-income and
sector activity in countries where contribu- informal sector workers. Some international
tory pension scheme coverage does not evidence suggests that matching contributions
exceed 10 percent of the labor force. These can increase coverage at the margin (see Hinz
workers simply do not have the choice to et al. 2013). In Korea, for example, matching
move in and out of the formal sector. contributions for fishermen and farmers were
174 LIVE LONG AND PROSPER
BOX 5.2 Integrating noncontributory and contributory pensions: Chile’s solidarity pillar
Total
pension
Self-financed pension
PENSION SYSTEMS AND AGING: REFORMS FOR WINNING THE RACE 175
found to have increased participation by pensions than might otherwise have been
around 10 percentage points. Under the paid. In China more than in most countries,
Malaysia Retirement Savings Scheme, estab- strong arguments can be made for a larger
lished as an incentive scheme for the self- social pension, given the huge disparities in
employed and for individuals without fixed pension levels across different schemes as well
monthly income they could save for retire- as the economic effects of excess savings.55
ment, the government provides a matching In the poorest countries in the region,
contribution of 10 percent of the contribution, matching contributions are unlikely to be the
subject to a maximum. However, only just best use of scarce fiscal resources. In these
over 73,500 individuals have participated countries, the trade-off is more acute when it
since the scheme was introduced in 2010, con- involves spending more on social assistance
sistent with the modesty of the match (see or social pensions or diverting funds into indi-
Dorfman 2014; Samad and Mansor 2013). vidual accounts. This trade-off is particularly
Thailand has introduced two pension pro- difficult if subsidies cannot be targeted effec-
grams for informal workers, which are admin- tively, since higher-income informal sector
istered by separate ministries (see World Bank workers will benefit most. Furthermore, the
2012b). One is an extension of the DB scheme lack of a financial and institutional infrastruc-
run by Thailand’s SSO, and the other is a DC ture that can channel savings is more prob-
scheme that provides matching contributions lematic in countries like Cambodia and
ranging from 50 percent below age 30, to 80 Myanmar than in China.
percent between ages 30 and 50, to 100 Matching contributions may work better
percent at age 50 and above, with a required in countries where extensive, well-targeted
contribution of less than US$2 per month. programs already exist and where the infra-
Almost 1.4 million people had signed up for structure for a DC scheme is already function-
the SSO scheme as of March 2013. ing. The conditional cash transfers program
China’s experience in combining a social in the Philippines provides a database that
pension with a matching contribution to could be used to target subsidized contribu-
workers illustrates both the potential and lim- tions, and it has a sophisticated system for
itations of matching contributions to expand delivering transfers that could be exploited to
coverage. A flat amount, initially around collect contributions. However, the country
US$9 per month (US$11.30 per month from lacks the infrastructure for administering and
mid-2014), is paid as a basic pension to those regulating other elements of a DC scheme.
ages 60 and above living in rural areas and to Legislation in Vietnam envisions a DC scheme
informal urban workers with a local residence (with a match) for informal sector workers,
permit (hukou). Eligibility for those already but it has not yet been implemented. The out-
over 60 is contingent on having working-age comes will depend on the incentives provided
spouses or children who contribute to their as well as on the administrative infrastructure
own pension accounts (a process referred to available to collect, track, and invest these
as family binding, though implementation funds.
demands appear to have resulted in weak
enforcement of this requirement) (Dorfman Options for changing the paradigm
et al. 2013). The massive enrollment that has The attempt to address the coverage gap in
taken place demonstrates the feasibility of col- East Asia and Pacific is part of a global pat-
lecting small amounts from informal sector tern occurring over the past decade or so for
workers; however, without better returns and policies that move away from a sole reliance
more direct incentives (for example, a higher on payroll taxes for financing pensions. As
match), many will remain outside the system discussed in the previous section, financing
for much of their working lives, and benefits the expansion of social pensions and match-
will be inadequate.54 At the same time, subsi- ing contributions for informal sector workers
dizing individual accounts results in an oppor- from the budget is a response to the persistent
tunity cost, notably in the form of lower social failure of contributory pension schemes to
176 LIVE LONG AND PROSPER
close the coverage gap. Meanwhile, the use of ties or phased withdrawals based on individ-
consumption taxes in Japan and budget sup- ual accounts.
port for the contributory pension scheme in This approach could be tailored to the ini-
China is partly driven by reluctance to raise tial conditions of each country. In a country
payroll tax rates further. The trend is not uni- with low mandated contributions and benefit
form, however, as evidenced by the introduc- targets (such as Thailand), individual accounts
tion of a social insurance scheme in Lao PDR could supplement the contributory pension.
and by legislation in Indonesia and Myanmar. Because the individual account–based pension
Although these countries are among the last would be the same for all workers, the effect
to have followed the social insurance–based on replacement rates would be greater for
path to provide pensions, the experience of low-income workers. In a country with higher
dozens of countries that once faced the same benefit targets and required contribution rates
choice suggests that it may not be the appro- (such as Vietnam), the latter could be reduced
priate path. proportionately as the new universal individ-
A way forward in pension reform can ual account fills the gap, achieving the same
instead be based on the premise that the replacement rates but with less reliance on
redistributive element of the pension system payroll taxes. Similarly, in countries that oper-
should be financed from general revenues ate DC schemes, this approach would simply
and clearly distinguished from the insurance involve a reduction in employees’ contribu-
or savings component (World Bank 1994). tion by the flat amount contributed by gov-
The latter would ideally be financed by indi- ernment. This method would increase
viduals on an actuarially fair basis. The sim- take-home pay and reduce the tax wedge
plest approach would be to eschew the more for low-income workers by acting as a
mandated scheme altogether and instead progressive wage supplement and encourag-
have a social pension with broad coverage ing workers with low wages into the formal
(for example, universal or affluence-tested sector, as done in Singapore. In the high-cov-
coverage), complemented by a voluntary erage DB schemes in Japan and Korea, financ-
retirement savings scheme. This approach is ing the basic pension with other taxes would
the situation in New Zealand, which also achieve the same objective without the need
features an opt-in default for a voluntary DC for individual accounts. Japan’s earmarking of
scheme that has led to relatively high partici- consumption taxes to finance the basic pen-
pation rates. However, this option is not sion is already a step in this direction.
likely to be accepted in countries that are In some East Asian and Pacific economies,
about to introduce mandates, and it may fiscal constraints or governance challenges
face resistance in countries with existing may make prefunding difficult or impractical.
mandates. A government contribution to a funded
Another option, first proposed in the con- scheme for all adult citizens would eventually
text of Mexico, is for the government to con- require an allocation of several percentage
tribute a minimum amount to individual points of GDP. The resources would have to be
accounts that would be opened for every identified and effectively earmarked. In
adult citizen. The government’s contribution resource-rich economies where sovereign funds
would be calibrated to produce an accumula- exist (such as Mongolia and Timor-Leste),
tion of funds sufficient to pay an annuity financing could involve earmarking of reve-
equivalent to the social pension, phased in nues for this purpose. Regardless of the source,
gradually.56 Social pensions or broad social allocating scarce resources to this purpose
assistance could provide for the current involves significant fiscal trade-offs. However,
elderly and transition cohorts until the indi- good equity and efficiency arguments can be
vidual account system matures. The result made for shifting from payroll taxes to other
would be a gradual transition from an revenue sources, especially where the tax
unfunded social pension to minimum annui- wedge is large and where the relative taxation
PENSION SYSTEMS AND AGING: REFORMS FOR WINNING THE RACE 177
of factors of production is currently overly reli- but the same approach is evident in Korea and
ant on labor (as in China and Vietnam). Thailand. The new scheme in China that links
Important challenges related to governance social pensions and a government contribution
and potential investments would also have to to individual accounts is the most important
be addressed. If centrally managed, address- example of what may represent a paradigm
ing those challenges would result in the accu- shift in pension provision, not only in East
mulation of a fund that was very large relative Asia but also globally.
to the economy. Many economies have lim- The reform agenda for the low- and
ited capital markets and may find investing middle-income countries of East Asia includes
abroad difficult. It is also difficult to avoid improving existing pension systems, such as
conflicts of interest, ensuring that investments parametric changes and harmonization or
are not motivated by reasons other than good integration of parallel schemes, as well as
pension outcomes. The international experi- expanding coverage of the informal sector.
ence with the management of large public The design of pension schemes and especially
pension funds does not provide many positive their retirement incentives can significantly
experiences in this regard (Iglesias and affect labor force participation rates and
Palacios 2001). If the individual accounts should be reformed to encourage longer
were managed by competing private manag- working lives as healthy life expectancy rises.
ers, as in Hong Kong SAR, China, the govern- However, historical experience suggests that
ment’s role as regulator and supervisor would these reforms can provide only part of the
be crucial, and few economies in the region solution and that a comprehensive plan will
have the required capacity. include programs that can reach the majority
Governments have two alternatives to pre- of people who operate in the informal sector.
funding. The first is the NDC approach, in This observation is especially true in those
which accounts are credited with contribu- countries where the race between coverage
tions but have no segregated assets.57 The sec- and aging is already well under way.
ond would be for the government to simply
issue government bonds that would be depos- Notes
ited into individual accounts. These would be
special, nontradable bonds and could be 1. However, the civil service pension schemes
linked to growth of GDP or income per cap- of Indonesia and Malaysia have significant
ita. The impact of the latter approach on unfunded liabilities.
2. Zuo (2013) calls these “empty individual
national savings depends on the overall fiscal
accounts” and estimates them at 90 percent of
adjustment, which could range from complete recorded contributions.
tax financing to complete debt financing. 3. These accounts mimic DC schemes but
are not funded. Individual accounts are
Conclusion credited with contributions that then earn a
notional interest rate that is typically linked
A paradigm shift is needed (and already under to the growth of wages or the wage bill. See
way) in how pension systems in East Asia and Holzmann and Palmer (2006) for a review of
Pacific are designed and financed. As discussed international experience.
throughout this chapter, a growing number of 4. Amendments in 2014 in Vietnam introduce
countries, frustrated with the traditional model automatic price indexation.
5. The Mandatory Provident Fund Authority
of contributory pensions that dominated the
is proactive in terms of financial literacy and
last century, are finding ways to plug the cov- making improvements to the system, for
erage and adequacy gaps with programs example, introducing life-cycle investment
financed by general revenues. These programs options.
break the link between formal employment 6. Conceptually, the unfunded liability is the
and pension (and health insurance) coverage. present value of future pension obligations.
Timor-Leste is the most dramatic example, The financing gap is the difference between
178 LIVE LONG AND PROSPER
the present value of contributions and pension cites March 2014 figures of about 70,000
payments. Unfortunately, international registered members (about 0.5 percent of the
standards for reporting these key measures of labor force) in 44 private funds.
sustainability for public pension schemes do 18. A World Bank pension database covering
not exist. 78 low-income countries revealed about
7. China has had a distinct scheme for workers 100 million contributors from a labor force
in public sector units, although the State of almost 900 million (about 11 percent).
Council in early 2015 announced its intention 19. Because of the way public DB pensions are
to merge civil service, public sector units, and treated in the fiscal accounts, this challenge
urban worker schemes. The system relies on of sustainability is true even during the period
occupational supplements to avoid a dramatic when reserves are being drawn upon, at least
fall in replacement rates for civil servants. to the extent that investments consist of
8. Although Vietnam has one pension scheme government bonds. When a public DB pension
for all workers, a case is made for including scheme invests in government debt, this is seen
it in this first group because of the practice as one part of government lending to another
of calculating benefits of the two groups and is netted out of general government debt.
of workers using a different methodology, To the extent that contribution revenues
though 2014 reforms seek to reduce the exceed benefit spending, this reduces the
disparity over time. budget deficit in that time period. Conversely,
9. Although civil servants participate in the when contributions are lower than spending,
same scheme as private sector workers, the general government deficits increase even
government contributes more than private where earmarked bonds are redeemed to pay
sector employers are required to contribute. pension benefits (see Goebel, van de Ven, and
10. In Japan, the systems for civil servants and Zwijnenburg 2015).
private school employees are to be merged 20. Mongolia also allowed workers to “buy
with the system for private employees on back” contribution years during the 1990s.
October 1, 2015. 21. Although pension coverage is almost universal
11. As of April 2015, a new social security law in Japan, average effective retirement ages are
that would cover all formal sector workers higher than age 60.
was still being discussed. 22. Data from 2011 for China are from CHARLS
12. Note that the median age of civil servants (2011).
tends to be higher than private sector workers. 23. This approach has been used in OECD and
13. The Philippines is an exception in this regard, Latin America and the Caribbean. They use
having passed a portability law in 1994 that prospective simulations for workers entering
makes it possible to combine service periods the scheme today, not what is actually paid
(Mesa-Lago, Viajar, and Castillo 2011). out today. The aim is to compare the implied
14. Mesa-Lago, Viajar, and Castillo (2011) also objectives of different designs.
advocate harmonization and integration of 24. The difference between gross and net
multiple schemes in the Philippines. replacement rates is that the net replacement
15. In addition, Korea’s mandated retirement rate takes taxation into account. A standardized
benefit system provides a lump sum payment set of assumptions was used for wage growth,
equivalent to 1/12th of annual salary per interest, and mortality rates. The figures are for
year of employment. However, this can men entering the system today with average
be withdrawn by the worker prior to wages throughout their careers and retiring
retirement and is therefore not defined here at normal retirement age. For DC schemes,
as part of the pension system. See the Korean simulations convert account balances into
law at http://www.moleg.go.kr/english annuities based on country- and sex-specific
/korLawEng?pstSeq=47471. mortality rates.
16. See the Thai Securities and Exchange 25. Because of the different life expectancies at
Commission’s website at http://capital.sec. retirement age across East Asian and Pacific
or.th/webapp/thaipvd/apps/pvdreports2 countries, the present value of the target
/stat12en.php. benefit, referred to as pension wealth, can
17. See those programs at Malaysia’s country be calculated. This method would change
profile at Wilmington PLC, http://www the rankings only slightly, with China
.pensionfundsonline.co.uk/content /country- becoming the most generous, followed by the
profiles/malaysia/102. Dorfman (2014) Philippines. Pension wealth for women would
PENSION SYSTEMS AND AGING: REFORMS FOR WINNING THE RACE 179
also generally be higher in DB countries compensation to late retirees than do all but a
because women live longer than men and may few OECD countries. Also see Giles, Hu, and
have lower retirement ages. Huang (2015) regarding East Asia and Pacific.
26. In Indonesia and Mongolia, reforms are being 35. Bloom, Canning, and Sevilla (2004) also find
implemented that represent a major departure that the expectation that longevity gains will
from current rules. In Indonesia, new lead to longer working lives is not borne out
parameters will certainly increase the overall by evidence.
replacement rate target reflected in figure 5.7. 36. The analysis relies on national-level data, so
In Mongolia, the replacement rates shown it is not possible to separate LFPRs for urban
here are much lower than those for workers and rural elderly for this sample. Controlling
born before 1960, with the change from one for the rural share in each country increases
cohort to the next being so dramatic that it R-squared significantly (0.66) and reduces the
has been cited as a potential problem. coefficient of the spending variable.
27. The guide to diversifying retirement income 37. There are, for example, various studies from
states, “Your EPF savings alone are not likely Italy, the United Kingdom, and the United
to be enough to support your retirement.” See States. The only study known to the authors
KWSP (2012, 5). of this report that includes East Asia and
28. The conditions for withdrawal are five years of Pacific is Shome and Saito (1980). There is
contribution and one month of unemployment. also evidence of crowding out when voluntary
29. The calculation of replacement rates in savings for retirement are provided tax
Singapore has added complexity owing to a incentives. See Kohl and O’Brien (1998) for a
second account that can be used for housing review of international studies.
or retirement and a third account that is 38. The authors also estimated a positive savings
used for medical services. Choon and Tsui effect for the period 1960–80 resulting from
(2012) show that replacement rates would be additional savings for longer retirement
substantially higher if housing were properly duration. The net effect was 25 to 30 percent
taken into account. displacement of savings.
30. As noted earlier, countries without social 39. See Onder and Pestieau (2014) for theoretical
pension programs include Cambodia, evidence.
Indonesia, Lao PDR, and Myanmar. Aside 40. The four countries are Chile (Coronado 1997;
from the subsidized food program in Fuentes 2013; Schmidt-Hebbel and Servén
Indonesia, national social assistance programs 1998); Colombia (Villar et al. 2013); Mexico
are almost nonexistent in these countries. (Villagómez and Antón 2013); and Peru
31. Indonesia provides free rice to more than (Secada, Cusato, and Zapata 2013).
half of households (although the new 41. In Thailand, the Government Pension Fund
administration has announced that the scheme was introduced in 1997 to partially replace
will be rationalized) and has a conditional cash the DB scheme for civil servants. In the same
transfer program for around 3 percent of the year, China introduced individual accounts
population. It also subsidizes health insurance that were nominally funded but soon became
for poor households. In the Philippines, more “empty” accounts.
than 20 percent of households benefit from 42. A concentration in government bonds,
a conditional cash transfer program, and especially nontradable bonds, can inhibit
a larger proportion is covered by a targeted potential effects of pension funds on capital
health insurance scheme. markets and savings. Both Malaysia and
32. These crude comparisons hide significant Singapore invest their provident funds in
differences in the scope of the wage bill to special government bonds. In Singapore,
which they are applied, including the low CPF purchases of nonmarketable bonds were
earnings ceilings in Hong Kong SAR, China, issued specifically to the CPF. In contrast,
and in Thailand. For the most part, however, Malaysia’s EPF holds state and corporate
the parameters do not affect the relative bonds, equities, and property, with only about
position of these economies. one-quarter in government bonds.
33. Japan’s current total contribution rate is 43. Preliminary projections based on the
16.7 percent but is scheduled to rise to 18.3 parameters being considered in Indonesia
percent by 2017. suggest that reserves could exceed 10 percent
34. D’Addio, Keese, and Whitehouse (2013) of GDP in 2030 and 25 percent of GDP by
find that Japan and Korea give higher 2050.
180 LIVE LONG AND PROSPER
44. The index focuses on the availability of Furthermore, although rules for portability
reliable information required for independent have been issued, the highly decentralized
monitoring as well as composition, system has no mechanism for ensuring this
independence, and qualifications of Board portability, which poses a major challenge in
members. the face of mass rural-to-urban migration.
45. Accountability is further promoted by 55. For example, Baldacci et al. (2010) estimated
competition between funds. The MPFA is that an increase in government spending on
especially concerned with the poor choice of pensions of 1 percent of GDP would raise
investment options and the lack of awareness permanent income by 1.25 percent of GDP.
about the impact of administrative fees, so it 56. A version of the proposal is still under
has embarked on an ambitious educational consideration in Mexico.
campaign in the past few years. 57. See Cai et al. (2012) for a discussion of this
46. For a recent example in Fiji, see Hofman and approach in the context of China’s rural
Chand (2012). pension scheme.
47. A key measure is the automatic adjustment
mechanism, which reduces pensions
depending on deviation from baseline
References
projections. Asher, Mukul G. 2012. “Singapore.” In Pension
48. In addition, disability pension reform has also Systems in East and Southeast Asia:
been important in some countries, such as Promoting Fairness and Sustainability, edited
Poland. by Donghyun Park, 85–98. Manila: Asian
49. As discussed in chapter 4, resistance to Development Bank.
increasing the retirement age often stems Asher, Mukul G., and Azad Singh Bali. 2013.
from concern about the effect on employment “Fairness and Sustainability of Pension
of younger workers (known as the “lump Arrangements in Singapore: An Assessment.”
of labor fallacy”), but regional and global Research Paper 13-21, Lee Kuan Yew School
evidence suggests that such concern is of Public Policy, Singapore.
misplaced. See, for example, Zhang (2012) Asher, Mukul G., and Amarendu Nandy. 2011.
for China. “Singapore: Pension System Overview and
50. Indonesia plans to integrate civil servants Reform Directions.” In Pension Systems
into a hybrid DB/DC pension scheme. and Old-Age Income Support in East and
However, the process may begin only with Southeast Asia, edited by Donghyun Park,
civil servants hired after 2028. Observers see 152–75. Oxon, U.K.: Routledge.
little justification for such a long transition Aterido, Reyes, Mary Hallward-Driemeier, and
period. Carmen Pagés. 2011. “Does Expanding
51. Despite this impressive achievement, even Health Insurance beyond Formal-Sector
after the coverage expansion, close to one- Workers Encourage Informality? Measuring
third of the self-employed in Korea, mostly the Impact of Mexico’s Seguro Popular.” Policy
low-income workers, do not regularly Research Working Paper 5785, World Bank,
contribute to the pension system. Washington, DC.
52. See Barrientos (2012) for a thorough Baldacci, Emmanuele, Giovanni Callegari, David
discussion. Coady, Ding Ding, Manmohan Kumar,
53. Aterido, Hallward-Driemeier, and Pagés Pietro Tommasino, and Jaejoon Woo. 2010.
(2011) document similar effects on informal “Public Expenditures on Social Programs
activity for noncontributory health insurance and Household Consumption in China.”
coverage in Mexico. Davis and Henrekson IMF Working Paper 10/69, International
(2004) present similar data from richer Monetary Fund, Washington, DC. http://
countries. www.imf.org / external/pubs/ft/wp/2010
54. The scheme suffers from some of the same /wp1069.pdf.
implementation problems as individual Barrel, Ray, Ian Hurst, and Simon Kirby. 2009.
accounts for formal sector workers. “How to Pay for the Crisis or Macroeconomic
Specifically, rates of return are set at bank Implications of Pension Reform.” National
deposit rates, resulting in poor incentive to Institute of Economic and Social Research,
contribute and low balances at retirement. London.
PENSION SYSTEMS AND AGING: REFORMS FOR WINNING THE RACE 181
Government of Korea. 2012. “Fund Management Kenneth Bercuson and Robert Carling, 42–51.
Report.” National Pension Service, Seoul. Washington, DC: International Monetary Fund.
Gruber, Jonathan. 1997. “The Incidence of Payroll IFLS (Indonesia Family Life Survey). 2007. RAND
Taxation: Evidence from Chile.” Journal of Corporation, Santa Monica, CA. http://www
Labor Economics 15 (3, pt. 2): S72–101. .rand.org/labor/FLS/IFLS.html.
Gruber, Jonathan, and David A. Wise. 1999. Iglesias, Augusto, and Robert Palacios. 2001.
“Social Security Programs and Retirement “Managing Public Pension Reserves: Evidence
around the World: Micro-estimation.” NBER from the International Experience.” In New
Working Paper 9407, National Bureau of Ideas about Old-Age Security: Toward
Economic Research, Cambridge, MA. Sustainable Pension Systems in the 21st Century,
Heckman, James, and Carmen Pagés, eds. edited by Robert Holzmann and Joseph Stiglitz,
2004. Law and Employment: Lessons from 213–53. Washington, DC: World Bank.
Latin America and the Caribbean. Chicago: IHME (Institute for Health Metrics and
University of Chicago Press. Evaluation). 2013. The Global Burden of
Herd, Richard, Hu-Wei Hu, and Vincent Koen. Disease: Generating Evidence, Guiding Policy.
2010. “Providing Greater Old-Age Security Seattle, WA: IHME.
in China.” OECD Economics Department ILO FACTS (International Labour Organization
Working Paper 750, Organisation for Economic Financial and Actuarial Services). 2013.
Co-operation and Development, Paris. http:// “Actuarial Valuation of the Public Pension
dx.doi.org/10.1787/5kmlh4x7pc7k-en. Scheme of the Vietnam Social Security Fund.”
Hinz, Richard, Robert Holzmann, David Tuesta, Social Protection Department, International
and Noriyuki Takayama, eds. 2013. Matching Labour Organization, Geneva.
Defined Contributions for Pensions: A Review IMF (International Monetary Fund). 1994.
of International Experience. Washington, DC: “Malaysia: Recent Economic Developments.”
World Bank. IMF, Washington, DC.
Hofman, Helene, and Satish Chand. 2012. “Fiji ———. 2011. “The Challenge of Public Pension
National Provident Fund Writes Off Millions Reform in Advanced and Emerging
in Member’s Funds.” Radio Australia, Economies.” Fiscal Affairs Department, IMF,
February 15. http://www.radioaustralia.net Washington, DC.
.au/international/radio/onairhighlights Jackson, Richard, Neil Howe and Tobias Peter.
/ fiji-national-provident-fund-writes-off 2013. The Global Aging Preparedness Index,
-millions-in-members-funds. Second Edition. Washington, DC: Center for
Holzmann, Robert, Robert Palacios, and Asta Strategic and International Studies.
Zviniene. 2004. “Implicit Pension Debt: Issues, JSTAR (Japanese Study of Aging and Retirement).
Measurement, and Scope in International 2011. http://www.rieti.go.jp/en/projects/jstar/.
Perspective.” Social Protection Discussion Karam, Philippe, Dirk Muir, Joana Pereira, and
Paper 0403, World Bank, Washington, DC. Anita Tuladhar. 2010. “Macroeconomic
Holzmann, Robert, and Edward Palmer, eds. 2006. Effects of Public Pension Reforms.” IMF
Pension Reform: Issues and Prospects for Working Paper 10/2, International Monetary
Non-financial Defined Contribution (NDC) Fund, Washington, DC. http://www.imf.org
Schemes. Washington, DC: World Bank. /external/pubs/ft/wp/2010/wp10297.pdf.
H o p f , G r e g o r. 2 0 0 9 . “ Wa s S i n g a p o r e Kashiwase, Kenichiro, Masahiro Nozaki, and Kiichi
Extraordinary? A Comparative View of Tokuoka. 2012. “Pension Reforms in Japan.”
Singapore’s Saving Performance 1965–99.” IMF Working Paper 12/285, International
HSBA Working Paper 02/2009, Hamburg Monetary Fund, Washington, DC.
School of Business Administration, Hamburg, KLoSA (Korean Longitudinal Study of Aging).
Germany. 2010. Korea Labor Institute, Seoul. http://
Hu, Yuwei. 2012. “Growth of Asian Pension www.kli.re.kr/klosa/en/about/introduce.jsp.
Assets: Implications for Financial and Capital Koettl, Johannes, and Michael Weber. 2012.
Markets.” ADBI Working Paper 360, Asian “Does Formal Work Pay? The Role of Labor
Development Bank Institute, Tokyo. Taxation and Social Benefit Design in the New
Husain, Aasim M. 1995. “Determinants of Private EU Member States.” IZA Discussion Paper
Saving in Singapore.” In Singapore: A Case 6313, Institute for the Study of Labor, Bonn,
Study in Rapid Development, edited by Germany.
PENSION SYSTEMS AND AGING: REFORMS FOR WINNING THE RACE 183
Kohl, Richard, and Paul O’Brien. 1998. “The ———. 2013. “Pensions at a Glance.” OECD,
Macroeconomics of Aging, Pensions, and Paris.
Savings: A Survey.” OECD Economics Onder, Harun, and Pierre Pestieau. 2014. “Is Aging
Department Working Paper 200, Organisation Bad for the Economy? Maybe.” Economic
for Economic Co-operation and Development, Premise 144, World Bank, Washington, DC.
Paris. Palacios, Robert. 2002. “Managing Public
Kugler, Adriana, and Maurice Kugler. 2009. Pension Reserves: Part II—Lessons from Five
“Labor Market Effects of Payroll Taxes Recent OECD Initiatives.” Social Protection
in Developing Countries: Evidence from Discussion Paper 0219, World Bank,
Colombia.” Economic Development and Washington, DC.
Cultural Change 57 (2): 335–58. ———. 2015. “Pensions in an Ageing East Asia
KWSP (Kumpulan Wang Simpanan Pekerja). and Pacific.” Background paper for the East
2012. “EPF Savings and Your Retirement.” Asia and Pacific regional report on aging,
KWSP, Kuala Lumpur. World Bank, Washington, DC.
Lu, Bei, Wenjiong He, and John Piggott. 2014. Palacios, Robert, and Charles Knox-Vydmanov.
“Should China Introduce a Social Pension?” 2014. “The Growing Role of Social Pensions:
Journal of the Economics of Ageing 4: History, Taxonomy and Key Performance
76–87. Indicators.” Public Administration and
Ma, Jun, Xiaorong Zhang, and Zhiguo Li. 2012. Development 34 (4): 251–64.
A Study of China’s National Balance Sheet. Palacios, Robert, and Edward Whitehouse. 2006.
Beijing: Social Sciences Academic Press. “Civil Service Pension Schemes around the
Mahalingam, Eugene, and Wei-Shen Wong. World.” Social Protection Discussion Paper
2013. “Malaysians Don’t Save Enough to 0602, World Bank, Washington, DC.
See Them through Their Retirement Years.” Rodan, Garry. 2014. “Singapore’s Vaunted CPF
Star, August 10. http://www.thestar.com.my under Fire.” Asia Sentinel, September 8.
/Business/Business-News/2013/08/10 /Saving http://www.asiasentinel.com/econ-business
-for-retirement-Malaysians-dont-save-enough /singapore-central-provident-fund-under-fire/.
-to-see-them-through-their-retirement-years/. Rofman, Rafael, Ignacio Apella, and Evelyn Vezza,
Mercer and Australian Centre for Financial Studies. eds. 2014. Beyond Contributory Pensions:
2015. Melbourne Mercer Global Pension Fourteen Experiences in Latin America.
Index, 2015 Report. Melbourne. http://www Washington, DC: World Bank.
.mercer.com/content/dam/mercer/attachments Samad, Suhaimi Abd, and Norma Mansor. 2013.
/global/Retirement/Melbourne-Mercer-Global “Population Aging and Social Protection in
-Pension-Index-2015/Report.pdf. Malaysia.” Malaysian Journal of Economic
Mesa-Lago, Carmelo, Verna Dina Q. Viajar, and Studies 50 (2): 139–56. http://ssrc.um.edu
Rolly Czar Joseph Castillo. 2011. Pensions in .my/images/ssrc/Books/Journal%20papers
the Philippines: Challenges and Ways Forward. /Population%20ageing%20and%20
Manila: Friedrich Ebert Stiftung. social%20protection%20in%20Malaysia
Moon, Hyungpyo. 2013. “Matching Contributions _suhaimi_norma.pdf.
and Compliance in Korea’s National Pension Schmidt-Hebbel, Klaus, and Luis Servén. 1998.
Scheme.” In Matching Contributions for Pensions, “World Saving: Trends and Theories.”
edited by Richard Hinz, Robert Holzmann, Estudios de Economía 25 (2): 191–215.
David Tuesta, and Noriyuki Takayama, 161–76. Schwarz, Anita M., and Omar S. Arias. 2014. The
Washington, DC: World Bank. Inverting Pyramid: Pension Systems Facing
MPFA (Mandatory Provident Fund Authority). Demographic Challenges in Europe and
2014. Annual Report. Hong Kong SAR, Central Asia. Washington, DC: World Bank.
China: MPFA. Secada, Pablo, Antonio Cusato, and Verónica
NPS (National Pension Service). 2012a. “Fund Zapata. 2013. “Efectos en Perú.” In
Management Report.” NPS, Seoul. Contribución del sistema privado de pensiones
———. 2012b. National Pensions Statistics: Silver al desarrollo económico de Latinoamérica,
Book. Seoul: NPS. edited by Rodrigo Acuña, 239–77. Medellín,
OECD (Organisation for Economic Co-operation Colombia: SURA Asset Management.
and Development). 2011. “Pensions at a Shome, Parthasarathi, and Katrine Saito. 1980.
Glance: Asia and the Pacific.” OECD, Paris. “Creating Capital through Social Security
184 LIVE LONG AND PROSPER
Institutions: The Asian Experience.” Domestic Wise, David A. 2004. “Social Security Provisions
Finance Study 61, World Bank, Washington, DC. and the Labor Force Participation of Older
Sluchynskyy, Oleksiy. 2015. “Defining, Workers.” Population and Development
Measuring, and Benchmarking Administrative Review Aging, Health, and Public Policy 30
Expenditures of Mandatory Social Security (Suppl.): 176–205.
Programs.” Social Protection Discussion Paper World Bank. 1994. Averting the Old Age Crisis:
1501, World Bank, Washington, DC. Policies to Protect the Old and Promote
Souto, Pablo, and Alberto R. Musalem. 2012. Growth. Washington, DC: World Bank.
“Assessing the Governance and Transparency ———. 2011. Mongolia: Policy Options for
of National Public Pension Funds.” Review of Pension Reform. Washington, DC: World
European Studies 4 (2): 148–67. Bank.
SSO (Social Security Office), Thailand. 2009. ———. 2012a. “Malaysia: Elderly Protection
Annual Report. Nonthaburi, Thailand: SSO. Study.” World Bank, Washington, DC.
UN (United Nations). 2013. “World Population ———. 2012b. “Reducing Elderly Poverty in
Prospects: The 2012 Revision, Highlights, Thailand: The Role of Thailand’s Pension and
and Advance Tables.” Working Paper ESA/P/ Social Assistance Programs.” World Bank,
WP228, Department of Economic and Social Bangkok.
Affairs, Population Division, UN, New York. ———. 2013. International Income Distribution
U.S. Social Security Administration. 2012. Social Database (I2D2). World Bank, Washington,
Security Programs throughout the World: Asia DC.
Edition. Washington, DC: U.S. Government ———. 2014. East Asia Pacific at Work:
Printing Office. Employment, Enterprise, and Well-Being.
Villagómez, Alejandro, and Arturo Antón. 2013. Washington, DC: World Bank.
“El caso de México.” In Contribución del Yamabana, Hiroshi. 2011. “Actuarial Studies
Sistema privado de pensiones al desarrollo of SSO Thailand.” Presentation of the
económico de Latinoamérica, edited by International Labour Organization Financial
Rodrigo Acuña, 105–80. Medellín, Colombia: and Actuarial Service, Bangkok, November 30.
SURA Asset Management. Yamada, Tetsuji, Tadashi Yamada, and Guoen
Villar, Leonardo, Jonathan Malagón, Julio César Liu. 1992. “Interdependency of Personal
Vaca, and Carlos Ruiz. 2013. “Experiencia Savings and Labour Force Participation
de Colombia.” In Contribución del Sistema of the Elderly, and Social Security Wealth:
privado de pensiones al desarrollo económico A Time Series Analysis.” Applied Economics
de Latinoamérica, edited by Rodrigo Acuña, 24 (4): 379–88.
55–104. Medellín, Colombia: SURA Asset Zhang, Chuanchuan. 2012. “The Relationship
Management. between Elderly Employment and Youth
Walker, Eduardo, and Fernando Lefort. 2002. Employment: Evidence from China.” MPRA
“Pension Reform and Capital Markets: Are Paper 37221, Munich Personal RePEc Archive,
There Any (Hard) Links?” Social Protection Munich, Germany.
Discussion Paper 0201, World Bank, Zhang, Chuanchuan, John Giles, and Yaohui
Washington, DC. Zhao. 2014. “Policy Evaluation of China’s
Wang, Lijiang, Daniel Béland, and Sifeng Zhang. New Rural Pension Program: Income, Poverty,
2014. “Pension Fairness in China.” China Expenditure, Subjective Wellbeing and Labor
Economic Review 28 (C): 25–36. Supply.” China Economic Quarterly 14 (1):
WHO (World Health Organization). 2010. Global 203–30.
Health Observatory Database. http://www.who Zuo, Xuejin. 2013. “Designing Fiscally Sustainable
.int/gho/mortality_burden_disease/life_tables/en/. and Equitable Pension Systems in China.”
Wickramanayake, Jayasinghe. 1998. “Savings and Paper presented at the International Monetary
Development.” Central Provident Fund and Fund Office for Asia and the Pacific–Fiscal
Savings in Singapore 22 (4): 435–54. Affairs Department Conference on Designing
Wiese, Patrick. 2006. “An Evaluation of Pension Fiscally Sustainable and Equitable Pension
Reform Options for Mongolia.” World Bank, Systems in Asia in the Post Crisis World, Tokyo
Washington, DC. January 9–10.
Population Aging and Health
Services in East Asia and Pacific 6
This chapter is based on a background paper prepared by Aparnaa Somanathan (2015) and a literature
review conducted by Harry Moroz, Naoko Miake, and Ying Ho.
185
186 LIVE LONG AND PROSPER
FIGURE 6.1 Health care in East Asia and Pacific faces major health system and service delivery challenges
People-
Limited emphasis on
provider Discontinuity of care
patient self-management
interaction
system where cost containment is a growing most economies (figure 6.2). For the majority
problem. Health care costs are a major of people living in the region, the first point
financial burden on both governments and of contact with the health system is usually
households. Moreover, the delivery of health at a hospital (around 53 percent in China).
services is characterized by inequalities, and Furthermore, throughout the region, rela-
quality of health services is poor. This section tively simple health conditions such as hyper-
discusses each of the three challenges in turn tension and uncomplicated diabetes result
and the way their effects are likely to be exac- in hospitalization. In China, according to
erbated in a rapidly aging population with a estimates, between 30 and 50 percent of
rising burden of NCDs. admissions could have been handled at an
ambulatory care setting (World Bank 2014a).
Yet, tertiary hospitals have registered higher
Weaknesses in delivery system design
growth for both inpatient and outpatient
Health systems in East Asia and Pacific are visits compared to primary and secondary
hospital centric. Hospital costs account for 40 hospitals (figure 6.3). As shown in table 6.1,
to 70 percent of total health expenditures in of all outpatient visits to central general
P O P U L AT I O N AG I N G A N D H E A LT H S E R V I C E S I N E A S T A S I A A N D PAC I F I C 187
hospitals in Vietnam, 35 percent could have FIGURE 6.2 Hospital share of health expenditures is often greater
been handled at provincial hospitals and than nonhospital share
20 percent at district hospitals. Of all out- Health expenditures by level of care, selected economies, 2012
patient visits to provincial general hospitals,
41 percent could have been handled at district
Australia
hospitals.
High average lengths of stay further con- Bangladesh
tribute to the hospital share of total spending.
China
As shown in figure 6.4, average lengths of
stay in China, Japan, the Republic of Korea, Taiwan, China
and Mongolia are higher than the average in
Hong Kong SAR, China
Organisation for Economic Co-operation and
Development (OECD) countries. Japan
The hospital bias is even stronger among
Korea, Rep.
the elderly. Health care utilization rates
among the elderly are dominated by hospi- Malaysia
tal inpatient care. Household survey data
Mongolia
from China, Indonesia, Japan, and Vietnam
(figure 6.5) show that inpatient admission New Zealand
rates increase with the mean age of house-
Sri Lanka
hold from about age 50 for both men and
women, with a spike after age 65 in all Thailand
countries. Analysis for China also shows
0 20 40 60 80 100
that inpatient admissions for older house-
holds are far more likely to be in general and Share of health expenditures (%)
township hospitals and that most outpatient Hospitals Providers of ambulatory care Other providers
care is sought at hospitals. Outpatient utili-
zation rates are much lower for the elderly Source: Jeong and Rannan-Eliya 2010.
TABLE 6.1 Hospital patients in Vietnam are being treated at higher levels than is necessary
Treatment of patients in hospitals in Vietnam, 2010 (percent)
FIGURE 6.4 Average length of stay in East Asian and Pacific acute The hospital-centric nature of care provi-
care hospitals is higher than or comparable to OECD averages sion combined with weak and poor-quality
Average length of stay (days) in acute care hospitals, selected economies primary care has resulted in fragmented
health systems throughout East Asia and
20
18
Pacific. Few of the region’s health systems
16 have formalized links between hospitals and
14 ambulatory providers, including structured
12
referral systems, patient discharge and hando-
Days
10
8 ver mechanisms, and patient outreach services
6 (McCollum et al. 2014). In almost all cases,
4
2 referral mechanisms exist on paper but not
0 in practice. Policies requiring patients to pay
much higher fees for self-referrals have had
C . ( 0)
ew anm (20 )
ala ( 2 )
H o u n S i n b od ( 2 0 )
R , m 1)
In Nep (20 )
i L (2 )
M CD ng a (2 0)
Ch ( 2 )
pu M e t n (20 )
ng uss e ( 11)
ala ( 0)
lad ( 2 )
(2 )
in (20 )
(2 8)
ne (2 )
B a a i l a ( 20 )
8)
SA ntr (20 )
ng ei D ga ia ( 09)
Ca s t r a ( 2 0 )
a N y am 08
m lia 10
in 10
R, ies 11
ng nd 10
Sr e s h 0 0 5
Gu ar 03
ka 0 8
Z e Fiji 0 0 8
do a l 1 0
T h y s i a 011
Vi ina 009
ao ou ia 11
d 1
Re 0 1
SA ala 201
OE Mo hin 201
M si a 0 1
ea 0
00
A u n 01
Ch (20
K o a r p o r 20
an 0
a c c ol 0
a, (2
a
p
Ko J a p
w
Ne
Pa
Br
Weaknesses in people-provider
interactions
Source: OECD Health Data 2012 database, http://dx.doi.org/10.1787/888932723266.
Note: OECD = Organisation for Economic Co-operation and Development. Data for the Republic of
Korea refer to average length of stay for inpatient care (including long-term care beds in hospitals).
Fragmented health care delivery systems
merely exacerbate an overly acute and epi-
sodic pattern of health care use by the elderly.
However, quality of care in primary care Poor older people with minor ailments in
facilities is poor, particularly with regard to Cambodia, Indonesia, and Vietnam reported
staff skills and competencies. Clinicians at seeking care from a range of practitioners such
the community level lack the skills to pro- as private health practitioners, local pharma-
vide effective diagnosis and treatment because cists, and traditional practitioners instead
of inadequate training and limited access to of government primary health care centers,
continuing medical education (box 6.1). Staff which they found to be less convenient and
turnover is high in primary care settings as a unreliable. Notably, in most East Asian and
result of poor living conditions in rural areas, Pacific countries, annual health check-ups
which further impedes quality of care. In a and screenings are infrequent. Countries such
survey in Singapore, older people complained as China (Strauss et al. 2012) and Thailand
they saw a different doctor on each visit. (Knodel, Prachuabmoh, and Chayovan 2013)
P O P U L AT I O N AG I N G A N D H E A LT H S E R V I C E S I N E A S T A S I A A N D PAC I F I C 189
a. China b. Indonesia
40
40
30
30
20
20
10
10
0 0
25–40 41–50 51–60 61–70 71–80 81+ 25–40 41–50 51–60 61–70 71–80 81+
c. Japan d. Vietnam
20 40
18
16
30
Inpatient rate (%)
Inpatient rate (%)
14
12
10 20
8
6
10
4
2
0 0
25–40 41–50 51–60 61–70 71–80 81+ 25–40 41–50 51–60 61–70 71–80 81+
have made efforts to encourage annual physi- East Asian and Pacific countries, the chroni-
cal exams and outpatient visits by ensuring cally ill tend to have multiple contacts with the
they are covered by insurance. In Thailand, delivery system, such as diagnostic services,
over half of those ages 60 and above reported specialists, emergency rooms, outpatient clin-
having an annual physical exam, with little ics, and pharmacies. Moreover, as discussed
variation by gender, age, or area of residence, in chapter 1 of this report, the elderly tend to
a trend associated with high levels of outpa- have multiple chronic conditions and comor-
tient care coverage provided by the Universal bidities, and the lack of integration of care
Coverage Scheme.1 However, a study of pri- means that eventually, they seek care from
mary health care in Cambodia, Indonesia, multiple providers for multiple conditions.
Singapore, and Vietnam found that annual Without coordinated care approaches to help
health exams were rare and that the elderly the elderly navigate the delivery system, fol-
cited cost as the major reason for not seeking low the course of their diseases, and provide
them (Dodge 2008). continuous management and follow-up, they
Fragmentation is also directly associated often do not receive effective or timely care.
with discontinuity of care and duplication of This inadequacy results in acute complica-
services, particularly for those—such as the tions, repeat visits, or duplicated care—often
elderly—with chronic conditions. In most for the same illness episode—and higher costs
190 LIVE LONG AND PROSPER
BOX 6.1 Poor qualifications of staff to provide effective diagnosis and treatment
for individuals, insurers, and government efforts are unconnected. This issue is dis-
(World Bank 2014a). cussed further in chapter 7.
An added challenge specific to the elderly is
integrating emerging long-term care arrange-
Weaknesses in the organizational
ments with the health delivery system, which
setting
can be institutional, community-based, or
home-based care or care provided by fam- Distorted provider incentives leading
ily members or formal aged care workers. In to overprovision
most East Asian and Pacific countries, sepa- Providers in many East Asian and Pacific
rate ministries are responsible for health and countries face strong incentives to overpro-
nonhealth services for the elderly, and their vide services because fee-for-service (FFS)
P O P U L AT I O N AG I N G A N D H E A LT H S E R V I C E S I N E A S T A S I A A N D PAC I F I C 191
payment methods are widely used without and prioritization at present (Leung, Tin, and
adequate measures of volume and price con- Chan 2007).
trol. Admittedly, FFS provider payment sys- The distortion of provider incentives
tems are used in Japan; Korea; and Taiwan, and consequent overprovision of expen-
China, and they have been successful in con- sive services is driving up costs, particularly
trolling cost growth because FFS is used in because strategic purchasing capacity is lim-
combination with global budgets and a uni- ited in most East Asian and Pacific econo-
fied fee schedule. The significant inflationary mies. Strategic purchasing capacity refers to
effects of FFS systems are well established the ability of the purchaser (insurer, ministry
in Canada; the Czech Republic; parts of the of health, and so on) to incentivize provid-
Russian Federation; and the United States ers to improve quality processes and results
(Langenbrunner and Somanathan 2011). In while contributing to cost containment (that
China and Vietnam, the FFS approach is used is, reducing readmissions and unnecessary
in a context where health facilities essentially admissions, long stays, repeat procedures,
operate as commercial enterprises, with an and adverse events). In most systems financed
emphasis on self-financing of facilities (for by general revenues (Cambodia; Hong Kong
example, sales or business income comprised SAR, China; Malaysia; and most Pacific
over 60 percent of hospital revenue in both Island states), the purchaser and provider are
countries in 2010). The link between hospi- essentially the same (the ministry of health),
tal business revenue and physician income which leaves little room for the development
has encouraged overservicing, unnecessary of strategic purchasing capacity. Even in
care, and inefficiency in service produc- insurance-based systems (China, Indonesia,
tion (Somanathan et al. 2014; World Bank Mongolia, the Philippines, and Vietnam),
2014a). social insurance agencies have yet to estab-
In addition, the lack of appropriate mea- lish the informational, managerial, and finan-
sures to prioritize among new and available cial accounting capacities to exercise robust,
health interventions has resulted in rapid arm’s-length service purchasing and monitor-
growth in health technologies and phar- ing of providers, whether public or private. In
maceuticals. Distorted pricing for treat- many cases, the fragmentation of governance
ments has given health care providers strong and administration to subnational levels has
incentives to generate demand for profit- further weakened the purchasing power of
able high-technology services and drugs in the insurance agency.
place of less-profitable basic alternatives. In
China and Vietnam, hospitals’—and there- Insufficient focus on chronic disease
fore doctors’—revenues are enhanced by fees management and population health
earned through prescriptions and medical Primary health care (PHC) facilities tend to
tests and procedures and are supplemented focus on treatment rather than health promo-
by kickbacks from pharmaceutical compa- tion and education. Health promotion and
nies (Somanathan et al. 2014; Yip and Hsiao education interventions are crucial for the
2008). In China, doctors’ prescribing behav- management of NCDs, particularly among the
ior has been found to vary with the insurance elderly. In Cambodia, Indonesia, Singapore,
status of patients and the incentives offered and Vietnam, elderly survey respondents
by pharmaceutical companies (Currie, Lin, reported that PHC providers do not spend
and Meng 2014; Lu 2014). More generally in time discussing behavioral changes such as
the region, few measures are in place to pri- healthy diets and exercise (Dodge 2008).
oritize and restrict the diffusion of new tech- Moreover, the readiness of PHC facilities
nologies and their incorporation into routine to manage chronic NCD conditions is limited.
care. In many cases, the underlying health Box 6.2 describes the limited readiness of pub-
care system is practically or philosophically lic sector primary care facilities (puskesmas)
incompatible with mechanisms for evaluation in Indonesia to diagnose and treat three key
192 LIVE LONG AND PROSPER
BOX 6.2 Limited supply-side readiness to manage and treat noncommunicable diseases at
the primary health care level in Indonesia
TABLE 6.2 Significant shares of elderly are experiencing catastrophic out-of-pocket payments for health care
relative to total household spending
Percentage of households by age groupings with health spending of 25 percent or more of total household expenditure
a. China
Household mean age
25–49 (%) 50–69 (%) 70 and above (%) Total (%)
Sector
Rural 9.36 12.48 18.57 11.23
Urban 8.21 10.13 13.55 9.38
Total 8.78 11.35 15.82 10.30
Gender
Male 8.03 10.10 14.64 9.34
Female 9.44 12.53 16.78 11.16
Total 8.78 11.35 15.82 10.30
Sector and gender
Rural men 8.75 11.87 17.91 10.61
Rural women 9.91 13.08 19.13 11.80
Urban men 7.33 8.14 11.83 8.06
Urban women 8.99 11.96 14.90 10.54
Total 8.78 11.35 15.82 10.30
b. Indonesia
Household mean age
25–49 (%) 50–69 (%) 70 and above (%) Total (%)
Sector
Rural 1.60 2.86 3.56 1.85
Urban 1.82 3.58 4.16 2.06
Total 1.70 3.15 3.76 1.95
Gender
Men 1.60 3.15 3.98 1.84
Women 1.80 3.15 3.65 2.04
Total 1.70 3.15 3.76 1.95
Sector and gender
Rural men 1.57 2.79 3.64 1.79
Rural women 1.63 2.92 3.53 1.91
Urban men 1.63 3.67 4.80 1.90
Urban women 1.98 3.51 3.90 2.20
Total 1.70 3.15 3.76 1.95
c. Japan
Respondent’s age
Gender 50–69 (%) 70 and above (%) Total (%)
Men 0.25 0.12 0.20
Women 0.25 0.48 0.33
Total 0.25 0.31 0.27
d. Thailand
Household mean age
Sector 25–49 (%) 50–69 (%) 70 and above (%) Total (%)
Rural 1.48 2.52 3.68 1.88
Urban 1.55 2.92 4.61 2.04
Total 1.50 2.65 3.95 1.94
table continues on next page
194 LIVE LONG AND PROSPER
TABLE 6.2 Significant shares of elderly are experiencing catastrophic out-of-pocket payments for health care
relative to total household spending (continued)
Percentage of households by age groupings with health spending of 25 percent or more of total household expenditure
e. Vietnam
Household mean age
25–49 (%) 50–69 (%) 70 and above (%) Total (%)
Sector
Rural 1.84 4.16 5.88 2.37
Urban 1.59 3.76 6.78 2.00
Total 1.75 4.06 6.09 2.25
Gender
Men 1.64 4.10 6.22 2.10
Women 1.90 4.02 5.97 2.43
Total 1.75 4.06 6.09 2.25
Sector and gender
Rural men 1.68 4.28 5.64 2.16
Rural women 2.05 4.04 6.08 2.63
Urban men 1.54 3.58 8.01 1.97
Urban women 1.64 3.94 5.60 2.03
Total 1.75 4.06 6.09 2.25
Sources: CHARLS 2011; IFLS 2007; JSTAR 2011; ThaiSES 2011; VHLSS 2012.
TABLE 6.3 Significant shares of elderly are experiencing catastrophic out-of-pocket payments for health care
relative to total nonfood household spending
Percentage of households by age groupings with health spending of 25 percent or more of total nonfood expenditure
a. China
Household mean age
Sector 50–69 (%) 70–79 (%) 80 and above (%) Total (%)
Rural 33.41 33.16 40.83 34.00
Urban 29.62 32.78 25.70 30.57
Total 31.69 32.96 33.57 32.33
b. Indonesia
Household mean age
Sector 50–69 (%) 70–79 (%) 80 and above (%) Total (%)
Rural 15.86 18.50 17.94 16.74
Urban 12.15 14.79 4.57 12.29
Total 14.44 17.24 12.83 15.09
c. Vietnam
Household mean age
Sector 50–69 (%) 70–79 (%) 80 and above (%) Total (%)
Rural 27.99 29.98 35.20 29.88
Urban 26.70 25.57 53.01 29.91
Total 27.72 28.98 38.60 29.88
Sources: CHARLS 2001; IFLS 2007; VHLSS 2012.
P O P U L AT I O N AG I N G A N D H E A LT H S E R V I C E S I N E A S T A S I A A N D PAC I F I C 195
Health care access for the elderly is also health system delivery costs and weak cost-
characterized by large inequalities between containment measures. Despite increases in
rich and poor elderly. Household surveys from central government subsidies of 1 to 2 percent
China, Indonesia, Japan, and Vietnam show a of gross domestic product (GDP) during the
two- to threefold difference between the health past decade or so, payers (insurance orga-
care utilization rates of the richest 20 percent nizations and local governments) and pro-
of the population compared to the poor- viders are increasingly under financial stress
est 20 percent, with the exception of Japan. because benefits packages are overly gener-
In Thailand, civil servants who have more ous and not fully financed in many cases.
generous coverage use outpatient care two to Meanwhile, as discussed earlier, the com-
three times as often and inpatient care four to bination of distorted provider incentives,
five times as often as the remaining popula- lack of strategic purchasing capacity, and
tion. Earlier analysis has shown that higher weaknesses in the design of the service deliv-
household per capita expenditure is positively ery system make cost containment a major
associated with use of inpatient and outpatient challenge. As costs rise, providers increas-
care in China (Strauss et al. 2012). ingly resort to cost recovery from patients,
Thus, effective health coverage—people such as the common practice of extra bill-
who need health care receive it in a timely ing for services in the Philippines and Vietnam.
and affordable manner and at an accept- With or without regard to aging, current
able level of quality—remains elusive for the trends in service delivery patterns and costs
elderly as much as or more than for the over- will not only increase the OOP burden on
all population. The only exceptions are the households, but also undermine the sustain-
high-income East Asian and Pacific countries. ability of public financing for health. Even
However, affordability of care is not the only without population aging, political commit-
issue; a growing body of evidence indicates ments toward universal health coverage and
that higher levels of insurance coverage do the consequent expansion in health care cov-
not always lead to use of appropriate, good- erage are expected to lead to quite substantial
quality services (Currie, Lin, and Meng 2014; increases in the demand for health care. Such
Somanathan et al. 2014). increases are likely to put additional pressure
on health systems that are already supply con-
strained. Moreover, NCDs pose a new set of
Universal health coverage in challenges to health systems because of the
the context of population aging different models of care delivery required to
and NCDs manage chronic diseases.
Coverage—defined in terms of enrollment Adding an aging population to this toxic
in universal coverage programs and ben- mix of cost drivers is widely expected to lead to
efit levels—has increased quite significantly, rapid cost escalation. Exactly how large will the
accompanied by an expansion in the over- aging effect be relative to other drivers of costs?
all supply of health services. The drive for The next section examines this in more detail.
universal health coverage, backed by large
increases in public spending on health, is
the single most unifying feature of health
Population aging and health
policy in East Asia and Pacific today. Rapid
expenditures in East Asia
increases in enrollment coverage occurred in
and Pacific
five of the largest countries in the region since Overall health spending is relatively high
the mid-1990s. Benefit packages have also in East Asia and Pacific and has risen quite
been expanded to include chronic conditions sharply in recent years, fueling concerns
and outpatient care. about fiscal sustainability, particularly in
Despite increased coverage levels, OOP the context of aging. This section examines
payments remain high because of rising whether aging itself or other cost pressures
196 LIVE LONG AND PROSPER
FIGURE 6.6 Health share of GDP has been rising in East Asia and exhibit a J-shape: per capita spending is
Pacific with income growth somewhat high at very young ages, declines
Total and public expenditure on health as a percentage of GDP, 1995–2012 and remains stable for most of the prime-age
period, and increases again from about age
7.0
55 for men and age 60 for women. Figure 6.7
6.5 illustrates this age-expenditure profile in the
6.0 case of Japan. Given this trend, the frequent
Share of GDP (%)
that the age profile of health expenditures will FIGURE 6.7 Age-specific expenditure profile exhibits J-shape
remain constant when life expectancy increases. for Japan
Decreases in age-specific mortality rates may Health cost per capita by age relative to average index = 100, 2001, 2011,
lead to decreases in age-specific costs, because and 2012
declining mortality rates reduce the proportion
500
of high-cost users (that is, those close to death).
In figure 6.7, the rightward shift of the J-curve 450
+
50 49
60 9
70 69
10 9
20 19
40 39
80 79
almost negligible (figure 6.8). The moder-
30 29
25 4
55 4
75 4
4
15 14
35 4
45 4
65 4
4
85
–5
5–
–2
–5
0–
–3
–4
–6
–7
–8
–
–
–
–
–
–
ately old orange countries (China, Indonesia,
Malaysia, Mongolia, Thailand, and Vietnam) Age group
will have to absorb about 0.28 percent more 2001 2011 2012
of the population dying annually by 2040.
Source: MHLW 2011.
The very old red economies (Hong Kong SAR,
China; Japan; Korea; and Singapore) face the
biggest challenge, with 0.36 percent more of FIGURE 6.8 Spikes in death rates vary with the speed of aging
across East Asia and Pacific
the population dying annually by 2040.
Projected percentage of deaths in population by country grouping,
various years
Health status and life expectancy
If proximity to death is a key driver of costs,
0.60
then so is the progression of illness and dis-
ability leading up to death. If the bulk of 0.50
health care costs is incurred within a few
months or years of severe illness before death 0.40
and if this timespan shrinks and shifts over
time to ever-higher ages (morbidity compres- 0.30
Percent
cohorts of 10 to 20 years ago (figure 6.9). The never been a major driver of health care
middle-age cohort of today has higher rates spending (Oxley and MacFarlane 1994).
of hypertension and diabetes and higher rates Growth in health spending has been driven
of smoking and exposure to other risk fac- by other demand effects (per capita income
tors. Therefore, one can reasonably expect growth and extension of coverage by
morbidity to expand as the current middle- national health systems), but primarily by
age cohorts reach older ages. Putting in place supply factors (technological progress, infla-
health policies and strategies that take a tion, and supply-induced demand effects in
whole-of-life approach to healthy aging will a sector where information asymmetries are
help curb these trends in the future. pervasive). These non-demographic drivers
are considered below.
As national income or wealth increases,
The main cost drivers: Non-demographic
both expectations and health spending rise,
effects on health expenditures
regardless of changes in needs (Getzen 2000).
The idea that aging has a significant effect East Asian and Pacific governments’ priori-
on health expenditures is often referred to tization of universal health coverage goals
as a “red herring” because it distracts from and willingness to accompany those policy
more important underlying cost drivers commitments with large increases in public
(Zweifel, Felder, and Meiers 1999). From spending in recent years, especially in China,
a historical perspective, demography has Indonesia, the Philippines, and Vietnam, are
evidence of this rise. In these countries, rapid
economic growth during the 1990s and early
FIGURE 6.9 Years lived with disability are increasing in East Asia 2000s was associated with rising expecta-
and Pacific as life expectancy increases tions and demand for greater access to health
Percentage change in years lived with disability for the cohort, by gender, care and with governments’ acknowledge-
ages 55 to 59, 1990–2010 ment of the need to address those demands by
increasing public spending on health, as evi-
Japan denced by elasticities of spending of greater
Singapore than or equal to one in almost all countries
Korea, Rep.
China (figure 6.10).
Taiwan, China All other non-demographic drivers aside
Cambodia from income are referred to as excess cost
Indonesia
Lao PDR growth—essentially, the excess of growth
Malaysia in per capita health expenditures over the
Myanmar growth in per capita GDP after controlling
Philippines
Sri Lanka for the effect of demographic change. This
Thailand approach captures the effects of two cost
Timor-Leste
Vietnam
drivers: technological change and policy
Oceania choices regarding key levers of health financ-
Micronesia, Fed. Sts. ing. The available evidence on the effect of
Fiji
Kiribati these two drivers and the significance for East
Marshall Islands Asia and Pacific are discussed below. A third
Papua New Guinea element of excess cost growth worth noting
Samoa
Solomon Islands is Baumol’s cost disease: low-productivity
Vanuatu increases in medical care relative to other
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 less labor-intensive sectors shift the relative
% change in years lived with a disability prices of medical care upward (Baumol 1996,
Men Women 2012).
Technological change has accounted for
Source: IHME 2013. the bulk of medical care cost increases over
P O P U L AT I O N AG I N G A N D H E A LT H S E R V I C E S I N E A S T A S I A A N D PAC I F I C 199
time (Newhouse 1992; Smith, Newhouse, FIGURE 6.10 Total and public health spending has increased
and Freeland 2009). Recent estimates sug- more rapidly than GDP in much of East Asia and Pacific in the 1990s
gest that medical technology explains 27 to and 2000s
48 percent of growth in health care spend- Elasticity of total and public health spending to GDP, selected East Asian
and Pacific countries, 1995–2012
ing in OECD countries since 1960 (Smith,
Newhouse, and Freeland 2009). Although
technological change often contributes to Timor-Leste
efficiency improvements, in recent years, new
technologies have been dominated by costly Lao PDR
product innovations to alleviate disease rather
Myanmar
than to improve efficiency or strengthen pre-
ventive care. Expensive end-stage cancer Vietnam
drugs that do little to cure the disease but may
improve outcomes temporarily are one exam- Thailand
ple. This trend has also contributed to the Papua New Guinea
growth in death-related costs (Baumol 2012;
Zweifel 2003). In countries where cost con- China
trol mechanisms are weak and efforts to rein
in technology growth are limited, technologi- Indonesia
cal change can lead to rapid cost escalation Mongolia
in association with aging. The contribution of
the costly product innovations to per capita Philippines
expenditures at different ages, rather than
Malaysia
aging itself, is the driver in health spending.
Rising incomes and expectations will likely 0.0 0.5 1.0 1.5 2.0
add further pressure to technology growth.
Elasticity
As noted earlier in the chapter, the absence
of appropriate measures to prioritize the dif- Total Government
fusion of new technologies combined with Source: World Bank 2015.
distorted FFS-driven incentives have already
resulted in rapid growth in health technology
spending in East Asia and Pacific. In China, not only the introduction of new mechanisms
for instance, the medical device market grew for evaluation and prioritization (discussed in
from US$8 billion in 2006 to US$20 billion the next section), but also the political will to
in 2011—a compound annual growth rate withstand demand pressures for new technol-
of 20.1 percent, exceeding even the pace of ogies. In many countries, the elderly are more
total health expenditures (US$156 billion to vocal politically than others and may drive up
US$357 billion, compound annual growth rate the adoption of new technologies.
of 18.0 percent). Box 6.3 discusses the role of As shown by Hagist and Kotlikoff (2005),
technological change and interactions with in addition to costly product innovations, the
aging in the escalation of health care costs. contribution of benefit levels at different ages
In East Asia and Pacific, efforts to rein can drive up spending. Because health care
in the growth of technologies are critical for benefit levels (health care expenditures per
controlling health expenditure growth, partic- beneficiary at a given age) are much higher
ularly in the context of an aging population. for the elderly than for the young, benefit
Health policy needs to focus on ensuring that levels increase as the proportion of older age
health technologies are adopted on the basis groups in the population increases. However,
of evidence and need and are diffused appro- considerable heterogeneity exists across
priately and that the costs of adoption are countries in health spending at older ages.
managed carefully. This approach will require As figure 6.11 shows, in Austria, Germany,
200 LIVE LONG AND PROSPER
BOX 6.3 The role of technological change and interactions with aging in health care cost
escalation
Technological change—the adoption and utilization The combination of population aging and
of new pharmaceuticals and medical technologies—is advances in medical technology provides the rationale
not necessarily problematic. For instance, Cutler and to spend more, with more medical technology being
McClellan (2001) found that the use of new technol- oriented toward the elderly (Dormont, Grignon, and
ogy helped increase the average coronary patient’s Huber 2006; Jones 2002). A Dutch study that looked
life expectancy by one year (valued at US$70,000 at utilization trends and the state of medical technol-
per case) while treatment costs increased US$10,000 ogy (medical patents) found that large advances in
per case (4.2 percent per year), for a net benefit of medical technology were significantly associated with
US$60,000 per case. Whether a particular technolog- higher growth in the probability of hospitalization,
ical development increases or decreases costs depends particularly for the older population (Wong et al.
on its effect on unit cost, its level of use, and the like- 2011).
lihood that the treatment complements or replaces
existing methods.
FIGURE 6.11 Age-specific expenditure profiles vary for OECD Spain, and Sweden, per capita expenditures
countries for those ages 75 to 79 and 80 and above are
Health spending by age relative to reference group 50–64 = 1, selected only twice the level of per capita expenditures
countries, latest available year, early 2000s of the reference age group—ages 50 to 64. In
contrast, the oldest-old (ages 80 and above)
14 in the United States receive benefits that aver-
age 8 to12 times those of the reference group,
Relative health spending, reference group 50–64 = 1
TABLE 6.4 Effective strategies to reduce exposure to key NCD-related risks in East Asia and Pacific
TABLE 6.5 Tobacco prices and taxes in East Asia and Pacific are not high enough
TABLE 6.5 Tobacco prices and taxes in East Asia and Pacific are not high enough (continued)
TABLE 6.6 Effective strategies for the management of key risks such as cardiovascular disease
Areas needing
intervention Recommended strategy
Screening for and • Improved diagnosis rates of hypertension and high cholesterol
management of high • Treatment with an essential package of four readily available and off-patent medicines:
blood pressure and high aspirin, beta blockers, other antihypertensives such as angiotensin-converting enzyme
cholesterol inhibitors or calcium channel blockers, and statins
Screening for and • Improved diagnosis rates of diabetes in the population
management of diabetes • Lifestyle interventions focused on diet and exercise patterns, including group-based
activities
• Drug therapy (metformin)
Screening for cancer risks • Screening for and treatment of precancerous lesions to prevent cervical cancer
and early treatment • Raising of awareness among women, promotion of self-exams, and mammograms for those
with elevated risks
Self-management • Patient education delivered in small groups by health professionals
• Peer education programs to improve diet and physical activity
• Structured phone support and telemonitoring
• Online self-management programs and resources
Medication adherence • Reduced out-of-pocket costs of medications
• Introduction of programs such as reminder or calendar packaging combined with education
and reminder strategies (Mahtani et al. 2011); use of mobile phones to send out reminders
(Ettenhofer et al. 2009); and provision of new pill bottlesa
a. New technologies such as pill bottles that glow if pills are not taken are designed to increase adherence rates by combating inattention. A randomized
control trial is underway to evaluate their effectiveness (U.S. National Institutes of Health, Clinical Trials Registry, http://www.clinicaltrials.gov/ct2/show
/NCT01798784).
It should be considered a necessary strategy affordable under existing insurance and tax-
today in responding to NCDs, in addition to financed programs. Because of inadequate
other preventive measures. Proactive screen- health financing coverage, many elderly may
ing, prompt treatment, and regular moni- go without the needed drugs.
toring by qualified medical professionals, Effective diagnosis and management of
combined with patient self-management, are diabetes are limited in the region for simi-
all important. lar reasons. The complications arising from
Yet such practices remain limited in East untreated diabetes require acute medical
Asia and Pacific because of inadequate screen- care interventions (for example, dialysis and
ing and monitoring in the health system, amputations) and result in severe disabilities.
limited efforts at self-management, and high The evidence suggests that effective screening
OOP costs. The available evidence points to and monitoring, early diagnosis and manage-
high proportions of all of these conditions ment of diabetes with lifestyle interventions,
going undiagnosed in most countries in the and improved self-management can bring
region, and the likelihood of being diagnosed enormous cost savings. Even in cases where
correlates with socioeconomic status and drugs are required to control blood sugar lev-
education. Among those who are diagnosed, els, the costs are higher but still significantly
medication adherence is low. Moreover, most less than the costs of treating the disease when
countries have limited supply-side capacity to it progresses further. As an example, the costs
screen people and carry out such secondary of treating diabetes in Vanuatu are presented
prevention activities effectively (Mendis et al. in table 6.8.
2012). Although aspirin, antihypertensives,
and statins are included in the standard pack- Cancer
age of services in many countries (table 6.7), As the proportion of elderly increases, the
their high cost means that many patients must majority of East Asian and Pacific countries
pay out of pocket. Such OOP payments occur face the prospect of rapid increases in new
even though more efficient procurement cancer cases over the next 10 to 20 years, not
of these drugs could make them far more least because too little focus on or investment
TABLE 6.7 Status of availability of essential package of cardiovascular drugsa in selected East Asian and Pacific
countries
Are drugs included in the standard benefits Are drugs free of charge in the public
Country package? sector?
Indonesia Drugs are included in the benefits package for Yes, in principle, but drugs are not always
National Health Insurance. available.
Malaysia Yes Yes
Mongolia Yes If obtained at primary care level, drugs are
discounted through Health Insurance Fund.
Philippines Department of Health provides treatment packs Yes, no copayments are charged.
(ComPacks) for hypertension and diabetes mellitus,
but when the new health insurance benefit package
takes effect, ComPacks will be discontinued.
Sri Lanka Yes Yes
Thailand Yes Yes
Tonga Yes Yes
Vanuatu Yes Yes
Vietnam Drugs are included in principle, but they not always No, co-payments are charged (≥ 5 percent
available, particularly at primary care level. depending on type of insurance).
a. The package includes four key drugs: aspirin, beta blockers, other antihypertensives such as angiotensin-converting enzyme inhibitors or calcium
channel blockers, and statins for reducing LDL (low-density lipoprotein) cholesterol.
P O P U L AT I O N AG I N G A N D H E A LT H S E R V I C E S I N E A S T A S I A A N D PAC I F I C 207
in cancer prevention and screening has increasing age, female gender, number of
occurred to date (figure 6.12). Cancer screen- diagnoses, recent hospitalization, and depres-
ing and early diagnosis and treatment are sion (Elliott 2006). In countries where cover-
particularly important in aging populations. age is not universal, OOP payments for drugs
However, most countries in the region have pose the biggest threat to medication adher-
not yet put effective strategies in place. ence; evidence shows that OOP payments, no
matter how small, reduce adherence to medi-
cation regimes (Sidorenko and Zaidi 2013).
Improving self-management and
Anecdotal evidence from Mongolia shows
medication adherence
that medication for hypertension is often
Improving self-management of conditions is taken only for a high reading of blood pres-
an integral part of managing the progression sure and not otherwise. A range of interven-
of chronic conditions, particularly in older tions to improve medical adherence through
populations. The ability to self-manage—that self-management is also summarized in
is, the patient and his or her family take a more table 6.6.
proactive role in contributing to the patient’s
health outcome in partnership with the health
Transforming the service delivery
care providers—is affected by access to health
system
care, education, socioeconomic status, and
especially health literacy.4 In the United States Transforming service delivery to better meet
and Australia, 70 to 80 percent of older peo- population needs calls for stronger care coor-
ple are estimated to be at risk for low health dination throughout the health system. The
literacy, while in the United Kingdom, one in reformed model would anticipate and shape
three have problems reading and understand- patterns of care according to the projected
ing basic health information (Renehan et al. health and medical needs of the popula-
2012). Self-management interventions for tion while placing considerable emphasis on
chronic disease that are targeted at the elderly strengthening the role and raising the quality
have proved effective in developed countries of primary care. It would include significant
and are summarized in table 6.6. strengthening of community-focused care and
Efforts to improve medication adherence vertical and horizontal integration (physical
are particularly important for the manage- layout, information, and communication) of
ment of hypertension and high cholesterol. facilities to provide comprehensive services
In aging populations, cognitive decline and along a continuum or chain of care. The
polypharmacy (having to take four or more model also would encompass formal outreach
medications) usually lead to poor medication to long-term care providers5 as well as atten-
adherence. For instance, older Australians tion to acute care management and rehabilita-
take nine medications on average. The pre- tion and palliative care needs for populations
dictors of multiple medication use include with a growing proportion of elderly.
208 LIVE LONG AND PROSPER
FIGURE 6.12 The rising tide of cancer in East Asia and Pacific
Projected percentage of new cases of cancers in the elderly, ages 65 and above, selected East Asian and Pacific countries
and country groupings, 2012–35
Cambodia
Singapore
Vietnam
Micronesia, Fed. Sts.
Malaysia
Vanuatu
Phillippines
Korea, Rep.
Mongolia
China
Myanmar
Thailand
Papua New Guinia
Indonesia
Lao PDR
Fiji
Samoa
Solomon Islands
New Zealand
Australia
Timore-Leste
Japan
Strong primary care is the backbone of primary care would be the supervisor for
an effective delivery system and is the plat- directing patients through the delivery system,
form for active coordination of care. Clear as shown in figure 6.13. In OECD countries,
evidence exists that primary care contributes the primary and secondary prevention inter-
to better health outcomes, higher-quality ventions described earlier (pharmacological
and more appropriate care, improved access, management of hypertension and high cho-
better prevention and early management of lesterol and screening and management of
diseases, and cost containment (Macinko et diabetes) are increasingly located at the pri-
al. 2010; Macinko, Starfield, and Shi 2003; mary level. Some countries have fostered new
Saltman, Rico, and Boerma 2006; WHO family medicine and general practice models
2008). In a reformed service delivery sys- to facilitate their central role in coordinated
tem that emphasizes coordination of care, care programs. Others, such as Germany,
P O P U L AT I O N AG I N G A N D H E A LT H S E R V I C E S I N E A S T A S I A A N D PAC I F I C 209
FIGURE 6.13 A health system is transformed from a pyramidal model into a primary care–driven
coordinated care model
Diagnosis
Primary
Secondary Hospital Hospital
health center
Nursing
Primary Materials
home
Transport
incentivize patients to register with primary design features of the reformed service deliv-
care providers. ery model advocated in this report.
Coordination of care involves linking pro- Coordination of care emerged in the
fessionals and organizations at all levels of the OECD countries in response to pressures
health system, emphasizing patient-centered similar to those currently faced by many
care integration, managing patient referral East Asian and Pacific countries. These
through the delivery system, and promoting pressures included cost inflation, growing
follow-up care and the continuity of long- burden of chronic conditions, aging popu-
term service provision. Table 6.9 lists the key lations, technological change, and rising
210 LIVE LONG AND PROSPER
population expectations (Curry and Ham healthy life expectancy, and age at exit from
2010; Hofmarcher, Oxley, and Rusticelli the workforce.
2007; OECD 2010, 2012, 2013). Box 6.4 The good news is that a range of interven-
outlines five of the main coordinated care tions can help slow functional decline among
strategies adopted in OECD countries. The the elderly and save costs (table 6.10). For
strategies are aimed at containing costs by instance, despite increases in aging, vision has
reducing unnecessary use, frequent hospital- improved over time in Spain, Sweden, and
ization, inappropriate admissions, and avoid- the United States as a result of effective and
able readmissions. Box 6.5 summarizes the timely interventions (Christensen et al. 2009).
evidence to date on the effect of coordination Interventions to prevent falls among the
of care on key health system outcomes. elderly have been shown to produce positive
public health benefits such as reduced costs
of care and hospitalization, and treatment of
Managing acute care episodes more
osteoarthritis is known to improve quality of
effectively
life and functional capacity (Renehan et al.
Even as the emphasis is correctly placed on 2012; WHO 2014b).
shifting the balance of care from treatment Functional decline and NCD-related dis-
to prevention and management of disease, abilities will mean increased demand for reha-
the need to better manage the consequences bilitation and longer-term follow-up care,
of CVD should not be ignored. Treatment which makes effective coordination of care
of the consequences of CVDs accounted for critical. In East Asia and Pacific at present,
40 to 50 percent of the reduction in mortal- the majority of these services are provided
ity in OECD countries. Acute medical treat- in acute care hospital settings. This pattern
ment of the consequences of CVD will always of care is not only inefficient but also ineq-
be needed, particularly in aging populations. uitable, because hospitals are used mainly
Efforts must be made to ensure adequate by those who can afford such care. Effective
staffing; well-equipped facilities; and adop- coordination of care can provide for post-
tion of new, cost-effective treatment technolo- discharge planning and follow-up care for
gies such as clot-dissolving drugs that are part chronic disease patients with complex needs
of high-quality and effective cardiovascular and reduce the likelihood of hospital readmis-
care. sions. By ensuring that chronically ill elderly
patients receive the follow-up care they need
in a primary care setting or at home instead of
Preparing for functional and cognitive
at the hospital, countries achieve cost savings,
decline among the elderly
and the patient’s quality of life is improved.
The demographic shift in East Asia and Likewise, demand for palliative or end-of-
Pacific will give rise to many of the degen- life care for cancers and other painful chronic
erative diseases of older age, highlighting the conditions will increase, requiring coordi-
importance of introducing timely and effec- nation of care between acute care and pal-
tive interventions to slow the progression of liative care settings. In high-income countries,
these conditions and ameliorate their effects. aggressive therapy beyond the point of a rea-
These diseases include the sensory organ sonable chance of recovery is often expensive
disorders of old age such as worsening eye- and unnecessary and not always consistent
sight (cataracts and macular degeneration with the wishes of the patients themselves.
emerge around age 60) and hearing, as well Although high-cost aggressive therapies to
as weakening of bones and joints leading to treat cancer and other chronic conditions
osteoarthritis and osteoporosis. Although not at the end of life are not a feasible option
affecting mortality directly, these conditions for most people in middle- and low-income
can have important effects on disability levels, East Asian and Pacific countries, spending
P O P U L AT I O N AG I N G A N D H E A LT H S E R V I C E S I N E A S T A S I A A N D PAC I F I C 211
BOX 6.4 Care coordination strategies in Organisation for Economic Co-operation and
Development countries
Five main coordinated care strategies observed multidisciplinary team, is responsible for locat-
in Organisation for Economic Co-operation and ing, monitoring, and coordinating care based on
Development member countries are outlined below. a care plan.
Some overlap occurs in program features across cat- • Community-based day centers for frail elderly,
egories, and the list of strategies and initiatives is not which refers to special programs for frail elderly
exhaustive. Many countries have launched coordinated and patients with multiple and usually disabling
care efforts, but documentation is limited in terms of conditions. One-stop shops provide a full range of
describing the what and how of these programs. health and social services and operate somewhat
like geriatric day wards. Services are provided by
Shifting care out of the hospital to ambulatory set-
a multidisciplinary team of nurses, physicians,
tings for shared management of patients with chronic
nutritionists, social workers, therapists, and so
diseases
on and follow a care plan. Programs usually work
• Relocation of clinicians from hospitals to primary closely with home support and assisted living ini-
care settings. tiatives that aim to keep the elderly living in their
• Incorporation of specialists into multidisciplinary communities rather than in institutional care.
primary care teams: in some countries, this
Fostering transitional care practices
method is referred to as a multidisciplinary team
approach to care management. • Postdischarge planning and follow-up care for
• Educational outreach in which hospital-based spe- chronic patients with complex needs. These pro-
cialists provide instruction and advice on screen- grams center on patients with multiple conditions
ing, referrals, and treatment. who are discharged from hospitals. The model
provides comprehensive medication manage-
Shifting care from hospital and ambulatory settings ment, step-down rehabilitation, self-management
to the home education for patients and families, and timely
• Telemedicine and communication outreach between outpatient follow-up with primary care provid-
outpatient clinics or hospitals and households. ers. Hospital and primary care providers work
Interventions involve e-consultations, e-prescriptions, together to address clinical and social issues that
e-diagnostics and monitoring, and e-reminders (the can affect the patient’s recovery. The objective is
United Kingdom and the United States). to reduce readmissions and emergency room visits.
• Outreach home visits and hospital-at-home pro- Establishing accountable primary care organizations
grams, which consist of home visits by nurses or
community health advisers for both chronic dis- • Patient-centered medical homes, defined as “an
ease management and health promotion. This enhanced model of primary care that offers whole-
approach can be used in combination with tele- person, comprehensive, ongoing, and coordinated
medicine and electronic outreach interventions patient and family-centered care” (Kaye, Buxbaum,
(for example, Sweden, the United Kingdom, and and Takach 2011). Initiated in the United States,
the United States) (Health Foundation 2011). medical homes consist of an organized group of
ambulatory providers that furnish comprehensive
Targeting specific diseases or cases care to individuals, including the elderly and those
requiring complex care. Patients register with a
• Disease management programs that focus on primary care practice or center, which is respon-
disease-specific objectives, treatment goals, and sible for coordinating care, navigating the delivery
referral criteria for a defined population (for system, and arranging care as needed. Providers
example, diabetic patients). Programs generally are financially accountable for furnishing quality
target high-cost and high-risk patients with one care while containing costs. They generally present
chronic condition. Interventions are oriented to a comprehensive menu of services, including diag-
the entire course of a specific disease and usu- nostics and pharmacotherapies.
ally are founded on evidence-based guidelines.
A single care coordinator, usually part of a Source: World Bank 2013a.
212 LIVE LONG AND PROSPER
BOX 6.5 The effect of care coordination on key health system outcomes
Coordination of care has the following effects on key was conducted in person rather than by phone
health system outcomes. (Sochalski et al. 2009).
• When compared with the National Health Service
Positive effect on patient functional status and out-
(NHS) in the United Kingdom, Kaiser Permanente
comes, but ambiguous effect on mortality
performs better: of 11 medical conditions studied,
• The effect is positive, especially when mul- the United Kingdom’s NHS uses 3.5 times the
tiple strategies such as case conferences, care number of bed days as Kaiser for those ages 65
plans, and funding were used to coordinate care and above (Pike and Mongan 2014).
(Ouwens et al. 2005; Powell-Davies et al. 2008).
Mixed or insignificant effect on costs
• Kaiser Permanente members also experience more
comprehensive and convenient primary care ser- • A review of systematic reviews and meta-analysis
vices and much more rapid access to specialist of the effectiveness of care programs in the United
services and hospital when required (Pike and States and in the United Kingdom and other parts
Mongan 2014). of the European Union found mixed or insignifi-
cant evidence of integrated care resulting in cost
Positive effect on efficiency through reduced savings (OECD 2012; Ouwens et al. 2005).
hospitalizations • Medicare randomized trials showed that costs
• In Medicare randomized trials, programs in were generally higher for patients enrolled in
which case managers had substantial direct inter- the coordination arm of the trial compared with
action with physicians and significant in-person patients in the control arm, although these dif-
interaction with patients were more likely to ferences were not statistically significant in most
reduce hospital admissions than programs with- instances (Peikes et al. 2009).
out those features, but no program was likely to • Overall, evidence suggests that the effect on costs
reduce overall costs (Peikes et al. 2009). depends on the types of diseases being treated
• Review of effect of chronic care management and whether these conditions are associated with
found that patients with heart failure were less high-cost unnecessary treatment or with under-
likely to be hospitalized when care was pro- treatment before the integrated care initiatives
vided by a multidisciplinary team and when (RAND Europe and Ernst & Young 2012).
communication between providers and patients Source: OECD 2012.
TABLE 6.10 Effective strategies for slowing functional decline among the elderly
their last few days or months in hospital is Assessment instruments that create
often the only option available because hos- person-focused, longitudinal records that can
pice and palliative care services barely exist. be viewed, understood, and used across the
Eventually, systems in the region will need to continuum of care are vital for ensuring con-
consider expanded coverage for hospice and tinuity of care for the elderly. The use of the
palliative care, as well as enhanced case man- same protocols to support care planning in
agement and web-based tools and informa- sectors between which people may move (for
tion for end-of-life care. Good palliative care example, acute, community, and residential
would include the use of assessment tools (see care) allow for a collaborative approach to
below) that allow patients and their families service provision that spans the continuum of
to decide on the level of comfort versus cure care. The focus is on recording information
at the end of life (Lancet 2014). about the individual and his or her chang-
Delaying the onset of cognitive function ing strengths, abilities, and preferences (for
disorders and managing their symptoms will example, palliative care preferences) rather
also require attention through integrated than the information required only by a par-
health and social networks. There are few ticular care setting where that person hap-
effective treatments for dementia, although pens to be at any given time. Currently, the
the reduction of CVD risk factors has been interRAI suite of assessment instruments is
identified as having a protective effect (WHO the only example of this type. This integrated
2012). Current expenditures on dementia- system enables transfer of common informa-
related costs already amount to 1 percent of tion across care settings, thereby facilitating
GDP worldwide and 1.24 percent of GDP continuity of care across transitions. A num-
in high-income countries (Wimo and Prince ber of OECD countries are now introduc-
2010). Much of the care is provided through ing the interRAI instruments in response to
family and social networks, but these are growing pressures on care delivery services
straining with economic development. As (OECD and European Commission 2013).6
a result, ensuring that interventions are put The transformation of service delivery
in place to diagnose early, delay onset, and described above is likely to require substan-
manage symptoms through integrated health tial additional resources, particularly with
and social networks is even more important. regard to health workforce requirements. As
Similarly for depression, limiting psychosocial discussed, the current primary care system
stressors through support and long-term care in most East Asian and Pacific countries is
will become imperative. far from adequate to carry out the kinds of
Stronger coordination of care is critical for functions described and will need complete
improving access to care and saving costs in rebuilding. Strengthening the capacity of
relation to cognitive function disorders, as and adequately resourcing the primary care
well. In mental health services, a number of level will be essential for fostering and sus-
OECD countries have seen a recent shift from taining coordinated care approaches. Many
hospital-based care to coordinated commu- countries in the region such as China and
nity care (for example, Australia, the Nordic Vietnam have organizational roadmaps in
countries, and the United Kingdom) (OECD place for strengthening primary care as well
2013). Coordinated mental health care and as reform pilots, but scaling up of small-scale
integrated treatment have led to improved reforms is yet to begin. Staff skills will need
service delivery and better clinical outcomes. to be improved, and in some cases, a whole
Following this approach, fewer readmissions, new cadre of staff members such as general
reduced use of intensive care services, and practitioners and geriatricians may need to be
contact with community crisis teams have created.
resulted in cost savings (Rosenheck, Resnsick, The health infrastructure may eventually
and Morrissey 2003; Stewart et al. 2012; be transformed, too. For the service delivery
Woods and McCollam 2002). model described earlier, smaller hospitals and
214 LIVE LONG AND PROSPER
intermediate facilities (ambulatory surgical allocations to the health sector. Health system
units, day hospitals, and specialty clinics) and inefficiencies are already exerting pressure
larger primary care centers combined with on tax and social health insurance revenues.
greater use of communication technologies Service delivery reforms are thus critical to
for community, long-term care, and home ensure the sustainability of health systems
outreach would be more appropriate. As is for aging populations and justify additional
the case in several OECD countries, hospitals investment.
themselves may need to adapt more flexible In the short to medium term, significant
designs that respond to changing demand and efficiency savings could be gained in three
technological change (Rechel et al. 2009). All areas: (a) paying providers, (b) purchasing
of these changes will require substantially pharmaceuticals, and (c) prioritizing among
higher spending on health. available interventions and technologies.
Stronger purchasing capacity is a cross-
cutting issue. The key strategies in these areas
Getting better value for money from
are summarized in table 6.11.
health systems
Strengthening provider payment methods
Because mobilizing new resources to meet all to align providers’ incentives with deliver-
of the additional requirements of service deliv- ing good-quality, effective care is essential in
ery reforms would simply not be sustainable, East Asia and Pacific. Optimal systems do
generating efficiency savings to gain better not exist, and provider payment systems and
value for money from existing health sys- incentives need to be designed to address the
tems is imperative in the region. Although the specific policy issues and objectives inherent
reformed service delivery model will generate in a country’s health sector. However, regional
significant efficiency savings over time, the and global experience provide a few key
initial investment costs are likely to be sub- trends that the region’s countries can draw
stantial. Meanwhile, other policy priorities— on, as described below.
not least of which is the policy commitment For physician services, three provider
to achieve universal health coverage—will payment models tend to dominate: salary,
continue to place huge demands on East capitation, and FFS, or some combination
Asian and Pacific governments’ budgetary of the three. The specific FFS model used in
TABLE 6.11 Effective strategies for getting better value for money from health systems
Area needing intervention Recommended strategy
Paying providers • Movement away from line-item budgets and fee-for-service payments
• Introduction of capitation payment systems for physicians and primary care providers
and global budgets with case mix–adjusted payments for hospitals
• Consideration of bundled payments to providers for specific medical conditions (for
example, diabetes)
Purchasing pharmaceuticals • Reform of drug procurement and movement toward smart strategic purchasing
• Risk-sharing arrangements, such as price-volume arrangements, for high-cost drugs
• Sole-source tendering in which the winning bidder is sole supplier for a fixed term
• Therapeutic reference pricing
Prioritizing among available • Institution of clear, transparent processes to make decisions about new technologies
interventions and technologies and drugs
• Use of the findings from more established health technology assessment (HTA)
agencies to inform their own prioritization process (for example, de facto HTA)
Ensuring strategic purchasing • Movement toward a single purchaser
capacity • Building of strategic purchasing capacity
• Investment in health information management systems
P O P U L AT I O N AG I N G A N D H E A LT H S E R V I C E S I N E A S T A S I A A N D PAC I F I C 215
Japan; Korea; and Taiwan, China, has been payment categories within the hospital sec-
successful in avoiding the cost inflation asso- tor. In Western Europe, most countries have
ciated with FFS payment in many settings. moved to a performance-based approach using
Capitation payment has emerged as more some combination of payment per admission
effective in controlling costs, particularly based on case mix–adjusted diagnosis-related
in relation to primary care (see box 6.6 for groups. Turkey adopted a global budget for
Thailand’s experience with capitation). all ministry hospitals, which has helped con-
For inpatient care services, countries trol spending growth (see box 6.6).
have been moving toward some combina- High-income countries are now experi-
tion of global budgets and case mix–adjusted menting with bundled payments to providers
BOX 6.6 Cost containment through provider payment and purchasing reforms in Thailand
and Turkey
A single purchaser, virtually have continued to rise. The cost per member rose
from B 1,201.40 in 2002 to B 2,693.50 in 2011,
The National Health Security Office (NHSO) is the
or from a total of less than B 60 billion in 2002 to
largest purchaser in Thailand, purchasing services
over B 120 billion in 2011. This is equivalent to a
on behalf of 75 percent of the population (approxi-
70 percent increase in real terms over the period. The
mately 50 million Universal Coverage Scheme mem-
increase is attributed to rising remuneration of health
bers). NHSO has adequate strategic purchasing
care staff, particularly a rapid increase in extra incen-
capacity and has used its bargaining power to reduce
tive payments to keep highly skilled professionals in
the price of medicines, medical products, and inter-
the system.
ventions over time. For instance, the price of hemo-
dialysis decreased from US$67 to US$50 per cycle,
Performance-based supplementary payment system
leading to savings of US$170 million per year.
During the 2000s, Turkey undertook comprehen-
Capitation payment and diagnosis-related groups
sive hospital reforms in the context of the Health
with a global budget
Transformation Program, which was geared toward
The combination of mixed payment methods (capita- improving effectiveness, equity, and efficiency. A key
tion for outpatient care and diagnosis-related groups reform was the introduction of a performance-based
for inpatient care) with a global budget for inpatient supplementary payment system for employees of
care has further contributed to cost containment in Ministry of Health hospitals beginning in 2004,
Thailand. The closed-end payment system has pro- based on specified individual and institutional crite-
vided strong signals to providers to contain costs by ria. The changes in hospital financing were facilitated
prescribing generic medicines, dispensing medical by reforms that brought together all previous health
technologies appropriately, and encouraging pre- insurance schemes under the umbrella of the Social
ventive treatments. At the same time, to counteract Security Institution. Furthermore, expenditure caps
underprovision of services, NHSO has established for ministry, private, and university hospitals have
procedures such as (a) a 24-hour complaint manage- been introduced in recent years to ensure controlled
ment mechanism operated by NHSO staff, (b) quality growth in hospital spending in line with gross domes-
assurance through hospital accreditation processes, tic product. In 2007, a fixed global budget for all
(c) routine auditing by random medical audit teams ministry hospitals was implemented, which has been
with financial penalties, (d) utilization reviews to successful in containing further spending growth in
monitor utilization rates, and (e) an annual poll sur- these hospitals. Since 2010, the government has also
vey of consumers’ and providers’ satisfaction con- successfully maintained spending with agreed limits
ducted by independent polling institutes. for university and private hospitals.
Despite those measures, one should note that
overall Universal Coverage Scheme expenditures Source: Hanvoravongchai 2013; World Bank 2013b, 2013c.
216 LIVE LONG AND PROSPER
for specific procedures or medical conditions, pays for any volume over the agreed threshold
an approach that holds potential for aging (for example, Korea); sole-source tendering in
populations (Mandeville and Sinnott 2014; which the winning bidder is the sole supplier
Painter, Burns, and Bailit 2012). Under this for a fixed term; and therapeutic reference
method, fixed payments cover a set of services pricing in which the purchasers set fixed reim-
for a defined time period, usually including bursement levels for the price of a drug by ref-
both inpatient and outpatient care. Bundled erencing a base drug within that therapeutic
payments for chronic conditions such as dia- class (Mandeville and Sinnott 2014). Box 6.7
betes have created incentives for primary and compares purchasing strategies in Korea, Sri
secondary care providers to move care out of Lanka, and Thailand with the highly ineffi-
the hospital. High-quality primary care has cient purchasing arrangements in Vietnam.
prevented expensive inpatient stays for avoid- Strong purchasing capacity is as critical for
able complications. undertaking smart purchasing of drugs as it
Finally, the presence of a large purchaser is for undertaking provider payment reforms.
with adequate strategic purchasing capacity When new technologies or drugs become
has generally proved critical for cost contain- available, most countries—particularly
ment. Being a large, preferably single pur- those with social health insurance–financed
chaser gives the purchaser greater bargaining systems—grapple with decisions about
power to negotiate prices with providers and including them in the benefits package and
to control utilization and quality. at what level of patient cost sharing. As dis-
Inefficiencies in purchasing pharmaceu- cussed, the differential adoption and dif-
ticals result in high prices being paid for fusion of costly product innovations in the
drugs, suggesting opportunity costs in terms health sector in OECD countries explains the
of improving cost-effectiveness and reducing increase in per capita costs of elderly across
co-payments. Table 6.12 compares the prices countries, and not aging itself. Costs have
of common statins in selected countries. For been contained far more effectively in coun-
example, for off-patent common statins (ator- tries such as Japan that have been more suc-
vastatin and simvastatin), the price in Vietnam cessful in rationing the growth of technology
is more than twice that in New Zealand. compared to the United States, for example.
East Asian and Pacific countries can learn Ideally, decisions regarding new technologies
from pharmaceutical purchasing strategies and drugs should be based on objective cri-
that have been used in countries within and teria such as cost-effectiveness, efficacy, and
outside the region to reduce drug prices. These potential effect on the status of budgets and
strategies include risk-sharing arrangements health insurance funds and should be deter-
for high-cost drugs such as price- volume mined through a transparent process. In
arrangements, whereby the manufacturer practice, decision making is a priority-setting
BOX 6.7 The power of smart purchasing: Pharmaceutical procurement in the Republic of
Korea, Sri Lanka, and Thailand versus Vietnam
exercise that involves value judgments and is This approach is a systematic and transparent
influenced by a mix of political, social, and appraisal and deliberation process for making
economic factors. decisions on public reimbursement of medical
Many OECD countries have introduced technologies, devices, and procedures. It can
explicit prioritization systems—health technol- be very effective; for instance, New Zealand’s
ogy assessments (HTAs)—that transparently PHARMAC (Pharmaceutical Management
assess the value for money of new technologies. Agency) has been highly successful at keeping
218 LIVE LONG AND PROSPER
FIGURE 6.14 PHARMAC has positively affected drug expenditure over time in New Zealand
Actual drug spending versus projected drug spending at 2000 subsidy levels, 2000–2015
3,500
Forecast
Drug cost (millions of New Zealand dollars)
3,067
3,000 2,895
2,635
2,500 2,336
2,000 1,788
1,590
1,436
1,500 1,248
1,099
898 993
1,000 844
671 723
517 586
694 706 777 784 804 822
500 640 653
517 516 504 510 536 567 566 602
0
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
Fiscal year (ending June 30)
Actual expenditure Estimated expenditure at 2000 subsidies
costs down over a 12-year period by combin- services and setting co-payment rates for
ing HTAs with strengthened pharmaceutical different types of services, providers, and
procurement (figure 6.14) (Mandeville and patients. Thailand’s Universal Coverage
Sinnott 2014). Scheme and Korea’s National Health
In East Asia and Pacific, two strategies Insurance program provide good examples
can be pursued to support the prioritization of this process (box 6.8).
of new technologies and drugs: institute a
transparent process for prioritization and use Use findings from more established HTA
findings from more established HTA agen- agencies to inform prioritization processes
cies to inform prioritization processes. These Kosovo, Romania, and Serbia looked to
approaches can be followed even if HTA the experiences of other HTA agencies. For
capacity has not yet been established because example, they used a de facto HTA method by
developing national capacity to undertake drawing on recommendations of the United
HTAs is costly and can take several years. Kingdom’s HTA agency (National Institute
for Health and Clinical Excellence) to assess
Institute a transparent process for the cost-effectiveness of their national for-
prioritization mularies. Using this method, Serbia was able
This strategy typically involves a committee to renegotiate prices on a range of drugs and
consisting of, among others, providers and save millions of euros.
researchers who are charged with making
recommendations based on clear and agreed-
Preparing communities and institutions
upon criteria such as cost- effectiveness,
outside of the health system
medical necessity, and financial burden on
patients. The committee and decision process Several interventions outside of the formal
can be used not only for inclusion of new health care delivery system can help pre-
interventions, but also for delisting existing pare society for the emerging needs of an
P O P U L AT I O N AG I N G A N D H E A LT H S E R V I C E S I N E A S T A S I A A N D PAC I F I C 219
FIGURE B6.8.1 National Health Security Office’s decision process for new technologies involves
extensive criteria in Thailand
Recommendations
increasingly older population (table 6.13). For often have invested time and money and who
instance, ensuring ready availability of assis- hold a detailed understanding of systems and
tive devices and aids such as upper and lower practices—and attract others will have a com-
limb prostheses, assistive devices for toilet petitive advantage. An important factor in
use and bathing, walking sticks and frames, increasing the motivation and thus retention
wheelchairs, lifting hoists (for lifting people of mature workers will be the work environ-
in and out of bed, for example), and adap- ment itself. Even if workers have the ability
tive transport are key items that can assist in and motivation to work, a poorly adapted
mobility and increase independence of older workplace culture will discourage retention
people with disabilities. In addition, installing of mature workers (Eide, Krause, and Rosas
environmental features (for example, wheel- 2001; Seike, Biggs, and Sargent 2011).
chair ramps; wheelchair-friendly transport,
homes, offices, and buildings; ample seating;
and safe footpaths) can enable older people Conclusion
with disabilities to circulate within their com- The range of recommended policy measures
munities (WHO 2014b). Finally, improve- could help East Asian and Pacific countries
ments in public spaces could encourage address their health care challenges in a fis-
walking and other physical activities, which cally sustainable manner and improve out-
not only reduce the risk of chronic disease, comes for the elderly. As discussed earlier,
but also exert protective effects by strengthen- even without the effects of aging, the expected
ing the physiological systems of older adults increase in disease burden attributable to
and reducing functional limitation such as NCDs and the inefficiencies in the financ-
that from osteoarthritis (Beard et al. 2011). ing and delivery of health services are likely
Equally important are changes to organi- to drive up health care costs and undermine
zational culture in the workplace to accom- fiscal sustainability. The various measures
modate an increasingly older workforce. This proposed in this chapter would help address
includes the social environment within busi- NCD risks in a highly cost-effective manner
nesses, institutions, and service agencies. As and improve the effectiveness and efficiency
older workers form a larger proportion of of health care service delivery, leading to bet-
a shrinking workforce, businesses that can ter health outcomes for both the elderly and
keep these existing workers—in whom they the broader population. At the same time,
TABLE 6.13 Effective strategies outside of the health system to prepare for functional and cognitive decline
in old age
pressure on the health care system could also ———. 2012. The Cost Disease: Why Computers
be relieved by reducing the current default Get Cheaper and Health Care Doesn’t. New
reliance on health systems for long-term care Haven, CT: Yale University Press.
of the elderly, which is discussed in detail in Beaglehole, Robert, Ruth Bonita, Richard Horton,
Cary Adams, George Alleyne, Perviz Asaria,
chapter 7 of this report.
Vanessa Baugh, Hank Bekedam, Nils Billo,
Sally Casswell, Michele Cecchini, Ruth
Colagiuri, Stephen Colagiuri, Tea Collins,
Notes Shah Ebrahim, Michael Engelgau, Gauden
1. Interestingly, the percentage of elderly who Galea, Thomas Gaziano, Robert Geneau,
reported having an exam was considerably Andy Haines, James Hospedales, Prabhat Jha,
higher among the one-fourth of those ages Ann Keeling, Stephen Leeder, Paul Lincoln,
60 and older who participated in an elderly Martin McKee, Judith Mackay, Roger
club during the past year, especially at more Magnusson, Rob Moodie, Modi Mwatsama,
advanced ages. Sania Nishtar, Bo Norrving, David Patterson,
2. M. L. Sang provided information based on Peter Piot, Johanna Ralston, Manju Rani, K.
research carried out for preparation of the Srinath Reddy, Franco Sassi, Nick Sheron,
Vietnam Health Professional Education and David Stuckler, Il Suh, Julie Torode, Cherian
Training project. Varghese, and Judith Watt. 2011. “Priority
3. Members of the EU-27 are Austria, Belgium, Actions for the Non-communicable Disease
Bulgaria, Cyprus, the Czech Republic, Crisis.” Lancet 377 (9775): 1438–47.
Denmark, Estonia, Finland, France, doi:1016/S0140-6736(11)60393-0.
Germany, Greece, Hungary, Ireland, Italy, Beard, John R., Alexandra Kalache, Mario
Latvia, Lithuania, Luxembourg, Malta, the Delgado, and Terry Hill. 2011. “Aging and
Netherlands, Poland, Portugal, Romania, the Urbanization.” In Global Population Ageing:
Slovak Republic, Slovenia, Spain, Sweden, Peril or Promise?, edited by John R. Beard,
and the United Kingdom. Simon Biggs, David E. Bloom, Linda P. Fried,
4. Health literacy refers to an individual’s Paul Hogan, Alexandra Kalache, and S. Jay
capacity to seek, understand, and utilize health Olshansky, 93–96. Geneva: World Economic
information to make informed decisions Forum.
about his or her own health (U.S. Department Bloom, David E., Elizabeth T. Cafiero, Mark
of Health and Human Services 2000). E. McGovern, Klaus Prettner, Anderson
5. The service delivery reforms proposed here Stanciole, Jonathan Weiss, Samuel Bakkila,
draw extensively from work on service and Larry Rosenberg. 2013. “The Economic
delivery reform in China led by the World Impact of Non-communicable Disease in
Bank from 2013 to 2015. The final report of China and India: Estimates, Projections, and
this study will be published in 2016 by the Comparisons.” NBER Working Paper 19335,
World Bank. National Bureau of Economic Research,
6. For more information, see the interRAI Cambridge, MA.
website at http://interrai.org/index.php?id=7. CHARLS (China Health and Retirement
Longitudinal Study). Various years. National
School of Development, Peking University.
References http://charls.ccer.edu.cn/en.
Christensen, Kaare, Gabriele Doblhammer, Roland
Anderson, Ian. 2013. “The Economic Costs of Rau, and James W. Vaupel. 2009. “Aging
Non-communicable Diseases in the Pacific Populations: The Challenges Ahead.” Lancet
Islands: A Rapid Stocktake of the Situation 374 (9696): 1196–208.
in Samoa, Tonga, and Vanuatu.” Health, Currie, Janet, Wanchuan Lin, and Juanjuan Meng.
Nutrition, and Population Discussion Paper, 2014. “Addressing Antibiotic Abuse in China:
World Bank, Washington, DC. An Experimental Audit Study.” Journal of
Baumol, William J. 1996. “Children of Performing Development Economics 110: 39–51.
Arts, the Economic Dilemma: The Climbing Curry, Natasha, and Chris Ham. 2010. “Clinical
Costs of Health Care and Education.” Journal and Service Integration: The Route to
of Cultural Economics 20 (3): 183–206. Improved Outcomes.” King’s Fund, London.
222 LIVE LONG AND PROSPER
Cutler, David M., Angus S. Deaton, and Adriana Hanvoravongchai, Piya. 2013. “Health Financing
Lleras-Muney. 2006. “The Determinants of Reform in Thailand: Toward Universal
Mortality.” NBER Working Paper 11963, Coverage under Fiscal Constraints.” UNICO
National Bureau of Economic Research, Study 20, Universal Health Coverage, World
Cambridge, MA. http://www.nber.org/papers Bank, Washington, DC. http://documents.
/w11963. worldbank.org/curated/en/2013/01/17208145
Cutler, David M., and Mark McClellan. 2001. /thailand-health-financing-reform-thailand
“Is Technological Change in Medicine Worth -toward-universal-coverage-under-fiscal
It?” Health Affairs 20 (5): 11–29. doi:10.1377/ -constraints.
hlthaff.20.5.11. Health Foundation. 2011. “Evidence: Getting
Dodge, Brooks A. 2008. Primary Healthcare Out of Hospital? The Evidence for Shifting
for Older People: A Participatory Study in Acute Inpatient and Day Case Services from
Five Asian Countries. Chiang Mai, Thailand: Hospitals into the Community.” Health
HelpAge International Asia and Pacific Foundation, London.
Regional Development Centre. Higuchi, Michiyo. 2010. “Access to Diabetes Care
Dormont, Brigitte, Michel Grignon, and Helene and Medicines in the Philippines.” Asia-Pacific
Huber. 2006. “Health Expenditure Growth: Journal of Public Health 22 (3): 96S–102S.
Reassessing the Threat of Aging.” Health Hofmarcher, Maria M., Howard Oxley, and
Economics 15 (9): 947–63. Elena Rusticelli. 2007. “Improved Health
Eide, Asbøjrn, Catarina Krause, and Allan Rosas. System Performance through Better Care
2001. Economic, Social, and Cultural Rights: Coordination.” OECD Health Working Paper
A Textbook. 2nd ed. Dordrecht, Netherlands: 30, Organisation for Economic Co-operation
Martinus Nijhoff. and Development, Paris.
Elliott, Rohan A. 2006. “Problems with HSPI (Health Strategy and Policy Institute). 2010.
Medication Use in the Elderly: An Australian “Hospital Overcrowding and Undercapacity:
Perspective.” Journal of Pharmacy Practice Issues and Solutions.” Study report, Hanoi,
and Research 36 (1): 58–66. Vietnam.
Ettenhofer, Mark, Charles H. Hinkin, Steven A. IFLS (Indonesia Family Life Survey). 2007. RAND,
Castellon, Ramani Durvasula, Jodi Ullman, Santa Monica, CA. http://www.rand.org/labor
Mona Lam, Hector Myers, Matthew J. /FLS/IFLS.html.
Wright, and Jessica Foley. 2009. “Aging, IHME (Institute for Health Metrics and
Neurocognition, and Medication Adherence in Evaluation). 2013. The Global Burden of
HIV Infection.” American Journal of Geriatric Disease: Generating Evidence, Guiding Policy.
Psychiatry 17 (4): 281–90. Seattle, WA: IHME.
Feng, Jin, Pingyi Lou, and Yangyang Yu. 2015. Intersalt Cooperative Research Group. 1988.
“Health Care Expenditure over Life Cycle “Intersalt: An International Study of
in the People’s Republic of China.” Asian Electrolyte Excretion and Blood Pressure:
Development Review 32 (1): 167–95. Results for 24-Hour Urinary Sodium and
Fuchs, Victor. 1998. “Provide, Provide: The Potassium Excretion.” British Medical Journal
Economics of Aging.” NBER Working Paper 297 (6644): 319–28.
6642, National Bureau of Economic Research, Jeong, Hyoung-sun, and Ravi P. Rannan-
Cambridge, MA. Eliya. 2010. “SHA-Based Health Accounts
Getzen, Thomas E. 2000. “Health Care Is an in Twelve Asia-Pacific Economies: A
Individual Necessity and a National Luxury: Comparative Analysis.” SHA Technical Paper
Applying Multilevel Decision Models to 10, Health and Social Policy Programme,
the Analysis of Health Care Expenditures.” Organisation for Economic Co-operation
Journal of Health Economics 19 (2): and Development, Paris, and Korea Policy
259–70. Centre, Seoul.
Hagist, Christian, and Laurence Kotlikoff. Jones, Charles I. 2002. “Why Have Health
2005. “Who’s Going Broke? Comparing Expenditures as a Share of GDP Risen So
Growth in Healthcare Costs in Ten OECD Much?” NBER Working Paper 9325, National
Countries.” NBER Working Paper 11833, Bureau of Economic Research, Cambridge, MA.
National Bureau of Economic Research, JSTAR (Japanese Study of Aging and Retirement).
Cambridge, MA. 2011. Research Institute of Economy, Trade,
P O P U L AT I O N AG I N G A N D H E A LT H S E R V I C E S I N E A S T A S I A A N D PAC I F I C 223
and Industry, Hitotsubashi University and Packaging for Improving Adherence to Self-
University of Tokyo, Tokyo. http://www.rieti Administered Long-Term Medications.”
.go.jp/en/projects/jstar/. Cochrane Database of Systematic Reviews
Kaye, Neva, Jason Buxbaum, and Mary Takach. 2011, 7 (9): CD005025.
2011. “Building Medical Homes: Lessons Mandeville, Kate, and Emily Sinnott. 2014.
from Eight States with Emerging Programs.” “Healthier Lives in Aging Societies.”
Commonwealth Fund, New York, and Background paper for the East Asia and
National Academy for State Health Policy, Pacific regional report on aging, World Bank,
Washington, DC. Washington, DC.
Knodel, John, Vipan Prachuabmoh, and Napaporn McCollum, Rosalind, Lieping Chen, Tang
Chayovan. 2013. The Changing Well-Being of Chenxiang, Xiaoyun Lui, Barbara Starfield,
Thai Elderly: An Update from the 2011 Survey Zheng Jinhuan, and Rachel Tolhurst. 2014.
of Older Persons in Thailand. Chiang Mai, “Experiences with Primary Healthcare in
Thailand: HelpAge International. Fuzhou, Urban China, in the Context of Health
Lancet. 2014. “Palliative Care: More than One Sector Reform: A Mixed Methods Study.”
Chance to Get It Right.” 384 (9938): 103. International Journal of Health Planning and
L a n g e n b r u n n e r, J o h n C . , a n d A p a r n a a Management 29 (2): e107–26.
Somanathan. 2011. Financing Health Care Meerding, Willem J., Johan Polder, Luc Bonneux,
in East Asia and the Pacific: Best Practices Marc Koopmanschap, and Paul van der Maas.
and Remaining Challenges. Washington, DC: 1998. “Health-Care Costs of Ageing.” Lancet
World Bank. 351 (9096): 140–41.
Lee, T.-J. 2013. “Pharmaceutical Price Setting Mendelsohn, Daniel N., and William B. Schwartz.
in Vietnam.” Presented at the workshop on 1993. “The Effects of Aging and Population
Vietnamese Health Insurance Law Revision, Growth on Health Care Costs.” Health Affairs
Vinh Phuc, Vietnam, June 17–18. 12 (1): 119–25.
Leung, Gabriel, Keith Y. K. Tin, and Wai-Sum Mendis, Shanthi, Igbal Al Bashir, Lanka
Chan. 2007. “Hong Kong’s Health Spending Dissanayake, Cherian Varghese, Ibitihal
Projections through 2033.” Health Policy 81 Fadhil, Esha Marhe, Boureima Sambo, Firdosi
(1): 93–101. Mehta, Hind Elsayad, Idrisa Sow, Maltie
Lu, Fangwen. 2014. “Insurance Coverage and Algoe, Herbert Tennakoon, Lai Die Truong, Le
Agency Problems in Doctor Prescriptions: Thi Tuyet Lan, Dismond Huiuinato, Neelamni
Evidence from a Field Experiment in China.” Hewageegana, Naiema A. W. Fahal, Goitom
Journal of Development Economics 106: Mebrhatu, Gado Tshering, and Oleg Chestnov.
156–67. 2012. “Gaps in Capacity in Primary Care in
Lubitz, James D., and Gerald F. Riley. 1993. Low-Resource Settings for Implementation
“Trends in Medicare Payments in the Last year of Essential Noncommunicable Disease
of Life.” New England Journal of Medicine Interventions.” International Journal of
328 (15):1092–96. Hypertension 2012: article 584041.
Macinko, James, Inês Dourado, Rosana Aquino, MHLW (Ministry of Health, Labor, and Welfare),
Palmira de Fátima Bonolo, Maria F. Lima- Japan. 2011. “Estimates of National Medical
Costa, Maria G. Medina, Eduardo Mota, Care Expenditure.” MHLW, Tokyo.
Venexa B. de Oliveira, and Maria A. Turci. Newhouse, Joseph P. 1992. “Medical Care
2010. “Major Expansion of Primary Care Costs: How Much Welfare Loss?” Journal of
in Brazil Linked to Decline in Unnecessary Economic Perspectives 6 (3): 3–21.
Hospitalization.” Health Affairs 29 (12): OECD (Organisation for Economic Co-operation
2149–60. and Development). 2007. Health at a Glance
Macinko, James, Barbara Starfield, and Leiyu 2007: OECD Indicators. Paris: OECD.
Shi. 2003. “The Contribution of Primary ———. 2010. “Improving Co-ordination of Care
Care Systems to Health Outcomes within for Chronic Diseases to Achieve Better Value
Organization for Economic Cooperation and for Money.” In Value for Money in Health
Development (OECD) Countries, 1970–1998.” Spending, 125–50. OECD Health Policy
Health Services Research 38 (3): 831–65. Studies. Paris: OECD.
Mahtani, Kamal R., Carl J. Heneghan, Paul B. ———. 2012. “Ageing, Non-communicable
Glasziou, and Rafael Perera. 2011. “Reminder Diseases (NCDs), and Emerging Models
224 LIVE LONG AND PROSPER
Global Population Ageing: Peril or Promise?, Prepared for the Bloomberg Initiative to
edited by John R. Beard, Simon Biggs, David E. Reduce Tobacco Use. http://global.tobacco
Bloom, Linda P. Fried, Paul Hogan, Alexandre freekids.org / files/pdfs/en/Cigarette_tax
Kalache, and S. Jay Olshansky, 46–50. Geneva: _BIcountries.pdf.
World Economic Forum. Sylvia, Sean, Yaojing Shi, Hao Xue, Xin Tian,
Sidorenko, Alexandre, and Asghar Zaidi. 2013. Huan Wang, Qingmei Liu, Alexis Medina,
“Active Ageing in CIS Countries: Semantics, and Scott Rozelle. 2014. “Survey Using
Challenges, and Responses.” Current Incognito Standardized Patients Shows Poor
Gerontology and Geriatrics Research 2013: Quality Care in China’s Rural Clinics.”
article 261819. Working Paper 264, Rural Education Action
Smith, Sheila, Joseph P. Newhouse, and Mark Program, Stanford University, Stanford, CA.
S. Freeland. 2009. “Income, Insurance, and https://reap.fsi.stanford.edu/sites/default/files
Technology: Why Does Health Spending / Standardized_Patients_China_13Feb2014.
Outpace Economic Growth?” Health Affairs pdf.
28 (5): 1276–84. ThaiSES (Thai Household Social and Economic
Sochalski, Julie, Tiny Jaarsma, Harlan M. Surveys). 2011. National Statistical Office,
Krumholz, Ann Laramee, John J. V. Bangkok. http://web.nso.go.th/survey/house
McMurray, Mary D. Naylor, Michael W. Rich, _seco/socio.htm.
Barbara Riegel, and Simon Stewart. 2009. U.S. Department of Health and Human Services.
“What Works in Chronic Care Management: 2000. Healthy People 2010: Understanding
The Case of Heart Failure.” Health Affairs 28 and Improving Health. 2nd ed. Washington,
(1): 179–89. DC: U.S. Government Printing Office.
Somanathan, Aparnaa. 2015. “Ageing and Van Minh, Hoang, Young K. Do, Mary Ann C.
Health in the East Asia and Pacific Region.” Bautista, and Tran T. Anh. 2014. “Describing
Background paper for the East Asia and the Primary Care System Capacity for
Pacific regional report on aging, World Bank, the Prevention and Management of Non-
Washington, DC. communicable Diseases in Rural Vietnam.”
Somanathan, Aparnaa, Ajay Tandon, Huong L. International Journal of Health Planning and
Dao, Kari L. Hurt, and Hernan Fuenzalida- Management 29 (2): e159–73.
Puelma. 2014. Moving toward Universal VHLSS (Vietnam Household Living Standards
Coverage of Social Health Insurance Survey). 2012. General Statistics Office of
in Vietnam: Assessment and Options. Vietnam, Hanoi. http://www.gso.gov.vn
Washington, DC: World Bank. /Default_en.aspx?tabid=491.
Spillman, Brenda C., and James Lubitz. 2000. WHO (World Health Organization). 2008. World
“The Effect of Longevity on Spending for Health Report 2008: Primary Health Care
Acute and Long-Term Care.” New England Now More than Ever. Geneva: WHO.
Journal of Medicine 342 (19): 1409–15. ———. 2011b. Noncommunicable Diseases
Stewart, Malcolm V., Michael Wilson, Karla Country Profiles 2011. Geneva: WHO.
Bergquist, and John Thorburn. 2012. “Care ———. 2012. “Global Health and Aging.” WHO,
Coordinators: A Controlled Evaluation of an Geneva.
Inpatient Mental Health Service Innovation.” ———. 2013. WHO Report on the Global
International Journal of Mental Health Tobacco Epidemic 2013: Enforcing Bans
Nursing 21 (1): 82–91. on Tobacco Advertising, Promotion, and
Strauss, John, Hao Hong, Xiaoyan Lei, Lin Li, Sponsorship. Geneva: WHO. http://www.who
Albert Park, Li Yang, and Yaohui Zhao. 2012. .int/tobacco/global_report/2013/en.
“Healthcare and Insurance among the Elderly ———. 2014a. “Raising Tax on Tobacco: What
in China: Evidence from the CHARLS Pilot.” You Need to Know.” WHO, Geneva.
In Aging in Asia: Findings from New and ———. 2014b. The Health of Older People in
Emerging Data Initiatives, edited by James P. Selected Countries of the Western Pacific
Smith and Malay Majmundar, 342–70. Region. Geneva: WHO.
Washington, DC: National Academies Press. Wimo, Anders, and Martin Prince. 2010. World
Sunley, Emil M. 2009. “Taxation of Cigarettes in Alzheimer Report 2010: The Global Economic
the Bloomberg Initiative Countries: Overview Impact of Dementia. London: Alzheimer’s
of Policy Issues and Proposals for Reform.” Disease International.
226 LIVE LONG AND PROSPER
Wong, Albert, Bram Wouterse, Laurentius C. J. ———. 2014a. “Fostering Primary Care Centered
Slobbe, Hendriek C. Boshuizen, and Johan J. Integrated Care: Realizing High-Quality and
Polder. 2011. “Medical Innovation and Age- Value-Based Service Delivery in China.” Study
Specific Trends in Health Care Utilization: Concept Note, World Bank, Washington, DC.
Findings and Implications.” Social Science and ———. 2014b. Supply-Side Readiness for
Medicine 74 (2): 263–72. Universal Health Coverage: Assessing the
Woods, Kevin, and Allyson McCollam. 2002. Depth of Coverage for Non-communicable
“Progress in the Development of Integrated Diseases in Indonesia. Washington, DC: World
Mental Health Care in Scotland.” International Bank
Journal of Integrated Care 2: e21. ———. 2015. “Macrofiscal Implications of
World Bank. 2013a. “Reshaping Health Service Achieving Universal Health Coverage in East
Delivery through Strengthening Primary Care Asia and Pacific.” In Adjusting to a Changing
and Service Coordination in China: Practical World, 61–77. Washington, DC: World Bank.
Lessons and Models from National and doi:10.1596/978-1-4648-0618-6.
International Experience.” Draft concept note Yip, Winnie, and William C. Hsiao. 2008. “The
for the China Programmatic Health Analytical Chinese Health System at a Crossroads.”
and Advisory Activities study, World Bank, Health Affairs 27 (2): 2460–68. doi:10.1377/
Washington, DC, October 22. hlthaff.27.2.460.
———. 2013b. “Thailand Case Study: Japan– Zweifel, Peter. 2003. “Medical Innovation: A
World Bank Partnership Program for Challenge to Society and Insurance.” Geneva
Universal Health Coverage.” World Bank, Papers on Risk and Insurance 28 (2): 194–202.
Washington, DC. Zweifel, Peter, Stefan Felder, and Markus Meiers.
———. 2013c. “Turkey Case Study: Japan–World 1999. “Ageing of Population and Health
Bank Partnership Program for Universal Health Care Expenditure: A Red Herring?” Health
Coverage.” World Bank, Washington, DC. Economics 8 (6): 485–96.
Long-Term Care in Aging
East Asia and Pacific 7
This chapter is based on a background paper prepared by Aparnaa Somanathan (2015) and a literature
review conducted by Harry Moroz and Naoko Miake. It is also based on a set of background papers pre-
pared for China National Development and Reform Commission by Elena Glinskaya of the World Bank
and Joshua Wiener, Zhanlian Feng, and Nan Tracy Zheng of Research Triangle International. It has also
benefited from inputs from Professor John Campbell and Dewen Wang.
227
228 LIVE LONG AND PROSPER
planning with respect to their interaction with has resulted in longer, healthier lives, thus
informal care systems and existing formal postponing the need for LTC.
health and welfare systems. With the proportion of the population
This chapter outlines the key issues and ages 80 and above expected to double or
options for LTC in developing East Asian and quadruple in some East Asian and Pacific
Pacific countries. The first section describes countries, the demand for LTC is expected
current trends and projections for demand for to increase significantly. In Japan and the
LTC and LTC expenditures in the region. The Republic of Korea, the proportion of the
second section provides an overview of LTC population ages 80 and older is expected
provision within East Asia and Pacific and else- to increase three- to sixfold in the next four
where, and the third section describes financing decades (figure 7.1). Disability rates tend to
arrangements for LTC, including arrangements increase with age, generating demand for LTC
for provision and types of benefits. Challenges services. Figure 7.2 shows projected ADL
in coordinating health care and LTC and and IADL disabilities in China, holding cur-
ensuring sufficient capacity for LTC provision rent disability rates constant. In East Asia and
and quality of care are discussed in the fourth Pacific, particularly in lower-middle-income
and fifth sections, respectively. Given the and middle-income countries, disabilities
nascent state of formal LTC systems in most may not remain constant or even decline, as
developing East Asian and Pacific economies was the case in the OECD. As noted in ear-
the examples and policies cited in the chapter lier chapters, the population cohorts enter-
rely heavily on experience from richer East ing their 70s and 80s in the next two decades
Asian and Pacific economies and countries in have been more exposed to risks related to
the Organisation for Economic Co-operation noncommunicable diseases in middle age than
and Development (OECD). Furthermore, previous cohorts, making them more vulner-
because the policy choices for LTC remain able to disabilities in old age.
much more open at this early stage (compared The poor and rural elderly as well as the
to, for example, pensions and health care, as female elderly are more likely than oth-
discussed in previous chapters), the recommen- ers to need LTC services. The prevalence of
dations are less prescriptive on what the most family-based care can lead to older people
appropriate choices might be. with care needs being missed when they have
no family nearby (or their family is unable
or unwilling to help), no statutory or volun-
LTC demand and expenditures: tary services can fill the gap, or older people
Current trends and projections themselves cannot or choose not to ask for
The proportion of the population surviving to help (Research Base 2012). A recent study
very old ages (80 years and older) is a major by HelpAge International using data from
driver of the demand for LTC, but it is not the World Health Surveys2 looked at unmet
the only one. Demand for LTC is also greatly need for LTC among old people. On average,
influenced by the prevalence of functional 55 percent of elderly men and 45 percent of
and cognitive limitations among the elderly elderly women across the region faced care
as measured by trends in ADLs and IADLs gaps. Figure 7.3 shows that the care gaps are
and by the proportion of frail elderly in the greater for the rural elderly, the female elderly,
population. As already noted in this report, and the poor in some of the poorer East Asian
the extent to which aging is associated with and Pacific countries.
functional and cognitive decline and frailty Predicting growth in demand for LTC in
among older populations is not uniform East Asia and Pacific is complicated because
across countries. In many OECD countries, the region’s LTC is largely informal. The
the adoption of life-course approaches to level of unmet demand for LTC is presumed
preventing the onset of noncommunicable to be very high. Where it exists, institutional
diseases and associated chronic conditions care caters to a relatively small share of
LO N G - T E R M C A R E I N AG I N G E A S T A S I A A N D PAC I F I C 229
the population. The relative underprovision FIGURE 7.1 Long-term care users in OECD countries are expected
of LTC services at present, combined with the to increase with the aging population
prospect of an unhealthy population bulge Proportion of total long-term care users by age, 2009
entering old age with greater dependency and
disability than current elderly cohorts, has Hungary 31 42 27
created a sense of urgency among East Asian Poland 48 22 30
and Pacific policy makers about the need to Czech Republic 30 26 44
develop LTC systems.
Austria 18 34 48
As with health care, LTC expenditures
are driven by both demographic and non- Norway 32 19 49
demographic factors. The demographic driver Sweden 26 25 49
is related to the number of dependent people Netherlands 23 28 50
in the population, which depends on the evo-
Luxembourg 23 28 50
lution of life expectancy and health expendi-
ture. The non-demographic drivers are related Korea, Rep. 6 45 50
to income growth and changes in demand Germany 20 28 52
for publicly financed LTC services. Income Finland 21 25 54
has a direct effect through increases in liv- Switzerland 20 25 55
ing standards (gross domestic product [GDP]
Australia 15 28 57
per capita) and an indirect effect through
changes in relative productivity, or Baumol Denmark 11 28 61
effects. Because the LTC sector is highly labor New Zealand 3 36 62
intensive, room for productivity gains is more Japan 4 32 64
limited than elsewhere in the economy. With
0 10 20 30 40 50 60 70 80 90 100
equalization of wages across sectors, the rela-
tive price of LTC vis-à-vis other goods and ser- Share of population in long-term care (%)
vices in the economy tends to rise as aggregate 0–64 65–79 80+
productivity and GDP per capita increase.
Sources: OECD Health Statistics 2012 data (http://dx.doi.org/10.1787/health-data-en), augmented
With demand being price-inelastic, the share with additional Australian and Swedish data.
of LTC expenditure in GDP would tend to Note: OECD = Organisation for Economic Co-operation and Development.
increase over time. Given the importance of
home production of LTC services, the demand
for publicly financed LTC depends on devel- LTC insurance in 2008. This trend is expected
opments in formal labor force participation to continue.3
(De la Maisonneuve and Martins 2013). OECD projections of LTC expenditures
In 2010, OECD countries allocated an from 2010 to 2060 show that demographic
average of 1.56 percent of GDP to public changes are not likely to be the most impor-
spending on LTC, with private spending on tant determinants of future public LTC expen-
LTC absorbing another 0.67 percent of GDP. ditures. The projections included all OECD
Figure 7.4, panel a, shows total LTC expen- countries and Brazil, China, India, Indonesia,
ditures in OECD countries. Although still and the Russian Federation. Because the cost
relatively low, LTC expenditure—particularly of helping one person with ADL or IADL lim-
public LTC spending—has shown a faster itations should be more or less the same, irre-
upward trend in per capita terms than health spective of age, the pure age effect has only
care spending, with an annual average growth a moderate impact on spending. The OECD
of over 9 percent across 25 OECD countries, model also assumes “healthy aging,” which
compared with 4 percent for total public further mitigates the age effect. As with pro-
expenditure on health (figure 7.4, panel b). jections of public health expenditures, non-
In Korea, the dramatic increase in public LTC demographic factors account for the largest
expenditures followed the introduction of share of the increase.
230 LIVE LONG AND PROSPER
FIGURE 7.2 In China, ADL and IADL disabilities may increase rather than decline as in OECD countries
Projections of ADL and IADL disabilities by gender in China
a. Female b. Male
20 20
15 15
Percent
Percent
10 10
5 5
2010 2020 2030 2040 2050 2010 2020 2030 2040 2050
Year Year
Activities of daily living Independent activities of daily living
FIGURE 7.3 Care gaps are greater for the rural elderly, the female elderly, and those living in poverty in East Asian and
Pacific countries
Percent of total population with unmet care needs by gender, rural/urban location, and poverty status, latest available year, 2000s
a. Care gaps by gender b. Care gaps by rural/urban sector c. Care gaps by poverty level
Brunei Brunei Brunei
Darussalam Darussalam Darussalam
Cambodia Cambodia Cambodia
China China China
Timor-Leste Timor-Leste Timor-Leste
Share of population (%)
and Indonesia). Differences in the initial level FIGURE 7.4 Long-term care expenditures as a share of GDP are
of female labor force participation rates as relatively low but growing faster than health expenditures
well as policies in place also explain the dif- Public LTC expenditures as a share of GDP, 2010, and average annual real
ferences across countries. growth in public LTC expenditure, 2000–10
The analysis of LTC expenditures and pro-
a. LTC expenditures as a share of GDP, 2010
jections for OECD countries provides impor-
Mexico
tant lessons for the non-OECD countries in Slovak Republic
East Asia and Pacific, where LTC remains Australia
Portugal
a gray economy that is largely unquanti- Estonia
Czech Republic
fied. First, reducing levels of severe disability Hungary (2008)
combined with a scenario of healthy aging Poland
United States
can lead to better containment of LTC costs, Korea, Rep.
Spain
because older people remain at home for Luxembourg
longer with less expensive care needs. Second, Slovenia (2008)
Germany
a degree of path dependence exists in how Japan
Austria
LTC systems develop, which can serve to Canada
drive up costs well into the future or contain New Zealand
Switzerland (2009)
them. Responding to LTC needs with invest- OECD countries
Iceland
ment in LTC institutions rather than more France
holistic systems based in home, community, Belgium
Finland
and institutional settings may lock countries Norway
Denmark
into a high-cost system that is hard to restruc- Sweden
ture or shrink later. Countries such as Japan Netherlands
provide useful lessons on this scenario. 0 1 2 3 4
Share of GDP (%)
LTC is provided in three broadly defined b. Average annual real growth in public expenditure on
settings—home, community, and institu- LTC (health), 2000–10 (or nearest year)
Korea, Rep.
tions—and includes both health and social Estonia
care services. Social care services include Japan
Netherlands
assistance with performing ADLs and IADLs Hungary
Switzerland
as well as social support. Figure 7.5 illustrates OECD countries
who requires LTC and what types of services Poland
Portugal
are available. East Asian and Pacific econo- Japan
Australia
mies that have already established formal Belgium
LTC systems—or key elements—are China; Norway
Czech Republic
Japan; Korea; Singapore; Taiwan, China; Iceland
Luxembourg
and Thailand. In East Asia and Pacific more New Zealand
broadly, LTC is delivered informally by family France
Denmark
members. The majority of informal caregiv- United States
Slovenia
ers, both within and outside the household, Austria
Canada
are women who are informally employed. Sweden
Many informal caregivers are themselves Finland
Germany
older people.
0 10 20 30 40 50 60
Across OECD countries and elsewhere,
Average annual growth rate in real terms (%)
institutional care has tended to dominate
formal (that is, paid) LTC provision, despite Sources: OECD 2013, using data from Colombo et al. 2011.
research and polls invariably showing that Note: Based on the definition in the 2012 Joint Health Accounts Questionnaire of the System of
Health Accounts, long-term care (LTC) expenditure comprises long-term (health) care and social
the majority of older people with LTC needs services of long-term care.
232 LIVE LONG AND PROSPER
FIGURE 7.5 Long-term care services include both social and health care provided in a variety of settings
Types of LTC
Assisted
living
Home care
Unhealthy Nursing
elderly homes
Elderly
Healthy
elderly
Hospice
Sick elderly Health care care
prefer services in their homes or community- (Campbell 2014). Of Japan’s elderly (ages
based settings (Colombo et al. 2011; Keenan 65 and over) population, 4.7 percent live in
2010). Home- or community-based services publicly supported institutions, a higher pro-
are also considerably more cost-effective. portion than in many developed nations. The
These preferences appear to be universal, reason is historical and political: in the early
regardless of cultural differences. 1970s, the Japanese government reduced co-
In the United States, publicly financed LTC payments for medical care from 30 percent
has been shifting away from nursing home or 50 percent to zero for those over 70 years
care and toward home- and community- of age in response to political pressure. The
based services over the past several decades, a result over several years was a flood of older
process known as rebalancing (Feng, Fennell, people into hospitals, many newly built to
et al. 2011; Wiener, Anderson, and Brown meet this demand. The problem of these
2009). Currently, home- and community- “social admissions” into hospitals, with little
based services account for nearly half of total or no medical justification, has plagued gov-
Medicaid LTC expenditures (including spend- ernments ever since: 24-hour institutional
ing for people with intellectual and develop- care is perceived as attractive and normal,
mental disabilities), up from just 20 percent a perception that once ingrained, is hard to
in 1995 (Kaiser Family Foundation 2013). remove (Campbell 2014).
Health reforms under the Affordable Care Similarly, the initial heavy reliance on
Act of 2010 aim to further expand home- and LTC hospitals in Korea to meet the popula-
community-based services and to accelerate tion’s LTC needs has become an entrenched
the pace of rebalancing. Although progress pattern of care behavior now. As described
has been made, the bulk of total spending is in box 7.1, LTC hospitals and LTC facili-
still for institutional care. ties compete directly for LTC patients, even
In Japan, the predominance of institution- though the LTC insurance scheme creates
based LTC is at odds with family ideology, some (albeit weak) incentives to move LTC
which is thought to make families reluc- patients to LTC facilities. In addition, the
tant to put aging parents in an institution introduction of LTC insurance was followed
LO N G - T E R M C A R E I N AG I N G E A S T A S I A A N D PAC I F I C 233
BOX 7.1 LTC hospitals and LTC facilities in the Republic of Korea: In direct competition
In Korea, long-term care hospital (LTCH) services rehabilitative care services, whereas the LTCFs take
are covered by national health insurance, and long- the less clinically demanding patients requiring sup-
term care facilities (LTCFs) are covered by long-term port with physical and cognitive limitations. By com-
care (LTC) insurance, which was introduced in 2010. parison, in Korea, no clear distinction exists between
Before the introduction of LTC insurance, LTCHs patients in LTCHs and LTCFs.
were the main source of institutional care for old or
disabled patients, resulting in a high rate of “social
admissions.” Once patients were discharged from FIGURE B7.1.1 Average length of stay in long-term
LTCHs, they were no longer eligible for health insur- care hospitals in the Republic of Korea has declined
ance coverage. only marginally
Ideally, the introduction of LTC insurance should Length of stay in long-term care hospitals, 2010–12
have led to LTCFs taking over most of the LTC
patients, leaving the more clinically complex patients 35
to the LTCHs. As table B7.1.1 shows, expenditures
per person and the out-of-pocket share of financing 30
BOX 7.1 Long-term care hospitals and long-term care facilities in the Republic of Korea:
In direct competition (continued)
TABLE B7.1.2 Distribution of resource utilization groups in the Republic of Korea and Ontario by type of institution
by a substantial increase in LTC hospitals more popular among governments and local
and facilities. By 2012, Korea had 4,181 LTC authorities because building facilities and beds
facilities and 10,759 ambulatory LTC provid- is easier and yields more tangible and visible
ers. Over 60 percent of the LTC residential results than providing and managing services
care facilities in Korea cater to 30 or fewer in homes and communities. Application of
residents, making them too small to achieve appropriate policy instruments (including
economies of scale (Kwon 2014). tax exemption, subsidies, and other financial
Japan’s and Korea’s experiences are quite incentives) should be consistent with the high-
pertinent for other countries in East Asia level policy decisions to build a three-team
and Pacific, where a certain degree of pressure system of social services for the aged, with
exists to expand institutional care despite fil- home-based family care as its bedrock, sup-
ial piety–driven family ideologies and cultural ported by community-based care and under-
preferences to remain at home. For instance, pinned by institutional care. Both preferences
in China, older people prefer receiving LTC and high-level decisions clearly indicate that
at home or in their communities (Shanghai as far as possible, LTC should be delivered in
Daily 2013; Xinhua 2014). The government’s the least restrictive settings.
overarching policy framework for a three- A range of community-based care mod-
tiered LTC system aptly emphasizes home- els exists throughout the region and could
and community-based services. However, be built upon as part of the rebalancing
current policies and resource allocation pat- process (Research Base 2012). Singapore
terns incentivize institutional care more than Programme for Integrated Care for the
home- and community-based services (Feng Elderly (SPICE) centers operate integrated
et al. 2012). services but also operate in home settings
East Asian and Pacific policy makers need as required.4 Thailand has piloted a home
to use policy instruments at their disposal health care scheme in 26 hospitals, targeted at
to rebalance the mix of services as early as older people living at home. Services include
possible and thus avoid a bias for an insti- health promotion, treatment, and rehabilita-
tutional setting in the development of LTC tion. Thailand has also piloted service mod-
systems. Institutional solutions to LTC are els that integrate health and social care—the
LO N G - T E R M C A R E I N AG I N G E A S T A S I A A N D PAC I F I C 235
Bangkok 7 Model and the Community- about US$20 to community-based carers who
Based Integrated Services of Health Care and receive minimal training in supporting others
Social Welfare for Thai Older Persons—that in ADLs. Anecdotal evidence suggests that
involve collaboration among local authori- carers are often younger elderly people who
ties, volunteers, and older people. The ROK– have time and are likely to be more sensitized
Association of Southeast Asian Nations to the needs of older elderly. This program
(ASEAN) Home Care Programme, supported was nationally funded in the initial years, with
by HelpAge International, builds on the suc- the intention to shift financing to subnational
cess of a home care model involving volun- authorities over time. Perhaps even more
teers in Korea and is the most prevalent one innovative is the growth of “time banks”
in the region, having been adopted in all in China with support from local authori-
ASEAN countries. Community groups and ties in cities such as Hangzhou. Under these
networks and older people’s groups operate schemes, younger elderly provide home- and
or are planned in Brunei Darussalam, Japan, community-based support to other elderly
and Thailand, offering support networks, people in need of care and in return generate
health information dissemination, and fit- a “time credit” that compensates the caregiv-
ness activities. In all these initiatives, the role ers with equivalent care time when they reach
of community-based organizations is vital, the stage of needing care.
in many cases supported by older people’s In most OECD countries, provision of LTC
associations as providers, trainers, facilita- services is largely the responsibility of the pri-
tors, and advocates. In 2012, for example, vate sector, although the mix of for-profit and
410,000 older people’s associations existed nonprofit services varies. Notable exceptions
across the region, according to the National to the dominance of private provision are the
Commission on Aging, and more than 23,000 Nordic countries, where local municipalities
exist in Thailand under the Senior Citizens provide a substantial share of services directly.
Council of Thailand.5 In the United States, approximately 69 percent
Examining pilots in several Chinese cities of all nursing homes are for profit, 25 percent
to develop comprehensive community-based are nonprofit, and only 6 percent are govern-
care services is also useful. The Chinese gov- ment run (CMS 2012). Among residential
ernment is implementing pilots in 42 cities care facilities (for example, assisted-living
(including Beijing, Chengdu, Hangzhou, and facilities and board-and-care homes) serving
Shanghai) to provide comprehensive care ser- relatively less disabled elders than nursing
vices with a focus on innovation and appli- homes, more than 82 percent are private and
cability. Key elements of the pilot programs for profit. In England in the early 1980s, the
are establishing or strengthening community majority of care-home providers were publicly
care centers to provide comprehensive ser- owned, and community-based services were
vices; conducting ability and needs analy- limited. Today, however, the overwhelming
sis; purchasing services from the market or majority of adult social care service providers
through service vouchers; using management are in the private sector (Malley et al. 2014).
information systems and information com- Private but nonprofit providers play a larger
munication technology for monitoring and role in some countries, such as Germany and
evaluation, quality control, and management; Australia, with faith-based nonprofits playing
and encouraging involvement of volunteers in a significant role in the latter.
the provision of elderly care services. The mix of for-profit and nonprofit ser-
A further distinctive feature of LTC pro- vices also varies across East Asia and Pacific.
vision in East Asia and Pacific is the emer- Japan has a mixed model: for home- and
gence of state-supported informal caring community-based services, for-profit compa-
arrangements. The most notable example is nies compete with established social service
the Thai scheme for informal carers in rural organizations and other nonprofits, while
areas, which pays a monthly stipend equal to for-profit organizations are excluded from
236 LIVE LONG AND PROSPER
the institutional sector that provides 24-hour has an important role to play in facilitating
care. Under LTC insurance, institutions in and supporting self-care and family care, safe-
principle compete for clients, but in reality, guarding individuals from abuse, and guaran-
high demand plus government restrictions teeing quality and safety while encouraging
on building new facilities has created wait- cost containment. Simply setting basic con-
ing lists instead (Campbell 2014). Korea is sumer standards is often insufficient in a sec-
an interesting intermediate case with respect tor where market forces do not ensure quality
to LTC service provision. In principle, private and safety and where older people and fami-
LTC providers are nonprofit, and the health lies are disempowered consumers. The role of
law does not allow for-profit entities to own the state is therefore vital to safeguard qual-
health care or LTC entities. However, non- ity and safety, not just through establishing
profit providers, which are owned by health minimum standards for care but also through
care professionals or nonprofit foundations, providing information and grievance mecha-
often behave like for-profit entities in prac- nisms, training, setting up professional stan-
tice, with the health professionals who own dards, carrying out inspections, and enforcing
establishments as residual claimants on prof- standards. In Japan, instead of simply subsi-
its. For example, Samsung and Hyundai have dizing or encouraging private provision, the
hospitals, but they are owned by nonprofit government guides the program in the inter-
foundations established by the conglomerates ests of national coverage, equity and fairness,
(Jung et al. 2014). high quality, and efficiency in particular. As
In developing East Asia and Pacific more has long been evident in Japan’s health care
generally, formal or paid LTC services are field, a strong public role is crucial for keep-
still at a very early stage of development. ing down expenditures (Campbell 2014).
Traditionally, most countries have had a small
number of social welfare institutions run by
the state or local governments that served
Financing LTC
childless elders, orphans, and people with men- Informal LTC provided by family members
tal illnesses. As the demand for LTC services constitutes the bulk of LTC consumed and
increases, both state and private sector provi- is associated with significant uncompensated
sion are growing rapidly. Chinese policy mak- costs in almost all countries in the world. In
ers have recognized that the state is no longer the United States, 87 percent of the 12 mil-
able to meet pressing needs without enlist- lion people who need LTC receive it from
ing the private sector, and they have issued unpaid family caregivers, with the annual
a series of national policy directives over the economic value of unpaid care estimated at
past decade to speed up private sector devel- US$450 billion in 2009, which exceeded the
opment of social services for the aged. These total reported LTC expenditure for paid care
policy documents promote preferential policy of US$357 billion in 2011 (Kaiser Family
treatments such as tax exemptions, govern- Foundation 2013). Although provision of
ment subsidies for new and existing beds, land LTC by family members has significant nega-
appropriation or leasing for new construction, tive impacts on their welfare (Van den Berg,
and reduced rates for utilities. The policy doc- Fiebig, and Hall 2014), formal LTC services
ument on aged care issued by the State Council are expensive and beyond the means of most
in 2013 clearly signals that major reforms are families. Across the OECD, private LTC
under way to further privatize the manage- insurance is rarely purchased voluntarily and
ment of publicly built senior care facilities.6 is affordable only to a small number of fami-
As service provision gravitates toward the lies (Colombo et al. 2011).
private sector, concerns about quality of care Acknowledging the challenges in financ-
will emerge, as happened in the OECD, which ing LTC, governments in many countries have
calls for a stronger stewardship role by the stepped in to fill the void by developing pub-
state. In provision of LTC services, the state licly supported LTC services to complement
LO N G - T E R M C A R E I N AG I N G E A S T A S I A A N D PAC I F I C 237
informal care. In almost all developed coun- The philosophical premise in these systems
tries, the government typically pays most or is that the government should take the lead
a substantial share of total LTC costs for eli- in ensuring that all people with disabilities,
gible recipients. For this reason, in most coun- regardless of financial status, should be eli-
tries, much of the policy discussion revolves gible for the LTC services they need. Social
around public financing of LTC services. In solidarity is highly valued, and the right to
non-OECD East Asian and Pacific countries, LTC is viewed similarly to the right to medical
public financing for LTC is minimal, but pol- care. This view is reflected in the public LTC
icy makers are considering the introduction of insurance models in Germany, Japan, Korea,
LTC financing schemes. Luxembourg, and the Netherlands, and in the
Broadly, the approaches to public financ- personal care and nursing care through health
ing of LTC can be classified according to the coverage in Belgium (Colombo et al. 2011;
scope of entitlement to LTC benefits (universal Wiener 2011). In Germany, for example,
vs. means tested), the structure of the service the primary source of financing is through
delivery system at national and subnational a universal social insurance program for
levels, and the types of services and benefits. LTC that provides nursing home and home
These approaches vary greatly across OECD care benefits for people of all ages with dis-
countries. This section assesses the different abilities (Campbell, Ikegami, and Gibson
types of financing systems in OECD countries 2010; Cuellar and Wiener 2000; Gibson and
to draw lessons for East Asia and Pacific. Redfoot 2007). The program is administered
by sickness funds for 70 million Germans, and
private health insurers cover an additional
Scope of entitlement to LTC benefits
10 million, mostly upper-income individuals.
In those countries characterized as having uni- Table 7.1 describes the key features of uni-
versal coverage within a single program, LTC versal LTC insurance programs in Japan,
coverage is provided through a single system. Korea, and Taiwan, China, which share
TABLE 7.1 Features of long-term care insurance in Japan, the Republic of Korea, and Taiwan, China
Sources of financing
Government subsidy 45% 20% 90%
Premium 45% 60–65%
Co-payment 10% 15–20% 10% (poor are exempt)
Eligibility Ages 65 and older Ages 65 and older All ages with disability or age-
Ages 40–64 with mental or All ages with age-related related conditions
physical disability conditions
Service benefits
Home care Yes Yes Yes
Community-based care Yes No Yes
Nursing care No Yes No
Institutional services Yes No Yes
Cash benefits No Only exceptional cases Yes for family members and
home help
Management Municipal government but Under national health Under national health
with uniform fee schedule insurance but financed insurance but financed
independently independently
Source: Lu 2014.
238 LIVE LONG AND PROSPER
many similarities. In all three cases, financing Thus, a local approach can establish impor-
is shared by governments and households. tant links between LTC and other services
In those countries characterized as having often needed by people with disabilities.
means-tested safety-net schemes, strict income Second, LTC is an intensely personal issue
or asset tests are used to set financial thresh- involving decisions about how consumers
olds for eligibility for publicly funded LTC want to live their lives. Thus, the planning
services and benefits targeting the needy. Such and delivery of services can be influenced by
schemes exist in England, New Zealand, the local circumstances, norms, and values as
United States, and more recently Singapore.7 well as by the local preferences of the popu-
The philosophical premise in these countries lation with disabilities, their caregivers, and
is that the primary responsibility for care of providers. Finally, because subnational gov-
older people and younger persons with dis- ernments are less driven to routinize their
abilities rests with individuals and their fami- decision-making process and because indi-
lies, and the government should act only as a vidual cases loom larger in the policy process,
payer of last resort for those unable to provide locally administered programs are arguably
for themselves. The means-tested programs less rigid and bureaucratic than centrally run
thus limit public benefits to people who are programs.
poor or who become poor because of the high At the other end of the continuum are
costs of LTC and medical care. This system countries such as Germany and Japan, which
results in inefficient use of personal and state have a more nationalized and centralized
resources and inhibits personal savings. approach to LTC, although subnational gov-
In between these two relatively clear-cut ernmental entities are often still involved. For
schemes are mixed systems in other OECD example, under the LTC insurance program
countries. In these systems, LTC coverage is in Japan, 2,895 municipal governments or
provided through a mix of different universal alliances of municipalities are the insurers,
programs and benefits or a mix of universal and they have a generalized responsibility
and means-tested LTC entitlements. to provide adequate services (Campbell and
Ikegami 2000, 2003). However, because
almost all aspects of the program (eligibil-
Structure of the service delivery system
ity, most benefits, and reimbursement rates)
A key issue in the design of LTC systems are fixed at the national level, the ability of
is the level of government responsible for the municipalities to shape the program is
financing and delivery. Many developed coun- strictly limited. Thus, although premiums are
tries, including Canada, the Netherlands, set at the municipal level, almost all financing
Sweden, the United Kingdom, and the United parameters are set at the national level.
States, rely heavily on subnational govern- Two main arguments favor consolidation
ments to design and administer their LTC at the national level (Wiener and Tilly 2003).
systems, albeit often with substantial policy First, a uniform national program helps guar-
guidance from the national government. antee horizontal equity across geographic
For example, Sweden devolves almost all areas. In other words, national rules help
responsibility for financing, organization, and ensure that similarly situated individuals in
administration of LTC to municipalities, even different geographic areas are treated simi-
though it is a small country with fewer than larly. In England, for example, which relies
10 million people (SALAR 2007). on subnational governmental units, beneficia-
Advocates for devolution make three argu- ries often complain of a “postcode lottery” in
ments in favor of assigning responsibility for which persons with similar needs and finan-
LTC to smaller geographic governmental cial status are treated very differently because
units (Wiener 1996). First, states, provinces, they are subject to different local authorities.
and municipalities are heavily involved in a In countries with insurance approaches such
variety of social services in many countries. as Germany’s, regional variations are thought
LO N G - T E R M C A R E I N AG I N G E A S T A S I A A N D PAC I F I C 239
to be unfair, and efforts are made to eliminate few future prospects. Furthermore, even
them (Cuellar and Wiener 2000; Wiener and if the costs of the allowances are low, the
Cuellar 1999). Second, developing a single savings may be offset by the increased num-
national program may involve less adminis- ber of users (Wiener 1996, 2011). Services
trative expense because program rules and such as assistance with bathing are less
systems need to be developed only once. Each attractive except when really needed, and
subnational governmental unit need not rein- demand is likely to be limited, at least for sev-
vent the wheel. eral years, in countries where traditional fam-
ily caregiving is widely accepted as the norm
(Campbell 2014).
Types of services and benefits
Public financing for in-kind services
In most countries, publicly supported LTC is requires the development of infrastructure
provided in the form of in-kind services rather for nursing homes and for community-based
than cash benefits. Public payments and sub- services (for example, home help, day care,
sidies for LTC typically go to service provid- respite care for family caregivers, and assis-
ers rather than directly to care recipients. This tive devices). Nursing homes are expensive
structure is primarily driven by concerns over to build and operate, but because the gov-
the misuse of or lack of control over cash ben- ernment is likely to control capital expen-
efits or by concerns over potential coercion of diture, it can control the pace of expansion.
female family members to become LTC pro- Community care agencies do not require
viders and forgo market opportunities. Japan, much new construction, but they do need
for instance, operates a services-only model. trained staff and good management, which
However, demand for cash allowances to require investment. However, they too can
LTC recipients or family caregivers is increas- be developed gradually. A long-run benefit
ing in a number of OECD countries (such is that the training and work experience will
as Eastern European countries, England, raise the quality of the female workforce.
Germany, Italy, and the Netherlands) to Private LTC insurance does not play a
enhance consumer choice and flexibility. major role in financing LTC in any country
In the United States, several recent policy (Colombo et al. 2011). In the United States,
initiatives such as the Medicaid Cash and which has had an active market in private
Counseling Program aim to promote partic- LTC insurance since the mid-1980s, only
ipant-directed personal assistance services for about 12 percent of the population 65 years
frail older adults with disabilities and other of age and older has any LTC insurance
people with disabilities in the Medicaid pro- (almost all of which contains substantial
grams. Case by case, it can be less expensive restrictions on coverage), and it accounts for
than services: Germany offers a choice of cash only about 7 percent of total expenditures
and services, and even though the value of the (Frank, Cohen, and Mahoney 2013; Kaiser
services is about double the cash allowance at Family Foundation 2013). In France, which
various levels of disability, most people take has significant market penetration, LTC
the cash (Campbell 2014). The proposed LTC insurance is integrated with health insur-
insurance in Taiwan, China, includes both in- ance, but the benefits are limited and private
kind and cash benefits as a way of addressing insurance accounts for less than 1 percent
problems of waning family support. of LTC expenditures (Nadash, Doty, and
The disadvantages of cash allowances Racco 2013). Private insurance in Germany
may outweigh the advantages, particularly is primarily offered as an alternative to the
for middle-income countries. Budget con- mandatory statutory sickness funds for upper-
straints mean the cash amounts would have income people and government officials. In
to be low, insufficient to provide a decent short, private health insurance is not a viable
income but perhaps enough to trap caregiv- option for financing LTC for the large major-
ers (mostly women) into a situation with ity of the population.
240 LIVE LONG AND PROSPER
Private out-of-pocket payments by recipi- care, which sets a floor on the quality of care
ents are a feature of all public LTC coverage possible. Even though it is a means-tested
schemes regardless of the specific financing system in which not everyone qualifies for
mechanisms used (Colombo et al. 2011). government financial support, in the United
Countries generally believe that individu- States, the government pays for approximately
als should be financially responsible for the three-quarters of nursing home residents. In
costs of “normal living,” such as housing addition, for private providers, threatening
and meals. Even in universal coverage sys- to terminate participation in the government
tems, services are often targeted to those with financing program is the principal enforce-
the highest care needs, and not all countries ment mechanism for addressing poor quality
provide full and comprehensive coverage of of care (AHCA 2013; Wiener et al. 2007).
services. Individuals and families not infre- The degree of redistribution achieved
quently pay a significant share of LTC costs by public financing schemes is a function
out of pocket, as is the case in Switzerland of how funds are raised and how they are
(60 percent), Portugal (45 percent), Spain spent. LTC financing systems redistribute
(30 percent), and Korea (20 percent for insti- resources from people without disabilities to
tutional care and 15 percent for home care). people with disabilities and potentially from
In the United States, out-of-pocket and private younger (less likely to be disabled), working
payments accounted for 22 percent (15 percent people to older, retired people. Depending on
out of pocket, 7 percent private insurance) of the system, resources can also be redistrib-
the total US$357 billion in LTC spending in uted from higher-income to lower-income
2011 (Kaiser Family Foundation 2013). people. Although disability affects people of
Levels of public financing support and cost all ages and income or wealth levels, people
sharing are determined on the basis of eligibil- with disabilities disproportionately have
ity rules. Eligibility is based on objective and lower incomes and less wealth than people
uniform assessments of individuals’ care needs, without disabilities. For example, in 2001
typically measured by the extent of physi- in the United States, the median income of
cal and cognitive impairments (in addition to older people with no disabilities was about
means testing, where it exists). Mechanisms twice that of older people with severe dis-
for assessing eligibility for LTC services are abilities, while the median total household net
discussed in a separate section below. worth for people with no disabilities was four
The structure and details of public LTC times that of people with severe disabilities
financing have large impacts on the bal- (Johnson and Wiener 2006). As a result, even
ance between institutional and home- and universal coverage financing systems serve a
community-based services, quality of care, disproportionately low-income population.
supply and access to services, and cost con- Finally, cost containment goals can be
tainment. Without adequate financing, the affected by the structure of financing. For
need for LTC may not be translated into instance, the 2000 initiation of LTC insur-
effective demand. That is, people may go ance in Japan was partly aimed at impos-
without formal services (or rely on informal ing a more rational and efficient system for
services) because they have no way to pay LTC. However, because the government had
for them. This is already happening in China already been providing generous benefits to
to some extent, with substantial vacancies in people who sometimes had relatively light
some high-end residential care facilities. The needs, the government found backtracking on
lack of sustainable and predictable, adequate the earlier generosity politically impossible.
financing serves as a deterrent to private In contrast, when Germany started its quite
sector investments into the industry. systematic program in 1995, it was providing
For both publicly and privately provided rather low benefits to relatively few people,
services, the financing system largely estab- so it could start out with stricter eligibility
lishes the level of resources available to provide rules and lower benefits (Campbell 2002).
LO N G - T E R M C A R E I N AG I N G E A S T A S I A A N D PAC I F I C 241
In most East Asian and Pacific countries, communities and relatively less on financial
the policy challenge of providing broad-based support, also seem worth further exploration.
and equitable access to LTC while ensuring
financial sustainability looms large with the
rapid aging of their populations. The experi-
Coordinating LTC services
ences of many OECD countries suggest that a The lack of integration of services across
broad-based, universal LTC financing system different programs or settings, both within
offers the most equitable coverage to people in LTC and across medical care and LTC, is a
need of LTC. For East Asian and Pacific coun- common challenge across countries. Part of
tries with rapidly aging populations (China, this problem is rooted in the common split
Thailand, Vietnam), a targeted universalism between health care and social care. In most
approach, which builds on the principle of countries, the LTC system is separate from
universal coverage while targeting benefits for the health care system (a notable exception
individuals with relatively higher care needs, is Belgium), even though the lines are not
may be desirable. For younger countries, the always clear-cut. Bridging this gap has proved
community-based approaches being piloted difficult because of the multiple provid-
in countries such as Cambodia and Myanmar ers involved in service delivery, the complex
with support from HelpAge International seem financing arrangements, and differences in
promising in leveraging nonfinancial inputs working cultures (as discussed below).
from communities and families that are crucial Many different actors and levels of gover-
to achieving financial viability even with sup- nance are involved in service delivery, contrib-
port from governments or development part- uting to fragmentation of services between the
ners. Approaches such as the time banks in health and LTC sectors. Figure 7.6 illustrates
some cities of China, which rely primarily on the problem of fragmentation in England,
the convening power of government or local where multiple agencies are involved in the
FIGURE 7.6 The health and long-term care systems in England are complex and riddled with perverse
incentives
Parliament
Department of Department of
Communities and Health
Local Government (and social care)
Citizens
Funding Accountability
financing and delivery of health care and than social service employees. Differences
LTC. The fragmentation is greater in those in interests, motivations, and behaviors
countries where the acute care sector domi- at the provider level can lead to potential
nates, the primary care sector is weak, and inefficiencies or failures in service delivery.
community services are limited. Given that The workload of service providers is also
a move toward greater coordination of care likely to increase, at least temporarily, as a
and increased care of the elderly at home or result of integrating care services. Learning
in communities is likely to result in decreased new referral and information systems and
demand for hospital beds, a dominant acute adapting to new integrated ways of work-
care sector can indeed be an impediment to ing can take time away from handling their
reform (OECD 2013). existing workload (Maslin-Prothero and
Moreover, the financing arrangements Bennion 2010).
for health and LTC services are highly In East Asia and Pacific, health care and
complex, with multiple sources of financ- LTC are completely separate from each
ing associated with some indirect spillover other. In China, for instance, relatively few
effects. In Germany, Japan, and Korea, pub- senior care facilities have medical or profes-
lic LTC insurance operates separately from sional nursing staff available (Feng, Zhan,
the health insurance system. As described et al. 2011). In most facilities, residents are
in box 7.1, LTC hospitals and facilities in responsible for providing their own medica-
Korea, financed by national health insur- tions (although facility staff often help with
ance and LTC insurance, respectively, are in storing and administering them) and medical
direct competition with each other. In the care, which may be covered by various health
United States, the federal Medicare program insurance programs. In emergency situations
(which provides health insurance—including that warrant hospitalization of a resident, a
hospital insurance, medical insurance, and typical response by the facility is to call a fam-
prescription drug coverage—for people ages ily member who will take the resident to the
65 and older as well as younger people with hospital, which could lead to problems in care
certain disabilities) covers limited postacute transition and delays in transfer. The situa-
care by nursing homes and home health agen- tion is compounded by the lack of portability
cies, whereas Medicaid (a federal-state jointly of health insurance between regions in coun-
funded, means-tested insurance program for tries such as China and Vietnam.
low-income people) covers a broad array Bridging the gap between health care and
of LTC services, although coverage varies LTC holds promise in two ways: by reducing
by state (Grabowski 2007, 2012). In a set- the cost of acute and institutional care and
ting with two or more bodies responsible for by improving service users’ satisfaction by
funding a service, an incentive exists to pass enabling them to stay in their community
on the costs to the other (Hofmarcher, Oxley, (OECD 2013). Integrated care for the elderly
and Rusticelli 2007). It also increases the cost has been defined as “a coherent set of prod-
to the individuals and to the public purse. ucts and services, delivered by collaborating
In the United States, these concerns have led local and regional health care agencies through
to the creation of a new Medicare-Medicaid securing liaison or links within and between
Coordination Office under the Affordable the health and social care systems” (Hardy
Care Act, with the express purpose of align- et al. 1999). It implies providing a compre-
ing the two programs. hensive array of preventive, primary, acute,
Differences in working cultures and ser- and LTC services, coordinated and provided
vice providers’ roles in the two sectors have by an interdisciplinary team of professionals in
also hindered coordination between health a community-based center and in participants’
care and LTC in the OECD. Health care homes, helping participants delay or avoid
providers, particularly doctors, have tradi- long-term nursing home care. Coordinated
tionally been seen as holding a higher status care experiments in Australia, Japan, and the
LO N G - T E R M C A R E I N AG I N G E A S T A S I A A N D PAC I F I C 243
United States include efforts to align provid- Ensuring capacity and quality
ers’ and users’ incentives for more efficient and for LTC provision
quality-enhancing patterns of care. In time, they
are likely to provide useful models that can be LTC is labor intensive, and the majority of the
built upon in East Asian and Pacific countries. LTC workforce consists of front-line workers.
In some cases, eligibility for LTC services In all countries, women—typically in their
is determined by caseworkers employed middle ages—make up the vast majority of
(or contracted) by municipal governments. the LTC workforce. The front-line workers
These caseworkers evaluate the physical include certified nurses’ aides, home health
and mental condition of the potential clients care aides, and home and personal care work-
plus their income and assets and the extent ers who provide hands-on help for basic
to which family care is available. The case- ADLs, such as eating, bathing, dressing, and
worker draws up a personalized care plan using the toilet. In most countries, these are
for services to be provided by the munici- low-skilled workers with minimal training
pality, determining the amount of benefits requirements. In the United States, federal
to be offered within the constraints of the law requires a minimum 75 hours of training
municipal budget. An advantage of this or passing a certification exam for a certified
system is that it allows the benefits to be nurses’ aide, although some states have addi-
tailored flexibly to individual situations; tional requirements.
moreover, if a means test for income or At the higher end of the skill set, the LTC
assets is included, public resources do not workforce includes a group of licensed health
go to people who could afford care on their professionals. Such professionals include
own. A disadvantage of this system is that registered nurses, licensed practical and
differences in the financial situations among vocational nurses, social workers, physical
municipalities may result in regional varia- therapists, occupational therapists, physi-
tions in eligibility criteria and benefit levels cian assistants and aides, and LTC facility
that may be seen as unfair—or even as a administrators who often assume supervis-
geographic lottery (Campbell 2014). ing or managerial responsibilities rather than
The alternative is to decide on eligibility providing direct, hands-on care (Stone and
and degree of disability through a national Harahan 2010). In only a few countries, such
standardized instrument. The test usually as the Netherlands, are physicians directly
measures how well applicants can perform involved in the provision of LTC.
the standard ADLs. Japan uses a 74-item Many countries face a chronic shortage of
questionnaire that is administered by a mod- LTC workers. Recruiting and retaining direct-
erately trained public servant in a home visit care workers in LTC settings is particularly
of an hour or so. The numerical results are challenging because of a multiplicity of fac-
analyzed with a statistical algorithm to sort tors, such as low pay, low job prestige, few
people into eight categories: independent; fringe benefits, and lack of career paths. As
two grades of “needs support” at a relatively a result, these positions are often viewed as
low level; and five grades of “needs care.” dead-end jobs characterized by high turnover,
The computer’s decision is reviewed by a low retention, and job dissatisfaction. These
local committee, which also looks at a brief workforce challenges are common across
report on the applicant’s medical condition OECD countries, and policy efforts aimed to
from the family physician and may move address them abound (Colombo et al. 2011).
the grade up or down one level. Regardless Unfortunately, few successes exist thus far
of which process is adopted, the key is to that can be readily shared and replicated in
ensure that no one involved in making deci- different countries.
sions about eligibility stands to gain from In this situation, high-income countries
it and that the system has legitimacy in the such as Australia, Austria, Italy, and
eyes of the public. Singapore are increasingly relying on migrant
244 LIVE LONG AND PROSPER
in their fledgling LTC systems. The experi- enormously challenging. Investing early on
ences in other countries demonstrate that in a three-tiered formal LTC system that puts
effective oversight entails building an infor- the emphasis on home- and community-based
mation system to facilitate evidence-based care can avoid many of these challenges, and
policy making, quality assurance, and regula- in the East Asian and Pacific context, it is also
tory enforcement (Mor, Leone, and Maresso likely to be more culturally acceptable. An
2014). In the United States, all publicly certi- added advantage of the early establishment of
fied nursing homes must report both facility- a formal LTC system in middle-income nations
and resident-level data electronically, using a is that training and certification programs can
uniform annual facility inspection survey and be built up gradually as demand expands.
standardized resident assessment (Stevenson Whatever directions are chosen, piloting and
and Bramson 2014). Building an informa- careful evaluation seem appropriate.
tion infrastructure in East Asian and Pacific Even in younger countries, which do not
countries may take time and resources, but yet have a large frail, elderly population, ini-
the Internet and availability of cloud-based tiating a good LTC system early is fiscally
systems mean that countries such as China prudent and makes political sense. Even at an
can move more quickly than was possible in early stage in population aging, the burden of
the United States two decades ago (Feng et al. caring for the elderly is acutely felt by more
2012). This process can be accelerated because and more ordinary citizens who are concerned
the government is increasingly at risk—with about what will happen to their parents and
significant investments of public resources in ultimately themselves. The earlier an LTC sys-
the rapidly growing senior care sector—and tem is started, the lower the cost. When the
better positioned to wield its growing “buyer number of qualifying older people is relatively
power” to make regulations work. low, and traditional family supports are still
working fairly well, the demand for public pro-
grams will not be very high. Moreover, when
Conclusion no public services had previously been avail-
Middle-income and lower-middle-income able, even modest benefits will be welcomed.
countries in East Asia and Pacific should The government can decide later on whether
consider early adoption of systematic LTC coverage or benefits should be upgraded.
programs, ideally before the frail, elderly Initiating a good LTC system early is also
population becomes too large. Israel and good public policy. As the experience of high-
Korea provide two examples of how early income nations demonstrates, a well-designed
establishment of LTC systems can keep costs LTC system delivers good benefits at moder-
down.9 If a comprehensive system is estab- ate cost—a key yardstick for all public poli-
lished before other potential solutions to the cies. Good LTC will lead to real savings in
problem have evolved much, serious difficul- health care provision because frail older
ties can be avoided (Campbell 2014). As dis- people who get good day-to-day care are less
cussed, in the face of a rapidly growing frail, likely to get sick and less likely to relapse after
elderly population with inadequate family being hospitalized, and LTC is inherently less
support, local governments or charities and costly than medical care because it uses staff
other welfare groups generally resort to build- with lower wages and does not lead to open-
ing lots of nursing homes to house those indi- ended benefits. Moreover, if care for frail
viduals. Once institutional care becomes the older people can be handled by a dedicated
norm, and if an increasing number of public system, the health care system can focus on
or quasi-public nursing homes—or worse still, acute care and prevention while other social
hospital owners—band together to protect programs can concentrate on pressing needs
their interests politically, changing course and like poverty, urban housing, and helping
adopting more rational programs becomes young people to succeed.
246 LIVE LONG AND PROSPER
———. 2014. “Japan’s Long-Term Care Insurance Feng, Zhanlian, Chang Liu, Xinping Guan, and
Program as a Model for Middle-Income Vincent Mor. 2012. “China’s Rapidly Aging
Nations.” In Universal Health Coverage Population Creates Policy Challenges in
for Inclusive and Sustainable Development: Shaping a Viable Long-Term Care System.”
Lessons from Japan, edited by Naoki Ikegami, Health Affairs 31 (12): 2764–73. doi:10.1377/
57–68. Washington, DC: World Bank. hlthaff.2012.0535.
Campbell, John Creighton, and Naoki Ikegami. Feng, Zhanlian, Heying J. Zhan, Xiaotian Feng,
2000. “Long-Term Care Insurance Comes to Chang Liu, Mingyue Sun, and Vincent Mor.
Japan.” Health Affairs 19 (3): 26–39. 2011. “An Industry in the Making: The
———. 2003. “Japan’s Radical Reform of Long- Emergence of Institutional Elder Care in Urban
Term Care.” Social Policy and Administration China.” Journal of the American Geriatrics
37 (1): 21–34. Society 59 (4): 738–44.
Campbell, John Creighton, Naoki Ikegami, and Frank, Richard G., Marc Cohen, and Neale
Mary Jo Gibson. 2010. “Lessons from Public Mahoney. 2013. “Making Progress: Expanding
Long-Term Care Insurance in Germany and Risk Protection for Long-Term Services
Japan.” Health Affairs 29 (1): 87–95. and Supports through Private Long-Term
CMS (Centers for Medicare and Medicaid Services). Care Insurance.” Long Beach, CA: SCAN
2012. Nursing Home Data Compendium, Foundation. http://www.thescanfoundation
2012 Edition. Baltimore, MD: CMS. .org/sites/thescanfoundation.org/files/tsf_ltc
Colombo, Francesca, Ana Llena-Nozal, Jérôme -financing_private-options_frank_3-20-13.
Mercier, and Frits Tjadens. 2011. Help pdf.
Wanted? Providing and Paying for Long- Gibson, Mary Jo, and Donald L. Redfoot. 2007.
Term Care. Paris: OECD Publishing. “Comparing Long-Term Care in Germany
doi:10.1787/9789264097759-en. and the United States: What Can We Learn
Cuellar, Alison Evans, and Joshua M. Wiener. from Each Other?” Research Report 2007-
2000. “Can Social Insurance for Long-Term 19, AARP Public Policy Institute, Washington,
Care Work? The Case of Germany.” Health DC. http://assets.aarp.org/rgcenter/il/2007_19
Affairs 19 (3): 8–25. _usgerman_ltc.pdf.
De la Maisonneuve, Christine, and Joaquim Giles, John, Xinxin Chan, Yuqing Hu, and Yaohui
Oliveira Martins. 2013. “Public Spending on Zhao. 2015. “What are the Current and Future
Health and Long-Term Care: A New Set of Long-Term Care Needs of China’s Aging
Projections.” OECD Economic Policy Paper 6, Population?” Background paper, World Bank,
Organisation for Economic Co-operation and Washington, DC.
Development, Paris. Grabowski, David C. 2007. “Medicare and
Duk, Sunwoo. 2012. “The Present Situation Medicaid: Conflicting Incentives for Long-
and Problems of the Long-Term Care Insurance Term Care.” Milbank Quarterly 85 (4):
in South Korea: From Comparative Perspectives 579–610.
between South Korea and Japan.” Japanese ———. 2012. “Care Coordination for Dually
Journal of Social Security Policy 9 (1): 49–60. Eligible Medicare-Medicaid Beneficiaries under
Feng, Zhanlian, Mary L. Fennell, Denise A. the Affordable Care Act.” Journal of Aging
Tyler, Melissa Clark, and Vincent Mor. 2011. and Social Policy 24 (2): 221–32.
“Growth of Racial and Ethnic Minorities in Hardy, Brian, Ingrid Mur-Veemanu, Marijke
U.S. Nursing Homes Driven by Demographics Steenbergen, and Gerald Wistow. 1999.
and Possible Disparities in Options.” Health “Inter-agency Services in England and the
Affairs 30 (7): 1358–65. Netherlands: A Comparative Study of
Feng, Zhanlian, Xinping Guan, Xiaotian Feng, Integrated Care Development and Delivery.”
Chang Liu, Heying J. Zhan, and Vincent Health Policy 48 (2): 87–105.
Mor. 2014. “Long-Term Care in China: HelpAge International. 2015. “Community-Based
Reining in Market Forces through Regulatory Social Care in East and Southeast Asia.”
Oversight.” In Regulating Long-Term Care Briefing Note, HelpAge International East
Quality: An International Comparison, Asia/Pacific Regional Office, Chiang Mai,
edited by Vincent Mor, Tiziana Leone, and Thailand.
Anna Maresso, 409–443. Cambridge, U.K.: Hirdes, John P., Lori Mitchell, Colleen J. Maxwell,
Cambridge University Press. and Nancy White. 2011. “Beyond the ‘Iron
248 LIVE LONG AND PROSPER
253
254 LIVE LONG AND PROSPER
dynamics is thus a topic that would benefit policy choice, health technology, and aging
from increased attention as aging accelerates itself to cost pressures in health and long-
across the region. term care systems, particularly in light of
With regard to fiscal issues, regular projec- research on China that suggests that the
tions of age-related spending scenarios and respective contributions seen in OECD
improved approaches to disentangling the countries may vary in rapidly urbanizing
contribution of aging and other factors to and aging countries.
public spending are needed. Although pro-
jecting the effects of pension, health, long-
term care, and other age-related spending
Labor markets
has matured in the richer economies of the The importance of the labor market as a chan-
region, it remains in its infancy in the devel- nel through which the socioeconomic effects
oping countries. There is an ongoing agenda of aging are felt is well demonstrated. Thus,
of actuarial capacity building in this regard, gaining a deeper understanding of evolving
including education in actuarial science, labor market behavior across the life cycle
development of examination and licensing and of the most effective policies to mitigate
procedures, codes of conduct and profes- the effects of aging on the labor force is vital.
sional bodies, and standards of practice. One area needing better evidence is the
Strengthening actuarial capacity in relative productivity of workers across the
government and the private sector in devel- life cycle, as well as the factors that can affect
oping East Asian and Pacific countries is a worker productivity at older ages. In devel-
priority, as is integrating such analysis sys- oping East Asian and Pacific countries, lim-
tematically into public policy making (for ited information exists on productivity across
example, by mandating regular updates of the life cycle, including how the picture may
actuarial projections of pension and health be changing as more educated cohorts of
insurance programs and by requiring actu- workers begin to pass through middle age.
arial results be included in proposals for Assessing the productivity profile at the firm
reform of social insurance programs). Many level based on the workforce age mix within
OECD countries have a legal requirement firms is important in determining how older
of periodic publication of actuarial results and younger workers can best combine their
of their social security funds—for exam- relative strengths. This approach would help
ple, an annual actuarial valuation in the inform further reforms of seniority wage sys-
U.S. Social Security Administration (over a tems and, more generally, optimal combina-
75-year time horizon) and five yearly (over tions of workers of different ages in the work
90 years) and three yearly actuarial reports place. Related to this is assessment of work-
in Japan and Canada, respectively (Hoskins place adjustments to enhance the productivity
2010). More broadly, Australia is obliged of older workers, an area still in need of evi-
every five years to publish an intergenera- dence on the most cost-effective adjustments.
tional report (latest edition 2015), which Few studies document workplace adjustments
forecasts the effects of demographics on and organizational changes designed to facili-
growth, labor force and productivity, and tate and encourage longer working lives in
public finances, assuming unchanged pub- East Asia and Pacific. This is an area where
lic policies (Commonwealth of Australia studies could be a catalyst to wider adoption
Treasury 2015). It has proven to be an of such adjustments and changes and would
effective vehicle for generating public warrant public research support.
debate on the effects of fertility, longevity Understanding effective practices with
and aging, and migration on the economy respect to development and retention of
and fiscal position. At a basic research adult skills in middle age and beyond will
level, more Asia-specific work is needed be vital. This includes both on-the-job train-
to understand the relative contributions of ing and more structured adult learning and
TO LIVE LONG AND PROSPER: LOOKING AHEAD 257
would help identify market failures in the As countries in the region continue to
adult training market where public interven- reform their social security systems, having
tion may be warranted, as suggested by evi- better insights on the labor market effects
dence from Japan and Korea in chapter 4 of of incentives and disincentives to retire in
this report. With regard to life-long learning social security and tax systems is impor-
outside the workplace, this is an area where tant. Evidence from OECD countries is clear
advocacy sometimes runs ahead of evidence that retirement incentives in social security
with regard to workers in mid-career and systems substantially affect the behavior of
beyond. In countries such as China, signifi- older workers. In addition, current retire-
cant expansion has occurred in short-course ment rules and early retirement incentives
adult training in the TVET (technical and in middle-income East Asian and Pacific
vocational education and training) system in countries appear to lead urban formal sec-
response particularly to the needs of migrant tor workers to retire earlier than will be sus-
workers, though they are usually in their tainable in an aging society. However, more
20s and 30s. Such experience may be adapt- systematic analysis is needed of the observed
able for those in mid-career and beyond. effects in the region’s developing countries of
Continuing evaluations of the experience and how strongly pension rules alter labor force
cost-benefit of older worker training in richer behavior, how formal sector work and self-
East Asian and Pacific economies would also employment after withdrawal from wage
be important. employment interact, and what levels of gen-
In light of the proliferation of initiatives erosity of social pensions produce significant
to stimulate female labor force participation effects on the work of older people. These
across the region’s richer countries in recent questions are more challenging in environ-
years, evaluating the relative effects of dif- ments with high informality, but are impor-
ferent approaches will be important. Certain tant to analyze further.
public interventions such as increasing the An important dynamic in labor markets
supply of affordable child care and elder care of developing East Asian and Pacific coun-
appear to have positive effects on working tries that needs to be better understood is the
women in East Asia and Pacific and glob- interaction of the labor force participation of
ally. The effects of other interventions such working parents of young children, the work
as paid parental leave have less demonstrated and caring responsibilities of grandparents,
effect in the labor market (though it may have and the child care and long-term care mar-
other positive effects on children and family kets. The care provided by grandparents
life). A more rigorous cost-benefit analysis to across the region appears to facilitate the
inform future policy prioritization, including work of their adult children, and also par-
the potential interaction effects of various tially fills an important gap in long-term care
incentives and benefits, would be valuable. markets. One must understand the extent
To date in richer East Asian and Pacific to which caring duties push older people,
economies, little rigorous assessment has particularly women and especially those in
been done of the cost-effectiveness of various urban areas, out of the labor force earlier
financial incentives to promote extension of than they may otherwise choose and the for-
active working lives. An important question mal child and long-term care arrangements
related to financial incentives for hiring older that are needed to offset this effect if it is
workers is the extent to which they should occurring.
be targeted by age only or as part of broader A final and inadequately studied issue
schemes based on characteristics such as in the East Asian and Pacific context is
unemployment or low income. In addition, the degree to which extending the work-
a better understanding of what influences ing lives of older people has any effects
employer attitudes toward hiring or retaining on the employment prospects of younger
older workers is needed. people. Evidence from OECD countries is
258 LIVE LONG AND PROSPER
very clear that extending the working lives older people and their families across a range
of older people does not harm, and may of dimensions. For countries with ongoing
mildly help, the employment prospects of HRS-style surveys, the challenge is to sustain
younger people. However, the phenomenon funding for future rounds and, ideally, to
is not well studied in East Asia and Pacific, bring the surveys into the mainstream of offi-
though studies in China find a similar effect. cial statistical systems. Scope also exists to
Nevertheless, in economies where public further improve the aging surveys, through
employment accounts for a higher share wider sampling of working-age people and
than in OECD countries (for example, the the eldest elderly; introduction of a panel
Lao People’s Democratic Republic, the element to the samples lacking that element;
Pacific Island countries, and Vietnam) and and inclusion of more detailed questions
fiscal constraints are tight, assessing the on issues such as asset wealth, physical and
extent to which OECD findings hold would cognitive functioning, and caring arrange-
be important. This approach is part of a ments. 3 Countries without an HRS-style
global need to assess the lump-of-labor fal- survey should consider initiating one, poten-
lacy in non-OECD settings. tially with co-financing from global research
funding sources such as those that provided
The situation of older people support for CHARLS (China Health and
and households in East Asia and Retirement Longitudinal Study) in China
Pacific (from the National Institute on Aging at
the National Institutes of Health, U.S.
Understanding the financial well-being, Department of Health and Human Services).
health, family relations, and consumption In cases where funding or capacity is lim-
and savings behavior of different age groups ited, inclusion of aging modules in regular
requires better data than are currently avail- household living standards surveys would
able from most household living standards be useful. For example, Thailand provides
surveys in East Asia and Pacific. Current aging-related modules in its general house-
national household surveys of income and hold survey every four to five years.
consumption tend to be poorly geared toward Areas needing a better understanding of
studying aging, because most use the house- the dynamics of elderly and household wel-
hold as a unit of analysis and are thus not fare include the following:
well suited to understanding intrahousehold
well-being and behavior. • Total wealth of older people beyond cur-
The region has benefited in recent years rent income and expenditure. The mixed
from the spread of surveys dedicated to evidence from income and expenditure
the situation of older people, but scope to surveys on elderly welfare highlights the
expand geographic coverage and enhance importance of including wider aspects of
the aging focus of regular surveys remains. household wealth in welfare and inequal-
The expansion of aging-related surveys in the ity analysis, including housing and land
region includes health and retirement liv- ownership, savings, and other assets.
ing survey (HRS)–style surveys in China, • The extent that the labor force behavior
Indonesia, Japan, Korea, and Thailand— of older people changes as societies urban-
and also India in South Asia. 2 Countries ize and other factors such as health and
such as Myanmar, Thailand, and Vietnam longevity change. The very distinct nature
also have conducted useful surveys of older of rural and urban labor markets for
populations, and initiatives such as the older people in East Asia and Pacific sug-
World Health Organization’s SAGE (Study gests potential shifts in work and retire-
on Global Ageing and Adult Health) survey ment behavior as urbanization deepens.
include China. These surveys allow for more However, the behavior of urban older
detailed analysis of the living conditions of people is likely to change over time in the
TO LIVE LONG AND PROSPER: LOOKING AHEAD 259
BOX 8.1 The Active Aging Index: A tool for monitoring the situation and potential of
older people
The Active Aging Index (AAI) is a tool developed by 2. Social activity and participation of older people
the European Commission and the United Nations a. Voluntary activities
Economic Commission for Europe to measure both b. Care to children and grandchildren
the current living conditions and the untapped c. Care to older adults
potential of older people for active and healthy d. Political participation
aging across 28 European countries. It measures
the level at which older people live independent 3. Independent and autonomous living of older
lives and participate in paid employment, social persons
activities, and civic life, as well as their capacity to a. Physical exercise
actively age. b. Access to health services
The index is constructed from 22 individual indi- c. Independent living
cators that are grouped into four domains: employ- d. Financial security (three indicators)
ment, social participation, independent living, and e. Physical safety
capacity for active aging. The results of the AAI are f. Lifelong learning
presented in a country ranking for the overall AAI
4. Capacity and enabling environment for active
and for each of the domains. AAI also offers a break-
aging
down of results by gender. It was developed in part in
a. Remaining life expectancy at age 55
response to the review and appraisal of implementa-
b. Share of healthy life expectancy at age 55
tion of the Madrid International Plan of Action on
c. Mental well-being
Ageing and its Regional Implementation Strategy.
d. Use of information and com munication
The domains and their indicators are outlined as
technology
follows:
e. Social connectedness
1. Employment of older workers f. Educational attainment
a. Employment rate ages 55–59
b. Employment rate ages 60–64 For details of the questionnaire that underlies the
c. Employment rate ages 65–69 reporting and of the methodology, see Zaidi et al.
d. Employment rate ages 70–74 (2013).
poverty, whereas expansion of contributory focal agencies for a national aging strategy
pension schemes, including those with subsi- or action plan, and in some cases, they also
dies for the informal sector, can help address help monitor performance under specific laws
the situation in the long term. Combining on the welfare of the elderly. Both the agen-
these two policy instruments in a dynamic cies and the strategies perform a useful func-
plan should be part of the overall strategy for tion, though the primary policy development
a social protection system. function continues to rest with line ministries
A third set of questions relates to the effect such as labor, social security, health, and wel-
of pension systems on the broader economy. fare. Among those ministries, as in developed
In light of the common objectives of East countries, uncertainties about the boundaries
Asian and Pacific countries of sustaining eco- of sectoral leadership in areas such as the con-
nomic competitiveness and gradually increas- tinuum of care from aged care to long-term
ing formalization of the labor market, a need care services, or the interaction of labor and
exists to understand better the responses of social security policies with respect to older
firms and workers to different social contri- workers, sometimes exist. These are ongoing
bution rates. Do high rates deter labor mar- challenges even in countries that have had
ket formalization, as analysis from some aged populations for long periods. However,
Latin American and Caribbean countries, the there are several examples of so-called super-
European Union, and Turkey suggests? If so, ministries in OECD countries, which bundle
at what level of social contribution are nega- the key policy and service delivery functions
tive effects on labor market formality likely around aging populations (health and long-
to be felt? What segments of the labor market term care, pensions and social security, aged
are most likely to be affected, and what are care, and sometimes employment) into a sin-
the options to reduce the effect? gle public agency.
A second area in which sharing experience
across and within countries would be valu-
Cross-cutting issues able is the building of social consensus around
A challenging aspect of aging societies is insti- policy and behavioral change needed to man-
tutional change and the way public, market, age rapid aging. The experience of richer East
and private institutions need to change in the Asian and Pacific economies suggests that
face of rapid aging. In this respect, “institu- there may be a distinctive political economy
tions” can have several meanings: organiza- to regional efforts to navigate societal aging.
tions or stakeholders and their configuration; To date, the region’s countries have demon-
public policies that set the institutional frame- strated less conflictual political economy
work in specific areas (for example, labor dynamics across generations as they have
regulations); or more broadly, sociocultural sought to respond to the rising needs of grow-
norms with respect to the extended fam- ing elderly populations. At the same time,
ily or gender relations. With respect to pub- older people do not yet appear to have mobi-
lic institutions, developing countries in East lized as openly as in other parts of the world
Asia and Pacific are still coming to terms to promote their interests, and the voice of
with institutional arrangements to respond older people’s associations is still emerging.
to rapid aging. Several countries, including How much such trends reflect strong resid-
China, Indonesia, Thailand, and Vietnam, ual respect for older generations or different
have national commissions, councils, and political systems and dynamics is unclear, as is
committees on aging that act as the highest- how such dynamics may shift over time.
level bodies to raise the profile of aging issues,
perform some coordination functions, and
coordinate reporting on international obliga-
Conclusion
tions such as the Madrid International Plan of Aging will fundamentally reshape societies
Action on Ageing. Such bodies are often the and economies in East Asia and Pacific, but
TO LIVE LONG AND PROSPER: LOOKING AHEAD 263
the shape of the future is very much in the and IFLS (Indonesia Family Life Survey) in
hands of policy makers, communities, and Indonesia include a range of cognitive tests.
citizens. Demography is a powerful force in 4. This position and a debunking of age-related
development, but it is not destiny.4 Policy myths is efficiently summarized in Börsch-
Supan (2013).
makers in the region have the potential to
shape responses that increase the chances for
healthy and productive aging and promote References
societies in which the compact between gen-
Börsch-Supan, Axel. 2013. “Myths, Scientific
erations is fair and realizes people’s potential
Evidence, and Economic Policy in an Aging
at all ages. World.” Journal of the Economics of Ageing
In this process, East Asian and Pacific 1–2: 3–15.
economies can learn from the experiences— Commonwealth of Australia Treasury. 2015. 2015
good and bad—of richer and older coun- Intergenerational Report: Australia in 2055.
tries, as well as from one another. Crafting Canberra: Commonwealth of Australia. http://
appropriate policies will inevitably involve www.treasury.gov.au/PublicationsAndMedia
experimentation and course correction. This /Publications/2015/2015-Intergenerational
approach will also require strong leadership -Report.
in the face of inevitable reluctance to embrace Hoskins, Dalmer D. 2010. “U.S. Social Security at
75 Years: An International Perspective.” Social
change. Fortunately, the region’s strong eco-
Security Bulletin 70 (3): 79–87.
nomic and cultural traditions place it in a
Kwena, Rose Musonye, and John A. Turner.
favorable position to create environments in 2013. “Extending Pension and Savings Scheme
which citizens will live long and prosper. Coverage to the Informal Sector: Kenya’s
Mbao Pension Plan.” International Social
Security Review 66 (2): 79–99.
Notes McDonald, Peter. 2014. “Demographic Change in
1. For more information, see the Wittgenstein the Asian Century: Implications for Australia
C e n t r e D a t a E x p l o r e r, h t t p : / / w w w and the Region.” EABER Working Paper 95,
.wittgensteincentre.org/dataexplorer. See East Asian Bureau of Economic Research,
also McDonald (2014) for an analysis of Australian National University, Canberra,
demographics using the Wittgenstein Centre prepared for the Australian National
projections for a number of East Asian and University conference on Prosperity in Asia:
Pacific countries. The Intergenerational Dimensions, Canberra,
2. Originating with the U.S. Health and April 16–17, 2015.
Retirement Living Survey, HRS-style surveys OECD (Organisation for Economic Co-operation
are also conducted in a number of OECD and Development). 2011. Doing Better for
countries, most notably ELSA (English Families. Paris: OECD.
Longitudinal Study of Ageing) in the United Smith, James, and Malay Majmundar, eds.
Kingdom and SHARE (Survey of Health, 2012. Aging in Asia: Findings from New and
Ageing, and Retirement in Europe) in the Emerging Data Initiatives. Washington, DC:
European Union, which in its latest wave National Academies Press.
included 18 European countries (SHARE, Z a i d i , A s g h a r, K a t r i n G a s i o r, M a r i a
http://www.share-project.org). See Smith and M. Hofmarcher, Orsolya Lelkes, Bernd
Majmundar (2012) for a detailed review of Marin, Ricardo Rodrigues, Andrea Schmidt,
aging-related data initiatives in Asia. Pieter Vanhuysse, and Eszter Zolyomi.
3. Future waves of CHARLS (China Health and 2013. “Active Aging Index 2012: Concept,
Retirement Longitudinal Study) in China, LASI Methodology and Final Results.” European
(Longitudinal Aging Study in India) in India, Centre, Vienna.
ECO-AUDIT
Environmental Benefits Statement
The World Bank Group is committed to reducing its environmental
footprint. In support of this commitment, the Publishing and Knowl-
edge Division leverages electronic publishing options and print-on-
demand technology, which is located in regional hubs worldwide.
Together, these initiatives enable print runs to be lowered and shipping
distances decreased, resulting in reduced paper consumption, chemical
use, greenhouse gas emissions, and waste.
The Publishing and Knowledge Division follows the recommended
standards for paper use set by the Green Press Initiative. The majority
of our books are printed on Forest Stewardship Council (FSC)–certified
paper, with nearly all containing 50–100 percent recycled content. The
recycled fiber in our book paper is either unbleached or bleached using
totally chlorine-free (TCF), processed chlorine-free (PCF), or enhanced
elemental chlorine-free (EECF) processes.
More information about the Bank’s environmental philosophy can
be found at http://www.worldbank.org/en/about/what-we-do/crinfo.