What drives old age work in China?

Abstract

The aim of this paper is to contribute to an understanding of the decisions of the near old and older Chinese to work, and how work characteristics vary across genders, localities, and within overall income and security situations as ageing advances. The results of this analysis suggest that the decision to work and the amount of work among older persons constitute key elements of coping strategies, and these continue to differ across four demographic groups: rural, urban, male, and female. These differences are associated with variations in government policies targeted towards older populations as
well as with long established cultural and social norms. A large share of the near old and older individuals, particularly in rural areas, continue to work at least part time and predominantly in non-regular employment. Much of the rural work is likely based on
necessity rather than choice, since many of those with adequate sources of income, such as employment pensions, opt out of work by age 70. Those who continue to work well into old age report few alternative sources of income and security. Most put in long hours, and a small segment continue to do so up to a very old age. Women in rural areas appear to be the more vulnerable and more likely to work well into old age. Their informal work status, longer life expectancy, reliance on household agriculture, and likelihood to engage in care work all contribute to lower work-related old age security. For urban women, lower mandatory retirement ages translate into lower pension benefits and less opportunity to
build savings.

Full text

DECEMBER 2018
What drives old age work
in China?
CARLA HENRY
FEDERICO FRAGA
TANG YU
ISSN 2306-0875
RESEARCH DEPARTMENT WORKING PAPER NO. 40

Research Department Working Paper No. 40
What drives old age work in China?
Carla Henry*
Federico Fraga†
Tang Yu§
December 2018
International Labour Office
* Research Department, International Labour Organization (ILO). Corresponding author: henryc@ilo.org
† Junior Research Consultant. Email: fragafederico@gmail.com
§ Junior Research Consultant. Email: tangyu@gmail.com

Acknowledgements
The authors would like to thank several ILO colleagues for their very useful comments on this study.
We are grateful to Christian Viegelahn, Andre da Silva Gama, Sara Elder, Li Qingyi, Jeff Johnson and
Claire Courteille-Mulder for their suggestions on how to improve it.

Abstract
The aim of this paper is to contribute to an understanding of the decisions of the near old and older
Chinese to work, and how work characteristics vary across genders, localities, and within overall
income and security situations as ageing advances. The results of this analysis suggest that the decision
to work and the amount of work among older persons constitute key elements of coping strategies, and
these continue to differ across four demographic groups: rural, urban, male, and female. These
differences are associated with variations in government policies targeted towards older populations as
well as with long established cultural and social norms.
A large share of the near old and older individuals, particularly in rural areas, continue to work at least
part time and predominantly in non-regular employment. Much of the rural work is likely based on
necessity rather than choice, since many of those with adequate sources of income, such as employment
pensions, opt out of work by age 70. Those who continue to work well into old age report few alternative
sources of income and security. Most put in long hours, and a small segment continue to do so up to a
very old age. Women in rural areas appear to be the more vulnerable and more likely to work well into
old age. Their informal work status, longer life expectancy, reliance on household agriculture, and
likelihood to engage in care work all contribute to lower work-related old age security. For urban
women, lower mandatory retirement ages translate into lower pension benefits and less opportunity to
build savings.
For China to end the need to work in old age, more progress is needed in deepening social pensions for
the vulnerable as well as extending relatively young mandatory retirement ages. Within this context,
stronger incentives can be delivered to women to stay in the formal work force for longer durations.
However, without a reconsideration of restrictions on hukou driven benefits packages affecting rural
migrants, this aging group will miss out on accruing employment related social benefits associated with
their urban work history.

Table of contents
Acknowledgements ............................................................................................................................... v
Abstract ................................................................................................................................................. vi
Table of contents ................................................................................................................................. vii
List of figures & tables ...................................................................................................................... viii
1. Introduction ................................................................................................................................... 1
2. Context ........................................................................................................................................... 3
2.1 China’s old age demographics and dependency outlook ........................................................ 3
2.2 Rural poverty and vulnerability of older populations ............................................................. 4
2.3 Gender-based old age inequality ............................................................................................. 5
2.4 Pension and health insurance coverage ................................................................................... 5
3. Data and methodology .................................................................................................................. 7
4. Results: Factors influencing old age work .................................................................................. 7
4.1 Whether to work: Descriptive results ...................................................................................... 7
4.2 Sources and levels of income ................................................................................................ 10
4.2.1 Family support .................................................................................................. 13
4.3 Determinants of old age labour participation: Multivariate results ....................................... 13
4.4 Working time of older workers ............................................................................................. 15
4.5 Work activity ........................................................................................................................ 17
4.5.1 Multiple types of work ...................................................................................... 19
4.5.2 Care work .......................................................................................................... 19
4.6 Determinants of old age working time: Regression results .................................................. 20
5. Conclusion .................................................................................................................................... 21
References ............................................................................................................................................ 23

List of figures
Figure 1: Main sources of individual income by age groups for China, 45+ years old ......................... 2
Figure 2: Chinese population distribution by age group, 2017, and projections for 2050 ..................... 3
Figure 3: Labour participation rate of women and men by rural and urban status, age 45 and over ..... 9
Figure 4: Mean annual work income by age group, gender and locality type ..................................... 10
Figure 5: Reported income sourced from work, employment and social pensions, by age ................. 12
Figure 6: Annual number of working days by age; urban and rural women and men who work ........ 16
Figure 7: Levels of work reported by age, gender and locality, broken down by work types ............. 18
Figure 8: Average share of household agricultural work as a share of total annual working days ...... 19
Figure 9: Persons caring for grandchildren: Mean number of hours per year occupied ...................... 20
List of tables
Table 1: Labour participation rates, by age groups, gender disaggregated (2015) ................................ 8
Table 2: At what age do you plan to stop working? ............................................................................... 9
Table 3: Respondent expectations regarding primary financial source for old age support ................ 13
Table 4: Probit regression results on probability of persons over 45 working ..................................... 15
Table 5: Total working days in a year by gender, age group, and current residence status ................. 16
Table 6: Regression Table – OLS-2 ..................................................................................................... 21

What drives old age work in China? 1
1. Introduction
China is experiencing profound demographic changes on several levels that combine a rapidly aging
population with the effects emerging from over three decades of one-child social policies.1 These
demographic forces call for new approaches to employment and social protection policies to address
the needs of a burgeoning elderly population.
China has one of the highest labour force participation rates in the world, hovering at 80 per cent,
although it has declined slightly in recent years. Women as well as men work for much of their adult
lives.
China also maintains relatively young statutory retirement ages, ranging from 45 to 60 years old, which
differ based on gender, individual health status, job occupations, and sectors. Under the law, most
women retire from formal employment five years younger than men. Within the informal economy, and
particularly in rural areas, formal retirement is less common, and changes to work routines are believed
to be more affected by deteriorating health and changes in family situations.
Despite the importance of employment in maintaining the welfare of the near old (45–59 years) and
older (60 years and over) population in China, only a few studies have tried to understand the evolving
labour choices of the elderly as part of their coping strategies. Cai, F. et al., 2012 have compared the
labour supply of older workers across urban and rural China in the context of institutional differences.
The authors note the strong association between employment pension eligibility and exit from
employment, which is primarily an urban phenomenon. In rural areas, where employment pensions are
largely absent, a stronger association was found between physical health and employment status. In a
2015 study (Giles et al., 2015) showing similar results, the authors call for the government to further
develop and unify the Chinese old age support system, including delaying the formal retirement age
and providing more attractive inducements for participation of the rural population in basic pension and
health care old age security schemes. Noting that the care burden falls disproportionately on women,
the authors also call for better ways to support care work, so that women can extend their employment
periods to improve income and pensions during retirement.
Work practices of the older population can provide an insight into the extent to which older persons
will in reality become economic burdens, irrespective of their actual age (Sanderson and Scherbov,
2015). China’s 2010 census highlights the main sources of income for the near old and older population,
who were found to continue to rely heavily on labour income until they reached their mid-sixties, after
which support from other family members became their main source of income support (see Figure 1).
1 China’s one child policy, introduced in 1979, restricted the number of children a couple could have through
legal, economic and administrative measures. It was strictly implemented mainly in urban areas, but more
loosely applied in rural areas and less developed western regions of the country.
1

2 ILO Working Paper No. 40
Figure 1: Main sources of individual income by age groups for China, 45+ years old
Source: China 2010 Census
The aim of this paper is to contribute to an understanding of the decisions of the near old and older
Chinese to work, and how work characteristics change across genders and between rural and urban
localities as ageing advances. Using 2017 survey data, this study updates previous findings on this issue.
Section 2 reviews the major socio-economic contexts of old aged people in China. Section 3 profiles
the data and methodologies used in this study. Section 4 summarizes the results regarding labour force
participation and develops a profile of the characteristics of those persons who work and of those who
do not, in part based on probit analysis. The section also examines the characteristics of the work in
which older persons engage, including both formal and informal employment, under the categories of
own agricultural work, wage work, and earnings from the household’s family-run business. It illustrates
the working patterns of the near old and older population and explores the determinants of working time
to suggest evolving strategies used by the near old and older Chinese to support themselves as they age.
Section five offers conclusions and suggests next steps to be taken.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
45-49 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85-89 90-94 95-99 100+
Labour income Pension
Unemployment insurance Subsistence Allowance
Property income Support from other family members
Others

What drives old age work in China? 3
2. Context
2.1 China’s old age demographics and dependency outlook
China is in the midst of a long-term population trend regarding its age distribution. In 2017, an estimated
231 million people were aged 60 years and above. This number is projected to increase by 109 per cent
by 2050, reaching 483 million.2 At the same time, due to slowing population growth, low birth rates,
and extended life expectancies, the percentage of the working age population is projected to decline.
This in turn is expected to cause the old age dependency ratio to rise more than three-fold by 2050,
hitting 44 per cent, according to United Nations projections.3
As Figure 2 shows, a large segment of China’s population is currently middle aged but is expected to
expand drastically the share of near old and older in the overall population by 2050. A World Bank
report also predicts that the old age dependency ratio will rise more rapidly in rural areas, in large part
as a result of sustained outward migration of rural young adults and return migration of older workers
(F. Cai et al., 2012).
Figure 2: Chinese population distribution by age group, 2017, and projections for 2050
Source: UN DESA / Population Division, World Population Prospects 2017
Family structures have changed radically over recent decades, particularly for urban areas, as a result
of the one-child family policy. Today, the resulting configuration of four grandparents, two parents and
one child has become commonplace. The extended family is deteriorating further through migration of
grown children to find better work, usually in urban areas. It is also common for today’s elderly to
sacrifice their own income opportunities so that their children can work and earn income. This may take
the form of caring for grandchildren, which provides benefits to the household and the economy if it
enables adult children to maintain employment. Older persons also may forego asset accumulation in
favour of supporting investments for their children, such as building houses or buying durables, paying
for education or marriages. As a result, net savings of the elderly are frequently too meagre to cover
their own support. As they age, parents must rely on a dwindling number of children to provide for
2 UN DESA / Population Division, World Population Prospects 2017.
3 Ratio of population aged 65+ over population aged 15–64, ibid.
100 50 050 100
0-4
10-14
20-24
30-34
40-44
50-54
60-64
70-74
80-84
90-94
100+
Population (millions)
2017
Age (
years
)
Male
Female
100 50 050 100
0-4
10-14
20-24
30-34
40-44
50-54
60-64
70-74
80-84
90-94
100+
Population (millions)
2050 (projected)
Male
Female
3

4 ILO Working Paper No. 40
them. (J. Song, 2017)
The hukou system, which restricts household registration to specific birth localities and ties social
benefits to these, has been relied upon as a key means to restricting labour mobility from mainly rural
areas to urban areas with better jobs.4 Most of those who migrate must forgo claims to education,
pensions, and healthcare in their new locations. Despite these disincentives, migrant flows to urban
areas has been considerable, with an estimated one third of the total labour force and half of urban
employment consisting of migrants (Meng, 2012). The vast majority work in the private sector and
informally. In addition to foregoing benefits, migrant workers earn just over half of urban workers (Lam
et al., 2015; Meng, 2012).
Once retired, migrant workers often return to their place of origin but cannot claim entitlements that
would have otherwise accrued to them based on their urban work history. For those working in the
formal sector for a minimum period of 15 years, pension benefits are now made portable (Cho, 2016).
China has experienced large-scale migration of rural youth to urban areas. By 2015, the total number
of China’s migrant workers was 277.5 million, with an average age of 38.6 years, of whom 68.8 per
cent were male.5
2.2 Rural poverty and vulnerability of older populations
China is struggling to close a poverty gap that is widest in rural areas and among the elder population,
who have not benefited from growth to the same extent as younger people and the urban elderly.
Although average incomes in China have risen dramatically since the 1980s, concerns over the poverty
and vulnerability of the rural elderly have been increasing (F. Cai et al., 2012). Unemployment and
poverty have long been associated with one another, so it should be of little surprise that older persons
are more vulnerable to poverty. A 2012 study identified characteristics associated with higher rates of
old age poverty that included: being without pension, being illiterate, living alone, being female and
over 70, or being in poor health. The same study pointed to the ameliorating effects of savings and
public and private – mainly family – transfers (Park et al., 2012).
The proportion of China’s agricultural employment in rural areas has fallen rapidly in recent decades,
with roughly half of all rural workers active in non-agriculture (J. Song, 2017). Rural agricultural
employment is primarily self-employment, with land belonging to the State and allocated through local
leaders. Land leasing arrangements of these allocations remain largely informal and local.
Prior to the introduction of the household responsibility reforms in the late 1970s, rural work continued
under the collective system well into old age. Collectives then provided security to those unable to work
and without support from children. Access to land remains the backbone of rural work for the elderly,
who have been found to farm and maintain small-scale household plots. However, access to adequate
agricultural land is becoming more difficult for older rural residents. In many areas, urban encroachment
has reduced the availability of arable land, which has tended to reduce the size of family plots within
farming communities and also to contort established claims to land use rights (J. Song, 2017).
4 The hukou system is a classification system to keep record of all Chinese as either rural or urban citizens. It
has been a means to limit rural migration to urban areas, where higher paying jobs and better social benefits
are available. Many rural residents move to urban areas unofficially, while retraining their rural hukou status
and associated social benefits.
5 http://www.stats.gov.cn/tjsj/zxfb/201604/t20160428_1349713.html.

What drives old age work in China? 5
Commercial investments in consolidated farmland holdings are becoming more widespread (Long et
al., 2016).
Per capita, arable agricultural land is declining overall in China due to population growth and land
conversion. Land allocations to rural persons are also growing less certain, as rural localities reduce the
frequency of land reallocations, thereby increasing the risk of landlessness for more vulnerable persons.
Rural women have slightly smaller average land allocations (Hare et al., 2007), and their access to land
can be complicated by their marital status.
2.3 Gender-based old age inequality
Early retirement ages for women, which are set at five years younger than those for men, are the first
sign of major shifts in older age work, when urban women, in particular, are found to reduce or end
wage work. Because women work fewer years and earn less than men, they also have less opportunity
to contribute to pension funds, thereby limiting their pension benefits in later years. The retirement age
for men drawing a public pension benefit receipt is 60. For women working in certain professions, it is
also 60, but 55 for female managers and 50 for blue collar women workers, with a mandatory retirement
age of 50 for most women. The difference between women and men in average eligible years of work
is 7.5 years (Chen and Turner, 2015).
Women’s life expectancy in China exceeds that of men by around four years (UN ESCAP, 2017). In
addition, elderly women are less likely to be married – 56 per cent of women compared to 80 per cent
of men. Women are more likely to transition from outside work to care of family members, including
elderly parents, grown children, and grandchildren. As grown children migrate to urban areas, their own
children are often left in the care of elder family members. Where previously older parents would move
in with their grown children – traditionally with their sons – such practices are becoming less common.
2.4 Pension and health insurance coverage
China has established a social security system which, for many decades, has been based on social
insurance, social assistance, and social welfare and has incorporated basic old age pensions and basic
health care. Most of the protections disproportionately benefited urban resident workers, but, in 2008,
the New Rural Social Pension was introduced to provide basic coverage to the rural elderly.
Over the past few years, a significant extension of coverage in both contributory and non-contributory
pension schemes has been initiated by the Chinese government. In 2011, the Social Insurance Law
included provisions for basic pension and health care benefits across different regions – both rural and
urban – and the start of a nationwide unified personal social security ID system, as well as a national
pooled fund for the basic pension scheme. These are to be supplemented with provincial pooled funds
for other social insurance schemes (ISSA, 2013).
A shift from state-owned enterprise to private firms (from 65 per cent of all employment in 1995 to 24
per cent in 2012) has changed the pension portfolios of the working population compared to those now
in retirement (Giles et al., 2015).
The current configuration of social pension schemes provides for a basic income support from age 60
on, but it also calls for contributions from members; these contributions form the basis for augmenting
the base payouts after reaching the age of 60. However, many lower income individuals already close
5

6 ILO Working Paper No. 40
to age 60 either cannot meet the criteria or afford lump sum contributions to qualify for future payouts.
To achieve widespread coverage and participation, public financial contributions to the schemes have
been augmented. Coverage under several social assistance schemes is limited to the elderly who cannot
receive support from their children. China now is credited with having achieved near universal coverage
of at least minimum pension protection for its older population (ILO, 2018). Although improving, non-
contributory pension payouts have amounted to less than 40 per cent of the national poverty line (ibid.,
p. 24).
The Chinese government has also expanded its social health insurance system to include nearly all
people through the establishment of the New Rural Cooperative Medical Scheme, NCMS (2003), and
the Urban Resident Basic Medical Insurance, URMI (2007), which covers mainly those with
urban hukou. NCMS contributes the most to China’s universal coverage, with an estimated 95 per cent
of the Chinese mid-aged and older rural people participating (Zhu et al., 2017). The Urban Employee
Basic Medical Insurance (UEBMI), launched in 1998, is an employment based insurance programme
covering those eligible through their employment situation. Both NCMS and URMI are heavily
subsidized through government funding.
In 2011, government health expenditures accounted for 12.5 per cent of total government expenditures.6
This share is expected to increase as the population ages. High levels of subsidization may also account
for the fairly modest levels of coverage. At the national level, the proportion of out-of-pocket
expenditures to total national health expenditures was significantly reduced from 60 per cent in 2001 to
35 per cent in 2011.7 For lower income elderly, however, health care treatment continues to account for
a major expense, particularly for rural migrants and rural residents Long, et al 2016). Pressure to
subsidize will continue as alternative sources of support decline with the growing numbers of vulnerable
elderly transitioning out of work.
6 Ibid.
7 Ibid.

What drives old age work in China? 7
3. Data and methodology
This paper reports the results of an analysis of the incidence of old age work and its determinants across
urban and rural, male and female seniors in China. The analysis supports an improved understanding
of the coping strategies of those over 45 as they transition from work to non-work.
The primary database used in this work is drawn from the China Health and Retirement Longitudinal
Survey (CHARLS).8 These national samples of Chinese residents aged 45 and older are conducted by
Peking University every two years. The fourth wave of CHARLS was fielded in 2015, which surveyed
nearly 20,000 individuals, with data released in September 2017. Only data from this survey round were
analysed, and weighting was applied to make the sample nationally representative. The data contained
in the CHARLS datasets include demographic and family structure information, health status,
employment status and pension, income and assets, and other related information about the respondents.
Their rich information on employment and pension make them ideal for analysing labour force
participation, type of work, working time and earnings of the near old and older population.
In addition to providing a descriptive analysis of differences between rural and urban, male and female
populations, regression analysis is used in this paper to identify significant factors associated with the
decision to work from age 45 and the working time of older persons still active in employment. A probit
model is used to examine the factors that influence the decision to work or exit the labour market.
Stepwise regression is subsequently applied to identify relevant factors in determining the hours worked
per week for those reporting to work. To identify factors influencing the number of hours worked per
week, an OLS regression is run, testing the same set of descriptors for explanatory significance.
4. Results: Factors influencing old age work
4.1 Whether to work: Descriptive results
This section considers trends in the decision to work or to exit the labour force for persons 45 years of
age and older. It does not explicitly consider formal retirement status, because many of China’s
officially retired persons continue to work or return to work under new circumstances. The labour
participation rate of survey respondents is defined as whether persons consider themselves to be
working or to be looking for work during the survey period. The labour participation rate for each age
group is shown in Table 1 below.
The total average labour participation rate for people aged between 45 and 85 is calculated to be 60.7
per cent, with the participation rates gradually decreasing for older age groups up to around age 70, at
which age the share of those participating in work declines more rapidly. For all age categories,
women’s participation rates are lower than those of men, with the largest difference occurring between
the ages of 50 and 59.
8 http://charls.pku.edu.cn/
7

8 ILO Working Paper No. 40
Table 1: Labour participation rates, by age groups, gender disaggregated (2015)
Age groups
Gender
Frequency
Men
Women
45–49
91.5%
80.6%
87,800,000
50–59
86.9%
67.5%
182,600,000
60–69
68.2%
56.1%
165,800,000
70–79
43.5%
33.2%
79,400,000
80 & above
20.9%
11.4%
29,900,000
Total
72.4%
58.7%
545,500,000
Source: CHARLS 2017; authors’ calculations
Figure 3 shows the labour participation rate of women and men by rural or urban current residence
address. Women living in rural residences show higher work participation rates than women living in
urban residences, with the biggest gap appearing between the ages of 55 and 75 years old. At age 65,
the average labour participation rate of rural women is 60 per cent, that of urban women 30 per cent.
This difference in part reflects the effects of mandatory retirement ages for women in formal work,
which is still primarily urban based. Beyond the age of 65, urban men’s and women’s labour
participation rates show a similar steady downward pattern.
For men before the age of 55, the labour participation rate is just over 90 per cent. Men in rural
residences exhibit somewhat higher work participation rates after 55, with a widening difference
between 60 and 75, when most men in wage work reach their statutory retirement age. By age 65, nearly
60 per cent of urban based men have left work. In contrast, close to 70 per cent of rural women and 80
per cent of rural men still participate in work at age 65, and roughly one in four report still doing so at
age 80.

What drives old age work in China? 9
Figure 3: Labour participation rate of women and men by rural and urban status, age 45 and over
Source: CHARLS 2017; authors’ calculations
For those surveyed who report to be still in work from age 60 and older, 72 per cent of rural men and
73 per cent of rural women engaged in household agricultural work do not anticipate leaving work until
they physically are no longer able to continue (see Table 2). This contrasts with half of men and 36 per
cent of women in non-agricultural work in urban areas anticipating to stop work at a specific age.
Table 2: At what age do you plan to stop working?
Urban Male Urban Female Rural Male Rural Female
Of those who reported to engage in household agricultural work:
Will stop at a certain age
27%
24%
28%
27%
Until physically not able to
73%
76%
72%
73%
Of those who reported wage work or non-farm self-employment and unpaid family business:
Will stop at a certain age
50%
36%
42%
26%
Until physically not able to
50%
64%
58%
74%
Source: CHARLS 2017; authors’ calculations
0.2 .4 .6 .8 1
Work participation, female
45 50 55 60 65 70 75 80 85
Age
Female urban Predicted values, female urban
Female rural Predicted values, female rural
0.2 .4 .6 .8 1
Work participation, male
45 50 55 60 65 70 75 80 85
Age
Male urban Predicted values, male urban
Male rural Predicted values, male rural
9

10 ILO Working Paper No. 40
4.2 Sources and levels of income
For the purposes of this study, analysis of income as it corresponds with work patterns draws upon on
four broad measures covered in the CHARLS survey. These include earnings from work or equivalent
in-kind value of household consumption of products home-produced; pension payouts; transfer income;
and income received from family.
Near old and older persons rely on multiple sources of income, which vary in importance as age
progresses. Wage income is the dominant source of income for urban men until age 60, when
employment pension income gradually replaces the income previously earned from wage work. A
similar shift happens for urban women around age 55. Sources of income among the rural near old and
older men and women derive to a large extent from work throughout all age groups.
Figure 4 shows reported average annual earnings from work for different old age groups, by urban and
rural status.9 Reported income earned from work for women is lower than that for men in all age groups,
and income earned from work for rural women is consistently lower than that for urban women. Average
reported income among respondents suggests that earnings from work decline steadily, falling below
minimum annual wage levels for rural and urban areas.10
Figure 4: Mean annual work income by age group, gender and locality type
Source: CHARLS 2017; authors’ calculations
9 The quantity of work is calculated in the following way: Estimated hours worked per day is multiplied by days
per week worked, then annualized by estimated months per year worked.
10 For comparison purposes, in 2017, the minimum wage for Shanghai was set at 27,600 yuan (US$4000) a year,
and for lower income rural counties in Heilongjiang, Hunan and Liaoning provinces, the minimum wages were
as low as 12,000 yuan per year (Melnicoe, 2017).
010,000 20,000 30,000 40,000 50,000
Mean annual work income (RMB)
45-49 50-59 60-69 70-79
Urban Rural Urban Rural Urban Rural Urban Rural
Male Female

What drives old age work in China? 11
Figure 5 shows reported income sourced from work as well as employment and social pension payouts
by age. According to the analysis, income from work and employment pension payouts are highest for
urban men, and amounts do not drop greatly as age advances, in large part due to payouts from
employment pensions replacing work income. For urban women, a similar pattern can be observed.
However, combined income from the various sources is lower than that for urban men, and the transition
from work to employment pension income begins earlier, at around age 55. Rural men and women
report lower incomes from work and pensions. Rural women, in particular, show on average incomes
from these sources that are at, or fall below, China’s poverty line, as indicated in the fourth panel of
Figure 5.11
A breakdown of income by source also shows the limited protection that social pensions are providing
to the old, particularly in rural areas, where incomes are lowest. Of those receiving social pensions, less
than 4 per cent report payouts from employment pensions. For ages 60–79, median payouts from social
pensions are low for urban and rural populations, averaging just over 900 yuan annually. Less than 10
per cent report supplementing this income through work, but over 65 per cent of urban recipients and
nearly three out of four rural recipients report receiving support from their children. With the additional
payments from employment and other pensions added, major differences in non-work payouts from the
same sources emerge between urban and rural recipients; the average payouts for urban men
approximate more than fifteen times the average payouts for rural women.
For the New Rural Social Pension, 52 per cent of rural respondents from age 60 on report receiving
payouts, which average 105 yuan per month in rural areas. The old age subsidy payouts indicate far less
coverage (15 per cent of rural respondents aged 70 to 79 years and 20 per cent of urban respondents
report payouts). For those 80 and over, 44 per cent report receiving payouts. The average payout is 210
yuan in urban areas and 117 yuan in rural areas. The Urban and Rural Unified Resident Pension,
launched in 2014, provides the largest payouts – an average of 981 yuan – but is reported to be received
by only three per cent of the respondents aged 60 and over.
11 In 2012, China’s poverty line was set at 2,300 yuan (US$360) per year. This level was valid through 2015.
11

12 ILO Working Paper No. 40
Figure 5: Reported income sourced from work, employment and social pensions, by age
Source: CHARLS 2017; authors’ calculations

What drives old age work in China? 13
4.2.1 Family support
Support from the family, and particularly from children, remains a major source of old age security for
some. According to Table 3 below, 70 per cent of rural residents believe that they can rely on their
children for future support. For urban residents, the share drops to 29.7 per cent, with another 60 per
cent relying more on pensions. As the results suggest, only 6.1 per cent of urban residents and 4.7 per
cent of rural residents feel they can rely on savings. These results contrast with long-held traditions of
family caring for old age persons, at least in urban areas. Equally clear is that rural populations envision
limited alternatives of support besides their children, indicating their ongoing old age vulnerability.
Table 3: Respondent expectations regarding primary financial source for old age support
Urban Rural
On whom can you rely financially for old age support?
Children
29.7
70.0
Savings
6.1
4.7
Pension or retirement salary
60.0
20.1
Commercial pension insurance
1.1
0.4
Other
3.1
3.8
Source: CHARLS 2017; authors’ calculations
4.3 Determinants of old age labour participation: Multivariate results
Models to capture determinants of labour market participation differ for developing and developed
countries and also for prime age and old age people. In China’s case, by age 60, most of those who were
working in formal jobs have retired; some start new work. Personal characteristics and household
characteristics can be significant determinants of whether to work. Age, health, gender, marital status
and employment status or health status of a spouse, savings and assets, claim to pension, education and
wealth of children, presence of children and grandchildren in the home, and sector or type of work can
all influence old age labour participation. All of these explanatory variables were tested, but only some
showed significance. Data on savings and assets, including land and home ownership, were not found
to correlate with labour force participation among older persons. The education level of respondents
and the income received from children instead can be considered as proxies for wealth in China. For
education, the cut-off at the completion of middle school was found to capture significant differences
between those working and not working, and it also reflects differences between rural and urban
populations in older age groups.
To better understand the interplay between various factors that associate with older Chinese populations
leaving work behind, a binary logistic regression analysis is used. In this case, a probit analysis is used
to estimate the average change in probability of whether a person does or does not work (dependent
variable y) when there is a 1 unit change in the each of the regressors.
13

14 ILO Working Paper No. 40
Within the framework of the availability of data in the survey, the following explanatory variables are
selected based on their expected influence and overall correlation with the dependent variable:
Married
dummy taking value 1 if individual is married
Spouse working
dummy taking value 1 if individual is married and spouse is working
Age
age of individual in years
Age squared
age ^2
Rural
dummy taking value 1 if individual lives in rural area
Education
dummy taking value 1 if individual completed at least middle school
Health
self-assessed health, ranging from 1 to 10; the higher, the better health
Work pension
dummy taking value 1 if individual receives employment pension
Social pension
dummy taking value 1 if individual receives social pension
Income from children
dummy taking value 1 if individual receives supporting income from children
Own housing
dummy taking value 1 if individual owns the house in which he/she lives
A summary of the regression results is shown in Table 4 below. The ages range from 45 to 90+ years,
with the lower age being roughly five years before official retirement begins in order to capture early
retirements. Characteristics of spouse’s employment and health status were not used to avoid large
drops in sample size.
As noted in Table 4, rural or urban status, marital status, a spouse working, and gender are found to
strongly associate with a higher propensity to work. The probability of older women working is 16 per
cent lower than for men, and the probability of working is 17 per cent higher for rural individuals versus
those in urban areas. The probability of working decreases by 7 per cent for those married, ranging from
a decrease of 17 per cent for urban women to 9 per cent for rural women. Marital status was not found
to be significantly associated with urban and rural males’ decision to work. Age and age squared does
not change the probability of working when other factors are taken into account.
Receipt of an employment pension is associated with an 18 per cent lower probability of working,
strongly significant for men and urban women, but only at the 10 per cent level for rural women, among
whom employment pensions are rare. Overall, receipt of a social pension is associated with a 7 per cent
increased probability of working, possibly reflecting its poverty focus on vulnerable recipients.
The level of education attainment, which is higher among men and urban residents overall in China,
associates with a slightly lower probability of working (2 per cent), if at least a middle school education
has been achieved. However, within all groups but rural men, the association is not strong. There is –
with education among urban respondents in particular – cross-correlation with pension status, as the
better educated are more likely to have had formal employment. Better health is expected to positively
correlate with labour force participation. A higher self-assessment of respondents’ health associates
with a 3 per cent increase in probability of working, with almost no variation across groups.
Variables reflecting children and grandchildren numbers, whether living at home or not, and children’s
dependency status were not found to change the probability of working, despite the established belief
that family care responsibilities and later family support often replace reliance on work income. A factor
that was also not found to differentiate whether one worked was the type of work, including agriculture,
which in rural areas becomes nearly the only form of work for those of very advanced age. The receipt
of income from children (dummy) likewise did not show significance related to the probability of
working, with a larger decrease for urban men and women (5 per cent and 4 per cent, respectively), but

What drives old age work in China? 15
near zero effect for rural residents. These intergenerational transfers may partially reflect differences in
earning capacities between rural hukou and urban hukou children.
Table 4: Probit regression results on probability of persons over 45 working
Dependent variable = Work participation
Explanatory variables
Urban
Rural
All
Male
Female
Male
Female
Married (yes=1)
0.03
-0.17
***
-0.02
-0.09
***
-0.07
***
(0.9)
(-5.2)
(-0.9)
(-4.2)
(-5.4)
Spouse working (yes=1)
0.17
***
0.20
***
0.17
***
0.24
***
0.19
***
(6.0)
(6.4)
(14.9)
(14.4)
(17.8)
Age
-0.03
**
0.02
0.02
**
0.03
***
0.01
*
(-2.0)
(1.0)
(2.2)
(3.4)
(1.8)
Age squared
0.0
0.0
**
0.0
***
0.0
***
0
***
(0.7)
(-2.1)
(-3.8)
(-5.0)
(-4.5)
Gender (Female=1)
-
-
-
-
-0.16
***
(-17.5)
Rural / Urban (Rural=1)
-
-
-
-
0.17
***
(14.7)
Education
-0.03
-0.03
-0.02
*
-0.02
-0.02
**
(-1.2)
(-1.0)
(-1.8)
(-0.8)
(-2.0)
Health
0.02
***
0.02
***
0.03
***
0.03
***
0.03
***
(4.2)
(2.5)
(10.0)
(7.0)
(10.9)
Receives employment pension (yes = 1)
-0.15
***
-0.15
***
-0.13
***
-0.10
*
-0.18
***
(-5.5)
(-5.2)
(-4.8)
(-1.7)
(-11.4)
Receives social pension (yes =1) 0.12*** 0.15*** 0.04*** 0.04*** 0.07***
(3.2)
(4.6)
(3.0)
(2.7)
(7.0)
Receives supporting income from children
(yes=1)
0.03 0.01 -0.01 0.0 0.01
(1.2)
(0.4)
(-0.8)
(0.7)
(1.2)
Own housing (yes=1)
0.04
-0.05
0.0
-0.02
-0.02
(0.8)
(-1.6)
(-0.4)
(-1.3)
(-1.2)
Number of obs.
1,830
2,050
4,589
5,094
13,563
Adjusted R-squared
0.38
0.25
0.25
0.17
0.30
Notes: Average marginal effects are reported; t-statistics in parentheses.
***, **, * indicate statistical significance at the 1, 5, and 10 percent levels, respectively.
4.4 Working time of older workers
Work as a source of income remains critical to a large share of China’s older people as they advance in
age. For those who participate in work, the analysis of hours worked does not suggest a widespread
winding down of the number of hours worked as people age but instead a steady exit from work as age
advances. Even at advanced ages, the small share who continue to work report working relatively long
hours.
Figure 6 and Table 5 below show the average number of working days per year by locality type and
15

16 ILO Working Paper No. 40
gender for those who work. In general, women participating in work and living in rural areas report
working an average of 190 days per year, which does not decline until advanced age (e.g. 80+ years).12
Urban working women show a steady decrease in days worked as age progresses. Men’s total equivalent
working days also fall in the same range as those of women, with urban men averaging slightly more
work days. Table 5 also shows the declining sample size of those who work as age progresses.
Figure 6: Annual number of working days by age; urban and rural women and men who work
Table 5: Total working days in a year by gender, age group, and current residence status
Male Female Total
Days worked
Obs.
Days worked
Obs.
Days worked
Obs.
For urban residences
45–49
262
455
268
380
265
835
50–59
240
804
227
453
236
1,257
60–69
226
338
198
229
214
567
70–79
207
80
203
38
206
118
80 & +
198
8
336
1
209
9
Subtotal
242
1,685
236
1,101
240
2,786
For rural residences
45–49
214
815
197
876
206
1, 691
50–59
210
1,773
188
1,581
199
3,354
60–69
197
1,568
188
1,308
193
2,876
70–79
179
498
194
337
185
835
80–89
165
60
173
42
169
103
Subtotal
203
4,714
190
4,144
197
8,858
Total
217
6,416
203
5,268
211
11,684
12 The data show a high variability that increases with age. This is in part due to the steady decline in sample size
as greater shares leave work altogether.
100 150 200 250 300
Annual worki ng days, female
45 50 55 60 65 70 75 80 85
Age
Female urban Predicted values, female urban
Female rural Predicted values, female rural
150 200 250 300
Annual worki ng days, male
45 50 55 60 65 70 75 80 85
Age
Male urban Predicted values, male urban
Male rural Predicted values, male rural

What drives old age work in China? 17
4.5 Work activity
Not only is the amount worked informative about old age coping strategies, but also the types of work
performed. The data support the analysis of three types of working time measured in the survey:
agricultural work in own household or on others’ farms; wage work; and non-farm self-employed and
unpaid family business. The questionnaire does not distinguish between small-scale household
agricultural work and farming, nor does it measure household work more broadly; moreover, unpaid
care of others is not treated as productive work. For these reasons, working time, particularly for
women, is likely to be under-reported. Figure 7 charts the average hours worked per year, broken down
by the three broad work categories for near old and older persons by gender, locality, and age. Unlike
the previous analysis, the averages include those working no hours to show the relative exit patterns by
work type.
As the charts indicate, wage work is the predominant form of work for urban men and women up to
their mid-sixties, tracking largely common retirement ages. Rural residents engage in relatively little
wage work compared to urban residents, with rural older women relying the least on wage income.
Non-farm self-employment and unpaid family business account for a relatively small share of
equivalent working days. For men, the equivalent working days decline continually as age progresses,
while for women, there is a drop from age 55. Rural residents work much less than urban residents in
non-farm self-employment and unpaid family business.
There is reliance on agricultural work at all ages for rural men and women. In fact, nearly all work
reported by those in rural areas above the age of 70 years is agricultural, suggesting that for some it
continues to be a safety net in extreme old age. Figure 8 shows the percentages of agricultural work in
own household in equivalent working days divided by total equivalent working days, for female and
male rural residents before the age of 85. For both men and women in rural residences, the share of
agricultural work over the total equivalent working days increases with age to the point that, for the
oldest, it becomes almost the only form of work.
17

18 ILO Working Paper No. 40
Figure 7: Levels of work reported by age, gender and locality, broken down by work types
Source: CHARLS 2017; authors’ calculations

What drives old age work in China? 19
Figure 8: Average share of household agricultural work as a share of total annual working days
Source: CHARLS 2017; authors’ calculations
4.5.1 Multiple types of work
A final lens for considering how China’s near old and older population rely on work as part of their
coping strategies is their reliance on multiple sources of work. Roughly 30 per cent of rural males rely
on more than one type of work between ages 45 and 56, with wage and agricultural work being the most
common combination. Less than 10 per cent rely on multiple work sources after age 65. Roughly 10
per cent of rural women carry out work of more than one type – primarily a combination of wage and
agricultural work until age 55, after which the incidence also tapers down. Urban women are the least
reliant on multiple sources of work, with less than 5 per cent depending on them, and then primarily
between the ages of 45 to 54.
4.5.2 Care work
The data do not measure hours spent doing household work or family care, both of which would likely
augment rural and urban old age workloads. More than 60 per cent of respondents between the ages of
45 and 70 report caring for grandchildren during the year. For those married and living with their spouse,
most report that this care work is a shared responsibility, with equal amounts of time used, as suggested
in the gender disaggregated results in Figure 9. After age 70, the average share caring for grandchildren
drops to less than one in three. Roughly 25 to 30 hours per week of care work are reported by those
providing care.
.6 .7 .8 .9 1
Share of work. days on household agric. work in total work. days
45 50 55 60 65 70 75 80 85
Age
Male rural Predicted values, male rural
Female rural Predicted values, female rural
19

20 ILO Working Paper No. 40
Figure 9: Persons caring for grandchildren: Mean number of hours per year occupied
4.6 Determinants of old age working time: Regression results
So far, the descriptive analysis suggests that there may be different factors influencing decisions about
how much to work. In order to find out, ordinary least squares regression was used to test the
significance of the various characteristics considered for decisions related to labour participation. The
sample used was reduced to only those indicating some hours of productive work over the past year.
As shown in Table 6, explanatory factors that influence the number of hours worked for those reporting
to participate in work differ only somewhat from those found to associate with the decision to work. As
would be expected, age is negatively associated with the number of hours worked. Likewise, having a
working spouse and being in relatively better health are positively associated with more work hours per
week, which is consistent with earlier results. Others found inversely related to higher amounts worked
are being female, being rural, receiving an employment pension, and being financially supported by
children. Showing little effect on the amount worked are education level and being a recipient of at least
one social pension.
0500 1,000 1,500 2,000
Average amount of hours devoted to taking care of grandchildren in previous year
45-49 50-59 60-69 70-79 80 & above
Urban Rural Urban Rural Urban Rural Urban Rural Urban Rural
Male Female

What drives old age work in China? 21
Table 6: Regression Table – OLS-2
Dependent variable = Hours worked in a week
Explanatory variables Urban Rural All
Male Female Male Female
Married (yes=1)
-4.93
4.08
3.04
*
1.18
0.68
(-1.3)
(0.9)
(1.9)
(0.6)
(0.6)
Spouse working (yes=1)
2.49
0.88
1.72
*
-0.14
1.65
**
(1.5)
(0.2)
(1.7)
(-0.1)
(2.0)
Age
-0.48
***
-0.42
**
-0.40
***
-0.30
***
-0.39
***
(-3.6)
(-2.4)
(-6.4)
(-4.6)
(-9.1)
Gender (Female=1)
-
-
-
-
-3.46
***
(-5.9)
Rural / Urban (Rural=1)
-
-
-
-
-6.97
***
(-9.0)
Education
0.99
1.04
-1.18
1.16
0.40
(0.6)
(0.4)
(-1.4)
(1.0)
(0.6)
Health
0.48
0.14
0.70
**
0.43
*
0.52
***
(0.9)
(0.2)
(3.1)
(1.7)
(3.2)
Receives employment pension
(yes = 1)
-5.78* -9.63*** -5.73** 6.65 -5.23***
(-1.9)
(-3.0)
(-2.3)
(1.4)
(-3.2)
Receives social pension (yes =1) -3.67 -6.23* 0.28 0.74 0.11
(-1.1)
(-1.7)
(0.3)
(0.7)
(0.14)
Receives supporting income from
children (yes=1)
-1.55 -5.57** -1.82** -2.77*** -2.57***
(-1.1)
(-2.5)
(-2.3)
(-3.2)
(-4.6)
Number of obs.
977
726
3,396
3,190
8,289
Adjusted R-squared
0.07
0.09
0.04
0.02
0.21
5. Conclusion
This research has shed light on how old age work in China configures within the overall income and
security situations of older persons. This study does not advocate old age work, nor does it advocate an
end to it. Within the context of ending old age poverty, it tried to analyse the incidence of work and
identify factors associated with it. The results of this analysis suggest that the decision to work and the
amount of work among older persons constitute key elements of coping strategies, and these continue
to differ across four demographic groups: rural, urban, male, and female. Part of these differences are
associated with variations in government policies targeted towards older populations based on gender,
age, locality, and family circumstances, as well as with long established cultural and social norms.
Specifically, the analysis has pointed to several insights.
For those households surveyed, a large share of the near old and older individuals, particularly in rural
areas, opt to work at least part time and predominantly in non-regular employment as part of their old
age coping strategies.
Much of the rural work is likely based on necessity rather than choice, since many of those with
21

22 ILO Working Paper No. 40
adequate sources of income, such as employment pensions, opt out of work by age 70. Those who
continue to work well into old age report few alternative sources of income and security. Most put in
long hours, and a small segment continue to do so up to a very old age.
Women in rural areas appear to be the more vulnerable and more likely to work well into old age. Their
informal work status, longer life expectancy, reliance on household agriculture, and likelihood to
engage in care work all contribute to lower work-related old age security. For urban women, lower
mandatory retirement ages translate into lower pension benefits and less opportunity to build savings.
The current configuration of basic health and social pension scheme payouts also appears to be
inadequate to offset the need to work even into advanced old age, particularly for those in rural areas
who do not have other sources of income. Recipients will likely require supplemental income and/or
family support to avoid poverty. In the absence of the latter, a significant share of China’s aging rural
population, in particular, will face deep poverty, especially when combined with poor health. In
contrast, based on the work and retirement patterns of those eligible for employment pensions, payouts
appear to deliver income security to the point that recipients no longer need to rely on work as a main
source of income as they advance in age.
A final point should be made on those persons under 70 and still working. The findings of this study
call into question the reliance on old age dependency ratios as a means of estimating the demographic
burden of China’s aging population on an economy, based on the assumption that people beyond
retirement age cease to work and rely on other means of support. Although the data suggest this to be
largely the case at advanced ages, many Chinese work well past their retirement ages.
Policy implications
Ensuring that old age work is available to those able to work and needing income from it will be an
important means of fighting old age poverty worldwide. However, for China to end the need to work in
old age, more progress is needed in deepening social pensions for the vulnerable. This study has found
evidence of relatively low payouts, which leave many falling short of covering their basic needs. There
is a need to deepen payouts for the current generation, who have not had the opportunity to build up
years of contributions.
China is currently considering changes to its relatively young mandatory retirement ages to enable older
workers to remain longer in productive employment. This in turn would extend their old age income
security. Such a move would likely benefit to a large extent many urban workers. Within this context,
stronger incentives can be delivered to women to stay in the formal work force for longer durations.
For informal work, the policy changes would probably have limited effect, particularly in rural areas,
except to perhaps extend the period of contribution to national health insurance and social pension
schemes.
Neither deepening social pension payouts nor extending retirement ages, however, will address the
barriers to pension claims of rural migrants working urban jobs. Without a reconsideration of hukou
driven benefits packages, this aging group will miss out on accruing employment related social benefits
associated with their urban work history. Older returning migrants will make up a rapidly growing share
of the rural elderly. This study did not delve into the issues surrounding this phenomenon, but more
research is needed to inform better policy configurations for this growing demographic group.

What drives old age work in China? 23
References
Cai, F.; Giles, J.; O’Keefe, P.; Wang, D. 2012. The elderly and old age support in rural China:
Challenges and prospects. Directions in Development: Human Development (Washington, DC, The
World Bank).
Chen, T.; Turner J. A. 2015. “Gender and public pensions in China: Do pensions reduce the gender gap
in compensation?” in Sustainability, Vol. 7, No. 2, pp. 1355–1369.
China Health and Retirement Longitudinal Survey (CHARLS), Wave 1–4. Available at:
http://charls.pku.edu.cn/.
Cho, Y. “Equality elusive for China’s urban migrants”, Nikkei Asian Review, 3 Aug., 2016.
Giles, J.; Lei, X.; Wang, Y.; Zhao, Y. 2015. One country, two systems: Evidence on retirement patterns
in China, Asian Bureau of Finance and Economic Research (ABFER) Working Paper.
http://abfer.org/working-papers/search-result?filter_search=Giles&layout=columns
Giles, J.; Wang, D.; Cai, W. 2011. The labor supply and retirement behavior of China’s older workers
and elderly in comparative perspective, IZA Discussion Paper No. 6088 (Bonn, Institute for the
Study of Labor).
Hare, D.; Yang, L.; Englander, D. (2007). “Land management in rural China and its gender
implications”, in Feminist Economics, Vol. 13, Issue 3–4, pp. 35–61.
International Labour Organization (2017). World Social Protection Report 2017–19: Universal social
protection to achieve the Sustainable Development Goals, (Geneva, ILO), pp. 75–99.
—. (2018). Social protection for older persons: Policy trends and statistics 2017–19. Social Protection
Policy Papers. Paper 17 (Geneva, ILO).
International Social Security Association (ISSA). 2013. Social security coverage extension in the
BRICS: A comparative study on the extension of coverage in Brazil, the Russian Federation, India,
China and South Africa (Geneva, ISSA).
Lam, W. R.; Liu, X; Schipke, A. 2015. China’s labor market in the ‘New Normal’, IMF Working Paper,
WP/15/151 (Washington, DC, International Monetary Fund).
Long, H.; Tu, S.; Ge, D.; Li, T.; Liu, Y. 2016. “The allocation and management of critical resources in
rural China under restructuring: Problems and prospects”, in Journal of Rural Studies, Vol. 47, Part
B, October 2016, pp. 392–412.
Melincoe, Mark, blog posted in 2017 for Bloomberg News. https://www.bna.com/
Meng, X. 2012, “Labor Market Outcomes and Reforms in China,” in Journal of Economic Perspectives,
Vol. 26, No. 4, pp .75–102.
National Bureau of Statistics of China. Census Data: 2010 Population Census. Available at:
http://www.stats.gov.cn/english/Statisticaldata/CensusData/.
23

24 ILO Working Paper No. 40
National Bureau of Statistics of China. China Statistical Yearbook 2016. Available at:
http://www.stats.gov.cn/tjsj/ndsj/2016/indexeh.htm.
National Bureau of Statistics of China. Statistical Database. Annual Data. Available at:
http://www.stats.gov.cn/english/Statisticaldata/AnnualData/.
Park, A; Shen, Y.; Strauss, J.; Zhao, Y. 2012. “Relying on whom? Poverty and consumption financing
of China’s elderly”, in J.P. Smith and M. Majmundar (eds): Aging in Asia: Findings from new and
emerging data initiatives (Washington, DC, National Research Council), pp. 148–172.
Sanderson, W.C.; Scherbov, S. 2015. “Are we overly dependent on conventional dependency ratios?”
in Population and Development Review, Vol. 41, No. 4, pp. 687–708.
Song, J. 2017. Gender and employment in rural China (Abingdon: Routledge).
Song, Y. 2017. “Six central features of the Chinese labour market: A literature survey”, International
Labour Review, Vol. 156, No. 2, pp. 213–242.
United Nations Department of Economic and Social Affairs (UN DESA) / Population Division. 2015.
World Population Ageing 2015 (New York, United Nations).
—. World Population Prospects 2017. Available at: https://esa.un.org/unpd/wpp/.
United Nations Economic and Social Commission for Asia and the Pacific (ESCAP). 2017. Ageing in
the Asian and Pacific Region: An overview (Bangkok, ESCAP).
Wu, X.; Li, L. 2014. “The motives of intergenerational transfer to the elderly parents in China:
Consequences of high medical expenditure”, in Health Economics, Vol. 23, No. 6, pp. 631–652.
Zhao, Y.; Hu, Y.; Smith, J.P.; Strauss, J.; Yang, G. 2014. “Cohort profile: The China Health and
Retirement Longitudinal Study (CHARLS)”, in International Journal of Epidemiology, Vol. 43,
Issue 1, pp. 61–68.
Zhao, Y.; Smith, J.P.; Strauss J. 2014. “Can China age healthily?” in The Lancet, Vol. 384, Issue 9945,
pp. 723–724.
Zhu, K., Zhang, L. Yuan, S., Zhang X. (2017). “Health financing and integration of urban and rural
residents’ basic medical insurance systems in China”, in International Journal for Equity in Health,
Vol. 16, pp. 194ff. (https://equityhealthj.biomedcentral.com/articles/10.1186/s12939-017-0690-z).