University of Southern California Dissertations and Theses

Children 's migration and the financial, social, and psychological well-being of older adults in rural China

(USC Thesis Other)

Children 's migration and the financial, social, and psychological well-being of older adults in rural China

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

CHILDREN'S MIGRATION AND THE FINANCIAL, SOCIAL, AND PSYCHOLOGICAL
WELL-BEING OF OLDER ADULTS IN RURAL CHINA

by

Zhen Cong

________________________________________________________

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(GERONTOLOGY)

May 2008

Copyright 2008 Zhen Cong

ii
Dedication
To my husband, Jianjun Luo, for his enormous love and encouragement. Without his
strong belief in my success, it would be impossible for me to pursue a degree ten thousand
miles from him.
To my parents, Pengsu and Lixin Cong, for their unreserved love. My mother provided
me tremendous help in household chores and child care during my dissertation writing,
particularly when I needed her most before and after the birth of my twin babies.
To my twin babies, who attended my dissertation defense before their birth and
introduced me into a completely new and amazing role.

iii
Acknowledgements
I would like to thank Drs. Iris Chi, Eileen Crimmins, and Merril Silverstein, to serve in
my dissertation committee. In addition to guiding me to finish my dissertation, they guide me
through the Ph.D. program, and give me every possible help for me to prosper in my career. I
am deeply grateful to every one of them. I am very lucky to have these exceptional scholars in
my dissertation committee.
I would like to give special thanks to Dr. Merril Silverstein, my advisor, for his
remarkable mentorship. He sets up high academic standard and encourages me to stretch
myself to reach the standard. He is extremely generous to provide me opportunities and
support whenever I need.
I would like to thank Dr. Vern Bengtson for his support all along the way in my Ph.D.
program, and my first mentor, Dr. Shuzhuo Li, from Xi’an Jiaotong University, for
introducing me to the field of research and for providing me endless support.
I would also like to thank my friends In Hee Choi, Aaron Hagedorn, and Sarah Ruiz, for
their encouragement and kind help in the Ph.D. program.
Finally, I would like to acknowledge the support of the Fogarty International Center,
National Institutes of Health (# R03TW01060).

iv
Table of Contents

Dedication

ii
Acknowledgements

iii
List of Tables

v
Abstract

vi
Chapter 1: Introduction, Background, and Conceptual Framework

1
Chapter 2: How do Chinese Rural Elders Share the Wealth with their Migrant Children?

15
Chapter 3: Which Sons Live with their Parents: How do Sons’ and their Male Siblings’
Exchanges with Parents Matter?

39
Chapter 4: Intergenerational Support and Depression Among Elders in Rural China: Do
Daughters-In-Laws Matter?

60
Chapter 5: Discussion and Conclusions

82
References 86

v
List of Tables

Table 2.1: Description of Analytic Variables

27
Table 2.2: Intergenerational Exchanges by Children’s Migration Status

29
Table 2.3: Random Effects Models Predicting T2 Financial Transfers from Adult
Children to Their Older Parents

31
Table 3.1: Description of Analytic Variables

51
Table 3.2: Intergenerational Exchanges between Parents and Non-Migrant Children

52
Table 3.3: Random Effects Logistic Regression Predicting Transiting into Coresidence

54
Table 3.4: Random Effects Logistic Regression Predicting Odds of Coresidence

55
Table 4.1: Description of Analytic Variables

69
Table 4.2: Percentage of Older Parents Receiving Baseline Household Support and
Personal Care by Gender and Living Arrangement with Daughter-in-Law (%)

73
Table 4.3: Unstandardized Ordinary Least Squares Regression Estimates Predicting
Depressive Symptoms
74

vi
Abstract
This dissertation consists of three independent papers, investigating how migration of
working age adults from rural to urban areas in China influences intergenerational transfers,
living arrangements, and the psychological well-being of elders who were raised and are
embedded in the patrilineal family system.
Analyses were performed using data from a three-wave longitudinal survey, the
Longitudinal Study of Older People in Anhui Province, China. Random effects regression
analysis from the first study (4,101 parent-child dyads from 1,147 parents) showed that for the
same amount of child care and financial help that elders provided, they reaped higher financial
return from their adult migrant children than from their adult non-migrant children. In Chinese
families, non-migrant children did not need to reciprocate for parents’ help in the short run.
However, because parents had higher bargaining power when they provided child care to
migrant children and because parents’ financial help to migrant children was a strategic
investment, migrant children were obliged to pay back parents’ help. In the second study,
random effects logistic regression (803 to 1154 parent-child dyads corresponding to 513 to
735 parents) showed that whether an elder coresided with a child depended on elders’ previous
exchange with that child and elders’ exchanges with the other siblings of that child. In the
third study, analyses based on cross-lagged effects model (1281 parents) indicated that
depressive symptoms were usually reduced by assistance from daughters-in-law, and increased
sometimes when such support was from sons. These relationships held most strongly when
mothers coresided with their daughters-in-law. This research suggested that the benefits of
intergenerational support were conditional on culturally prescribed expectations.
Taken together, this dissertation demonstrated that the migration of adult children had
substantial influence on the intergenerational transfer behaviors, the living arrangements of
elders, and the effects of intergenerational transfers on elders’ psychological well-being.

1
Chapter 1: Introduction, Background, and Conceptual Framework
Introduction
This dissertation consists of three independent papers, investigating how migration of
working age adults from rural to urban areas in China influences intergenerational transfers,
living arrangements, and the psychological well-being of elders who were raised and are
embedded in the patrilineal family system. Specifically, this dissertation examines the roles
that parents’ help in money and child care have played in these processes, and different
influences of support from sons, daughters, and daughters-in-law on the psychological
well-being of elders against the backdrop of reduced intergenerational coresidence.
Background
Rural Chinese elders and their families: Intergenerational relationships
China has the largest population and the largest older population in the world. The
majority of its elders live in rural rather than urban areas. In China, rural elders differ from
their urban counterparts because of the separation between rural and urban areas in order to
guarantee the development of urban areas and to keep the surplus of labor from flowing out of
the rural areas (Benjamin, Brandt, & Giles, 2005; Oi, 1999; Zimmer & Kwong, 2003).
Although urban elders usually have some pension after their retirement as a result of their
life long contributions to government or state-owned enterprises, rural elders generally
continue to work and support themselves because there is no pension to compensate for their
life long working. This leads to the economic vulnerability of rural elders, who have to
continue working until they can no longer work (Pang, 2004). Although there is “medical
cooperation”, a social insurance system which requires a little payment and covers a large
proportion of medical expenses when needed, many elders in poverty can not afford the
premium or co-payment (Zhu, 2000). In some extreme cases, elders have committed suicide
when they could no longer work and become a burden to their families.

2
A nationally representative survey in 2000 showed that only 1.5% of rural elders received
social insurance payments, 0.6% of them had some pension from former employers, and 9.5%
had saved for old age. Therefore it is not surprising that only 54.5% of rural elders felt that
they were economically secure and 79.1% chose investing in their children rather than
personal savings, insurance, or pensions as the first choice to ensure their economic safety in
old age (China Research Center on Aging, 2003). On the one hand, the reliance on children
reflects the economic vulnerability of elders, and on the other hand reflects filial piety and
culturally embedded Confucianism, which imposes strong obligations on adult children to
support their parents in various ways, including providing financial help (Shi, 1993; Sung,
1995).
Although rural elders’ financial vulnerability is recognized, it is regarded as the problem
of a family. This means that when elders can not work any more, their only dependable
resources are their assets, savings and their children. But life-long poverty does not allow
elders to accumulate enough assets for their old age. In addition, rural elders have to make
every effort to meet their sons’ financial needs for marriage. The filial piety from children is
conditional, and usually depends on whether parents can fulfill their responsibilities. In rural
areas, there is a saying: “Parents owe a son a wife, the son owes his parents a coffin”. In order
to recruit a bride and meet her demands for independent housing, elders have to use all
available means to help a son build his own house for independent living, e.g., to work even
harder, or to borrow money from others and promise to return in the future. Inability to do so
will justify children not providing any help to their parents (Cohen, 1998; Yan, 2003). This
kind of inter vivo asset transfers will have depleted parents completely when they reach old
age, particularly if they have several sons. Their reliance on children becomes even more
salient when they do not have resources for themselves (Lee & Xiao, 1998).

3
Although gender egalitarian is widely accepted by urban elders, son preference is still
strong in rural area (Graham, Larsen, & Xu, 1998). Because of the practice of patrilineal
marriage, daughters usually marry into places other than their natal village and are not entitled
to any land in that village after marriage. Theoretically she belongs to another family and
another village. Only married sons are entitled to land for farming and housing, and are
expected to take care of their parents in their old age (Feldman, Tuljapurkar, Li, Jin, & Li,
2006). Therefore, most elders coreside with married sons if they coreside with any married
child at all (Knodel & Ofstedal, 2002). In spite of this, elders get more and more help from
their daughters because of improved transportations that facilitate daughters’ contact with their
parents, as well as daughters’ economic independence gained through migration and work off
the farm, and (Li, Feldman, & Jin, 2004).
Migration of adult children
China is experiencing the worlds’ largest scale of internal migration. Most migrants are
from rural areas, where a surplus of rural labor produced by the lack of arable land has forced
working age adults to look for opportunities in places other than their natal villages. On the
one hand, the economic development in urban areas has stimulated the migration. On the other
hand, the development of rural enterprise and modernized farms also attract farmers from
areas with low income and/or low average land to move to work (Giles & Mu, 2007; Liu &
Chan, 2001).
Characteristics of migrants
Some characteristics of migrants influence intergenerational relationships with their
parents. First, they have strong bonds with their homelands, and remit substantial amount to
home. A survey in Jinan, capital of Shandong province, in 1995 found that 85% of migrants
remitted some income in the previous 12 months which accounted for over a third of their
urban labor earnings (Liu & Reilly, 2004).

4
Second, most migrants are temporary. A survey in Shanghai showed that most migrants
were not planning to stay permanently, and showed a pattern of circular migration (Wang &
Zuo, 1999). After 1997, the flow of labor back to villages accelerated (De Brauw & Rozelle,
2003). The major underlying reasons are the segregation of urban and rural societies by
“Hukou” (household registration system), which has established barriers for rural migrants to
live permanently in urban areas. In addition, other factors, such as low education and income,
poor benefits, segregated labor market and occupational opportunities, temporary housing,
barriers to children obtaining reasonably priced education, also lead to rural migrants’
temporarily staying (Bai & Song, 2002; Li & Zahniser, 2002; Wang & Zuo, 1999; Wang &
Fan, 2006).
In addition, there is increased demand for child care back home. This is because adult
children have barriers to live permanently in urban areas and expect to come back. Surveys in
Shanghai and Beijing showed that most rural migrants lived alone or in “collective
households” (which referred to many singles sharing one household, e.g. students in a
university, or single employees in a company), without their families and children (Pang, 1996;
Wang & Zuo, 1999). Without a good public child care system in rural China, grandparents are
valuable resources for child care (Chen, Short, & Entwistle, 2000; Hermalin, Roan, & Perez,
1998). When their adult children migrate for better economic opportunities, elders take either
a surrogate parenting role or co-parenting role for their grandchildren, which strengthens the
intergenerational interdependence between migrants and their (extended) families.
The fourth characteristic is the income pattern of the migrants. Migrants on average at
least double their income, and households with current or return migrants are richer than those
with only non-migrants (De Brauw & Rozelle, 2003; Wang & Zuo, 1999). This pattern of
remittances from migrants is described by an inverted U-shape remittance curve, which
suggests that the level of remittances is low at the beginning of migration and then increases

5
gradually and reaches its peak several years after migration. Immediately after migration, at
least in the short run, migrants may have difficulty increasing their income because of the risks
of a new environment and uncertainty of income; households with migrants also face reduced
profits from farming because of the loss of labor (Liu & Reilly, 2004; Rozelle, Taylor, &
deBrauw, 1999). Thus, migrants will turn to their families for a loan to initiate migration, e.g.
to cover transportation expenses or to get protection against the unstable income at the
beginning; and older people will serve as supporters and providers of their children’s needs
(Shi, 1993).
Influences of migration on intergenerational transfers
Migration has strengthened intergenerational interdependence and brought substantial
changes to intergenerational relationships for a number of reasons. First, elders are expected to
take custody of their grandchildren. Grandparents in China traditionally were not obliged to
take primary responsibility for the care of grandchildren. Grandparents were expected to assist
their adult children in providing care but were not expected to take on a custodial role.
Grandmothers intervened if they found that daughters-in-law were not caring for their
grandchildren in a proper way, and grandfathers became involved only in very rare situations
(Chao, 1983). However, because of the high demand for child care resulted from the booming
labor migration in recent decades, both grandmothers and grandfathers have increasingly
taken on full-time care for their grandchildren (Parish & Whyte, 1978; Silverstein, Cong, & Li,
2006).
Second, elders are more likely to receive financial support from children. With migrant
children’s increased economic capacity, elders are more likely to benefit financially (Liu &
Reilly, 2004; Zhang, 2002), because filial responsibility obliges children to care for their
parents’ well-being when they are able to do so (Lee, Netzer, & Coward, 1994a). Additionally,
if elders invest in their children’s migration, their children often pay back the family loan

6
when they become established, that is, migrants who received financial resources from parents
before migration, give more to their parents (Liu & Reilly, 2004). Finally, when elders provide
custodial care of grandchildren to facilitate their children’s migration, they may have
expectations for compensation when such care comes at a greater cost and is perceived as a
favor to children rather than a natural duty (Silverstein et al., 2006).
Third, elders are less likely to live in a stable environment with married sons. Availability
of kin is an important predictor for elders’ opportunity for coresidence, and children’s
migration has reduced the availability of children to coreside. Migrants have a hard time
bringing their nuclear family including children and wives with them and are even less likely
to bring their parents with them when they migrate. Migrant children are not likely to provide
coresidence for their parents. In addition, although the early phase of China’s internal
migration mainly involved men and unmarried women, more and more married women have
become migrants, either with or without their husbands. Therefore, although in the past, when
sons migrated, they left their wives to take care of older parents, now, daughters-in-law are
less likely to be available to provide coresidence and hands-on support to their parents-in-law
(Gaetano & Jacka, 2004).
Theories of Intergenerational Transfers in Chinese Families
The topic of intergenerational transfers is important in the field of gerontology, because
intergenerational transfer motivations and behaviors have implications for how public policies
can be most effective in supporting older people without crowding out private transfers. In the
basic framework there are two kinds of motivations: altruism and exchange (Cox, 1987).
In the framework of altruism, financial transfers from children to parents reflect their
willingness to help their comparatively poor parents without expecting anything in return.
Altruistic motivation predicts that the transfer is positively related to donor’s resources, while
negatively related to recipient’s resources (Becker, 1974; Lucas & Stark, 1985).

7
In the framework of exchanges, elders provide services, such as grandchildren care and
household help, or elders invest in their children’s capability to increase human capital, and
children pay back their investment in various ways (Cox, 1987; Lee & Xiao, 1998; Lillard &
Willis, 1997; Secondi, 1997).
Particularly in societies where children are important resources of old age support, the
corporate group /mutual aid model is used to explain the short or long term relationship
between parents and children (Lee & Xiao, 1998). The corporate group model tries to explain
the long-term arrangements that maximize the family’s well-being, e.g. parents invest in
children’s education, and children are expected to pay back in the future to provide for their
parents’ old age. On the other hand, the mutual aid model focuses on the short-term exchanges
which benefit each side, e.g., those who support children’s migration or are involved in
grandchildren care for the migrant children get more remittances. The corporate group model
and mutual aid model are similar to the tempered altruism model, which avoids the
bipolarization of altruism vs. self-interested exchanges, and emphasizes the cooperation and
benefits of both parties in the relationship (Lucas & Stark, 1985).
Empirical evidence in China with regard to intergenerational financial transfer from
children to parents basically supports the corporate group /mutual aid model. Studies from the
parent’s perspective usually support the idea that the financial support is need-based. Parents
who are female, older, with lower occupation status, lower income, poorer health get more
financial support from children. But there is also evidence that financial support is provided in
exchange for parents’ services and help, e.g., parents can provide housing, child care, and help
in housework (Lee & Xiao, 1998; Shi, 1993; Sun, 2002; Yang, 1996). Among children’s
characteristics, income is consistently an important factor in predicting financial transfer from
children (Shi, 1993; Sun, 2002). Intergenerational support exchanges also occur in other ways,
e.g., elders who provide more help in grandchildren care, household work, and farm labor to

8
children receive more help in household work and farm labor (Li et al., 2004; Liu & Reilly,
2004; Shi, 1993; Sun, 2002; Yang, 1996).
Help from Children and Elders’ Depressive Symptoms
Migration of children substantially reduces elders’ likelihood of living in a stable
environment with married sons, in which daughters-in-law are supposed to provide care. For
Chinese elders, coresidence with children is a strong predictor of elders’ potential for receiving
instrumental support from them (Agree, Biddlecom, Chang, & Perez, 2002; Yan, Chen, &
Yang, 2003). Consequently, coresidence raises elders’ expectation for getting help from their
children. In addition to reducing elders’ possibility of receiving instrumental help, migration of
sons and daughters-in-law may substantially change elders’ expectations of help from sons’
families, both of which may influence their appraisal of the support they receive and their
psychological well-being.
Depression
Depression is a psychological manifestation of stress, which may arise from either life
events or chronic strain, and it is an important risk factor for decline in cognitive ability,
increased functional limitations, need for long-term care, and eventually mortality (Arnsberger,
Fox, Zhang, & Gui, 2000; Chi & Chou, 2000; Mehta et al., 2003; Pearlin, 1989). Compared
with western elders, Chinese elders have lower levels of depression with an estimated 4%
prevalence. This lower level in depression is partly attributed to stronger family support
(Parker, Gladstone, & Chee, 2001). Despite the low prevalence, Chinese rural elders have a
prevalence of 5-6%, about twice that of urban elders (2-3%) because of the poverty and the
lack of a formal safety net (Chen, Hu, Qin, Xu, & Copeland, 2004; Chen et al., 2005).
Intergenerational support and depression
Social support is usually cited as an important factors in reducing depression either as a
resource for general health maintenance or as a resource to counteract the effect of stress to

9
produce a positive psychological outcome (Pearlin, 1989; Wheaton, 1985). Intergenerational
support is a particularly important source of social support for rural elders. Inability to secure
children’s help presents a very stressful situation, and suggests unmet demands and lack of
help in stressful situations. In addition, it may be regarded as shameful and represent a failure
in one’s childrearing. In a culture which values others’ appraisal of one’s life, failure to attract
children’s attention and help is often viewed as a failure which can be a cause of depression.
The consequences of fulfilling or violating normative rules governing the allocation of support
duties would seem to be pronounced in the Chinese cultural system, which has well-defined
family roles and few formal alternatives.
Different prescribed responsibilities among children
Although intergenerational transfers and help are important to elders, their consequences
are equivocal. Although it is assumed the support will reduce depression, empirical evidence
does not always support this hypothesis. Many contextual factors may moderate the effect of
support on elders’ depression. In Confucian countries such as China and Korea, children are
legally and morally obliged to be financially responsible for their parents (Shi, 1993; Sung,
1995). However, this belies the fact that expectations vary according to the type of child
considered. In most Asian societies, a strict hierarchical order usually exists among children
depending on family and gender norms (Knodel & Ofstedal, 2002; Lin et al., 2003). Raised in
a patrilineal culture, over 80% of Chinese rural elders believe that sons are the best providers
for old age, which reflected their strong son preference and high expectations for receiving
help from sons’ families (Chen & Silverstein, 2000).
Concerning instrumental support from children, the traditional belief that sons are the
best providers for parents’ old age reveals three nested expectations. First is the expectation
that children should be responsible for their parents. The second is the gendered expectation
that among children, sons and their families, rather than daughters and their families, should

10
provide for parents’ old age. In this family system, sons and daughters-in-law are clearly
defined as central family members, while daughters and sons-in-law are peripheral to the
family core. Their exchange is almost more characteristic of that of friends, such that their
exchanges with parents are more likely to be short-term. Sons-in-law are not active
participants in providing instrumental support in a patrilineal family system (Antonucci &
Jackson, 2003; Yang, 1996; Zhan, 2004). The third is the underlying gender norms, specifying
that within sons’ families female members, i.e., daughters-in-law, are the expected providers
of instrumental support because giving care or hands-on help is regarded as women’ work. In
this way, the obligations of daughters-in-law will preempt that of sons and daughters in
providing instrumental help to older people (Cohen, 1998; Lin et al., 2003; Youn, Knight,
Jeong, & Benton, 1999).
Migration and changes of coresidence
Although expectations are stipulated by cultures, they are contextual, as literature
relevant to U.S. elders indicates, to elders and children’s gender, proximity, needs and
availability of resources, as well whether a relationship is affinal or consanguineal (Ganong &
Coleman, 1999; Killian & Ganong, 2002; Rossi & Rossi, 1990; Silverstein & Angelelli, 1998).
In rural China, coresidence with children, particularly sons, is strongly endorsed by
cultural values and forms a basis for promoting intergenerational exchange (Yan et al., 2003;
Zhang, 2004). Coresidence with a married son strongly increases expectations of receiving
household assistance from daughters-in-law, overriding all other sources of support. Thus,
elders who live with daughters-in-law but lack their household assistance might find the
situation to be disappointing and distressing. In addition, when daughters-in-law are available,
support from sons or daughters may be viewed as culturally inappropriate, as it would put
intransigent daughters-in-law into sharp relief as violators of the cultural order. However, in
the absence of a daughter-in-law in the household, elders may adapt their expectations to

11
include alternatives that occupy a lower rank in the hierarchical order, as research has shown
among similarly situated Chinese and Korean immigrants to the U.S. (Pang, Jordan-Marsh,
Silverstein, & Cody, 2003; Wong, Yoo, & Stewart, 2006). Therefore, Chinese elders who do
not coreside with daughters-in-law will have lower expectations of them and regard support
from sons and daughters more favorably. The impact of intergenerational support will likely
be contingent on coresidence with daughters-in-law. Migration has reduced the availability of
daughters-in-law in the household, and will possibly influence the way elders’ appraise
support from sons, daughters and daughters-in-law.
Research Questions
The first paper addresses the financial support from children to their older parents against
the background of high migration rate. Specifically, I ask the research question: Will migrant
children pay back their parents’ for help more than non-migrant children. The second paper of
this dissertation addresses the opportunity of elders’ to coreside with a child. I ask the research
question: Will a parent’s previous help to a non-migrant child increase this parent’s odds of
living with that child? And will parents’ previous exchanges with migrant siblings’ of a
non-migrant child increase the odds of parents’ living with this non-migrant child? Will
children who received parents’ help when they were migrants be more likely to coreside with
parents? The third paper addresses the impact of instrumental help from sons, daughters, and
daughters-in-law on their parents’ depressive symptoms within different residential contexts.
In this paper, I ask the question: Will the impact of sons, daughters and daughters-in-law differ
in influencing their parents’ depression? Are these effects conditional on the availability of
daughters-in-law in the household? Are these effects conditional on the elder’s gender?
Contributions
This dissertation emphasizes the importance of studying families from two perspectives.
They are elaborated in the following.

12
Reciprocity and strategic investment
The strategic investment and bargaining power perspective expand the reciprocity
principle usually discussed in the literature, particularly in the corporate/mutual aid model, by
recognizing parents as strategic investors who can improve their situations by strategically
engaging in intergenerational exchanges with certain children (Becker, 1986). In the first and
second papers, guided by this reciprocity, strategic investment and bargaining power
perspective, I examine how previous help from parents may benefit parents in the form of
financial support from children and opportunities to coreside with children.
The literature usually shows that migrants give more financial transfers to their parents,
but no research has addressed whether migrants give more to their parents because they have
received more care for grandchildren and financial help or because they provide a higher
return for the same amount of previous help from their parents. If parents get a higher return
from migrant children, then elders at least have some power to guarantee their sharing in the
wealth of their migrant children. But if not, i.e., older parents have provide more resources to
gain more financial support from their children, the cost older people have to pay for
providing grandchildren care, such as strain in physical and psychological health as suggested
by many U.S. grandparenting studies, will deserve attention (Kelley, Whitley, Sipe, & Yorker,
2000; Szinovacz, DeViney, & Atkinson, 1999). Consequently, the extent that elders may
benefit from children’s migration will be lower than expected. And parents are not really
benefit from children’s migration. These kinds of intergenerational exchanges may benefit
children more than elders.
With regard to coresidence, few studies have linked intergenerational exchanges with
parents’ odds of coresiding with children (Attias-Donfut, 2003). This dissertation emphasizes
the importance of intergenerational exchange in predicting elders’ opportunities to coreside
with a child. Following the same strategic investment and bargaining power perspective, this

13
dissertation studies whether help to migrant children will be more effective in increasing the
odds that parents coreside with these children when they become feasible providers for
coresidence, compared with the help to non-migrant children. This section uses coresidence
rather than financial support as a measure of the benefit elders can expect by providing help to
their migrant children.
Linked lives within siblings: An extended family perspective and network interactions
Despite the usefulness of the corporate group /mutual aid model in explaining
intergenerational transfers, little research has analyzed the entire extended family, i.e.,
including parents and all their children’s nuclear families.
Interactions between a parent and a child can not be isolated from interactions between
that parent and the siblings of that child, and the interactions among all the children. For
example, Checkovich& Stern (2002) shows that the care provided by siblings is correlated
with each other. Enger (2002) shows that when there are daughters providing care, siblings
will reduce their provision of care, and Pezzin and his colleagues (2006) conceptualizes the
decision of parents’ living arrangements and the consequent care allocation as a two-stage
game involving an elder parent and two children, and illustrates that the living arrangement
and care provided by each child is a joint decision by family members with different
preferences and facing different constraints.
In China and other countries where complicated interactions and social exchanges are
carried out within extended families, it is more necessary to take an extended family
perspective in studying intergenerational transfers. The current cohort of elders in rural China
on average have 4 children (China Research Center on Aging, 2003). Dyadic parent-child
relationships are only one part of this network of interactions. It influences and at the same
time is influenced by other components of this network. To study exchanges between a parent
and a child within the context of exchanges between that parent and all of his or her children

14
will substantially enhance our knowledge about familial mechanisms which have largely been
ignored.
In the second and third paper, I emphasize the theme that children’s relationships and
interactions with their parents can not be limited to the dyadic relationship. In the second
paper, I examine how migrant children’s exchanges with parents influence parents’ odds of
coresiding with a non-migrant child, guided by the perspective that the parents’ living
arrangement is a joint decision derived from bargaining and negotiation within the extended
family (Pezzin, Pollak, & Schone, 2006). To single out a dyadic relationship will have
substantially simplified what happens in rural Chinese families, and ignored the linked lives
among siblings.
In the third paper, in addition to explaining a long debated effect of instrumental support
on elders’ psychological well-being from the cultural perspective, I emphasize that the effect
of children’s support on elders’ psychological well-being is not independent of other children’s
availability. Particularly, in rural China, daughters-in-law are expected providers of hands-on
support. Whether daughters-in-law are available in the household to provide support
significantly influences how elders appraise support from their sons and daughters, and
consequently influences the impact of intergenerational support on elders’ psychological
well-being. The migration of children has substantially changed the availability of
daughters-in-law in the household, and elders may adjust their expectation and their appraisals
of help from their sons, daughters and daughters-in-law.
In the following three chapters, I present these three papers which address the
overarching question about how rapid social change combined with traditional beliefs in rural
China shapes and changes intergenerational relationships and well-being of the rural elders.

15
Chapter 2: How do Chinese Rural Elders Share the Wealth with their Migrant Children?
Abstract
This investigation examines in rural China whether parents’ help in money and care for
grandchildren lead to higher financial return from migrant children than from non-migrant
children. The data were from a two-wave (2001, 2003) longitudinal study of 4,101
parent-child dyads from 1,147 parents, aged 60 and older, living in rural areas of Anhui
Province, China. Random effects regression analysis showed that elders reaped higher
financial return when they provided grandchildren care and financial help to migrant children.
In fact, parents’ help in money and care for grandchildren to non-migrant children did not
oblige them to increase their financial support to parents, but parents’ help to migrant children
lead to higher financial return from them. We explained that in the Chinese family, children
did not need to reciprocate parents’ help in short run. But the higher bargaining power parents
had when they provided care for grandchildren to migrant children, and parents’ financial help
to migrant children in the form of strategic investment, obliged children to pay back parents’
help.

16
In rural China, elders in most areas do not have a formal support safety net and more than
two thirds of them depend exclusively on their children for financial support (Joseph &
Phillips, 1999; Lee & Xiao, 1998; Shi, 1993; Zimmer & Kwong, 2003), consequently the
majority of them regard investing in their children as the first method for ensuring their
economic security in their old age (Chinese Center of Research on Aging, 2003). Instead of
the unidirectional upward flow of resources from children to parents in their old age as
endorsed by Confucianism (Sung, 1998), the intergenerational relationships in current rural
China resemble a bidirectional reciprocal relationship. Particularly important in promoting
parents’ old age support from children are parents’ recent and continuing assistance to children
including providing care for grandchildren and maintaining households to promote children’s
economic capacity, and providing financial help when children get married or are in other
need (Chen, 2001a; Liang, Gu, & Krause, 1992; Yan et al., 2003).
The large scale migration of working-age adults from rural to urban China stimulated by
fast urban economic development has strengthened migrants’ reliance on their parents for free
child care and family loans to initiate migration (Bai & Song, 2002). Compared with
non-migrants, migrant children, who move to urban areas for higher wages, usually provide
more financial support to their parents (Shi, 1993), but the underlying reasons are not clear. It
may be because elders provide more help to their migrant children, or it may also be because
elders get higher marginal rate of return from their help to migrant children than from help to
non-migrant children.
Guided by theory of intergenerational transfers and the perspective that facilitating family
members’ migration and receiving remittances is a family strategy for promoting the
well-being of the whole family (Stark, 1991), we compare the behavior of migrant and
non-migrant children in their provision of financial support to their parents, and examine
whether parents who previously provided help to children reap more return from migrant than

17
non-migrant children. This has theoretical implications regards the motivations used by
children to give monetary support to their parents.
Literature Review: Theoretical Approach
The idea that intergenerational transfers strive toward balance or symmetry has been used
to explain many forms of reciprocal exchanges in the fields of family economics (Bernheim,
Shleifer, & Summers, 1985; Cox, 1987);, social demography (Agree et al., 2002; Henretta,
Hill, Li, Soldo, & Wolf, 1997; Silverstein, Conroy, Wang, Giarrusso, & Bengtson, 2002) and
social psychology (Antonucci, Akiyama, & Birditt, 2004). The corporate/mutual aid model
regards the Chinese family as a corporate group following the principle of reciprocity, which
takes care of each member’s needs and maximizes their combined benefits based on the
capacity of each member; particularly, the mutual aid model focuses on the short-term
exchanges of intergenerational transfers that benefit each side (Lee & Xiao, 1998).
In societies with few public supports, mutual aid is an essential adaptation of families to
optimize the satisfaction of needs through the diffusion of resources (Agree et al., 2002;
Hermalin, 2002; Lee, Parish, & Willis, 1994b; Sun, 2002; Yang, 1996). Empirical evidence
usually provides support for corporate/mutual aid model in explaining why and how
intergenerational transfers are carried out in China. Children give financial support in
exchange for services and help from parents, including help in housing, housework, and
grandchildren care (Lee & Xiao, 1998; Shi, 1993; Sun, 2002; Yang, 1996). In addition, elders
receive more help in farm labor and household help if they provide help in household work,
child care, and farm labor to children (Li et al., 2004).
This article focuses on two important resources elders can provide to their children,
grandchildren care and financial support. Although elders provide both of them to non-migrant
well as migrant children (Shi, 1993), these two resources are particularly important to migrant
children because these resources help them to overcome barriers to migration.

18
In Asian families the “time-for-money” hypothesis is often advanced under the mutual aid
model, where parents provide child-care labor to the families of their adult children, in
exchange for transfers of money or food (Frankenberg, Lillard, & Willis, 2002; Lee et al.,
1994b). Migrant children also follow the same norms and migrants whose children are taken
care of by their parents give more money to their parents (Secondi, 1997; Silverstein et al.,
2006). In addition, this kind of exchange is usually simultaneous, i.e., financial support is
provided at the same time as grandchildren receiving care (Silverstein et al., 2006; Yang,
1996).
But do migrant and non-migrant children reciprocate for their parents’ provision of
grandchildren care to the same degree? The bargaining power model would say yes. This
model usually suggests that parents who have more resources and therefore higher bargaining
power in the exchange receive more from their children (Lee et al., 1994b). Usually, resources
refer to the property ownership which seduces children to contribute to their parents because
of their intention to inherit (Bernheim et al., 1985; Cox, 1987; Lucas & Stark, 1985). In China,
because of the collective ownership of land, land can not be inherited after the death of parents.
In addition the serial division of the family, i.e.., transferring assets to each son when they gets
married, depletes elders’ resources in advance (Yan et al., 2003). Therefore, the bargaining
power of elders resulting from children’s expectation to inherit is largely diminished. But the
bargaining power may also result from elders’ ability to provide certain services that are not
easily available elsewhere, or they are too expensive to access. Parents and children’s relative
bargaining power and position in the exchange help to explain why some children reward their
parents for similar help more than other children.
In China, grandchildren care is a valuable resource that older parents can provide
(Secondi, 1997; Sun, 2002; Yang, 1996), which confers elders greater bargaining power over
migrant children than non-migrant children. Most migrants from rural China are temporary

19
migrants because of the segregation of urban and rural societies by ‘Hukou’(household
registration system) and have difficulties in taking children with them because of the high
living, educating and child care expenses (Bai & Song, 2002; Zhao, 2005). Because of the
higher demand for care for grandchildren, which often involves grandparents’ full time
custody such as in skipped-generation households (Silverstein et al., 2006), and higher
expenses to replace parents’ help, elders have higher bargaining power over migrant children
than over non-migrant children when they are providing grandchildren care. Because of this,
we expect that migrant children will reward their parents better than non-migrant siblings for
the same amount of grandchildren care they receive from parents.
Whether elders can benefit from providing financial support to children depends on
whether the children are migrants or not. Although it is observed that elders who provide
financial support to children are not more likely to receive financial support from children
(Chen, 2001a; Li et al., 2004; Shi, 1997), migrant children who receive more financial help
from parents, such as help in training expenses and other expenses before migration, provide
more financially to their parents (Liu & Reilly, 2004).
The different response to elder’s financial help among migrants and non-migrants may be
related to different purposes for financial help from parents. Financial support to non-migrant
children may be a result of transfers to less capable children, or may be transfers that are
regarded as parents’ obligation such as wedding costs, or may be a loan to cover some urgent
need. In each of these cases, parents’ financial support is normative and will not be repaid, or
repaid only in long-term when children fulfill their filial responsibilities to provide for their
parents old age security (Frankenberg et al., 2002). This repayment is hard to observe with
short time span research design.
But financial support to migrant children helps to cover start up expenses including
transportation fees, training expenses etc., and insure against uncertainties and unstable

20
income at the beginning of migration when migrants are financially vulnerable. Financial
support of this kind is either conceptualized as a family loan, investment or insurance when
there is a lack of formal money market, and sometimes is explicitly understood as a family
decision that aims to maximize the benefits for the whole family, i.e., not only the benefit of
migrants, but also those who provide help and then benefit from increased remittances from
migrants (Liu & Reilly, 2004; Poirine, 1997; Stark, 1991). When it is applied to
intergenerational support, parents will strategically be more involved in helping migrant
children or children who have the potential to migrate and consequently benefit from children’
remittances enabled by children’s increased income in urban areas. This mechanism is similar
to that parents are more likely to invest in the education of children who are more likely to
succeed and benefit from education (Becker & Tomes, 1986; Poirine, 1997).
Although this strategic investment model looks similar to a mutual aid model in that those
children who receive more financial help from parents will give back more, this model further
predicts that parents are expected to get a higher return from migrant than non-migrant
children. We expect a time lag between the financial assistance provided by parents and
financial return from children because of its nature of paying back debt or investment.
Following the rule of reciprocity, we expect that previous help from parents will also
result in daughters’ paying back. But we do not expect that the migration status will
distinguish daughters as we have discussed above, because most married daughters are
intrinsic migrants, defined here as leaving one’s natal birth village in rural China, as a result of
exogamous marriage (Das Gupta & Li, 1999).
Hypothesis
Based on the classification of migration status, our hypotheses are:
1) Support from children is positively related to parents’ help in providing care for
grandchildren and financial support.

21
2) Migrant sons give more financial transfer to their parents than non-migrant sons.
3) The marginal return of parents’ help in grandchildren care and financial help is higher for
migrant sons than for non-migrant sons.
Methods
Sample
The sample for this investigation was derived from the Anhui Province of China, a mostly
rural province and the fifth most populous province in China. Currently, 12% of the rural
population is 60 years of age and older (compared to only 8.5% of the nation) making it one of
the most elderly provinces in China. This region was chosen specifically for its relatively high
density of older adults and high levels of out-migration of working age adults. Between 1995
and 2000, Anhui Province had the third highest rate of out-migration among all provinces in
China, and a higher than average rate of labor-related migration. Data were collected from a
sample of adults age 60 and over living in rural townships within Chaohu, a city of 4.5 million
people located on the north bank of the Yangtze River in the central part of Anhui Province.
This rural area of the province is generally known for its high rates of labor migration to the
cities of Hefei, Nanjing, and Shanghai.
The sample was identified using a stratified multistage method to randomly select 1,800
potential respondents. First, 12 rural townships were randomly selected from all 126
townships in Chaohu. Second, 6 administrative villages were randomly selected in each
township. Third, within each selected village, all people aged 60 and older were stratified to
form two sampling frames based on age: (1) those aged 60-74, and (2) those 75 and above,
providing an intentional over-sample of the 75+ population. Of 1,800 individuals randomly
selected for the study, 1,715 completed the survey, yielding a response rate of 95.3%. The
completed sample included 829 men (48.8%)and 869 women (51.2%). In terms of age, 61.2%
were 65-74 years old and 38.8% were 75 years and older. In October 2003 the follow-up

22
survey was conducted with 1,368 respondents, or 79.8% of the original participants. Of those
respondents who were not located, 76 had moved out of the village, and 240 died.
Twenty-three former respondents were located but refused to participate, terminated their
interviews, and/or were too ill to be interviewed. The analytic sample consists of 1,324 older
people who participated in both wave of interviews and who had at least one living children.
After deleting those with missing values in relevant variables, we have 1,147 elderly with
4,101 corresponding children.
Dependent variable
Financial transfers from children in the second wave were based on the total amount of
money that the parent received from each child during the past 12 months. Respondents
(parents) were asked to provide the exact amount of money first, and if they could not give a
exact number, they were asked to choose among the following categories based on Chinese
RMB currency (100 RMB = $12US): 0= “none”, 1= “less than 50”, 2= “50-99”, 3= “100
-199”, 4= “200-499”, 5= “500-999”, 6= “1000-2999”, 7= “3000- 4999”, 8= “5000 to 9999”,
9= “More than 10,000”. In the analysis, we took the actual amount if it was available and then
used the median amount of the category if the exact amount was not supplied. Then each
category correspond to 0=0, 1=25, 2=75, 3=150, 4=350, 5=750, 6=2000, 7=4000, 8=7500,
9=10,000.
To minimizing the risk of endogeneity, we controlled for first wave financial support
from children, which was measured in the same way as second wave financial support from
children. In this way, coefficients of other variables in the model indicated their effects on
residualized change in financial changes from children between waves (Nepomnyaschy,
2007).

23
Independent variables
The migration status of children was represented by three dummy variables. The
reference group was non-migrants, i.e. those who stayed in the same village with their parents
(either coresided with parents or not) over two interview periods. The three dummies
indicating children’s migration status referred respectively to recent migrants, i.e. who lived in
the same village with parents in the first wave but transited into living outside the village in
the second wave, established migrants, i.e., those who lived outside the village for both waves,
and return migrants, i.e., those who lived outside the village in the first wave but returned to
the village for the second wave.
This categorization reflected two important differences among those who shared the same
proximity with their parents at the second wave. First, we divided those who lived in the same
village with their parents in the second wave into non-migrants and return migrants. In China,
most rural to urban migrants were temporary migrants because of the barriers they faced in
staying permanently (Bai & Song, 2002; Wang & Zuo, 1999). A study of Anhui migrants
found a high rate of return, and that return migrants were negatively selected from migrants
and might represent failure (Wang & Fan, 2006), which would make them different from
non-migrants in their financial transfer behavior. Second, we divided those who were not in
the village in the second wave into recent migrants and established migrants. The pattern of
remittances of migrants is describe by an inverted U-shape remittance curve, which suggests
that the level of remittances is low at the beginning of migration and then increases gradually
and reaches its peak several years after migration. Immediately after migration, at least in the
short run, migrants may have difficulty in increasing their income because of the risks in new
environment and uncertainty of the income, and households with migrants also face reduced
profit from farming because of the loss of labor (Liu & Reilly, 2004; Rozelle et al., 1999).

24
The variable measuring the frequency of providing care for grandchildren in both waves ranges
from 0 to 6. The values of these variables were defined as follows: 0= “not taking care of
grandchildren”, 1= “seldom”, 2= “about once a month”, 3= “several times a month”, 4= “at least
once a week”, 5= “A period of a day (not the whole day)”, 6= “The whole day, from morning to
evening.” We included variables representing baseline level grandchildren care and change in care
for grandchildren between the two waves. This approach minimized the risk of endogeneity in
the event that financial support from children encouraged parents to provide grandchildren
care.
Similar to financial transfers from children, financial transfers to children were based on
the exact amount that each child received from the parent during the past 12 months. If this
number was not available, we took the median value of the categorical responses. We also
included variables of the baseline financial support from parents and its changes over time.
Because grandchildren care was more likely to be reciprocated simultaneously, we interacted
changes in grandchildren care with migration status of children to examine whether parents reaped
higher marginally return from migrant children by providing grandchildren care. Because there was
usually a time lag for loaned or invested money to receive pay back, we interacted baseline financial
support from parents with migration status of children to examine whether parents’ financial
investment on migrant children led to higher marginally financial return from migrant children.
We controlled for important parents’ characteristics and children’s characteristics, which
were shown to influence financial transfers between them (Li et al., 2004; Liu & Reilly, 2004;
Shi, 1993). Parents’ characteristics at baseline included age in chronological years, number of
children, and dummy variables for gender (1 = female), marital status (1 = married), education
(1 = some formal education), and occupation (1 = agricultural work). In addition, parents’
health status was measured as the extent of functional impairments, calculated as the sum of
15 items reflecting difficulties in performing personal activities of daily living, instrumental

25
activities of daily living and activities requiring physical strength, mobility, and flexibility.
Respondents indicated the level of difficulty performing each task: 0 (no difficulty), 1 (some
difficulty), or 2 (cannot do it without help). Because these items had high reliability (alpha
= .96), we calculated a summed scale that ranged from 0 (no difficulties) to 30 (unable to
perform all tasks). Income was measured as the logged RMB value of the total annual income
of respondent and spouse from work or pension (+1).
Children’s characteristics at baseline included their age, gender, education, and marital
status. Age was represented as age in years at the time the survey was carried out. Gender was
codes as 1= “female”; martial status (1= “currently married”); education (0= “no education”,
1=“primary school“, 2=“junior middle school“, 3=“senior middle school, vocational training,
college, university or above“).
We also controlled for the baseline instrumental support to children since parents’
instrumental support had also shown to induce financial pay back from children (Li et al.,
2004; Sun, 2002). In addition, we controlled for children’s help to parents in farm labor and
instrumental support in the second wave to avoid the confounding effects that might result
when parents’ previous help was reciprocated by children in farm labor and instrumental help,
and when children who gave money were also those who gave time because they had stronger
instilled filial piety norms (Altonji, Hayashi, & Kotlikoff, 2000; Lee et al., 1994a).
Parents’ instrumental support to each child was measured by a dummy variable with 1
meant that the parent gave this child and/or the spouse of this child any help during the past 12
months in two areas: (1) household tasks, such as cleaning the house and washing clothes, and
(2) personal care tasks such as bathing and dressing.
Children’s instrumental support to their parents was measured in a similar dummy
variable with 1 meant that the parent received any of the two dimensions of help from a child
and the spouse of this child during the past 12 months. Children’s farm labor help was

26
measured also as a dummy variable with 1 meant this child gave any farm labor help to the
parent during the past 12 months.
Model
We used Stata to estimate the random effects model with parents’ characteristics as well
as children’s characteristics in the model to examine children’s financial support to their
parents conditional on children’s migration status and parents’ help in providing grandchildren
care and financial support. We did the analysis for sons and daughters separately.
Results
In Table 2.1, we show the descriptives of variables for sons, daughters and sons and
daughters combined. For the total sample, at baseline, the average age of parents was 69.5
years, slightly more than half (52%) was female, 63% was married, 23% had some formal
education, and a large majority (93%) was currently or previously engaged in some form of
agricultural work. The average functional impairment score was 4.5 out of a possible 30. The
average respondent had more than four living children.
As shown in Table 2.1, daughters and sons resembled each others in age (averaged 40),
marital status (96% and 95% were married), but daughters had substantially lower education
than sons. In fact, eighty percent of sons received some level of education, whereas only 45%
of daughters received some education.
Seventeen percent of sons provided farm labor help and 20% of sons provided
instrumental help to their parents in the second wave. Twenty one percent of sons received
instrumental support from their parents at baseline. Although similar proportion of daughters
provided farm labor (15%) and instrumental support (16%) to their parents in the second wave,
only 4% of them received any instrumental support from their parents at baseline.
Sons’ financial support to parents averaged 4.65, which was transformed to RMB 104.
When compared with parents’ average income, 4.07, which was transformed to RMB 58,

27
Table 2.1
Description of Analytic Variables
Sons (n = 2102)
Parents (n=1035)
Daughters (n = 1999)
Parents (n=986)
Total (N = 4011)
Parents (n=1147)
Coding scheme and range
M SD M SD M SD
Second wave financial transfers from children
(ln+1)
4.65 2.26 4.61 1.77 4.63 2.03 0-9.393
Baseline financial transfers from children (ln+1) 4.32 2.18 4.50 1.54 4.41 1.90 0-9.210
Parents’ characteristics
Age 69.64 6.70 69.27 6.72 69.46 6.71 58-89
Female 0.53 0.50 0.51 0.50 0.52 0.50 0 (male), 1 (female)
Married 0.62 0.49 0.64 0.48 0.63 0.48 0 (not married), 1 (married)
Education 0.22 0.41 0.24 0.42 0.23 0.42 0 (no education), 1 (some education)
Income 3.94 3.52 4.21 3.47 4.07 3.50 0.000-9.393
Occupation 0.92 0.27 0.94 0.24 0.93 0.26 0 (other occupation), 1 (agricultural work)
Functional limitations 4.68 6.23 4.38 6.02 4.53 6.13 0 (none)-30 (most)
Number of children 4.53 1.51 4.69 1.50 4.61 1.51 1-10
Children’s characteristics
Age 40.15 8.30 39.54 8.75 39.85 8.53 22-70
Female 0.49 0.50 0 (male), 1 (female)
Married 0.96 0.20 0.95 0.21 0.96 0.20 0 (not married), 1 (married)
Education 1.37 0.90 0.62 0.79 1.01 0.93 0-3
Farm labor help from children (T2) 0.20 0.40 0.22 0.41 0.21 0.41 0 (no help) 1 (some help)
Instrumental help from children (T2) 0.20 0.40 0.16 0.36 0.18 0.38 0 (no help) 1 (some help)
Instrumental help to children (T1) 0.21 0.40 0.04 0.20 0.13 0.33 0 (no help) 1 (some help)
Migration Status of Children
Non-migrant (reference)
Recent migrant (M1) 0.12 0.32 0.05 0.22 0.08 0.28 0, 1
Established migrant (M2) 0.41 0.49 0.75 0.43 0.58 0.49 0, 1
Return migrant (M3) 0.07 0.25 0.05 0.21 0.06 0.23 0, 1
Parents’ help to children
Baseline grandchildren care 1.60 2.31 0.43 1.26 1.03 1.96 0-6
Change in grandchildren care -0.54 2.18 -0.18 1.28 -0.37 1.81 -6-6
Baseline financial transfers (ln+1) 1.24 2.00 0.95 1.69 1.10 1.86 0-8.923
Change in financial transfers (ln+1) -0.15 2.26 -0.22 1.93 -0.18 2.11 -7.601 - 8.700

28
children contributed a large part to parents’ economic well-being. Across two interviews,
40.3% of sons stayed in the village (non-migrants), 12% of sons began their migration (recent
migrants), 41% stayed outside the village (established migrants), and 6.8% returned from
outside (return migrants). The average score for grandchildren care that sons received was 1.6.
The low value might partly be attributable to the nature of the data in which one elder had
several children but only provided grandchildren care for some of them. The financial
transfers from parents to sons averaged 1.24 which was transformed to RMB 2.46. This small
number reflected that only a small portion of children received financial support from their
parents. From 2001 to 2003, parents reduced their financial support and grandchildren care to
children, reflecting their deteriorated health and diminished ability to provide help to children.
Seventy five percent of daughters are classified as established migrants because they
married out of the village; 5.1% and 4.7% are recent and return migrants, and 15.1% are
non-migrants. When compared with sons, daughters provided similar amount of financial
support to parents in both waves, but received much less in grandchildren care and financial
support at baseline. Daughters also received less grandchildren care and financial support from
parents in the second wave than in the first wave. In the whole, daughters provided similar
support to their parents in every dimension, but received far less transfers form parents both in
time and money.
We show in Table 2.2 the intergenerational exchanges between parents and children of
different migration status. For sons, established migrants gave most to their parents both at
baseline and in the second wave; they also received most financial help from their parents at
baseline. Sons who were in the village (i.e., non-migrants and return migrants) in the second
wave provided more support in farm labor and instrumental support to their parents in the
second wave. Sons who were in the village at baseline (non-migrants and recent migrants)
received more instrumental help from parents at baseline. This reflected a proximity effect that

29
Table 2.2
Intergenerational Exchanges by Children’ s Migration Status
Sons (n=2102) Daughters (n=1999)
Non-migrant
(n=849)
Recent
migrant
(n=245)
Established migrant
(n=864)
Return migrant
(n=144)
Non-migrant
(n=302)
Recent migrant
(n=101)
Established
migrant (n=1503)
Return
migrant
(n=93)
Financial transfers
from children (T2)
4.431 4.509 4.969 4.273 4.506 4.435 4.676 4.169
Financial transfers
from children (T1)
4.016 4.113 4.708 4.153 4.114 4.381 4.597 4.362
Farm labor help from
children (T2)
0.259 0.176 0.084 0.194 0.199 0.119 0.135 0.237
Instrumental help from
children (T2)
0.313 0.180 0.069 0.278 0.325 0.119 0.123 0.183
Instrumental help to
children (T1)
0.287 0.290 0.108 0.181 0.126 0.109 0.022 0.032
Grandchildren care to
children (T1)
1.125 2.171 1.815 2.146 0.775 0.683 0.348 0.355
Change in
grandchildren care to
children
-0.550 -0.531 -0.549 -0.438 -0.397 -0.287 -0.134 -0.194
Financial transfers to
children (T1)
0.900 1.101 1.590 1.380 0.862 1.061 0.944 1.129
Change in financial
transfers to children
-0.191 0.085 -0.166 -0.187 -0.238 -0.288 -0.206 -0.336

30
distance was an important barrier for children to exchange hands-on help, including
instrumental help and farm labor with their parents (Bian, Logan, & Bian, 1998). Recent
migrant sons received most grandchildren care from their parents at baseline, which might be
because parents provided grandchildren care and recent migrants felt they were free to migrate.
The comparatively lower intensity of grandchildren care to established migrant sons might
reflect established migrants’ improved capacity to bring their children together with them.
Non-migrant sons got least grandchildren care from their parents because they might need
grandchildren care only when they were at work, which substantially reduced their parents’
grandchildren care intensity. From first to second wave, parents reduced their services in
grandchildren care to sons. Since the reduction was similar in scale across sons’ migration
status, this might be an effect of age that elders had to reduce their grandchildren care because
of deterioration of health. At the same time, elders also reduced their support to their sons
financially with one exception that they increased financial support to their recent migrant
sons a little bit, which might reflect the unstable position of recent migrants who needed
parents’ support at the beginning of their migration.
Therefore, generally speaking, migrant sons received more child care, financial help, less
instrumental help from parents at baseline, and provide more financial help in the follow-up.
For daughters of different migration status, the differences in exchanges with parents
were smaller in scale. Similar as sons, those who were in the village for a particular wave were
more likely to exchange hands-on support with their parents. Daughters who were in the
village at baseline (non-migrants and recent migrants) received more grandchildren care at
baseline than established and return migrants, possibly because those who married into the
same village received more grandchildren care than those who married into other places.
Table 2.3 showed the estimates for coefficients for the influence of migration status and
parents’ help in providing grandchildren care and financial support. Parents’ income was

31
Table 2.3
Random Effects Models Predicting T2 Financial Transfers from Adult Children to Their Older Parents
Sons (2102/1035) Daughters (1999/986) Total (4101/1147)
Model 1 Model 2 Model 3 Model 1 Model 2 Model 3 Model 1 Model 2 Model 3
Baseline financial transfers from children 0.314*** 0.305*** 0.305*** 0.277*** 0.275*** 0.273*** 0.313*** 0.307*** 0.305***
Parents’ characteristics
Age 0.012 0.007 0.007 -0.012 -0.011 -0.010 0.001 0.001 0.001
Female -0.061 -0.045 -0.060 -0.060 -0.062 -0.064 -0.057 -0.052 -0.059
Married -0.196 -0.103 -0.099 -0.008 0.002 0.000 -0.120 -0.074 -0.070
Education 0.169 0.156 0.154 -0.012 -0.013 -0.011 0.044 0.040 0.036
Income -0.034 -0.041+ -0.038+ -0.042* -0.042* -0.041* -0.037* -0.040* -0.039*
Occupation -0.133 -0.144 -0.145 -0.142 -0.139 -0.150 -0.162 -0.162 -0.163
Functional limitations 0.011 0.003 0.004 -0.008 -0.009 -0.009 0.001 -0.004 -0.004
Number of children 0.075+ 0.102* 0.104* 0.011 0.016 0.017 0.037 0.053+ 0.054+
Children’s characteristics
Age 0.006 0.007 0.006 -0.006 -0.006 -0.007 -0.004 -0.005 -0.005
Female 0.081 0.040 0.050
Married 0.136 0.066 0.022 0.152 0.166 0.173 0.182 0.177 0.160
Education 0.152** 0.147** 0.142** 0.096+ 0.098* 0.094+ 0.145*** 0.143*** 0.143***
Farm labor help from children (T2) 0.281* 0.285* 0.078 0.100 0.226* 0.232**
Instrumental help from children (T2) 0.758*** 0.772*** 0.232* 0.228* 0.552*** 0.552***
Instrumental help to children (T1) 0.441*** 0.325** 0.344** 0.369+ 0.332+ 0.359+ 0.332** 0.245* 0.279**
Migration status
a

Recent migrant (M1) 0.009 0.130 0.013 -0.149 -0.101 0.122 -0.109 -0.012 0.006
Established migrant (M2) 0.424*** 0.624** 0.554*** -0.011 0.034 0.044 0.197** 0.330*** 0.259**
Return migrant (M3) -0.138 -0.103 -0.084 -0.356+ -0.336+ -0.247 -0.247+ -0.211+ -0.172
Parents’ help
Baseline grandchildren care 0.053+ 0.036 0.031 0.015 0.013 0.014 0.033 0.025 0.024
Change in grandchildren care 0.014 0.004 -0.036 0.078* 0.076* 0.100 0.029 0.024 -0.003
Baseline financial transfers 0.105** 0.104** 0.037 0.161*** 0.155*** 0.186** 0.104*** 0.099*** 0.038
Change in financial transfers 0.089** 0.089** 0.087** 0.175*** 0.170*** 0.174*** 0.104*** 0.100*** 0.098***
Interactions
M1* Change in grandchildren care 0.025 0.166 0.063
M2* Change in grandchildren care 0.095* -0.074 0.049
M3* Change in grandchildren care -0.149* 0.219 -0.096
M1* Baseline financial transfers 0.138+ -0.166 0.026

32
Table 2.3, Continued
Sons (2102/1035) Daughters (1999/986) Total (4101/1147)
Model 1 Model 2 Model 3 Model 1 Model 2 Model 3 Model 1 Model 2 Model 3
M2* Baseline financial transfers 0.106* -0.023 0.093*
M3* Baseline financial transfers -0.029 -0.044 -0.033
Constant 1.396 1.449 1.550 4.495*** 4.339*** 4.259*** 2.936*** 2.766*** 2.799***

χ
2

370.60 420.57 442.27 252.43 258.41 274.09 597.14 659.01 675.94
DF 20 22 28 20 22 28 21 23 29
Intraclass correlation coefficient 0.491 0.484 0.480 0.370 0.366 0.365 0.330 0.329 0.329
Within R
2
0.141 0.154 0.161 0.086 0.084 0.092 0.126 0.137 0.140
Between R
2
0.169 0.194 0.204 0.146 0.153 0.158 0.130 0.141 0.144
Overall R
2
0.149 0.166 0.173 0.119 0.123 0.129 0.131 0.142 0.146

33
associated with sons and daughters’ increases in financial help significantly and marginally
significantly respectively. Children’s number contributed significantly to sons’ increases in
financial help to parents. Both sons and daughters who had higher education increased their
financial support to parents over time. Sons and daughters who received instrumental support
from parents at baseline, and provided instrumental support and farm labor support to parents
in the second wave increased their financial support to parents. This not only showed the
reciprocity nature of intergenerational support but also the high correlation among dimensions
of support provided to parents, which could not be explained by the substitution hypothesis
that children who contributed more in one kind would contribute less in others, but rather
children who provided in one kind also provided help in others.
In Table 2.3, for sons, only established migrants gave more than those who were
independent of other variables in the model. Baseline grandchildren care had a marginally
significant effect to increase financial support from their children (as in Model 1), but this
effect was mediated by adding children’s other types of support in the second wave (as in
Model 2). This suggested that children paid back their parents’ grandchildren care in other
ways than financial support, so when other forms of support was controlled, grandchildren
care did not had significant effect on the changes in financials support from sons. Both
baseline and increase in financial support induced increases in financial support from sons in
Model 1 and Model 2. However, the effect of baseline financial support was obviously
conditional on children’s migration status as shown in Model 3. The insignificant main effect
of baseline financial support from parents meant that non-migrant sons did not pay back their
parents’ baseline financial support at all. Established migrants increased, recent migrants
marginally increased, their financial support to parents if they received more baseline financial
support. This supported our strategic investment hypothesis that elders reaped higher marginal

34
financial return from migrant sons (a positive return), than from non-migrant sons (a return not
significant from zero).
The insignificant main effect of changes in grandchildren care also resulted from the
contingency on migration status. Although non-migrant sons did not increase their financial
support to their parents because their parents increased grandchildren care to them, established
migrants did reward their parents for each additional unit of grandchildren care they received.
Interestingly return migrants reward their parents negatively for each additional unit of
grandchildren care they received during two waves, suggesting their greater need with
grandchildren care was associated with their status as a failed migrants.
Migration status did not influence daughter’s financial transfers to parents. Increases in
grandchildren care and both baseline and increases in financial support from parents
significantly increased daughters’ financial pay back to their parents. Greater grandchildren
care contributed to greater increases in financial support from daughters, but this effect
disappeared after interactions were added. This might partly result from the non-significant
countervailing effects among migrants and non-migrants.
For the total sample, being established migrant increased financial support to parents, so
did increases in financial support from parents. Although non-migrant children did not pay
back parents’ financial help in the second wave, established migrants rewarded their parents
for providing financial help.
Discussion
The large scale labor force migration from rural to urban China has brought changes to
intergenerational support that elders have access to and raised concerns about whether elders
can still rely on children’s support. Although children’s migration substantially reduces their
chances to provide hands-on support to older parents, their increased economic capacity and
enhanced reliance on their parents to facilitate their migration provides a prospect that elders

35
can benefit from children’s migration financially. Financial contributions to rural elders are
particularly important because the majority of rural elders feel that they do not have enough
financial resources and face serious financial constraints (Chinese Research Center on Aging,
2003).
A usually overlooked theme in migration remittance literature is that elders are widely
involved in intergenerational exchanges with migrant children and non-migrant children.
Elders provide grandchildren care, for both migrant and non-migrant children. And they not
only provide financial support to migrant children to facilitate their migration and benefit from
remittances, they also provide and receive financial support from non-migrant children. This
paper tries to bring non-migrant and migrant children into the same intergenerational transfer
framework, and compare their behaviors in returning their parents’ previous help.
Guided by the reciprocity, bargaining power and strategic investment perspective, our
major focus of this article is whether parents’ help to migrant children brings back higher
returns to parents, compared with parents’ help to non-migrant children.
For non-migrant sons, the baseline financial support from parents did not matter to their
contribution to their parents later, which might suggest the flow of some ritual money or
money that was supposed to help those who were less successful. But increases in financial
support from parents apparently stimulated sons’ financial support back to parents. It was
likely that non-migrant sons pay back loans for some particular purposes, which was supposed
to be paid back shortly. But the well-established migrant sons significantly increased financial
support to their parents if they received more financial help from parents at baseline. This
effect held marginally significant for recent migrants, which reflected the income vulnerability
of migrants at the beginning of their migration, as well as a time lag of paying back of family
loans or investment on migration. This supported our hypothesis that parents’ financial

36
investment on migrant sons reaps higher marginal return later from them than from
non-migrant sons.
An alternative explanation is that the differences between migrants and non-migrants
should be attributed to children’s income, which we do not have a measurement. The effect of
migration would in fact reflect the effect of migrants’ higher income, in that lower income
children do not pay back their parents’ financial help, while higher income children do pay
back their parent’s financial help, therefore that non-migrant sons who received more from
parents and gave less may be attributed to money flowing to those children who are less
successful, and hence are less likely to pay back (Shi, 1993). But intergenerational transfers
from parents to children are usually aimed at poorer children (Attias-Donfut, 2003; McGarry
& Schoeni, 1995), which apparently does not agree with the fact that migrants get more
financial help from their parents. Because of this, non-migrants and migrants’ difference in
their giving financial support to their parents can not be totally attributed to their income
differences, instead it really catches the fact that the exchange contract for migrant children
and non-migrant children are different. Migrant children enter into a contract which expected
them to pay back their parents financial help either more or more rapidly, while non-migrant
children have another contract which does not expect them to pay back debts in short-run at
least, because the money that parents give to non-migrants and migrants are given with a
different use in mind, i.e., helping non-migrants because they are in need but facilitating
migration for migrants (McGarry & Schoeni, 1995; Stark & Lucas, 1988).
Although providing more grandchildren did not seem to increase financial support from
children, the detailed inspection showed that there were countervailing effects between
children of different migration status. Non-migrant sons did not pay back their parents’
increases in grandchildren care, particularly when other types of support from children were
controlled. A qualitative study conducted for a subsample of this study found that most elders

37
regarded taking care of their grandchildren as their irrefutable responsibility, and although
children felt some degree of gratitude to parents’ help, they did not necessarily feel obliged to
provide more financial help to their parents, particularly when most rural people had strong
financial constraints. Under this circumstance, providing hands-on support and exchanging
time for time is a better way to reciprocate parents’ grandchildren care efforts instead of
providing financial pay back. But for migrant children, our bargaining power hypothesis was
supported that parents receive more from migrant children because of their higher bargaining
power. The period when grandchildren care was increased was also the period when financial
support from migrant was increased. It is possible to argue that migrant pay back financially
because they can not pay back by devoting time. But we controlled for instrumental and farm
labor help from migrant children, so we can exclude this substitution effect. Hence, we still
favor the bargaining power hypothesis.
Similar to children’s different responses to parents’ previous financial support, the
differences among migrant and non-migrant siblings in their responses to parents’
grandchildren care efforts reveal that different reciprocal rule of intergenerational exchanges is
observed among migrant children and non-migrant children. Short-term exchanges between
time and money were norms with migrant children, but not necessarily with non-migrant
children. Even when migrant children give parents instrumental support to the same extent,
they still reward their parents financially for taking care of their own children. Although we
regard this as a support to bargaining power hypothesis, we can not totally exclude the
possibility that remittances are for the purpose to cover expenses of kids themselves.
Finally, migrants gave more than non-migrants and this was not explained by previous
support their parents had provided to them, either in the form of financial support,
grandchildren care, or instrumental help. This is possibly related to migrants’ higher income,
which supports the altruistic motivation that those able children give more to their parents

38
perhaps out of the underlying filial norm (Becker, 1974; Lee et al., 1994b). It is also possible
that families renegotiate the financial arrangement among siblings, and migrants take more
financial responsibility for their older parents. In either case, altruism still plays an important
role in children’s providing financial support to their parents. On the other hand, we can also
see altruism on the parents’ side. Parents’ help is reciprocated financially only by certain
migrant children, who rely more on parents’ providing grandchildren care and who receive
parents’ financial support to facilitate migration. Even though, parents’ still provide financial
help and grandchildren care to their non-migrant children.
In conclusion, the power older parents have in sharing the wealth with their migrant
children on the one hand depend on parents’ ability to provide help to their migrant children
and on the other hand rely on children’s altruistic remittance behavior. For non-migrant
children, reciprocity is not strictly observed in the short-term, but for migrant children,
reciprocity is strongly observed. Therefore parents’ help to migrant children is more beneficial
for elders in short run. Despite of its beneficial effects, parents may have difficulties in
providing support to their migrant children, for example, taking custody role of grandchildren
may increase strain in physical and psychological health, as suggested by many U.S.
grandparenting studies (Szinovacz et al., 1999). In addition, financial resources are also not
abundant for rural elders, so even providing limited financial resources may be challenging for
the elders. Effectively and appropriately empowering elders to be active providers for their
migrant children will not only contribute to the economic development of China, but also
benefit elders through an intergenerational transfer mechanism which singles out migrant
children as more gratuitous children who rely on older parents’ help.

39
Chapter 3: Which Sons Live with their Parents: How do Sons’ and their Male Siblings’
Exchanges with Parents Matter?
Abstract
This article investigates in rural China how sons’ migration, through their exchanges with
parents, influences the odds that parents coreside with other non-migrant sons, and how
non-migrant sons’ previous migration experience and exchanges with parents influences their
own odds of coresiding with parents. Based on a three-wave longitudinal study in Anhui
Province, China, collected in 2001, 2003, and 2006, I conducted analyses to predict the odds
for an non-migrant son to coreside with the parent and to transit into coreside with the parent
with two lagged design which include 2.5 years lag model (2001-2003), and 5.5 years lag
model (2001-2006). The 2.5 years lag model include 735 elders who had at least one
non-migrant son, corresponding to 1154 non-migrant children, whereas the 5.5 years lag
model include 617 elders who had at least one non-migrant son, corresponding to 992
non-migrant children. Our unit of analysis is non-migrant children. Each observation
represents one non-migrant son, but with migrant siblings’ information embedded as family
level variables. Using random effect logistic regression, we found financial support from
migrant sons reduced elders’ odds to coreside with any non-migrant son. Previous help to a
non-migrant son increased elders’ odds to coreside with that child, but whether that child had
previous migration experience did not make a difference in the odds increased.

40
In rural China, coresidence with children, particularly married sons, is strongly endorsed
by cultural values and forms a basis for promoting intergenerational exchanges. Coresidence
increases the chance that a child help parents with family chores, even compared to children
nearby, and when parents get older they show greater preference to live with a married child,
rather than having children living close by (Unger, 1993; Whyte, 2003). In addition, when
intergenerational coresidence is an important manifestation of filial piety, it substantially
improves elders’ psychological well-being as well as physical health (Silverstein et al., 2006).
But migration has substantially reduced elders’ opportunity to live with a married son in a
stable household environment(Giles & Mu, 2007). The migration of children on the one hand
reduces the availability of children who can provide coresidence, and on the other hand the
migration of one child may influence the possibility of elders’ living with other non-migrant
children when the coresidence decision is made among the extended families. Guided by the
“corporate/mutual aid” framework, which describes cooperation within families based on
reciprocity to maximize each member’s benefit, this article investigates in rural China how
sons’ migration, through their exchanges with parents, influences the odds that parents
coreside with other non-migrant sons, and how non-migrant sons’ previous migration
experience and exchanges with parents influences their own odds of coresiding with parents.
Coresidence rates with adult children have decreased substantially in rural China between
1991 to 2000 from 70% to 60% (Giles & Mu, 2007), which was consistent with the reduced
coresidence rate around the world during the process of modernization (Knodel & Ofstedal,
2002). In rural China, changed norms of coresidence and children’s individual entitlement to
housing substantially reduced elders’ opportunity to coreside with a married son. In addition,
the labor force migration from rural to urban areas had geographically separated many adult
children from their aging parents and substantially reduced the number of children who stayed
in the same village with their parents. In fact, only these children were feasible providers of

41
coresidence, because of the household registration system which separated rural and urban
areas and restricted rural residents’ chances of having a permanent residence in urban areas,
which challenged substantially limited the possibility of elders reunited with their children in a
place other than their own resident place (Wang, Ping, Zhan, & Shen, 2005; Yan, 2003; Zeng
& George, 2001).
In parents’ old age, there were substantial changes in living arrangements and in
coresident partners (Zimmer, 2005). Studies usually found that elders’ needs such as
deteriorated health, cognitive impairment, depression, and marital status, were important
predictors for living arrangement transitions among elders, few studies regarded transition of
living arrangement as the strategic arrangement of elders through their previous exchanges
with their parents (Brown et al., 2002; Giles & Mu, 2007; Zimmer, 2005). Some evidence
showed that exchanges with others predicted living arrangement transitions. For instance,
Wolf (2000) found unbalanced exchanges predicted transitioning out of living with others in
the U.S., and Attias-Donfut (2003) found that taking care of grandchildren increased elders’
chances of living with children in the old age in France (Attias-Donfut, 2003).
Exchanges were quite common in the Chinese families (Cheng & Chan, 2006; Li et al.,
2004; Shi, 1993; Sun, 2002). The corporate group /mutual aid model explained the short or
long term transfers between parents and children in Chinese families from the perspective of
reciprocity (Lee & Xiao, 1998). Specifically the corporate group model explained
intergenerational exchanges as motivated by long-term arrangements that maximized the
family’s well-being, e.g. parents invested on children’s education, and children were expected
to pay back in the future to provide for their parents’ old age, whereas the mutual aid model
focused on the short-term exchanges that benefited each side. Empirical evidences in China
concerning intergenerational transfers from children to parents basically supported the
corporate group /mutual aid model. Elders who supported children’s by providing

42
grandchildren care, housing and housework got more financial help from children, and elders
who provided more help in grandchildren care, household work, and farm labor help to
children received more help in household work and farm labor help (Li et al., 2004; Liu &
Reilly, 2004; Shi, 1993; Sun, 2002; Yang, 1996).
Although some explorative studies pointed out that coresidence with married sons was a
desired ideal in old age, and elders strategically positioned themselves to gain coresidence in
their interactions with children such as remaining independence as long as possible and asking
for as less help as possible from children to gain their right for coresidence in their old age,
few studies have empirically and explicitly addressed how children’s previous exchanges with
parents influenced the odds of elders’ coresidence with them (Davis-Freidmann, 1991; Miller,
2004; Yan, 2003).
The purpose of this paper is to examine how the previous exchanges with children related
to coresiding and transiting into coresiding with children, under the condition of children’s
high migration rate. In particular, migration has brought substantial changes to
intergenerational exchanges, which may influence elders’ living arrangement and its transition.
Children’ s own migration and exchanges with parents matter
In rural China, migration from rural to urban areas substantially reduced the availability
of children, and reduced elders’ coresidence opportunities. But, studies also showed that
children might return from migration because of their parents’ deteriorated health (Giles & Mu,
2007; Zimmer, 2005). In addition, most rural to urban migrants were temporary migrants
because Hukou’(household registration system), low education and income, poor benefits,
segregated labor market and occupation, temporary housing, and barriers for children to get
reasonable price education, prevented them from staying in cities permanently (Bai & Song,
2002; Wang & Zuo, 1999; Wang & Fan, 2006). Particularly, since 1997, the flow of labor back
to villages accelerated (De Brauw & Rozelle, 2003).

43
Migration strengthened migrant children’s dependence on their parents to provide
grandchildren care and sometimes financial help to cover initial migration costs, because of
their inability to bring their children together with them and because of their financially
vulnerability at the beginning of their migration (Poirine, 1997; Stark, 1991; Zhang, 2002).
Based on principle of reciprocity of “corporate/mutual aid” framework, the services and
money elders’ provided may form the basis for children’s felt obligation to coreside with their
parents after they return. In addition, elders did not only provide help to migrant children, they
also provide help to non-migrant children (Shi, 1993). Compared with non-migrant children,
migrants’ stronger dependence on parents for help may increase elders’ bargaining power for
future coresidence when migrant children finally return from urban areas, as suggested by the
bargaining power theory which posed that parents who have more resources and therefore
higher bargaining power in the exchange receive more from their children (Bernheim et al.,
1985; Cox, 1987; Lee et al., 1994b; Lucas & Stark, 1985).
Siblings’ migration also matters
Coresidence with a specific child, in fact, was a joint decision among the extended family.
When a parent had more than one child, each of them were potential player in the decision
making process, e.g., Pezzin (2007) modeled the decision of living arrangement and the
consequent care allocation as a two-stage game involving an elder parent and two children,
and illustrated that living arrangement and care provided by each child was a joint decision by
family members with different preferences facing different constraints. And some studies in
the allocation of caregiving time also show that care provided by each child is correlated with
the others, and depends on the characteristics of other children available to provide care
(Checkovich & Stern, 2002; Engers & Stern, 2002).
In China and other countries where complicated interactions and social exchanges are
carried out within extended families, it is more than necessary to consider the possibility of

44
how other children’s characteristics influence elders’ coresidence with a specific child. The
current cohort of elders in rural China on average has 4 children (China Research Center on
Aging, 2003; Silverstein et al., 2006). The dyadic parent-child relationship was only one piece
of interactions among this network of interactions, as it influenced and was influenced by
other components of this network, e.g., daughters with brothers were significantly less likely
to coreside with parents (Logan, Bian, & Bian, 1998).
In China, there was evidence of children’s cooperation in providing for their parents, e.g.,
children living far way tend to provide financial support whereas coresident children are more
likely to provide instrumental support (Bian et al., 1998; Shi, 1993; Sun, 2002). Although this
cooperation is conditional on elders’ living arrangement, it is possible that within Chinese
extended families, the parents’ living arrangements result from the negotiation among children
that some children “buy” out of living with their parents, and some children are “bought-in” to
live with parents.
With the high rate of migration, this kind of coordination and cooperation may be more
salient. Rural migrants on average at least doubled their income, and migrants usually
provided more financial support to their parents, particularly when their parents were taking
care of migrants’ minor children, described as “time-for-money” exchanges (Hermalin, 2002;
Lee & Xiao, 1998). If the “corporate/mutual aid” model is applicable in the extended family,
then the ideal situation will be that children cooperate to provide different resources to their
parents depending on what they can provide. Migrant children’s increased financial support
may alleviate the burden for non-migrant children to provide financial support to their parents,
and oblige non-migrant children to fulfill their responsibility to parents by providing
coresidence.

45
Patrilineal family system and coresidence choices
Another relevant issue in elders’ coresidence with children is the patrilineal family system.
Because of the strong patrilineal family tradition, only 3% of elders lived with any daughter in
this sample (see Chapter 4), and very few daughters received grandchildren care from their
parents (see Chapter 2). Practically, we need only to consider parents’ possibility to coreside
with a son and “time-for-money” exchanges with a son. Based on the above discussions, we
propose the following hypotheses:
1. Receiving financial support from migrant sons increases parents’ odds of
coresiding with a non-migrant son.
2. Parents’ help to a child in grandchildren care and money increases parents’ odds
to coreside with that son.
3. Sons who returned from migration have greater odds of coresiding with parents
than never migrated sons for the same amount help they previously received from parents.
Methods
Sample
The sample for this investigation was derived from the Anhui Province of China, a mostly
rural province and the fifth most populous province in China. Currently, 12% of the rural
population is 60 years of age and older (compared to only 8.5% of the nation) making it one of
the most elderly provinces in China. This region was chosen specifically for its relatively high
density of older adults and high levels of out-migration of working age adults. Between 1995
and 2000, Anhui Province had the third highest rate of out-migration among all provinces in
China, and a higher than average rate of labor-related migration. Data were collected from a
sample of adults age 60 and over living in rural townships within Chaohu, a city of 141,000
people located on the north bank of Yangtze River in the central part of Anhui Province. This

46
poor, rural area of the province is generally known for its high rates of labor migration to the
cities of Hefei, Nanjing, and Shanghai (Chaohu Statistical Bureau, 2001).
The sample was identified using a stratified multistage method to randomly select 1,800
potential respondents. First, 12 rural townships were randomly selected from all 126
townships in Chaohu. Second, 6 administrative villages were randomly selected in each
township. Third, within each selected village, all people aged 60 and older were stratified to
form two sampling frames based on age: (1) those aged 60-74, and (2) those 75 and above,
providing an intentional over-sample of the 75+ population. In April 2001, of 1,800
individuals randomly selected for the study, 1,715 completed the survey, yielding a response
rate of 95.3%. The completed sample included 829 men (48.8%) and 869 women (51.2%). In
terms of age, 61.2% were 65-74 years old and 38.8% were 75 years and older. In October
2003 the follow-up survey was conducted with 1,368 respondents, or 79.8% of the original
participants. Of those respondents who were not located, 76 had moved out of the village, and
240 died. Twenty-three former respondents were located but refused to participate, terminated
their interviews, and/or were too ill to be interviewed.
In our analysis, each observation corresponded to one parent-child dyad. Therefore a
complete dataset included 1368 parents and their correspondent children. But we constrained
our working data to accommodate to the purpose of our analysis. First, we constrained our
working data to parent-child dyads when the child was a son. In addition, because only
proximate sons can provide coresidence, we further constrained our sample to include
parent-child dyads when the son was living in the same village with parents in the second
interview. These sons are defined as non-migrant sons. After these constraints, the resulting
analytic file constituted 1032 parent & non-migrant son dyads, corresponding to 687 parents.
But migrant son’s information was included as family level variables, which were matched
into each parent & non-migrant son dyad.

47
Random effect logistic regression was used to predict the odds of elders’ transitioning
into coresiding with a specific non-migrant son. Software STATA was used to carry out this
analysis. We had two levels of variables, i.e., individual-level variables referring to each
non-migrant son’s characteristics and family-level variables including characteristics of
parents and parents’ exchanges with all migrant sons. The family-level variables were shared
for non-migrant siblings with the same parent.
Dependent variable
Our dependent variable was an individual level variable indicating whether a non-migrant
son transited into living with the parent (1 = transiting into living with the parent) in the
second wave. To satisfy this condition, the non-migrant son was not living with the parent in
the first interview, but was living with the parent in the second interview. The idea of
“transitioning into coresiding” instead of “coresiding” helps to identify the cause-effect
relationship between parents’ exchanges with children and parents’ living arrangement, and it
reflects the dynamic nature of elders’ living arrangement in old age.
Independent variables
Our first group of variables reflects parents’ exchanges with each non-migrant son in the
first interview, including parents’ previous help in grandchildren care, monetary support and
children’s financial help to parents. These variables were individual level variables. The
variable measuring the frequencies of taking care of grandchildren ranges from 0 to 6. The values
of these variables were defined as follows: 0= “not taking care of grandchildren”, 1= “seldom”, 2=
“about once a month”, 3= “several times for a month”, 4= “at least once a week”, 5= “A period of a
day (not the whole day)”, 6= “The whole day, from morning to evening.” Financial transfers from
parents to children at baseline are based on the total amount of money that the child received
from the parent during the past 12 months. Parents were asked to provide the exact amount of
money first, and if they could not give a exact number, they were asked to choose among the

48
following categories based on Chinese RMB currency (100 RMB = $12US): 0= “none”, 1=
“less than 50”, 2= “50-99”, 3= “100 -199”, 4= “200-499”, 5= “500-999”, 6= “1000-2999”, 7=
“3000- 4999”, 8= “5000 to 9999”, 9= “More than 10,000”. In the analysis, we use the
categorical variable, ranging from 0 to 9. Similar to financial transfers to children, financial
transfers from children were based on the amount of money that each child provided to the
parent during the past 12 months. We also use the categorical variable, ranging from 0 to 9.
Our second group of independent variables reflected parents’ previous exchanges with
migrant sons. These variables included the sum of grandchildren care parents provided to all
migrant sons, and the sum of monetary support from all migrant sons in the second wave.
These variables are family level variables, and measured in the same way as exchanges
between parents and their non-migrant sons.
We measured non-migrant son’s previous migration experience with a dummy variable,
representing whether this non-migrant son was previously a migrant at baseline, i.e., return
migrant, or this non-migrant son was non-migrant at baseline, i.e., never migrated child. We
used “1” to represent a return migrant son, “0” represent a never migrated son. This variable
was an individual level variable.
We interacted each non-migrant son’s migration status and previous help from parents to
test whether being return migrants increased the odds of coresiding with parents for the same
amount of previous help they received.
Controls:
We controlled for important parents’ characteristics as family level variables. Parents’
characteristics at baseline included age in chronological years, number of children, and
dummy variables for gender (1 = female), marital status (1 = married), education (1 = some
formal education). In addition, parents’ health status was measured as the extent of functional
impairments, calculated as the sum of 15 items reflecting difficulties in performing personal

49
activities of daily living, instrumental activities of daily living and activities requiring physical
strength, mobility, and flexibility. Respondents indicated the level of difficulty performing
each task: 0 (no difficulty), 1 (some difficulty), or 2 (cannot do it without help). Since these
items had high reliability (alpha = .96), we calculated a summed scale that ranged from 0 (no
difficulties) to 30 (unable to perform all tasks). Income was measured as the logged RMB
value of the total annual income of respondent and spouse from work or pension (+1).
We also controlled for important children’s characteristics as individual level variables.
Children’s characteristics at baseline included chronological age, gender, education, and marital
status. Age was represented as age in years at the time the survey was carried out. Gender was
codes as 1= “female”; martial status (1= “currently married”); education (0= “no education”,
1=“primary school“, 2=“junior middle school“, 3=“senior middle school, vocational training,
college, university or above“). We also controlled for each child’s emotional closeness with the
parent by using three questions that assessed the quality of each parent–child relationship. We
adapted these questions from the Affectual Solidarity inventory (Mangen, Bengtson, & Landry,
1988), which assesses emotional cohesion between generations. The questions were: (a) “Taking
everything into consideration, how close do you feel to (this child)?” (b) “How much do you feel
that (this child) would be willing to listen when you need to talk about your worries and
problems?” and (c) “Overall, how well do you and (this child) get along together?” We coded the
items as follows: 0 (not at all close/not at all/not at all well), 1 (somewhat
close/somewhat/somewhat well), or 2 (very close/very much/very well). We computed an additive
scale, ranging from 0–6, for each child. For each parent we took the highest total score across all
children to indicate this construct. The reliability coefficient for these items was .82.
Results
Table 3.1 presents the mean values and percent distributions of all variables used in our
analyses. Socio-demographic characteristics reported in Table 3.1 revealed that parents in the
sample averaged 71 years of age (SD = 6.0), slightly more than half (58 %) were female, 54%

50
were not married, 18% had some formal education. The average log income was 3.5 (SD =
3.5), corresponding to about 32 RMB (the large majority of having had no household income).
The average functional impairment score was 5.5 (SD = 5.6) out of a possible 30.
Grandchildren care to migrant children is 0.8, which was not high because majority elders did
not provide grandchildren care to their migrant children. Financial support from migration
children averaged 2.86, which was between RMB 100-200, significantly higher than elders’
own income. However, financial support to migrant children averaged 0.4, which was below
RMB 50.
Table 3.1 also shows characteristics of non-migrant adult children, revealing that they
averaged 42 years of age (SD = 9.16). Their average education roughly corresponded to a
primary school level. The average emotional closeness with parents was 4 out of possible 6.
The average grandchildren care they got is 1.1, a little bit more than average grandchildren
care migrant children got. Their financial support to parents were 2.4, which was between
RMB 100-200. Among these non-migrant children, 16% were return migrants. Comparing
characteristics of parents and non-migrant children in the 2.5 years lag analysis and 5.5 years
lag analysis, we found the characteristics were very similar. But worthy of note, in the 5.5
years lag analysis, 24% of non-migrant children were return migrants, which signified the
increased returning of migrants in rural China.
As shown in Table 3.2, in the subsample used for living arrangement transitions,
compared with non-migrant children, return migrants received significantly more
grandchildren care and financial support from their parents, but provided marginally
significant (for 2.5 years lag model) or significantly (5.5 years lag model) higher level of
financial help to their parents at baseline. In the subsample used for prevalence analysis, the
exactly same conclusions hold.

51
Table 3.1
Description of Analytic Variables
2.5 years lag (1154/735) 5.5 years lag (992/617) Coding
M SD M SD
Family level variables
Age 71.04 7.07 70.45 6.79 57-88
Female 0.58 0.49 0.58 0.49 0 (male), 1 (female)
Married 0.54 0.50 0.58 0.49 0 (not married), 1 (married)
Education 0.18 0.38 0.22 0.41 0 (no education), 1 (some education)
Income 3.48 3.50 3.70 3.52 0-9.393
Functional limitations 5.48 6.47 5.16 6.47 0 (none)-30 (most)
Grandchildren care provided to migrant children 0.81 2.06 0.70 1.89 0-12
Financial support received from migrant children 2.86 3.77 3.60 4.35 0-25
Financial support provided to from migrant
children
0.39 1.32 0.32 1.46 0-15
Characteristics of non-migrant sons
Married 0.87 0.34 0.87 0.33 0 (not married), 1 (married)
Age 42.17 9.16 41.29 8.93 20-70
Education 1.16 0.87 1.21 0.85 0-3
Emotional closeness with parents 4.09 1.57 4.07 1.58 0-6
Grandchildren care from the parent 1.13 2.03 1.28 2.14 0-6
Financial support from the parent 0.48 1.10 0.49 1.14 0-8
Financial support to the parent 2.36 1.77 2.33 1.75 0-9
Return migrant 0.16 0.37 0.24 0.43 0 (never migrated), 1 (return migrant)

52
Table 3.2
Intergenerational Exchanges between Parents and Non-Migrant Children
Predicting Transiting into coresidence Predicting coresidence
2.5 years lag 5.5 years lag 2.5 years lag 5.5 years lag
Never
migrated
Return
Migrants
Never
migrated
Return
Migrants
Never
migrated
Return
Migrants
Never
migrated
Return
Migrants
Grandchildren care
from the parent
0.82 1.68* 0.92 1.63* 1.03 1.68* 1.17 1.63*
Financial support from
the parent
0.35 0.78* 0.32 0.77* 0.42 0.78* 0.40 0.77*
Financial support to
the parent
2.29 2.52+ 2.24 2.55* 2.33 2.52+ 2.26 2.55*
*Denote differences between return migrants and never migrated non-migrants at p < 0.5.

53
In Table 3.3, I present the multilevel logistic regression model predicting the odds of
transiting into living with parents vs. staying non-coresident with parents for a son who live in
the same village with their parents at the follow up. In the 2.5 years lag sample, female elders
are more likely to transit into coreside with children and married parents are less likely to
transit into coresiding with children. Non-migrant children who were unmarried and who had
higher emotional closeness with parents were more likely to transit into living with parents.
Migrant siblings’ financial support substantially reduced the odds for a non-migrant son to
transit into living with the parent. In Model 1 or 2.5 years lag model, grandchildren care from
a parent to a non-migrant child 2.5 years ago significantly increased the odds of this
non-migrant children transiting into living with this parent 2.5 years later. Return migrants
were significantly more likely to transit into living with parents compared with their never
migrated siblings. This may reflect a selection effect that return migrants came back for the
purpose of taking care of their parents. We further analyze whether being a return migrant
have increased odds of children’s coresidence for same help they have received from parents
in the past, but we failed to detect any interaction effect between being a return migrant and
receiving parents’ help on increasing parents’ odds of transiting into living with that child.
To catch the lagged effect of intergenerational exchanges on parents’ living arrangement,
I use the 5.5-years lag subsample to run the analysis again based on the same
conceptualization. The result showed that, financial help from migrant children 5.5 years ago
still reduce parents’ odds of transiting into living with a non-migrant child. However
grandchildren care provided by a parent to a non-migrant child 5.5 years ago did not increase
parents’ odd of transiting into living with that non-migrant child. Instead, financial support
provided by a parent to a non-migrant child 5.5 years ago increased parents’ odd of transiting
into living with that non-migrant child. In the 5.5-years lag subsample, we still did not detect

54
Table 3.3
Random Effects Logistic Regression Predicting Transiting into Coresidence
Predicting Transiting 2.5 years Later
(910/588)
Predicting Transiting 5.5 years later
(803/513)
Model 1 Model 2 Model 1 Model 2
Family level variables
Age 0.001 0.002 0.046 0.046*
Female 0.758* 0.739* 0.301 0.304
Married -1.255** -1.247** -0.406 -0.439
Education 0.443 0.452 0.212 0.226
Income -0.038 -0.039 -0.002 0.001
Functional limitations 0.022 0.022 0.029* 0.029+
Grandchildren care provided to
migrant children
-0.088 -0.087 0.085 0.084
Financial support received from
migrant children
-0.166** -0.167** -0.101** -0.097**
Financial support provided to from
migrant children
-0.100 -0.093 0.049 0.043
Characteristics of non-migrant sons
Married -2.355*** -2.400*** -1.109** -1.118***
Age -0.013 -0.015 -0.010 -0.010
Education 0.155 0.152 0.035 0.037
Emotional closeness with parents 0.149 0.145 -0.008 -0.009
Grandchildren care from the parent 0.189** 0.123 0.084 0.058
Financial support from the parent 0.133 0.263 0.239* 0.315+
Financial support to the parent -0.084 -0.080 0.067 0.137
Return migrant 1.862*** 1.780** 1.117* 1.424***
Interactions
Return migrant x Grandchildren
care from the parent
0.118 0.049
Return migrant x Financial support
from the parent
-0.196 -0.094
Return migrant x Financial support
to the parent
-0.002 -0.142
Constant -0.979 -0.865 -4.066** -4.192

χ
2

87.23 86.0 60.74 75.28
DF 17 20 17 20
Intraclass correlation coefficient 0.276 0.270 0.019 0.017

55
Table 3.4
Random Effects Logistic Regression Predicting Odds of Coresidence
Predicting Coresidence 2.5 years Later
(1154/735)
Predicting Coresidence 5.5 Later
(992/617)
Model 1 Model 2 Model 1 Model 2
Family level variables
Age 0.032 0.029 0.053** 0.053**
Female 0.304 0.296 0.180 0.182
Married -1.172*** -1.207*** -0.508* -0.532*
Education 0.641* 0.623+ 0.129 0.133
Income -0.096* -0.093* -0.046 -0.044
Functional limitations 0.040* 0.040* 0.028* 0.028+
Grandchildren care provided to
migrant children
-0.163+ -0.159+ 0.005 0.007
Financial support received from
migrant children
-0.101** -0.103*8 -0.059* -0.058*
Financial support provided to from
migrant children
-0.045 -0.054 0.084 0.080
Characteristics of non-migrant sons
Married -2.95*** -2.943*** -1.693*** -1.687***
Age -0.019 -0.018 -0.018 -0.017
Education 0.191 0.192 0.117 0.118
Emotional closeness with parents 0.287*** 0.289*** 0.132* 0.134*
Grandchildren care from the parent 0.298*** 0.293*** 0.160*** 0.164**
Financial support from the parent 0.267** 0.352** 0.243*** 0.250*
Financial support to the parent 0.004 0.026 0.003 0.024
Return migrant 0.143 0.570 0.204 0.424
Interactions
Return migrant x Grandchildren
care from the parent
0.006 -0.014
Return migrant x Financial support
from the parent
-0.261 -0.003
Return migrant x Financial support
to the parent
-0.101 -0.081
Constant -1.651 -1.572 -3.313* -3.363*

χ
2

121.71 119.16 66.58 66.76
DF 17 20 17 20
Intraclass correlation coefficient 0.367 0.361 0.125 0.125

56
any interaction effects between being a return migrant and receiving parents’ help on
increasing parents’ odds of transiting into living with that child.
In Table 3.4, the prevalence analysis predicted the odds of parents’ living with a
non-migrant son vs. not living with a non-migrant son. The sample size was bigger than that in
the previous analyses, because we included those non-migrant sons who stayed coresiding
with their parents and who transited out of coresiding with their parents during two waves.
Previous financial support and grandchildren care from parents significantly predict the odds
of living with a non-migrant son. And previous financial support from migrant children
significantly reduced parents’ odds of living with a non-migrant son. Parents’ previous help to
a non-migrant child in money and grandchildren care significantly predicted their living with
that child.
Discussion
Intergenerational exchanges have many dimensions, and coresidence is a very important
way of resource transferring when parents are in their old age. This article studied how a
parent’s help to a child contributed to his or her chances of living with and transitioning into
living with that child, and how a parent’s exchanges with the migrant siblings of that child
contributed to parents’ odd of living with that child and transitioning into living with that same
child. This article regarded living arrangement decision not as a consequence of dyadic
exchanges, but as a result of multiple exchanges within extended families.
We found that coresidence depends on parents’ previous help to children, which reflected
the reciprocal principle observed in intergenerational relationships. Although parents’ previous
grandchildren care and financial help were important for elders to live with that non-migrant
child 2.5 years and 5.5 years later, only grandchildren care predicted transiting into living with
that child in the short run (2.5 years later), and only financial help predicted transiting into
living with that child in the long run (5.5 years later). The possible reason was that elders who

57
lived together with children tended have more exchanges including giving more grandchildren
care and financial support to the coresident children. Therefore, the prevalence analyses
caught the effect that parents provided more help to non-migrant child were more likely to
stay living with these children. In the analysis of living arrangement transition, we found
grandchildren care had a short term effect in predicting transiting into living with children, but
financial support to children had a longer term effect. This was similar to the findings
concerning children’s financial support back to parents in Chapter 2 that grandchildren care
from parents predicted simultaneous increases in financial support from children, whereas
financial support from parents predicted lagged increases in financial support from children.
Grandchildren care and financial help might be perceived as helps with different nature that
requires pay back with different time frames.
Return migrants were not more likely to live with parents than their never migrated
counterparts, but being return migrant predicted transiting into coresiding with parents. This
was consistent with findings that migrants return to home to accommodate to the needs of
their parents, such as declined health (Giles & Mu, 2007). In the prevalence analysis, return
migrants were not more likely to coreside with parents because the analysis included
non-migrants who always coresided with parents.
Compared to support provided to non-migrant children, that provided to migrant children
did not benefit parents further in gaining higher odds of coresidence or transiting into
coresidence with a return migrant son. This does not support the bargaining power perspective
that parents will benefit more from providing help to migrant children when these help is
much more needed. Although there was a lack of evidence that elders had no bargaining power
over migrant children in reaching coresidence by providing help to them, elders might have
higher bargaining power in other ways, e.g., receiving higher marginal financial return from
migrant children, as illustrated in Chapter 2.

58
Within extended families, contrary to our hypothesis, financial support from migrant
siblings did not “buy” non-migrant children into coresiding with their parents; instead, it
reduced the odds of elders’ living with a non-migrant child. This is consistent with findings
that increased resources predict parents’ independent living (Unger, 1993). The results did not
differ when parents’ financial help to non-migrant children was and was not controlled. To be
specific, that there were two mechanisms that financial support from migrant children would
influence their parents’ odds of coresiding or transiting into coresiding with a non-migrant
child. First, financial support had directly reduced the odds of parents’ coresiding with a
non-migrant child, possibly because the financial support enabled parents to live
independently. Second, financial support from migrant children and financial support to
non-migrant children were significantly correlated (r = 0.26, p < 0.001), and there were an
indirect mechanism that financial support from migrant children increased parents’ financial
help to non-migrant children, which increased parents’ odds of coresiding with a non-migrant
child. However, when financial support to non-migrant child was not controlled, the negative
effect of financial support from migrant children on elders’ coresidence odds revealed that the
negative direct effect of financial support from migrant child on coresidence odds was more
powerful than the positive indirect effect. In other words, financial support from migrant
children reduced elders’ coresidence odds with a non-migrant more than the positive effect
that it brought by increasing elders’ financial support to non-migrant child. In addition to the
ability to independent living arrangement, there might be other alternative explanations
This analysis centered on siblings’ exchanges with parents. Although I have incorporated
other siblings’ exchanges with parents into analysis, the data lack the direct exchanges among
siblings, which substantially limited my ability to examine whether migrant can “buy” their
non-migrant children into coresiding with their parents by providing direct transfers to their
non-migrant siblings. In the future, a network of exchanges within an extended family should

59
be considered in examining the consequences in elders’ living arrangement. In addition, I lack
the detailed measurement on how elders’ preference for coresidence changes.
Despite of its incompleteness, the results in this article provided evidences that in rural
China, the decision of coresidence was made within the extended family setting based on
parents’ exchanges with each of them.

60
Chapter 4: Intergenerational Support and Depression Among Elders in Rural China: Do
Daughters-In-Laws Matter?
Abstract
This study examined the influence of intergenerational assistance with household tasks
and personal care from sons, daughters and daughters-in-law on the depressive symptoms of
older adults in rural China. The sample derived from rural Anhui Province, a region with a
strong hierarchy of support preferences that leads with sons and their families. We used data
from a random sample of 1281 adults aged 60 and over, who were interviewed in 2001 and
2003. Analyses indicated that depressive symptoms were usually reduced by assistance from
daughters-in-law, and increased sometimes when such support was from sons. These
relationships held most strongly when mothers coresided with their daughters-in-law. This
research suggests that the benefits of intergenerational support are conditional on culturally
prescribed expectations.

61
Decreased rates of intergenerational coresidence of elders in rural China have reduced
their opportunities for receiving instrumental support from adult children (Knodel & Ofstedal,
2002). In a cultural system with high filial expectations and few formal alternatives, the
absence of support should have substantial psychological consequences on parents. However,
the empirical evidence is equivocal about whether the absence of instrumental help from
children elevates depression in older parents. Drawing on a survey from a rural area in Anhui
Province, where preferences for old age support tend to be traditionally focused on sons and
their families, we ask how instrumental support from sons, daughters, and daughters-in-law,
influence elders’ depression. We speculate that inconclusive results in the literature might
result from pooling sources of intergenerational support that have different expected filial
responsibilities. In addition, we also investigate how the impact of support from sons,
daughters and daughters-in-law depends on the availability of daughters-in-law in the
household, who are the traditional hands-on support providers.
In Chinese cultures, where the Confucian norm of filial piety forms the basis for support
expectations, children are expected support providers, even when a spouse is present
(Chappell & Kusch, 2007; Chen, 2001b). In rural China, formal support systems were
generally sparse and lack coverage, consequently many elders rely exclusively on children for
financial, instrumental and emotional support (Agree, Biddlecom, & Valente, 2005; Joseph &
Phillips, 1999; Zimmer, 2005). Incapability to secure children’s help is stressful because of
unmet needs, and senses of helplessness and hopelessness, and stress deriving from adverse
life events or strained interpersonal relations may result in depression (Krause, 2001; Pearlin,
1989). Where tradition still prevails and expectations for intergenerational support is high, as
in rural China, failing to get help from children should be related to negative psychological
outcomes (Lee, Netzer, & Coward, 1995; Lee & Xiao, 1998).

62
Compared with Western elders, Chinese elders have lower level of depression with an
estimated 4% prevalence, partly attributed to stronger family support (Parker et al., 2001).
Chinese rural elders have a higher prevalence of 5-6%, about twice that of urban elders
(2-3%) , chiefly because of the poverty and lack of public services and support (Chen et al.,
2004; Chen et al., 2005).
Instrumental support has been substantially weakened as a consequence of children’s
migration and reduced coresidence (Giles & Mu, 2007; Joseph & Phillips, 1999; Knodel &
Ofstedal, 2002). However, unlike emotional and financial support, instrumental support, in
spite of its apparent cultural and practical significance, has shown inconsistent effects on the
psychological well-being of Chinese elders. (Chen & Silverstein, 2000; Cui & Li, 1997;
Krause & Liang, 1993; Silverstein et al., 2006; Sun, 2004). We suggest that cultural
complexities form expectations for instrumental support that depend on the relationship,
gender, and availability of the providers.
Guided by the hierarchical compensatory model and task-specific theory, we speculate
that the hierarchical order among children and the type of services delivered are important to
whether intergenerational instrumental support affects elders’ depression. The hierarchical
compensatory model emphasizes a rank order of preferred sources of help, whereas the
task-specific theory emphasizes the match between care tasks and care sources. When these
principles are violated (i.e., support is received from a non-preferred source or the service is
viewed as not appropriate for the provider), support may be appraised negatively and possibly
reduce psychological well-being (Cantor, 1979; Dean, Kolody, & Wood, 1990; Felton & Berry,
1992; Friedman, 1993; Krause, 2001; Litwak, Silverstein, Bengtson, & Hirst, 2003). These
theories provide guidelines for assessing the effect of intergenerational support based on how
well characteristics of a provider match with the performance required of that provider.

63
Raised in a patrilineal culture, Chinese rural elders mostly believe that sons are the best
providers in their old age (Chen, 2001b). However, cultural preferences for sons belie a more
complex set of nested expectations that also includes daughters and daughters-in-law. The
most general cultural expectation is that adult children should be responsible for meeting the
domestic and personal needs of their elderly parents. Among adult children, a son and his
family are expected to be the primary providers of support and care, whereas a daughter and
her family do not have the obligation to provide support. Although daughters do provide
considerable support to parents out of affection, sons-in-law rarely get themselves involved
because they are not defined as children at all (Antonucci & Jackson, 2003; Knodel &
Ofstedal, 2002; Lin et al., 2003; Wong, 2005; Yang, 1996; Zhan, 2004). Finally, within the
son’s family, the daughter-in-law has the primary responsibility to provide hands-on support,
deriving from a traditional gendered norm that assigns domestic work to women and a family
ideology that assigns daughters-in-law to positions of near servitude with respect to their
parents-in-law (Cohen, 1998; Youn et al., 1999).
But the generally preferred source of help may vary depending on the types of tasks
required. Whereas household chores deal with objects, personal care involves intimate
personal contact and often exposure of the body. In an in-depth interview, Wong (2005) has
found that Chinese elders are more comfortable receiving personal care from same-gender
children (i.e., older women generally prefer that assistance with bathing, dressing, and
grooming come from their daughters and daughters-in-law, and older men favor sons as
providers of personal care). Although daughters-in-law are culturally expected care providers,
they do not necessarily gain intimate relationships with their parents-in-law, a prerequisite for
personal care to be comfortably delivered (Willson, Shuey, & Elder, 2003). In this situation
daughters have chances to compete with daughters-in-law as the most preferred care givers.
Therefore, we expect that older mothers prefer daughters or daughters-in-law over sons,

64
whereas older fathers prefer sons over daughters and daughters-in-law to provide personal
care that involves physically intimate contact. However, concerning household chores, there is
no apparent gender difference in that both older mothers and fathers prefer daughters-in-law
over sons and daughters.
In rural China, coresidence with children, particularly married sons, is strongly endorsed
by cultural values and forms a basis for promoting intergenerational exchanges, and forms a
basis of high expectation for support from daughters-in-law (Yan et al., 2003; Zhang, 2004).
But, availability of a daughter-in-law in a household has been significantly reduced because of
changed coresidence norms, individual entitlement to land on which to build housing, as well
as increased labor force migration of sons and wives from rural to urban area (Wang et al.,
2005; Yan, 2003; Zeng & George, 2001). In the absence of a daughter-in-law in the household,
elders may adapt their expectations to include alternatives that occupy a lower rank in the
hierarchical order, as research has shown among similarly situated Chinese and Korean
immigrants to the U.S. (Pang et al., 2003; Wong et al., 2006). Consequently, elders coresiding
with daughters-in-law will expect to receive assistance from daughters-in-law (excepting
personal care to fathers-in-law), and regard lack of their assistance as disappointing and
distressing. They may also appraise support from sons or daughters as culturally inappropriate
and even disgraceful. Conversely, those not coresiding with daughters-in-law may regard
support from sons and daughters more favorably because expectations for support from
daughters-in-law are lowered.
How intergenerational support influences elders’ psychological well-being is a less
explored area in Chinese studies, and even fewer studies have differentiated children with
respect to their different functions. By considering the special cultural meaning of each
category of children and directing attention to the special functions of daughters-in-law, we
propose that instrumental support to older parents in rural China will result in fewer

65
depressive symptoms depending on the type of support considered and whether that support is
delivered by sons, daughters, or daughters-in-law, in conjunction with parent’s gender and
household composition. Together these form the basis for how children’s support is appraised
within a patrilineal family system with consequences for psychological well-being.
Methods
Sample
We collected data from a random sample of adults aged 60 and older living in rural
townships within Chaohu, a primarily agricultural municipal district of 4.5 million people
located on the north bank of the Yangtze River in the central part of Anhui Province. Using a
stratified multistage method, we selected eligible respondents from 72 randomly selected
villages within six rural townships in the Chaohu region. Adults aged 60 years and older were
randomly selected from village rosters, with an over-sampling of people 75 years old and
older (Silverstein et al., 2006). The survey was originally fielded in April 2001 as a joint
project between Xi’an Jaotong University and the University of Southern California. Of 1,800
eligible participants selected at baseline, 1,715 completed the survey, yielding a response rate
of 95.3%. In November 2003, a follow-up survey was conducted with 1,368 respondents, or
79.8% of the original participants. The primary reason for sample attrition was mortality
(14%), leaving only 6% of the sample lost to follow-up for non-mortality related reasons. The
longitudinal sample we analyzed consisted of 1,281 older people who participated in both
surveys, who had at least one living child, and had no missing value of depressive symptoms
for the second interview.
The surveys were conducted in respondents’ homes and included assessments of family
relations, intergenerational transfers, physical health status, and psychological well-being. A
bilingual speaker translated the English questionnaire into Chinese, then, another bilingual
speaker translated that Chinese questionnaire back into English. Final questionnaire was based

66
on satisfactory consistency between the original questionnaire and back-translated
questionnaire.
Dependent variable
The dependent variable was the severity of depressive symptoms at the second wave of
measurement. This measure was adapted from the Center for Epidemiologic
Studies–Depression scale (Radloff, 1977) and tailored to the target population (Hermalin,
2002). Three items indicated feelings of positive affect (feeling happy, enjoying life, feeling
pleasure), two items indicated feelings of negative affect (feeling lonely, feeling upset), two
items indicated feelings of marginalization (feeling useless, having nothing to do), and two
items indicated somatic symptoms (having poor appetite, having trouble sleeping). We coded
the frequency with which the respondent had experienced each symptom in the past week as 0
(rarely or none of the time), 1 (some of the time), or 2 (most of the time). After we reversed
the coding of positive affect items, we summed the nine variables, which resulted in a
depression score ranging from 0 to 18, with a higher score indicating more depressive
symptoms. The reliability coefficient for the nine items was α = .80 at the second wave of
measurement.
Independent variables
Our key predictor variables represented two aspects of intergenerational instrumental
support (household chores and personal care) from each of three sources (sons, daughters, and
daughters-in-law). We asked whether during the past 12 months respondents received support
because of their poor health with (a) household chores, such as cleaning the house and
washing clothes, and (b) personal care tasks, such as bathing and dressing. If they received
help we further asked them to name the person or persons who provided each kind of help,
and how often it was provided. For each person cited, we coded support intensity for each
kind of help as: 0 = none, 1 = seldom, 2 = several times a month, 3 = at least once per week,

67
and 4 = every day. We then added support intensities across providers of each type to obtain
two support scores each for sons, daughters and daughters-in-law.
We included variables representing baseline support and change in support between the
two waves, yielding 12 support variables in total (2 types of support x 3 sources x 2
baseline/change). Because baseline support variables were non-normally distributed with high
percentages receiving no support, we used dummy coding to represent baseline support (1 =
received any help, 0 = received no help), and difference scores to represent change in support
from each source (support
wave2
- support
wave1
).
We included a dummy variable indicating coresidence with a daughter-in-law at baseline
(1 = coresides with at least one daughter-in-law, 0 = not coresides with any daughter-in-law),
and we interacted it with each of the twelve support variables to test its role in moderating the
effects of support.
We controlled for total financial support from sons and from daughters respectively,
measured as the total RMB value elders received from all sons and all daughters during the
previous 12 months of the interview. Because spouses were important sources of support and
might compensate for the lack of intergenerational support, we controlled for instrumental
support from spouses in the second wave calculated as the sum of spouses’ assistance with
household chores and personal care.
We also controlled for socio-demographic variables known to be associated with old age
depression (Cheng & Chan, 2006; Chou & Chi, 2005; Krause, Liang, & Gu, 1998). These
included baseline age in chronological years, number of children, and dummy variables for
gender (1 = female, 0 = male) and education (1 = some formal education, 0 = no formal
education). We controlled for elders’ second wave marital status (1 = not married, 0 =
married), health and income instead of their first wave counterparts to avoid the confounding
effects that children’s support changed to accommodate to parents’ changes of marital status,

68
deterioration of health and reduction in income, which were closely related to depression.
Health status was measured as the extent of functional impairments, calculated as the sum of
15 items reflecting difficulties in performing personal activities of daily living, instrumental
activities of daily living and activities requiring physical strength, mobility, and flexibility.
Respondents indicated the level of difficulty performing each task: 0 = no difficulty, 1 = some
difficulty, or 2 = cannot do it without help. Because these items were highly reliable (alpha
= .96), we calculated a summed scale that ranged from 0 (no difficulties) to 30 (unable to
perform all tasks). Income was measured as the logged RMB value of the total annual income
of respondent and spouse from work or pension (+1).
Because our analysis investigated change in depressive symptoms, we controlled for
baseline depressive symptoms, measured using the same nine items and operationalized with
the same scale score as in the second wave ( α = .76). Coefficients of other variables in the
model indicated their effects on residualized change in depressive symptoms between waves.
This approach minimized the risk of endogeneity in our specification in the event that
depression influences the receipt of support.
Analytic strategy
We used OLS multiple regression to examine the lagged and dynamic effect of support
from sons, daughters and daughters-in-law on older people's depressive symptoms in the
second wave of measurement. All models were estimated for the total sample and stratified by
gender of the respondent in order to capture differences in sensitivity between older mothers
and fathers. We estimated equations hierarchically, first including all direct effects and then
adding interaction terms to test the moderating influence of coresidence with daughters-in-law
on support from sons, daughters and daughters-in-law. Because several interaction terms
shared common component variables, multicollinearity was a concern. Therefore, we added
interaction terms separately in blocks based on type of support (household tasks vs. personal

69
Table 4.1
Description of Analytic Variables
Total
(N = 1281)
Women
(n = 674)
Men
(n = 607)
Coding scheme and range
Variables M SD M SD M SD
Depressive symptoms at T2 6.26 4.06 7.05 4.09 5.39* 3.84 0 (least depressed) – 18 (most depressed)
Depressive symptoms at T1 6.29 3.97 6.85 3.94 5.69* 3.93 0 (least depressed) - 18 (most depressed)
Age 69.89 6.97 71.20 7.23 68.43* 6.37 58 - 92 years
Gender 0.53 0.50 0 (male), 1 (female)
Education 0.22 0.41 0.06 0.24 0.39* 0.49 0 (no education), 1 (some education)
Detailed education
Illiterate (%) 78.38 94.07 60.96
Primary school (%) 17.95 05.19 32.13
Junior high or more (%) 03.67 0.74 06.92
Unmarried 0.44 0.50 0.59 0.49 0.29* 0.45 0 (married), 1 (not married)
Detailed marital status
Married (%) 55.58 41.25 71.50
Widowed (%) 43.79 58.31 27.68
Divorced (%) 0.62 0.45 0.82
Functional limitations 5.48 7.33 7.45 7.90 3.30* 5.94 0 (none)-30 (most)
Income (ln+1) 3.40 3.62 2.33 3.39 4.59* 3.49 0 - 9.9
Children number 4.02 1.60 4.00 1.59 4.05 1.61 1 - 10
Spouses’ instrumental support 0.78 1.37 0.60 1.29 0.99* 1.43 0 (no support) - 8 (highest observed intensity)
Money from sons 5.24 2.62 5.34 2.47 5.13 2.78
Money from daughters 4.81 2.37 4.62 2.45 5.03* 2.25
Coresidence with any daughter-in-law 0.23 0.42 0.26 0.44 0.19* 0.40 0 (no), 1 (yes)
Instrumental support
Baseline household chores
Sons 0.23 0.42 0.28 0.45 0.17* 0.37 0 (no support), 1 (some support)
Daughters 0.24 0.43 0.25 0.43 0.22 0.42 0 (no support), 1 (some support)
Daughters-in-law 0.29 0.46 0.35 0.48 0.23* 0.42 0 (no support), 1 (some support)
Changes in household chores
Sons -0.15 2.23 -0.16 2.21 -0.14 2.26 -12 - 15
Daughters -0.29 2.92 -0.27 2.78 -0.31 3.07 -20 - 12
Daughters-in-law -0.29 2.23 -0.28 2.32 -0.31 2.12 -12 - 12

70
Table 4.1, Continued
Baseline personal care
Sons 0.10 0.31 0.12 0.33 0.09* 0.28 0 (no support), 1 (some support)
Daughters 0.11 0.31 0.13 0.34 0.08* 0.27 0 (no support), 1 (some support)
Daughters-in-law 0.12 0.32 0.17 0.38 0.06* 0.23 0 (no support), 1 (some support)
Changes in personal care
Sons -0.01 1.60 0.00 1.48 -0.02 1.73 -12 - 15
Daughters -0.01 2.09 0.00 2.12 -0.01 2.06 -20 - 12
Daughters-in-law -0.03 1.63 -0.07 1.75 0.01 1.48 -16 - 16
*Denotes significant difference between females and males at p < .05

71
care), and then together to ascertain their unique effects. To compensate for low power
associated with multiplicative variables, we used p-value < .10 to identify statistically
significant effects for interactions.
Missing values were uncommon in this sample; on no variable did the percentage of
missing values exceed 5%. In order to maximize our sample size and use full information, we
imputed missing values of independent variables using multiple imputations with the
expectation maximization algorithm found in Solas 3.0 (Statistical Solutions, Inc., 2001). This
procedure generates unbiased estimates and correct standard errors subject to the assumption
that values are missing at random, conditional on observed values (Acock, 2005; Little &
Rubin, 2002).
Results
We show in Table 4.1 the mean values of study variables for the total sample and by
gender. For the total sample, depressive symptoms averaged 6.3 (SD = 4.0) in the first wave
and 6.3 (SD = 4.1) in the second wave. Although there are no diagnostic benchmarks in this
scale, we note that extrapolating to the distributions of similar scales reveals that these figures
indicate moderate levels of distress, which, given the rural poverty of the region, is not
surprising (Silverstein et al., 2006).
Socio-demographic characteristics of the respondents in Table 4.1 revealed that the
average age of the sample at baseline was 69.9 years (SD = 7.0), slightly more than half (53%)
was female. 78.4% of elders had no formal education, 18.0% got primary school education,
and only 3.7% got education of junior high school or higher. 55.6% of elders were currently
married, 43.8% were widowed, and only 0.6% were divorced. Altogether, 44% elders were not
currently married. The average logged income was 3.4, a figure converting to 28 RMB that
reflected the large number of elders in this population with no external sources of income. The
average functional impairment score was 5.8 out of a possible 30. The average respondent had

72
four living children. Average intensity of spouses’ help was low, corresponding to 0.8, which
was less than receiving at least one item help. Financial support from sons and daughters
averaged 5.3 and 4.8, corresponding to RMB 188 and 122, both of which were much higher
than elders’ own income. Twenty three percent of elders lived with at least one of their
daughters-in-law, of whom almost all lived with a son as well, whereas only 3.1% of elders
lived with any of their daughters (not shown in the table).
In terms of support with household chores from children at baseline, 23% received help
from a son, 24% from a daughter and 29% from a daughter-in-law. With regard to receiving
personal care, 10% received such help from a son, 11% from a daughter, and 12% from a
daughter-in-law. Between the first and second waves, the average amount of household
assistance and personal care decreased from all three sources, which might have resulted from
social changes like reduced coresidence and adult children’s migration.
Comparing female and male respondents, we found several differences between them that
highlighted disadvantages experienced by older women in rural China. Female elders were
more depressed than males at both waves, as well as older, less educated, more likely to be
unmarried, in poorer health, poorer in income, and received less instrumental support from
their spouses. Female elders were also more likely to coreside with daughters-in-law and
receive baseline instrumental support from children (except household chores from daughters).
In Table 4.2, we examined differences in baseline support based on living arrangements
with daughters-in-law, stratified by parents’ gender. Living with a daughter-in-law
dramatically increased the amount of both types of support received from daughters-in-law
and sons. Rates of each type of support to older men and older women consistently
increased--by about four-fold for daughters-in-law and two-fold for sons--as a result of the
proximity produced by coresidence. However, assistance from daughters changed little and
did not appreciably decline as a consequence of living with daughters-in-law (except personal

73
Table 4.2
Percentage of Older Parents Receiving Baseline Household Support and Personal Care by Gender and Living Arrangement with Daughter-in-Law
(%)
Total (N = 1281) Females (n = 674) Males (n = 607)
Coreside with no
daughters-in-law
Coreside with
daughters-in-law
Coreside with no
daughters-in-law
Coreside with
daughters-in-law
Coreside with no
daughters-in-law
Coreside with
daughters-in-law
Receiving household support
at baseline from:

Sons 18.1 37.7 21.8 45.4 14.3 26.3
Daughters 24.4 21.6 26.0 22.4 22.7 20.3
Daughters-in-law 17.6 69.5 20.0 77.6 15.1 57.6
Receiving personal care at
baseline from:

Sons 7.8 19.5 9.0 21.3 6.5 16.9
Daughters 10.0 12.3 12.6 14.4 7.4 9.3
Daughters-in-law 6.4 30.5 9.2 40.8 3.5 15.3

74
Table 4.3
Unstandardized Ordinary Least Squares Regression Estimates Predicting Depressive Symptoms
Total (N = 1281) Women (n = 674) Men (n = 607)
Variable Model 1 Model 2 Model 3 Model 4 Model 1 Model 2 Model 3 Model 4 Model 1 Model 2 Model 3 Model 4
Constant 6.600*** 6.641*** 6.300*** 6.571*** 7.221*** 7.212*** 6.894*** 7.010*** 5.828*** 5.420*** 5.521*** 5.422***
Controls
Depressive
symptoms at T1
0.201*** 0.201*** 0.202*** 0.201*** 0.219*** 0.223*** 0.218*** 0.220*** 0.175*** 0.171*** 0.180*** 0.175***
Age -0.020 -0.020 -0.016 -0.019 -0.024 -0.024 -0.019 -0.021 -0.010 -0.003 -0.005 -0.003
Gender 0.288 0.255 0.255 0.225
Education -0.135 -0.171 -0.209 -0.204 -0.768 -0.787 -0.863 -0.859 0.040 0.011 -0.023 -0.003
Unmarried 0.732** 0.732** 0.700** 0.720** 0.522 0.529 0.547 0.566 0.874* 0.924* 0.831* 0.891*
Functional
limitations
0.220*** 0.221*** 0.223*** 0.224*** 0.197*** 0.198*** 0.198*** 0.199*** 0.259*** 0.263*** 0.256*** 0.262***
Income (ln+1) -0.114*** -0.115*** -0.113*** -0.116*** -0.168*** -0.168*** -0.165*** -0.166*** -0.069 -0.068 -0.069 -0.064
Children number 0.060 0.077 0.070 0.076 0.042 0.051 0.034 0.040 0.107 0.138 0.121 0.137
Spouses’
instrumental
support
0.128 0.114 0.118 0.112 0.185 0.180 0.181 0.180 0.106 0.088 0.093 0.085
Money from
sons
-0.151*** -0.149*** -0.147*** -0.144*** -0.143* -0.137* -0.135* -0.130* -0.168** -0.172** -0.173 -0.178**
Money from
daughters
-0.158*** -0.160*** -0.159*** -0.158*** -0.134* -0.135* -0.129* -0.128* -0.209** -0.219** -0.206 -0.216***
Coresidence with
any
daughter-in-law
-0.301 -0.940* -0.363 -0.765+ -0.143 -0.560 -0.081 -0.434 -0.317 -0.927+ -0.495 -0.609
Instrumental
support

Baseline
household chores

Sons 0.078 -0.413 0.105 -0.087 0.109 -0.485 0.211 -0.068 0.228 -0.131 0.127 -0.123
Daughters 0.690* 0.640+ 0.700* 0.643+ 1.131* 1.159* 1.052* 1.157* 0.508 0.494 0.461 0.393
Daughters-in-law -0.316 -0.456 -0.449 -0.952* -0.208 -0.039 -0.427 -0.611 -0.692 -1.427* -0.663 -1.327+

75
Table 4.3, Continued
Total (N = 1281) Women (n = 674) Men (n = 607)
Variable Model 1 Model 2 Model 3 Model 4 Model 1 Model 2 Model 3 Model 4 Model 1 Model 2 Model 3 Model 4
Changes in
household chores

Sons 0.091 0.004 0.093 0.088 0.036 -0.046 0.043 0.025 0.220* 0.125 0.197+ 0.157
Daughters 0.113* 0.101+ 0.114* 0.075 0.211** 0.230** 0.203** 0.219** 0.073 0.027 0.072 -0.020
Daughters-in-law -0.124+ -0.135+ -0.125+ -0.208* -0.200* -0.170+ -0.198* -0.224* -0.117 -0.237+ -0.104 -0.238+
Baseline
personal care

Sons 0.242 0.294 -0.571 -0.360 0.691 0.670 -0.575 -0.340 -0.525 -0.480 -0.339 -0.094
Daughters -1.251** -1.296** -1.355** -1.336* -1.483* -1.555* -1.623* -1.713* -1.198 -1.198 -1.161 -1.199
Daughters-in-law -1.134* -1.240** -0.492 -0.076 -1.742** -1.794** -0.576 -0.423 0.217 0.529 -0.744 0.078
Changes in
personal care

Sons -0.004 -0.001 -0.228* -0.211+ 0.249 0.237 -0.054 -0.030 -0.307* -0.271+ -0.373* -0.321*
Daughters -0.244*** -0.241*** -0.213** -0.180* -0.239* -0.243* -0.238* -0.247* -0.272** -0.242* -0.205+ -0.122
Daughters-in-law -0.085 -0.103 0.024 0.076 -0.199+ -0.202 -0.017 -0.004 0.168 0.168 0.065 0.147
Interactions with
coresidence with
daughters-in-law

Baseline
household chores

Sons 1.475* 0.617 1.716* 0.770 1.113 0.441
Daughters 0.236 0.421 -0.114 -0.377 -0.435 -0.158
Daughters-in-law 0.399 1.123 -0.350 0.497 1.582+ 1.380
Changes in
household chores

Sons 0.304* 0.022 0.254 0.052 0.268 -0.026
Daughters 0.053 0.181 -0.056 -0.082 0.075 0.343+
Daughters-in-law 0.008 0.208 -0.078 0.065 0.291 0.339
Baseline
personal care

Sons 2.493** 1.892+ 2.954** 2.343+ 0.321 -0.254
Daughters 0.223 -0.093 0.219 0.471 -0.352 -0.028
Daughters-in-law -1.512+ -2.298* -2.183* -2.417* 2.095 0.699

76
Table 4.3, Continued
Total (N = 1281) Women (n = 674) Men (n = 607)
Variable Model 1 Model 2 Model 3 Model 4 Model 1 Model 2 Model 3 Model 4 Model 1 Model 2 Model 3 Model 4
Changes in
personal care

Sons 0.736*** 0.664** 0.651** 0.588* 0.444 0.418
Daughters -0.115 -0.266 0.012 0.059 -0.246 -0.528*
Daughters-in-law -0.322* -0.461* -0.339+ -0.377 0.225 -0.062
R
2
0.346 0.352 0.356 0.361 0.336 0.340 0.346 0.348 0.327 0.342 0.338 0.352
†p < .10; *p < .05; **p <.01; ***p < .001

77
care from daughters increased for men). Consequently when daughters-in-law were not
available in the household, both older women and men relied more on their daughters than on
their sons and daughters-in-law for household support and personal care. When
daughters-in-law were available in the household, both older women and men relied
predominantly on their daughters-in-law for household support; older men relied more on sons
and daughters-in-law than on their daughters in these shared households for personal care, but
for older women, daughters-in-law were clearly the main providers of personal care by a large
margin.
Multivariate results predicting depression in older parents are shown in Table 4.3.All 4
Models for the total sample revealed that help in household chores from daughters increased,
whereas personal care from daughters reduced elders’ depression. When elders were living
with daughters-in-law, sons’ help in household chores (Model 2) and personal care (Model 3, 4)
increased depression. Daughters-in-law’s personal care reduced depression (Model 1, 2), and
even stronger in the coresident setting (Model 3, 4). When we did analysis separately by
gender, the general pattern observed in the total full sample appeared to be largely driven by
mothers.
For mothers, main effect model showed that baseline and increases in household chores
help from daughters were associated with more depressive symptoms, whereas increases in
help in household chores from daughters-in-law reduced mothers’ depression. Baseline and
increases in personal care from daughters, and baseline personal care from daughters-in-law
were associated with fewer depressive symptoms. When these effects were examined by
whether elder coresided with daughters-in-law as in Model 2, 3 and 4, the harmful effects of
baseline and increases in household help from daughters, beneficial effect of increases in
household help from daughters-in-law, and beneficial effects of baseline and increases in
personal care from daughters persisted across models and did not depends on elder’s living

78
arrangement. However the beneficial effect of receiving baseline personal care from
daughters-in-law’s was in fact driven by that in the coresident setting, as shown in Model 3
and 4. When elders were living with daughters-in-law, sons’ baseline household support
(Model 2), baseline and increases in personal care increased mothers’ depression (Model 3, 4).
For fathers, in the main effect model (Model 1), increases in household support from sons
were associated with more depressive symptoms, but increases in sons’ and daughters’
personal care were associated with fewer depressive symptoms. When these effects were
examined by whether elder coresided with daughters-in-law as in Model 2, 3 and 4, the effect
of increases in sons’ personal care persisted across 4 models and did not depend on elders’
living arrangement, whereas the effects of increases in daughters’ personal care in fact
reflected that in the coresident setting (Model 4). The effect of household support from sons
disappeared when interactions were added as shown in Model 2 and 4. Although increases in
household help from daughters did not show significant effects in the main model, in Model 4,
we detected that when daughters-in-law were available in the household, household help from
daughters contributed to increased depression. In addition, although baseline household help
from daughters-in-law did not show significant effects in the main model, we found in Model
2 that household support from coresident daughters-in-law increased depressive symptoms of
fathers, though this effect disappeared when all interactions were entered together (Model 4).
In terms of control variables, previous depressive symptoms and functional limitations
were positively related to depressive symptoms. Both sons and daughters’ financial support
reduced elders’ depression. Income was inversely related to depressive symptoms among
mothers, and unmarried fathers had more depressive symptoms than married fathers.
Discussion
In rural China, patrilineal norms prescribe that sons be primary providers of support to
their older parents. However, within this traditional cultural milieu, such a

79
formulation—though largely valid—is an oversimplification, particularly with regard to
instrumental support. Although previous studies about Chinese families have noted the labor
division between sons (and their families) and daughters (Lin et al., 2003; Zhan, 2004), we
noted the labor division within sons’ families by singling out support from daughters-in-law.
We tested this proposition in a sample of older adults from a rural region of China by
investigating whether receiving instrumental support from three intergenerational sources
influenced depressive symptoms. Although our results are not always consistent with
expectations, these results support the framework that beneficial psychological outcome is
related to support deriving from preferred resources, subject to the type of task, and the care
recipients’ gender and living arrangements.
A general pattern emerged that supported the prolific and meaningful contributions of
daughters-in-law in the support systems of older people in rural China--a circumstance in large
part attributable to strong cultural expectations for coresidence with married sons. Aversion to
household support from daughters and sons was sufficiently strong among older mothers and
fathers to cause negative psychological outcomes, affirming the adverse emotional
consequences that resulted when traditional expectations were violated.
In addition, when expectations of daughters-in-law were ostensibly heightened by
coresidence, preference to daughters-in-law’s support and aversion to sons’ support were
stronger among older mothers, with respect to personal care. The benefits of care from
coresident daughters-in-law might be better characterized by its inverse, namely that the
absence of personal care provisions from daughters-in-law who were best positioned to
provide support was particularly distressing to older mothers, as expectations were raised and
not met (Lee et al., 1995). Similarly with regard to care from sons, mothers in a household
context, in which expectations of daughters-in-law were heightened, tended to be happier
when personal care did not come from this unexpected source.

80
Cultural rules and the pragmatics of family relationships may sometimes lead to
contradictory expectations, such as when the requirements of modesty (favoring sons), the
convenience of proximity (favoring coresident sons and daughters-in-law), gender-expectation
traditionalism (favoring daughters and daughters-in-law) and the length of personal
relationships (favoring daughters and sons) create competing incentives where intimate
contact and body exposure are involved. Indeed, we found that personal care from daughters
was psychologically beneficial for older fathers in the presence of a coresident daughter-in-law.
This is consistent with findings in U.S. that expectations for support by elders are sensitive to
their children’s gender, proximity, needs, resources, and whether they are consanguineal or
affinal (Ganong & Coleman, 1999; Killian & Ganong, 2002; Rossi & Rossi, 1990; Silverstein
& Angelelli, 1998).
Our study confirmed a resounding theme in the literature that the appraisal of support is
more important than support itself (Chen & Silverstein, 2000; Krause, 2001; Sun, 2004). Our
results showed that elders appraise household chores support from sons and daughters
negatively even when daughters-in-law were not available in the household, reflecting their
refusal to compromise their expectations of daughters-in-law. Their belief lags behind societal
changes that have narrowed differences between support from daughters and daughters-in-law,
and have empowered women to maintain residential independence from their in-laws and
provide more support to their own parents (Cooney & Di, 1999; Goldenstein & Ku, 1993; Yan,
2003; Yang, 1996; Zhan & Montgomery, 2003). Consequently, elders will be psychologically
disadvantaged unless they contemporize their expectations to match the changing social
realities of Chinese society.
By directing attention to the different filial responsibilities prescribed to sons, daughters
and daughters-in-law, our findings help resolve contradictions in the literature about how
intergenerational instrumental support affects elders’ psychological well-being. The negative

81
effects of instrumental support are also observed in the U.S., where instrumental support
induces a sense of dependence or inability to reciprocate in intergenerational exchanges
(Silverstein, Chen, & Heller, 1996; Stoller, 1985; Wolff & Agree, 2004). In Chinese culture,
the negative effects of support has less to do with dependence than with the appropriate match
between the provider and the services delivered (Cheng & Chan, 2006).
Our study has several limitations that deserve mention. First, we note that our sample was
drawn from a poor, rural region in China, where strong son preference persists. Rapid social
change in China has altered traditional expectations of children as sources of support to their
older parents, alerting us to the possibility that our study population was unique in ways that
made our findings non-generalizable to urban elders who hold more egalitarian view of their
children, and who are less dependent on their children for instrumental support (Chen &
Silverstein, 2000). Second, we were not able to directly assess elders’ personal preference
about support providers that are necessarily inferred their outcomes. Third, we can not assess
the quality of relationships with daughters-in-law that may explain why daughters-in-law are
important psychologically. Older people will tend to live with daughters-in-law with whom
they feel closer. In addition, coresident daughters-in-law who do not exhibit the proper
respectful behavior will likely have strained and conflicted relationships that would prove to
be distressing.
In this investigation we have adapted hierarchical compensatory and task-specific
theories to a cultural and national context far different than the one on which they were
developed. Our results suggest that intergenerational instrumental support helps to reduce
depression in older parents if it comes from the cultural appropriate sources. This suggests that
attachment to traditional expectations for support may make elders more depressed in such a
rapidly changing society as China.

82
Chapter 5: Discussion and Conclusions
Summary of Findings
Each of the three papers in this dissertation has made a meaningful contribution to our
knowledge about intergenerational relationships in Chinese rural families. This dissertation
consists of three independent articles, each addressing one specific question under the
overarching theme of how fast social change and the migration of working age adults from
rural to urban areas shapes and changes intergenerational relationships and the well-being of
the rural elders. Particularly, in the first paper, random effects regression analysis showed that
elders reaped higher financial return from their adult migrant children than from their adult
non-migrant children for child care and financial help that elders provided to their children. In
Chinese families, children did not need to reciprocate for parents’ help in the short run, but
migrant children were obliged to pay back parents’ help. It is because when elder parents
provided child care for children, they had higher bargaining power over their migrant children
than over their non-migrant children, and parents’ financial help to migrant children was a
type of strategic investment. In the second paper, analyses based on random effects logistic
regressions showed that whether an elder coresided with a child depended on elders’ previous
exchange with that child as well as elders’ exchange with the other siblings of that child. In
the third paper, analyses indicated that depressive symptoms were usually reduced by
assistance from daughters-in-law, and sometimes increased when such support was from sons.
These relationships held most strongly when mothers coresided with their daughters-in-law.
This research suggested that the benefits of intergenerational support were conditional on
culturally prescribed expectations that could be adjusted to practical constraints.
Resounding Themes
These three papers combined emphasize the importance of two perspectives. They are the
perspective of reciprocity and strategic investment, and that of linked lives within siblings.

83
Reciprocity and strategic investment
This dissertation demonstrates that intergenerational relationships follow the principle of
reciprocity, and parents can strategically arrange for their old age security by providing
grandchildren care and financial support to migrant children who have more economic
potential, but parents will not benefit more from migrant children in the form of coresidence.
Linked lives within siblings: An extended family perspective and network interactions
In addition, intergenerational relationship should be studied within extended families
instead of being constrained to a specific parent-child dyad. In the second paper, results show
that coresidence decision with a specific child depends on parents’ interactions with other
siblings. The linked lives perspective is also apparent in the third paper, which shows that the
effect of children’s support on parents’ depression depends on whether the expected support
providers are available. This dissertation, therefore, provides evidence to support the
importance of understating intergenerational relationships within the extended family, instead
of focusing only on a dyadic relationship.
Limitations
There are several limitations of this dissertation. First, exchanges between parents and
children are life long processes (Silverstein et al., 2002). This study only covers 5.5 years of
time, and some mechanisms may not be observed in only 5 years. In addition, the study is
based on the perspective of parents because each parent reported their exchanges with each
child. The region has comparatively high fertility, with an average of four children, and some
elders have as many as 10 children. Consequently, parents may be challenged to report the
actual exchanges with each child. The incongruence that may exist between children and
parents’ support may bring new perspectives into this study. Third, although this dissertation
emphasizes the interdependence of each child’s relationship with the parent, I am not able to
examine how the direct interactions among siblings influence their individual interactions with

84
their parents. The mechanisms of siblings’ interactions are inferred from the data, but have not
been measured with data. Fourth, although my analysis of coresidence is at the individual
child level, exchanges between parents and migrant children are incorporated as a family level
variable to avoid technical difficulties. The combined measurement eliminates many
interesting interactions among siblings.
Future Studies
In the future, studies on Chinese intergenerational relationships should take an extended
family perspective, and collect information of interactions among siblings. Currently, many
family studies are constrained by the availability of family members, and consequently have to
follow a lineage design, which undermines the completeness of intergenerational interactions
and impedes theoretical development. To accommodate more development, more
sophisticated statistical methods are required to adequately examine every child’s influence on
other siblings’ interactions with their parents.
In addition, this dissertation revealed that the term “children” is too general a concept for
analyzing Chinese intergenerational relationships. This dissertation distinguishes migrant
children from non-migrant children, sons from daughters, and biological children from
children-in-law. It shows that all these distinctions are meaningful and important. Further
studies should be sensitive to the differences among children to yield rigorous and meaningful
empirical results.
Conclusions
This dissertation studies intergenerational relationships and their consequences on elders’
depression in rural China. By taking a strategic investment and bargaining power perspective,
and a holistic extended family perspective, this dissertation addressed several research
questions that have not been addressed before in the current literature. This dissertation studies
a period of time when migration is bringing enormous changes to intergenerational

85
relationships and a culture where tradition still prevails but begins to yield to societal changes.
The results of this dissertation will have theoretical as well as policy implications in a time
when global aging has brought tremendous changes to intergenerational relationships.

86
Reference
Acock, A. C. (2005). Working with missing values. Journal of Marriage and Family, 67,
1012-1028.
Agree, E. M., Biddlecom, A. E., Chang, M. C., & Perez, A. E. (2002). Transfers from older
parents to their children in Taiwan and the Philippines. Journal of Cross-Cultural
Gerontology, 17, 269–294.
Agree, E. M., Biddlecom, A. E., & Valente, T. W. (2005). Intergenerational transfers of
resources between older persons and extended kin in Taiwan and the Philippines.
Population Studies, 59, 181-195.
Altonji, J. G., Hayashi, F., & Kotlikoff, L. (2000). The effects of income and wealth on time
and money transfers between paretns and children. In A. Mason & G. Tapinos (Eds.),
Sharing the Wealth (pp. 306-345). New York: Oxford University Press.
Antonucci, T. C., Akiyama, H., & Birditt, K. (2004). Intergenerational exchange in the United
States and Japan. In M. Silverstein & K. W. Schaie (Eds.), Annual Review of
Gerontology and Geriatrics: Focus on Intergenerational Relations Across Time and
Place (V ol. 24, pp. 224-248). New York: Springer.
Antonucci, T. C., & Jackson, J. S. (2003). Ethnic and cultural differences in intergenerational
social support. In V . L. Bengtson & A. Lowenstein (Eds.), Global aging and
challenges to families (pp. 355-370). New York: Aldine de Gruyter.
Arnsberger, P., Fox, P., Zhang, X., & Gui, S. (2000). Population aging and the need for long
term care: A comparison of the United States and the People's Republic of China.
Journal of Cross-Cultural Gerontology, 15, 207–227.
Attias-Donfut, C. (2003). Family transfers and cultural transmissions between three
generations in France. In V . L. Bengtson & A. Lowenstein (Eds.), Global aging and
challenges to families (pp. 214-250). New York: Aldine de Gruyter.
Bai, N., & Song, H. (2002). Return or stay?: The return flow of out-migrants from rural China
(in Chinese). Beijing: China Finance and Economy Press.
Becker, G. (1974). A theory of social interactions. Journal of Political Economy, 82,
1063–1093.
Becker, G., & Tomes, N. (1986). Human capital and the rise and fall of families Journal of
Labor Economics, 4, S1-S39.
Benjamin, D., Brandt, L., & Giles, J. (2005). The evolution of income inequality in rural
China. Economic Development and Cultural Change, 53, 769-824.
Bernheim, D., Shleifer, A., & Summers, L. (1985). The Strategic Bequest Motive. Journal of
Political Economy, 33, 1045-1076.

87
Bian, F., Logan, J. R., & Bian, Y . (1998). Intergenerational relations in urban China: proximity,
contact, and help to parents. Demography, 35, 115-124.
Brown, J. W., Liang, J., Krause, N., Akiyama, H., Sugisawa, H., & Fukaya, T. T. i. L. A. A. E.
i. J. D. H. M. a. D. J. o. G., 58(B), S209-S220. . (2002). Transitions in Living
Arrangements Among Elders in Japan: Does Health Make a Difference? . Journals of
Gerontology, 58(B), S209-S220.
Cantor, M. H. (1979). Neighbors and friends: An overlooked resource in the informal support
system. Research on Aging, 1, 434-463.
Chao, P. (1983). Chinese kinship. Boston: Kegan Paul International.
Chappell, N. L., & Kusch, K. (2007). The gendered nature of filial piety: A study among
Chinese Canadians. Journal of Cross-Cultural Gerontology, 22, 29-45.
Checkovich, T. J., & Stern, S. (2002). Shared Caregiving Responsibilities of Adult Siblings
with Elderly Parents. The Journal of Human Resources, 37, 441-478.
Chen, F., Short, S. E., & Entwistle, B. (2000). Impact of grandparental proximity on maternal
childcare in China. Population Research and Policy Review, 19, 571-590.
Chen, J. (2001a). The effect of parental investment on old-age support in urban China. In M.
K. Whyte (Ed.), China's Revolutions and Intergenerational Relations. Ann Arbor:
Center for Chinese Studies, University of Michigan.
Chen, R., Hu, Z., Qin, X., Xu, X., & Copeland, J. R. M. (2004). A community-based study of
depression in older people in Hefei, China: The GMS-AGECAT prevalence, case
validation and socio-economic correlates. International Journal of Geriatric
Psychiatry, 19, 407 - 413.
Chen, R., Wei, L., Hu, Z., Qin, X., Copeland, J. R. M., & Hemingway, H. (2005). Depression
in Older People in Rural China. Archives of Internal Medicine, 165, 2019-2025.
Chen, X. (2001b). Social support, social change, and psychological well-being of the elderly
in China: Does the type and source of support matter? Unpublished Doctoral
dissertation, University of Southern California, Los Angeles.
Chen, X., & Silverstein, M. (2000). Intergenerational social support and the psychological
well-being of older parents in China. Research on Aging, 22, 43-65.
Cheng, S., & Chan, A. C. M. (2006). Filial piety and psychological well-being in well older
Chinese. Journal of Gerontology: Psychological Sciences, 61B, P262–P269.
Chi, I., & Chou, K. L. (2000). Depression predicts cognitive decline in Hong Kong Chinese
older adults. Aging & Mental Health, 4, 148-157.
China Research Center on Aging (2003). Data analysis of the sampling of survey of the aged
population in China. Beijing: China Biaozhun Publication.

88
Chou, K. L., & Chi, I. (2005). Prevalence and correlates of depression in Chinese oldest-old.
International Journal of Geriatric Psychiatry, 20, 41-50.
Cohen, M. L. (1998). North China rural families changes during the communist era. Etudes
Chinoises, 17, 59-154.
Cooney, R. S., & Di, J. X. (1999). Primary family caregivers of impaired elderly in Shanghai,
China. Research on Aging, 21, 739-761.
Cox, D. (1987). Motives for private income transfers. Journal of Political Economy, 95,
508-546.
Cui, L., & Li, H. (1997). Chenshi laonianren shehui zhichi wangluo yu shenghuo manyidu de
yanjiu [Urban elders' social support network and life satisfaction]. Psychological
Science, 20, 123-126.
Das Gupta, M., & Li, S. (1999). Gender bias in China, South Korea and India 1920-1990:
Effects of war, famine and fertility decline. Development and Change, 30, 619-652.
Davis-Freidmann, D. (1991). Long lives: Chinese elderly and the communist revolution (2nd
ed.). Cambridge: Harvard University Press.
De Brauw, A., & Rozelle, S. (2003). Household investment through migration in rural China.
Retrieved July 1, 2005, from
http://www.williams.edu/Economics/wp/debrauwmiginv_jce_initial.pdf
Dean, A., Kolody, B., & Wood, P. (1990). Effects of social support from various sources on
depression in elderly persons. Journal of Health and Social Behavior, 31, 148-161.
Engers, M., & Stern, S. (2002). Long-term care and family bargaining. Journal of Human
Resources, 43, 73-114.
Feldman, M. W., Tuljapurkar, S., Li, S., Jin, X., & Li, N. (2006). Son preference, marriage and
intergenerational transfer in rural China. In A. H. Gauthier, C. Y . C. Chu & S.
Tuljapurkar (Eds.), International Studies in Population: Allocating Public and Private
Resources across Generations (V ol. 3, pp. 139-162): Springer Netherlands.
Felton, B. J., & Berry, C. A. (1992). Do the sources of the urban elderly social support
determine its psychological consequences. Psychology and Aging, 7, 89-97.
Frankenberg, E., Lillard, L., & Willis, R. J. (2002). Patterns of Intergenerational Transfers in
Southeast Asia. Journal of Marriage and the Family, 64, 627-641.
Friedman, M. M. (1993). Social support sources and psychological well-being in older women
with heart-disease. Research in Nursing & Health, 16, 405-413.
Gaetano, A. M., & Jacka, T. (Eds.). (2004). On the move : women and rural-to-urban
migration in contemporary China / edited by New York Columbia University Press.
Ganong, L., & Coleman, M. (1999). Changing families, changing responsibilities. Hillsdale,
NJ: Erlbaum.

89
Giles, J., & Mu, R. (2007). Elderly parent health and the migration decisions of adult children:
Evidence from rural China. Demography, 44 265-288.
Goldenstein, M. C., & Ku, Y . (1993). Income and family support among rural elderly in
Zhejiang province. Journal of Cross-Cultural Gerontology, 8, 197-223.
Graham, M. J., Larsen, U., & Xu, X. (1998). Son preference in Anhui Province, China.
International Family Planning Perspectives, 24, 72-77.
Henretta, J. C., Hill, M. S., Li, W., Soldo, B. J., & Wolf, D. A. (1997). Selection of children to
provide care: the effect of earlier parental transfers. Journal of Gerontology: Social
Sciences, 52B (special issue), 110-119.
Hermalin, A. I. (2002). Well-being of the elderly in Asia: A four-country comparative study.
Ann Arbor, MI: University of Michigan Press.
Hermalin, A. I., Roan, C., & Perez, A. E. (1998). The emerging role of grandparents in Asia.
No. . Ann Arbor, MI: University of Michigan. (Report): Elderly in Asia Research.
Joseph, A. E., & Phillips, D. R. (1999). Ageing in rural China: Impacts of increasing diversity
in family and community resources. Journal of Cross-Cultural Gerontology, 14,
153-168.
Kelley, S. J., Whitley, D., Sipe, T. A., & Yorker, B., C. . (2000). Psychological distress in
grandmother kinship care providers: The role of resources, social support, and
physical health. Child Abuse & Neglect, 24, 311-321.
Killian, T., & Ganong, L. H. (2002). Ideology, context, and obligations to assist older persons.
Journal of Marriage and Family, 64, 1080-1088.
Knodel, J., & Ofstedal, M. B. (2002). Patterns and determinants of living arrangements. In A. I.
Hermalin (Ed.), The well-being of the elderly in Asian: A four-country comparative
study (pp. 143-184). Ann Arbor: The University of Michigan Press.
Krause, N. (2001). Social support. In R. L. Binstock & L. K. George (Eds.), Handbook of
aging and the social sciences (5th ed., pp. 273-294). San Diego, CA: Academic Press.
Krause, N., & Liang, J. (1993). Stress, social support, and psychological distress among the
Chinese elderly. Journal of Gerontology: Psychological Sciences, 48, P282-P291.
Krause, N., Liang, J., & Gu, S. (1998). Financial strain, received support, anticipated support,
and depressive symptoms in the People's Republic of China. Psychology and Aging,
13, 58-68.
Lee, G. R., Netzer, J. K., & Coward, R. T. (1994a). Filial Responsibility Expectations and
Patterns of Intergenerational Assistance. Journal of Marriage and the Family, 56,
559-565.
Lee, G. R., Netzer, J. K., & Coward, R. T. (1995). Depression among older parents: The role of
intergenerational exchange. Journal of Marriage and the Family, 57, 823-833.

90
Lee, Y ., & Xiao, Z. (1998). Children's support for elderly parents in urban and rural China:
Results from a national survey. Journal of Cross-Cultural Gerontology, 13, 39-62.
Lee, Y . J., Parish, W. L., & Willis, R. J. (1994b). Sons, Daughters, and Intergenerational
Support in Taiwan. American Journal of Sociology, 99, 1010-1041.
Li, H., & Zahniser, S. (2002). The Determinants of Temporary Rural-to-Urban Migration in
China. Urban Studies, 39, 2219-2235.
Li, S., Feldman, M. W., & Jin, X. (2004). Children, marriage form, and family support for the
elderly in contemporary rural China: The case of Songzi. Research on Aging, 26,
352-384.
Liang, J., Gu, S., & Krause, N. (1992). Social support among the aged in Wuhan, China.
Asia-Pacific Population Journal, 7, 33-62.
Lillard, L. A., & Willis, R. (1997). Motives for Intergenerational Transfers: Evidence from
Malaysia. Demography, 34, 115-134.
Lin, I. F., Goldman, N., Weinstein, M., Lin, Y ., Gorrindo, T., & Seeman, T. (2003). Gender
differences in adult children's support of their parents in Taiwan. Journal of Marriage
and Family, 65, 184-200.
Little, R. J. A., & Rubin, D. B. (2002). Statistical analysis with missing data. Hoboken, NJ:
Wiley.
Litwak, E., Silverstein, M., Bengtson, V . L., & Hirst, Y . W. (2003). Theories about families,
organizations, and social supports. In V . L. Bengtson & A. Lowenstein (Eds.), Global
aging and challenges to families (pp. 27-53). New York: Aldine de Gruyter.
Liu, Q., & Reilly, B. (2004). Income transfers of Chinese rural migrants: Some empirical
evidence from Jinan. Applied Economics, 36, 1295–1313.
Liu, T., & Chan, K. W. (2001). National statistics on internal migration in China:
Comparability problems. China Information, 15, 75-113.
Logan, J. R., Bian, F., & Bian, Y . (1998). Tradition and Change in the Urban Chinese Family:
The Case of Living Arrangements. Social forces, 76, 32.
Lucas, R. E., & Stark, O. (1985). Motivations to remit: evidence from Botswana. Journal of
Political Economy, 93, 901-918.
McGarry, K., & Schoeni, R. F. (1995). Transfer behavior in the Health and Retirement Study:
Measurement and the redistribution of resources within the family. Journal of Human
Resources, 30, s184–s226.
Mehta, K. M., Yaffe, K., Langa, K. M., Sands, L., Whooley, M. A., & Covinsky, K. E. (2003).
Additive Effects of Cognitive Function and Depressive Symptoms on Mortality in
Elderly Community-Living Adults. Journals of Gerontology: Medical Sciences, 58.

91
Miller, E. T. (2004). Compromises of old age: Power and independence in rural north China.
DAI, 65, 221.
Nepomnyaschy, L. (2007). Child support and father-child contact: Testing reciprocal pathways.
Demography, 44, 93-112.
Oi, J. C. (1999). Two Decades of Rural Reform in China: An Overview and Assessment. The
China quarterly, 13.
Pang, E. C., Jordan-Marsh, M., Silverstein, M., & Cody, M. (2003). Health-seeking behaviors
of elderly Chinese Americans: Shifts in expectations. Gerontologist, 43, 864-874.
Pang, J. (1996). A basic profile of marriage and family status among migrants in Beijing. In L.
Zou (Ed.), Floating Population of Beijing (pp. 128-130). Beijing: Chinese Population
Press.
Pang, L. (2004). Working till you drop: The elderly in rural China. China Journal, 52, 73-94.
Parish, W., & Whyte, M. K. (1978). Village and Family in Contemporary China. Chicago:
University of Chicago Press.
Parker, G., Gladstone, G., & Chee, K. T. (2001). Depression in the planet's largest ethnic group:
The Chinese. The American Journal of Psychiatry, 158, 857-864.
Pearlin, L. I. (1989). The sociological study of stress. Journal of Health and Social Behavior,
30, 241-256.
Pezzin, L. E., Pollak, R. A., & Schone, B. S. (2006). Efficiency in family bargaining: Living
arrangements and caregiving decisions of adult children and disabled elderly parents:
National Bureau of Economic Research.
Poirine, B. (1997). A Theory of Remittances as an Implicit Family Loan Arrangement. World
development, 25, 24.
Radloff, L. S. (1977). The CES-D Scale: A self-report depression scale for research in the
general population. Applied Psychological Measurement, 1, 385–401.
Rossi, A. S., & Rossi, P. H. (1990). Of human bonding: Parent-child relations across the life
course. New York: Aldine de Gruyter.
Rozelle, S., Taylor, J. E., & deBrauw, A. (1999). Migration, remittances, and agricultural
productivity in China. American Economic Review, 89, 287-291.
Secondi, G. (1997). Private monetary transfers in rural China: Are families altruistic? Journal
of Development Studies, 33, 487-511.
Shi, J. (1997). Living Arrangement of the Elderly in China, 1990: Modernization and
Demographic Factors. 227.
Shi, L. (1993). Family financial and household support exchange between generations: A
survey of Chinese rural elderly. Gerontologist, 33, 468-480.

92
Silverstein, M., & Angelelli, J. J. (1998). Older parents' expectations of moving closer to their
children. Journal of Gerontology: Social Sciences, 53B, S153-S163.
Silverstein, M., Chen, X., & Heller, K. (1996). Too much of a good thing? Intergenerational
social support and the psychological well-being of older parents. Journal of Marriage
and the Family, 58, 970-982.
Silverstein, M., Cong, Z., & Li, S. (2006). Intergenerational transfers and living arrangements
of older people in rural China: Consequences for psychological well-being. Journal of
Gerontology: Social Sciences, 61B, S256-S266.
Silverstein, M., Conroy, S. J., Wang, H., Giarrusso, R., & Bengtson, V . L. (2002). Reciprocity
in parent-child relations over the life course. Journal of Gerontology: Social Sciences,
57B, S3-S13.
Statistical Solutions, Inc. (2001). SOLAS for missing data analysis (release 3.0).
Saugus, MA: Statistical Solutions.
Stark, O. (1991). The Migration of Labor. Basil Blackwell: Oxford.
Stark, O., & Lucas, R. E. B. (1988). Migration, remittances, and the family.
Stoller, E. P. (1985). Exchange patterns in the informal support networks of the elderly: The
impact of reciprocity on morale. Journal of Marriage and the Family, 47, 335-342.
Sun, R. (2002). Old Age Support in Contemporary Urban China From Both Parents' and
Children's Perspectives. Research on Aging, 24, 337-359.
Sun, R. (2004). Worry about medical care, family support, and depression of the elders in
urban China. Research on Aging, 26, 559-585.
Sung, K. T. (1995). Measures and dimensions of filial piety in Korea. Gerontologist, 35,
240-247.
Sung, K. T. (1998). Exploration of actions of filial piety. Journal of Aging Studies, 12.
Szinovacz, M. E., DeViney, S., & Atkinson, M. P. (1999). Effects of surrogate parenting on
grandparents' well-being. J Gerontol B Psychol Sci Soc Sci, 54, S376-388.
Unger, J. (1993). Urban Families in the Eighties: An Analysis of Chinese Surveys. In D. Davis
& S. Harrell (Eds.), Chinese Families in the Post-Mao Era (pp. 25-49). Berkeley:
University of California Press.
Wang, F., Ping, R., Zhan, S., & Shen, A. (2005). Reproductive health status, knowledge, and
access to health care among female migrants in Shanghai, China. Journal of Biosocial
Science, 37, 603-622.
Wang, F., & Zuo, X. (1999). Inside China's cities: Institutional barriers and opportunities for
urban migrants. American Economic Review, 89, 276-280.

93
Wang, W. W., & Fan, C. C. (2006). Success or failure: selectivity and reasons of return
migration in Sichuan and Anhui, China. Environment and Planning, 38, 939-958.
Wheaton, B. (1985). Models for the stress-buffering functions of coping resources. Journal of
Health and Social Behavior, 26, 352-364.
Whyte, M. K. (2003). The persistence of family obligation in Baoding. In M. K. Whyte (Ed.),
China's revolutions and intergenerational relations (pp. 85–120). Ann Arbor, MI:
University of Michigan Center for Chinese Studies.
Willson, A. E., Shuey, K. M., & Elder, G. H. (2003). Ambivalence in the relationship of adult
children to aging parents and in-laws. Journal of Marriage and Family, 65,
1055-1072.
Wolff, J. L., & Agree, E. M. (2004). Depression among recipients of informal care: The effects
of reciprocity, respect, and adequacy of support. Journal of Gerontology: Social
Sciences, 59B, S173-S180.
Wong, O. M. (2005). Gender and intimate caregiving for the elderly in Hong Kong. Journal of
Aging Studies, 19, 375-391.
Wong, S. T., Yoo, G. J., & Stewart, A. L. (2006). The changing meaning of family support
among older Chinese and Korean immigrants. Journals of Gerontology: Social
Sciences, 61B, S4-S9.
Yan, S., Chen, J., & Yang, S. (2003). Living arrangements and old-age support. In M. K.
Whyte (Ed.), China's revolutions and intergenerational relations (pp. 143-166). Ann
Arbor, MI: University of Michigan Center for Chinese Studies.
Yan, Y . (2003). Private life under socialism: Love, intimacy, and family change in a Chinese
village, 1949-1999. Standford, CA: Stanford University Press.
Yang, H. (1996). The distributive norm of monetary support to older parents: A study in China.
Journal of Marriage and the Family, 58, 404-415.
Youn, G., Knight, B. G., Jeong, H., & Benton, D. (1999). Differences in familism values and
caregiving outcomes among Korean, Korean American, and white American dementia
caregivers. Psychology and Aging, 14, 355-364.
Zeng, Y ., & George, L. (2001). Extremely rapid ageing and the living arrangements of the
elderly: The case of China. Population Bulletin of the United Nations, 42/43, 255-287.
Zhan, H. J. (2004). Through gendered lens: Explaining Chinese caregivers' task performance
and care reward. Journal of Women & Aging, 16, 123-142.
Zhan, H. J., & Montgomery, R. J. V . (2003). Gender and elder care in China: The influence of
filial piety and structural constraints. Gender & Society, 17, 209-229.

94
Zhang, H. (2004). Living alone and the rural elderly: Strategy and agency in post-Mao rural
China. In C. Ikels (Ed.), Filial piety: Practice and discourse in contemporary East
Asia (pp. 63-87). Stanford, CA: Stanford University Press.
Zhang, Y . B. (2002). Urbanization and rural older people's family support. Sociology Study (in
Chinese), 5, 112-118.
Zhao, Z. (2005). Migration, labor market flexibility, and wage determination in China: A
review. The Developing Economies, XLIII, 285-312.
Zhu, L. (2000). Zhengfu yu nongcun jiben yiliao baojian baozhang zhidu xuanze [The
government and options for alternative medical and health systems in rural areas].
Social Sciences in China (in Chinese), 4, 89-99.
Zimmer, Z. (2005). Health and living arrangement transitions among Chinas oldest-old.
Research on Aging, 27, 526-555.
Zimmer, Z., & Kwong, J. (2003). Family size and support of older Adults in urban and rural
China: Current effects and future implications. Demography, 40, 23-44.

Abstract (if available)

Abstract This dissertation consists of three independent papers, investigating how migration of working age adults from rural to urban areas in China influences intergenerational transfers, living arrangements, and the psychological well-being of elders who were raised and are embedded in the patrilineal family system.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Intergenerational transfers in rural China: Do children's gender and birth order matter?

PDF

Intergenerational support between grandparents and grandchildren in rural China and its effect on the psychological well-being of older adults

PDF

Family relationships and their influence on health outcomes over time among older adults in rural China

PDF

The immediate and long term legacy of relationships with grandparents for the well-being of grandchildren

PDF

Intergenerational social support and the psychological well-being of older parents in China

PDF

The effects of time and space on health status of older adults in China

PDF

Falls among older adults: characteristics of fallers, predictors of falls, and the impact of falls on health care and long-term care utilization

PDF

Normative obligations and parental care in social context

PDF

Intergenerational coresidence of older adults in contemporary Japan: traditional cultural norms in divergent styles

PDF

The effect of body weight on mortality: different countries and age groups

PDF

Childlessness and psychological well-being across life course as manifested in significant life events

PDF

Intergenerational transmission of values and behaviors over the family life course

PDF

The measurement, life course patterns, and outcomes of intergenerational ambivalence among parent-adult child dyads

PDF

Internet communication use, psychological functioning and social connectedness at older ages

PDF

Translating two physical activity programs for older adults into home- and community-based settings: "Active Start" and "Healthy Moves for Aging Well"

PDF

The decisions of migration and remittances in rural China

PDF

Fall related injuries among older adults in the Los Angeles region

PDF

Using mixed methods to identify the characteristics of older fraud victims

PDF

Older adults with visual impairments: the role of health dimensions in predicting falls

PDF

Examination of the long-term psychosocial and functioning effects of a healthy living, life management behavior intervention for older adults

Asset Metadata

Creator Cong, Zhen (author)

Core Title Children 's migration and the financial, social, and psychological well-being of older adults in rural China

School Leonard Davis School of Gerontology

Degree Doctor of Philosophy

Degree Program Gerontology

Publication Date 04/21/2008

Defense Date 11/07/2007

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Family Studies,Gerontology,intergenerational relationships,OAI-PMH Harvest,rural China

Place Name Anhui Sheng (states), China (countries)

Language English

Advisor Silverstein, Merril (committee chair), Chi, Iris (committee member), Crimmins, Eileen M. (committee member)

Creator Email zcong@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m1161

Unique identifier UC1271714

Identifier etd-Cong-20080421 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-61398 (legacy record id),usctheses-m1161 (legacy record id)

Legacy Identifier etd-Cong-20080421.pdf

Dmrecord 61398

Document Type Dissertation

Rights Cong, Zhen

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu