Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Intergenerational coresidence of older adults in contemporary Japan: traditional cultural norms in divergent styles
(USC Thesis Other)
Intergenerational coresidence of older adults in contemporary Japan: traditional cultural norms in divergent styles
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
INTERGENERATIONAL CORESIDENCE OF OLDER ADULTS IN
CONTEMPORARY JAPAN: TRADITIONAL CULTURAL NORMS IN
DIVERGENT STYLES
by
Emiko Takagi
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
(SOCIOLOGY)
August 2007
Copyright 2007 Emiko Takagi
ii
DEDICATION
To my parents who have always believed in me with a loving heart.
To my parents-in-law who have given me another great family tie.
iii
ACKNOWLEDGEMENTS
This dissertation could not have been possible without the emotional and
intellectual support from my mentors, colleagues, friends, and families. First and
foremost, I would like to express my sincere gratitude to my advisor and dissertation
chair, Dr. Merril Silverstein, who, with his profound knowledge of aging and
families, has provided me with indispensable assistance on my dissertation project. I
am deeply grateful for the amazing learning experience that I gained from working
with him.
I also would like to thank my dissertation committee members, Dr. Eileen
Crimmins and Dr. Roseann Giarrusso. I am thankful for Dr. Crimmins’s advice and
support that enabled me to gain special access to the data in Nihon University, which
greatly elevated the level and quality of this dissertation. Her swift and precise
advice always helped me stay motivated throughout the process of dissertation
writing as well as helped me to pursue a career beyond graduate-school training. I
am also very grateful for Dr. Giarrusso being such an invaluable mentor for me. Her
positive attitudes toward my potential have always been a great encouragement for
me to believe in myself and continue my work.
My gratitude also goes to my guidance committee members, Dr. Vern
Bengtson, Dr. Timothy Biblarz, and Dr. Kelly Musick, who have greatly assisted me
to explore a variety of theoretical paths and to formulate research ideas for this
dissertation. The innovative ideas rooted in their expertise certainly taught me
multiple perspectives, all of which evolved into the critical parts of my dissertation.
I would especially like to thank Dr. Bengtson for his introducing me to the complex
iv
sociological world with his deep knowledge of theories, as well as for his welcoming
me to his 35-year project, The Longitudinal Studies of Generations (LSOG). I am
deeply appreciative for being able to be a part of the LSOG project, which has given
me another great learning experience both personally and intellectually.
I would also like to acknowledge my precious colleagues and friends that I
met at USC who gave me so much support throughout my time in the graduate
program. I would like to thank my colleagues at the LSOG project: Linda Hall,
Danielle Zucker, Betty Oswald, Joe Raksin, Rahul Srinivas, Katherine Tsai, Norella
Putney, Sara Ruiz, Casey Copen, Zhen Cong, and Daphna Gans for their generous
support and encouragement of both my academic work and my personal life. I am
also truly thankful for the time and knowledge shared with my friends at USC: Jong-
Hyun Kim, Jihye Yeom, In-Hee Choi, Aaron Hagedorn, Stephanie Nawyn, Banu
Birdal, Kristen Smith, and Anne Burton. They are, and always will be, my
inspiration.
I would like to express my deepest gratitude for my families who have
brought me to where I am. I very much appreciate my parents for trusting my
decisions and abilities and giving me so much love and support throughout my life. I
also wish to thank my grandparents who set a great example of how to thrive in this
aging society. I am also thankful for my new family in South Korea for their
heartwarming support and encouragement of my study in the United States. Finally,
I would like to acknowledge my husband, Tae Eun Choe, for always being on my
side with his loving and caring heart, and my son Jun for bringing me another great
joy of love.
v
TABLE OF CONTENTS
DEDICATION……………………………………………………………………….ii
ACKNOWLEDGEMENTS ........................................................................................iii
LIST OF TABLES .....................................................................................................vii
LIST OF FIGURES ..................................................................................................viii
ABSTRACT………………………………………………………………………....ix
CHAPTER I Introduction.....................................................................................1
1.1 Aging Population in Japan .................................................................................3
1.2 Traditional Cultural Norm of Filial Piety...........................................................5
1.3 Social Changes and Declining Traditional Coresidence in Japan......................6
1.4 Current State of Intergenerational Coresidence in Japan...................................8
1.5 Diversity in Coresidence Arrangements ..........................................................10
1.6 Diversity in Coresidence at the Individual-Level ............................................12
1.7 Diversity in Coresidence at the Family-Level .................................................14
1.8 Diversity in Coresidence at the Contextual-Level ...........................................17
1.9 Study Purpose and Research Questions ...........................................................20
1.10 Overview of this Dissertation.........................................................................23
CHAPTER II Methods.........................................................................................25
2.1 Study 1: Traditional Cultural Norm and Intergenerational Coresidence .........26
2.2 Study 2: Intergenerational Coresidence and Family Contexts.........................28
2.3 Study 3: Intergenerational Coresidence and Social Contexts ..........................30
CHAPTER III Traditional Cultural Norms and Coresidence ...............................33
3.1 Intergenerational Coresidence and Filial Norms in Japan ...............................34
3.2 Research Questions ..........................................................................................38
3.3 Method .............................................................................................................40
3.4 Results..............................................................................................................47
3.5 Discussion ........................................................................................................53
3.6 Conclusion .......................................................................................................57
Chapter IV Coresidence and Family Contexts..................................................58
4.1 Negotiation between Older Parents and Adult Children..................................61
4.2 Negotiations among Adult Children with their Siblings..................................62
4.3 Negotiations between Adult Children and their Spouses.................................63
4.4 Study Purpose and Research Questions ...........................................................64
4.5 Method .............................................................................................................64
4.6 Results..............................................................................................................75
4.7 Discussion ........................................................................................................83
vi
CHAPTER V Coresidence and Social Contexts..................................................90
5.1 Diversity of Coresidence across Individuals....................................................92
5.2 Diversity of Coresidence across Regions.........................................................94
5.3 Study Purpose and Research Questions ...........................................................96
5.4 Methods............................................................................................................97
5.4 Results............................................................................................................106
5.5 Discussion ......................................................................................................117
CHAPTER VI Discussion ....................................................................................123
6.1 Elasticity of Coresidence Culture ..................................................................125
6.2 Traditional and Non-Traditional Styles of Coresidence ................................126
6.3 Multi-Level Determinants of Coresidence Culture........................................128
6.4 Elastic and Divergent Family Culture............................................................131
6.5 Implications for Future Studies......................................................................134
6.6 Implications for Families in Aging Societies.................................................136
Bibliography………………………………………………………………………140
vii
LIST OF TABLES
Table 1.Outcome of Interest.......................................................................................25
Table 2. Sample Characteristics (n=855)...................................................................42
Table 3. Living Arrangements (n=855) .....................................................................43
Table 4. Cross-Classification between Living Arrangements and Normative
Belief in Traditional Multigenerational Coresidence..................................47
Table 5. Maximum Likelihood Estimates of Nonrecursive Model for
Coresidence with Married Children and Normative Belief in
Traditional Multigenerational Coresidence (n=855)..............................50
Table 6. Maximum Likelihood Estimates of Nonrecursive Model for
Coresidence with Unmarried Children and Normative Belief in
Traditional Multigenerational Coresidence (N=686).................................52
Table 7. Older Parents’ Characteristics......................................................................67
Table 8. Adult Children’s Characteristics..................................................................69
Table 9. Living Arrangements and Adult Children’s Gender....................................77
Table 10. Multinomial Logit (N=8432): Odds Ratios to Non-Coresidence
Arrangement................................................................................................78
Table 11. Sample Characteristics.............................................................................107
Table 12. Contextual Characteristics of 47 Prefectures in Japan.............................108
Table 13. Multinomial Hierarchical Linear Model with Individual Predictors
(N=3565)...................................................................................................110
Table 14. Multinomial Hierarchical Linear Models with Individual and
Prefecture Predictors (N=3565) ................................................................113
viii
LIST OF FIGURES
Figure 1. Coresidence Rates of Older Individuals Over 65 Years-Old in Japan .........9
Figure 2. Conceptual Framework of Study 1............................................................27
Figure 3. Conceptual Framework of Study 2............................................................29
Figure 4. Conceptual Framework of Study 3............................................................31
Figure 5. Research Model .........................................................................................40
Figure 6. Predicted Probability of Child-Headed Coresidence by Adult
Children’s Gender and Education ...............................................................81
Figure 7. Predicted Probability of Parent-Headed Coresidence by Adult
Children’s Gender and Education ...............................................................82
Figure 8. Predicted Probability of Boomerang Coresidence by Older Parent’s
Normative Attitude and Cultural Climate of Prefecture ...........................115
Figure 9. Predicted Probability of Boomerang Coresidence by Older Parent’s
Normative Attitude and Welfare Resource Availability of Prefecture .....116
ix
ABSTRACT
Intergenerational coresidence is one of the distinct types of living
arrangements of older adults that reflects their socioeconomic conditions and the
cultural values shared by their family and society at large. Although the prevalence
of coresidence is decreasing in most of the developed nations in the world, Japan
remains an anomalous case because of its relatively high prevalence of coresidence
and the intricate connections between coresidence and the traditional cultural norm
of filial piety.
The goal of this dissertation is to investigate the ways in which traditional
cultural norms are reflected in current intergenerational coresidence arrangements
between older parents and adult children in Japan – a society that has undergone
extensive modernization in the past decades. Particular focus is given to the
generally overlooked diversity within this type of living arrangement by examining
multiple types of coresidence. The three individual studies that make up this
dissertation involve analyses of two sets of national data on Japanese older
individuals: the Japanese General Social Survey (JGSS) in 2001 and the Nihon
University Japanese Longitudinal Study of Aging (NUJLSOA) in 1999.
The findings of this dissertation indicate the elastic and divergent nature of
the culture of coresidence in Japan. First, traditional cultural norms and values in
society are not static; rather they become modified and variegated in tandem with the
individual experiences of families, the socioeconomic conditions of older adults, and
the community contexts in which families are embedded. Second, the coresidence
culture in contemporary Japan takes on divergent styles, including both traditional
x
styles that mirror the traditional cultural norms, and non-traditional styles that are
contingent on the imminent needs and resources of older parents and adult children.
Finally, the sustenance of coresidence practice requires not only the embracement of
traditional cultural norms but also sufficient levels of socioeconomic resources that
enable families to maintain households shared across generations. These results
suggest that the sociopolitical environment in modern society needs to become more
versatile in order to accommodate diverse needs and preferences of older individuals
and their aging families.
1
CHAPTER I Introduction
Living arrangements are a basic life condition, which for older people,
represent their ability to maintain life with varying degrees of support from families,
friends, and service providers (United Nations, 2005). One distinct type of family
living arrangement is intergenerational coresidence wherein older parents live with
their children and their families in a shared household. This residential arrangement
shared across generations generally represents a structural manifestation of the
socioeconomic conditions and cultural values shared by the family and society at
large. Intergenerational coresidence is a type of help and support committed by one
generation to another because it allows parents and children to share an extensive
amount of resources with one another (Pezzin & Schone, 1997, 1999). The decision
to coreside potentially hinges upon factors at multiple levels, including embraced
values and socioeconomic conditions at the individual level, needs and resources at
the family level, and socioeconomic and cultural climates at the community and
national level.
Across the developed nations of the world, the prevalence of coresidence
between older individuals and their children has been steadily decreasing (United
Nations, 2005). Declining coresidence is generally interpreted as the gradual
disappearance of traditional family culture and diminished societal necessity for this
type of living arrangement due to modern nations’ improvement of socioeconomic
and welfare conditions for older people (e.g., Logan, Bian, & Bian, 1998; Ogawa &
Retherford, 1993; Ting & Chiu, 2002).
2
Japan is one such developed country that has been experiencing a steady
decrease of traditional coresidence households in society. In the past two decades,
the proportion of traditional three-generation households in the older population has
dropped from 50 percent to 24 percent (Cabinet Office, 2005). However, at the same
time, the country remains an anomalous case because of its still relatively high
prevalence of coresidence arrangements compared to other similarly developed
nations. For instance, as of 2001, more than half of the older population aged 60
years and over (58%) in Japan lived with at least one of their children, which is three
to five times greater than that of other Western societies, such as the United States
(17%), Germany (15%), and Sweden (5%) (Cabinet Office, 2001). In this context,
Japanese coresidence patterns suggest that societal modernization and the socio-
cultural practice of coresidence do not necessarily change in a simple linear mode;
and traditional family culture can remain effective even in a highly modernized
social environment (Maeda & Ishikawa, 2000; Palmore & Maeda, 1985). The ways
in which the traditional culture survives in the lives of families in modern society,
however, are not fully understood.
This dissertation examines the demographical phenomenon of
intergenerational coresidence between older parents and adult children in
contemporary Japanese society. In particular, the study explores the elastic nature of
traditional cultural norms that are reflected in different styles of coresidence between
older parents and adult children. The study also attempts to delineate the diversity of
intergenerational living arrangements that vary by the extent to which traditional
3
cultural values and socioeconomic conditions are involved across individuals as well
as broader social contexts where aging families are nested.
In this chapter, first, the socio-demographic background of coresidence
patterns in Japan will be discussed: 1) the rapidly growing aging population that has
significantly transformed family structure; 2) the traditional cultural norm of filial
piety that has significantly influenced intergenerational living arrangements; and 3)
the relevant social change that has contributed to the decline of the traditional
coresidence arrangement. After the Japanese contexts are introduced, the rest of the
chapter will describe the current state of intergenerational coresidence in Japan and
present the study aims of the dissertation. A general overview of the dissertation will
be briefly provided at the end of the chapter.
1.1 Aging Population in Japan
Japan is one of the most rapidly aging societies in the world; thus, it provides
an ideal social environment for examining social issues salient to older populations.
At present, 20 percent of the Japanese population is over 65 years old, which is
currently one of the highest proportions of elderly in the world (Cabinet Office,
2006). Also, the older population in Japan is growing in unprecedented rate. While
it took 115 years and 47 years for the older populations in France and England to
double from 7 percent to 14 percent, it only took Japan 24 years (Cabinet Office,
2004).
The Japanese population is aging in part because of the high life expectancy.
At present, the life expectancy of men and women in Japan is 78.6 and 85.6 years old
4
respectively, which are the highest life expectancies in the world (Cabinet Office,
2006). At the same time, the continuously declining fertility rates also contribute to
the growing proportion of older people in the population. From 1950 to 2002,
fertility rates in Japan have plunged from 3.52 to 1.32 (National Institute of
Population and Social Security Research, 2004). The shrinking size of the younger
population creates a shift in the demographic composition of Japanese society. In
particular, the rapid aging of the population makes the elderly quite visible in Japan,
raising critical issues about the allocation of social and family resources for older
people.
The aging of the Japanese population also has changed fundamental family
structures; from one in which older parents were followed by a large succeeding
younger generation to one that looks like a “beanpole”, a vertically extended family
lineage where older parents share their life with a smaller number of children, yet for
a longer period of time (Bengtson, Rosenthal, & Burton, 1995). The change in
family structures has transformed the familial contexts for intergenerational living
arrangements as well. Although the time shared by generations has become much
longer, the availability of younger generations for aging parents has diminished.
This has created dissonance between a cultural expectation for a traditional
coresidence arrangement and actual desires and abilities of families to carry out their
cultural practice of coresidence with their parents (Kono, 2000).
5
1.2 Traditional Cultural Norm of Filial Piety
Coresidence in Japan entails a culturally unique component represented as
the norm of filial piety, which defines adult children’s responsibility to take care of
their aging parents in a shared household (Bumpass, 1994; Cherlin, 1994; Koyano,
1996). The existence of a cultural norm that is still quite influential for current
intergenerational family arrangements adds another value to the study of living
arrangements in the Japanese context. Coresidence between older parents and adult
children in Japan is often interpreted as the structural manifestation of the traditional
cultural norm of filial piety. The norm of filial piety used to be the moral principle
behind the traditional family ideology called “ie”, which means “house” or
“household” in Japanese (Kumagai, 1996). The “ie” ideology emphasizes the
importance of maintaining the integrity of the generational lineage - a lineage
vertically connected through one nuclear family from each generation (Kumagai,
1996). The concept of family referred to in this ideology also used to imply the
nation as a whole; and it was repeatedly imposed as national propaganda at
educational and other public-speech settings until the end of the World War II
(Goodman & Peng, 1996). Within this ideological framework, the Confucian norm
of filial piety provided specific principles to abide by; one of which was the eldest
son’s responsibility to take care of his parents in a coresidence household (Koyano,
1996; Sung, 1998).
Although the legal effect of the “ie” ideology was abolished in the post-war
reform, the implied cultural norm remained in Japanese society. In this respect, the
relatively high rate of coresidence in Japan is viewed as the resilience of the
6
traditional family culture that continuously defines the importance of family for
supporting older parents (Palmore & Maeda, 1985). The sustenance of this cultural
norm in intergenerational coresidence makes Japan an ideal place to study the
intricate linkage between traditional culture and family living arrangements in
modern society.
1.3 Social Change and the Decline of Traditional Coresidence in Japan
Past demographic studies on coresidence explain the declining trend of the
traditional coresidence arrangement as the gradual disappearance of traditional
family culture induced by the social changes that the postwar modernization has
facilitated in Japan (Kono, 2000; Ogawa & Retherford, 1993; Sodei, 1995). With
respect to coresidence in Japan, the relevant social structural changes include; 1)
increased mobility of younger generations from rural to urban areas; 2) enhanced
educational and career advancement for women; and 3) declining social support for
the cultural norm of filial piety.
As the economy grows rapidly, many employment opportunities have
become concentrated in big cities in Japan. As a result, for better work and
educational opportunities, increasing numbers of younger generations have moved to
big cities, leaving behind their older parents in their hometown (e.g., Sodei, 1995;
Thompson, 2003). The establishment of a life away from parents has resulted in a
geographical distance between older parents and adult children, which is often hard
to reconcile to make a coresidence arrangement.
7
The improvement of educational and work opportunities for Japanese women
also has altered the structural environment for coresidence. With the proliferation of
egalitarian gender norms, more women in the postwar era of Japan pursued their own
education and career advancement by delaying their marriage and having fewer
children (Ochiai, 1997). In the past five decades, the proportion of Japanese women
who advanced to college or university education has increased from 4.9 percent to
49.8 percent (Ministry of Education; Culture; Sports; Science; and Technology,
2006). Also, across the past few decades, the age of first marriage of Japanese
women has increased from 24 years to 27 years old (Raymo, 2003). The change and
shift in the life-course of women in Japan has created dissonance with their
traditionally expected familial role, including their presumed caregiving
responsibilities for coresiding parents-in-law (Eto, 2001; Jenike, 1997; Ogawa &
Ermisch, 1996).
Furthermore, scholars argue that the modernization of social structures in
Japan has created negative social attitudes toward the traditional family norm. In
fact, parallel to the decrease of actual coresidence rate, longitudinal analyses indicate
that social support for the traditional family norm is in decline. For instance, the
repeated cross-sectional study by Ogawa and Retherford (1993) shows that the
percentage of middle-aged women in Japan (typically the primary caregiver for their
parents-in-law in a coresidence household) who considered the filial responsibility to
take care of their aging parents as a “good custom” dropped from 80 percent to less
than 50 percent in a few decades. Negative attitudes toward the traditional cultural
8
norm are interpreted as another main factor that has contributed to the disappearance
of traditional multigenerational household.
1.4 Current State of Intergenerational Coresidence in Japan
Traditionally, older parents in Japan coreside with their eldest son’s family in
preparation for their expected need for support (Hashimoto, 1996). However, recent
statistics show that some portion of coresidence arrangements in current Japanese
society contain a non-traditional style that may be significantly different from the
traditional three-generation household. For instance, the proportion of two-
generation households, where older parents and adult children live together without
the grandchild generation, recently started to increase (Cabinet Office, 2005). As a
result, despite the precipitous decrease of the traditional three-generation coresidence,
the overall coresidence rate among older adults is still about 50 percent (Figure 1).
9
0
10
20
30
40
50
60
70
80
1980 1985 1990 1995 1997 1998 1999 2000 2001 2002
Year
(% )
3-generation Household Coresidence with child
Figure 1. Coresidence Rates of Older Individuals Over 65 Years-Old in Japan
Source: Cabinet Office (2005)
Some potentially new trends of coresidence arrangements in Japan may also
exist in terms of which adult child lives with older parents and who benefits from
this type of living arrangement. For instance, the recent increase in adult children in
Japan who either delay their marriage or remain single (Raymo, 2003) suggests that
some non-traditional two-generation coresidence may involve single adult children,
rather than the married sons who are traditionally expected to be the coresiding
children.
Variation in coresidence households can also derive from the power
dynamics within the family (Mutchler & Burr, 2003; Speare & Avery, 1993).
Although most studies tend to assume that coresidence is primarily arranged for the
10
needs of older parents (Aquilino, 1990; Logan et al., 1998), educational and
economic changes in Japan such as the recent economic downturn and prolonged
educational training for younger generations indicate that there may be a substantial
benefit for adult children from coresiding with their parents as well (Raymo, 2003;
Ward, Logan, & Spitze, 1992).
Another trend of Japanese coresidence is the changing social functions
implied in this type of living arrangement. The change in intergenerational
coresidence in Japan is often described as shifting from a normatively determined,
life-long arrangement (Hashimoto, 1996) to a contingent arrangement where the
coresidence household is negotiated by families in consideration of their needs and
resources at a particular point in time (Izuhara, 2000; Martin & Tsuya, 1994). This
shift in the social functions of coresidence for parents and children suggests that the
primary function of non-traditional types of coresidence in Japan may be to
accommodate the needs of older parents or adult children, rather than to fulfill the
perceived cultural expectation.
1.5 Diversity in Coresidence Arrangements
What seems absent in past research on intergenerational coresidence in Japan
is the lack of focus on the elastic nature of traditional cultural norms and their
divergent reflections in coresidence arrangements. Past studies mostly discuss the
traditional culture in dichotomous terms, whether it is present/strong or absent/weak;
and they have not focused on the dynamic aspects of culture, in which the cultural
norm is elastically adapted as personal values and preferences change in response to
11
modern society. The adaptive feature of the traditional culture has been discussed by
some qualitative studies (Izuhara, 2000; O'Leary, 1993; Traphagan, 2003), but very
few studies have delineated the divergent reflections of traditional culture at the
national level. Also, so much discussion has been devoted to the declining three-
generation household that little attention has been given to other types of coresidence
patterns that do not necessarily mirror the traditionally expected arrangement but still
involve multiple generations sharing a household.
Therefore, this study examines multiple types of coresidence arrangements as
diversified structural manifestations of traditional cultural norms in Japanese society,
using nationally representative data of Japanese older adults. Among different types
of coresidence arrangements, some may remain “traditional” in a sense that the
living arrangement is primarily guided by traditional norms, but other types may be
considered “non-traditional” because of their relatively free stance from the
traditional family values.
The source of diversity in Japanese coresidence arrangements potentially
derives from multiple levels of social contexts, ranging from individual-level
conditions, familial contexts, and larger community contexts in which the coresiding
families are embedded. In the following sections, the potential determinants of
coresidence at each level are briefly discussed based on previous studies of
intergenerational living arrangements mainly in East Asian societies.
12
1.6 Diversity in Coresidence at the Individual-Level
Individual-level determinants of coresidence are represented by
characteristics of older parents and adult children, such as their normative attitudes,
demographic backgrounds, family relationships, and socioeconomic needs and
resources. Normative attitudes of older parents and adult children, the degree of
traditionalism in family values, tend to be associated with the likelihood of
coresidence. Generally, parents and children who have more traditional views of the
family norm are more likely to be found in a coresidence household (Logan & Bian,
1999; Logan et al., 1998; Rindfuss, Liao, & Tsuya, 1992). The level of educational
attainment also serves as a proxy for normative attitudes. Because of their exposure
to more individualistic ideas, parents and children with higher levels of education
tend to express a negative view of the traditional cultural norm; thus, they are less
likely to live in a coresidence household (Logan & Bian, 1999; Rindfuss, Liao, &
Tsuya, 1994).
Regarding the relationship between the parental normative attitudes and
coresidence, a study in China (Logan & Bian, 1999) suggests that older parents’
actual coresidence experience also shapes their normative attitudes toward
coresidence. In fact, this Chinese study showed that living with a child tends to
reinforce older parent’s negative attitudes toward coresidence. The examination of
such a reciprocal relationship in Japan will certainly extend our knowledge of the
sustenance of cultural norm in another modern society.
Other than normative attitudes, demographic factors such as older parents’
advanced age generally increases the likelihood of coresidence (Cabinet Office,
13
2004). While the gender of older parents is not a consistent determinant of
coresidence, there is a persistent culture of son preference in Japan (as well as in
other East Asian nations) as the coresiding child (Kumagai, 1996). Rural residency
of older parents and adult children also generally increases the likelihood of
traditional coresidence, presumably because of the immediate local culture that tends
to embrace the traditional family practice (Raymo & Kaneda, 2003; Tsuya & Martin,
1992).
The perceived quality of family relationships also affects the likelihood of
coresidence. As suggested by Koyano (1999; 2003), because of the recently
weakened power of the cultural norm of coresidence, affectional ties between
generations may play a greater role in making an intergenerational living
arrangement. In this context, parents and children who are relatively satisfied with
their family relationships may be more likely to coreside than those less satisfied. At
the same time, as some studies implicate (Koyano, 2003; Logan & Bian, 1999), the
everyday interactions in a coresidence household could strain family relationships
because of the extensive amount of time and responsibilities shared across
generations in the same household. The strained relationships in a coresidence
household are particularly relevant between parents and children-in-law, which may
subsequently create negative attitudes toward the cultural norms implied in
intergenerational coresidence (Koyano, 2003; Logan & Bian, 1999).
Socioeconomic needs and resources also play a critical role in the residential
decisions of older parents and adult children, particularly for the need-driven non-
traditional types of coresidence. One of the major needs of older parents is indicated
14
by their health conditions; relative to older parents in good health, those in poor
health are more likely to coreside with their children (Brown, 2003; Logan & Bian,
1999; Raymo & Kaneda, 2003). Also, the likelihood of coresidence may be
determined by the level of household income; for instance, low-income families
generally have a greater need to pool their household incomes to obtain better
economic outcomes; thus their likelihood of coresidence tends to be higher (Bawin-
Legros & Stassen, 2002). Housing resources play another critical role in arranging
coresidence. Older parents and adult children who own a home are more likely to
coreside than those in a rented home, presumably because of the relative space
limitations and poor housing quality of rented residences (Logan et al., 1998;
Rindfuss et al., 1992).
1.7 Diversity in Coresidence at the Family-Level
At the family-level, needs and resources for coresidence are dealt with by
multiple family members across generations. The conventional “common-
preference” approach to coresidence posits that the intergenerational living
arrangement is made out of the family’s unitary purpose shared by all members of
the family (Becker, 1991). However, the recent generational shift toward
increasingly negative attitudes toward the traditional coresidence arrangement
suggests that the desires and motivations for coresidence are not necessary agreed
upon by family members across generations (Cutler, Cho, Yust, & Franklin, 1999).
With the diminished effect of the cultural norm on actual family living arrangements,
the current pattern of intergenerational coresidence in Japan is more likely to be
15
based on the negotiations for particular needs and resources of the family (Izuhara,
2004; Koyano, 1999; Kumagai, 2005). In this respect, investigating the overall
picture of the ways in which needs and resources for coresidence are located and
recognized within the family contexts has a particular value.
One factor that represents the structure of the family context is the family
composition, which is defined as the relative size of the family and the degree of kin
availability. In general, older parents with a relatively large number of adult children
are more likely to live with one of the children. On the other hand, from the adult
child’s perspective, because it is rare for parents to simultaneously live with multiple
married children and their families, a relatively large number of siblings in the
family tends to decrease the likelihood of coresidence for each individual adult child
(Rindfuss & Raley, 1998). For older parents, the availability of a living spouse
serves as another critical determinant of coresidence. In general, families with a
widowed older parent are more likely to make a coresidence arrangement (Martin &
Tsuya, 1994; Raymo & Kaneda, 2003).
Within the family context, individual motivations and expected obligations
for coresidence vary among adult children. While the traditional coresidence
expectation is for the eldest son to live with his parents, non-traditional types of
coresidence may entail more flexibility in deciding which child lives with the parent
(Izuhara, 2004). For instance, if there is an adult child who remains single in the
family, the likelihood of coresidence for married siblings tends to decrease (Logan &
Bian, 1999; Rindfuss & Raley, 1998; Ward et al., 1992). This is presumably due to
the fact that single children tend to have a greater financial need; also having less
16
competing family responsibilities than married children, single children may have
more flexibility in making themselves available for older parents. In addition, the
presence of an adult child within the family who lives near the older parent’s
residence may decrease the chance of coresidence for other adult children. This is
based on the current rise of an alternative to coresidence called “intimacy-at-a-
distance,” where parents and children maintain separate households but live close to
each other (Martin & Tsuya, 1994).
Another potential factor considered among siblings is their socioeconomic
status represented by their educational level and occupational background. Since
coresidence tends to be more common for parents and children with lower levels of
education (Raymo & Kaneda, 2003), adult children with a relatively low levels of
educational training may be more likely to live with their parents than their sibling
counterparts. Similarly, the low-ranked occupational background of adult children
may increase their chance of coresidence compared to their siblings with higher-
ranked occupations. Also, self-employment tends to increase the likelihood of
coresidence (Raymo, Liang, Sugisawa, Kobayashi, & Sugihara, 2004). Presumably,
because self-employed people are generally involved in a family business that entails
relative flexibility in job location as well as instability in income, their likelihood of
coresidence tends to be higher than for those who are employed by someone else.
Thus, among siblings in the family, self-employed adult children may have a higher
likelihood of living with their parents than employed adult children.
Negotiations for coresidence can also take place between adult children and
their spouses. Past studies have shown that with a coresidence arrangement,
17
intergenerational conflicts often take places between older parents and children-in-
law, suggesting that compromise and reconciliation are expected from the spouse of
adult children in a coresidence household (Cutler et al., 1999; Koyano, 2003). In this
respect, the residential decision for coresidence requires negotiations between adult
children and their spouses; and presumably, who maintains the overall power for
household arrangements within the couple would affect their likelihood of
coresidence. One of the potential indicators of the power dynamics within the
couple is the educational difference between husbands and wives (Elder & Rudolph,
2003). In terms of coresidence arrangements, adult children-in-law who have a
lower educational background may yield to their spouse, leading to their acceptance
of a coresidence arrangement.
1.8 Diversity in Coresidence at the Contextual-Level
The theoretical approach of the political economy of aging posits that the
lives of older individuals and adult children are inevitably embedded in, and shaped
by, the broader social contexts of socioeconomic conditions and cultural ideology
(Daatland & Herlofson, 2003; Estes, 2003; Glaser & Tomassini, 2000; Rindfuss &
Raley, 1998; Walker, 1996). Limited numbers of studies, however, have discussed
the relationship between individual living arrangements and nesting social contexts
(Kono, 2000; Rindfuss & Raley, 1998); and very few have empirically examined
intergenerational living arrangements from a multi-level perspective (Burr &
Mutchler, 2003; Mutchler & Burr, 2003). The potentially relevant social contexts for
coresidence in Japan include the traditional cultural milieu, economic proliferation,
18
housing conditions, and the availability of welfare resources for older populations in
the community.
Within Japan, there is a significant regional variability in the rates of
coresidence among older people. At the aggregate regional level, the rate tends to be
higher in the northeastern region and relatively low in the northern island (Hokkaid)
and the southwestern region. This reflects the indigenous regional culture of
intergenerational family arrangements; while the northeastern regions maintain a
culture that encourages intergenerational coresidence for aging parents, the other
regions with low coresidence rates tend to have a culture that promotes independent
living arrangements for older parents and adult children (Shimizu, 2004). This
cultural milieu generally exists independently from the rural/urban contrast of the
area. In other words, the regional coresidence culture pervades across both rural and
urban areas within the larger unit of region, suggesting its unique cultural stance
independent of the socioeconomic conditions that differentiate the local rural
contexts from urban contexts (Raymo & Kaneda, 2003).
Community-level economic proliferation is another potential contextual
factor that shapes coresidence. In general, the lack of economic opportunities tends
to make it difficult for younger people to find work opportunities and remain in the
area, adding challenges for them to stay and form a coresidence household with their
parents (Thompson, 2003). Alternatively, it is also possible to hypothesize that
parents and children in an area with few economic opportunities may need to live
together to supplement their lack of financial resources. Household management is
inseparable from the local economy because of its dependency on financial means.
19
Therefore, the maintenance of the traditional living arrangement of coresidence also
hinges on the local economic situation for the residing families.
Housing conditions also influence intergenerational coresidence. Because the
household is shared by multiple generations, a coresidence household generally
requires a larger living space. Therefore, the overall shortage of large living spaces
in an area potentially lowers the individual likelihood of intergenerational
coresidence arrangements (Sodei, 1995). Housing conditions also can be influenced
by the affordability of housing in the area. If parents and children are to live in the
area with relatively high housing costs, on average, it may necessitate their living
together to share housing costs (Mutchler & Burr, 2003).
Finally, the availability of local welfare resources for older residents in the
area may influence older parents’ residential decisions for coresidence. Over the last
few decades, the Japanese government has instituted extensive political initiatives to
increase the level of welfare resources for older individuals in the community. These
political initiatives were implemented to accommodate increasingly negative
attitudes against the traditional family norm of filial responsibility and to create a
better balance between families and the welfare sector in fulfilling caregiving
responsibilities for older individuals (Eto, 2001; Maeda & Ishikawa, 2000; Peng,
2003). One of the major political shifts accompanied by this governmental effort is
to give local communities more authority and responsibility for welfare service
management, which subsequently has increased regional variability in welfare
resource availability for older people (Masuzoe, 2000). Because of its implied
intention to take over some of the caregiving responsibilities from families, the
20
relatively high availability of welfare resources in the community may assist adult
children who would otherwise have to coreside with their needy older parents to
maintain a separate household arrangement from their parents.
1.9 Study Purpose and Research Questions
This dissertation consists of three individual studies, all of which explore
distinctive aspects of traditional cultural norms and intergenerational coresidence
between older parents and adult children in Japan. These studies address questions
about the maintenance of traditional cultural norms and their divergent reflections in
current coresidence arrangements. The overall goal of the three studies is to examine
the value-laden nature of current coresidence arrangements in Japan and to
investigate the potential sources of diversity that differentiate traditional types of
coresidence from non-traditional types. The contrasts among multiple types of
coresidence throughout this dissertation are intended to explore the new forms in
which the traditional intergenerational family culture has evolved and their meanings
for aging societies at large. Specifically, the three studies in this dissertation will
investigate; 1) the nature of the traditional family norm that shapes, and is shaped by,
coresidence experiences of older individuals; 2) family contexts where multiple
needs and resources are mobilized for coresidence decisions; and 3) community-
level social contexts that create individual and regional variability in coresidence
arrangements.
The first study investigates traditional cultural norms in current Japanese
coresidence arrangements. The study specifically examines the reciprocal
21
relationship between older parents’ normative beliefs and actual coresidence,
addressing the question of how the cultural norm influences, and is influenced by,
actual coresidence experiences. This investigation also explores divergent styles of
coresidence in contemporary Japan by examining the difference between the
traditional style of coresidence with married children and the non-traditional type of
coresidence with unmarried children.
The second study focuses on family contexts as critical determinants of
coresidence. This part of the dissertation considers coresidence as a product of
negotiations among multiple family members across generations and investigates
power dynamics within the family that differentiate two types of coresidence: parent-
headed coresidence and child-headed coresidence. The family contexts examined in
this study consist of multiple family lineages; vertical lineage between older parents
and adult children; horizontal lineage among adult children as siblings; as well as
another horizontal link between adult children and their spouses. The differentiation
of coresidence by its household headship is intended to explore the variation in
potential benefits and costs of coresidence experienced by both older parents and
adult children.
Finally, the third study simultaneously examines both individual-level and
community-level contextual factors that enable and/or restrict two types of
intergenerational coresidence arrangements between older parents and adult children:
“life-long” coresidence and “boomerang” coresidence. The two styles of
coresidence are differentiated by their timing; while the traditional style of “life-
long” coresidence is defined as an arrangement where older parents and adult
22
children have lived together continuously since the birth of the child, the non-
traditional type of “boomerang” coresidence is characterized as a household where
older parents and adult children have reunited after some time apart. The two types
of coresidence are embedded in the community-level social contexts that vary by
cultural climate, economic well-being, housing conditions, and welfare resource
availability for older residents.
The set of three research studies as a whole investigate intergenerational
coresidence arrangements as structural reflections of traditional cultural norms in
Japan – a society that has experienced an extensive process of modernization in
recent decades. As contended by the modified modernization theory (Inglehart &
Baker, 2000), this study considers traditional family culture as an elastic entity that
remains effective in modern societies in divergent styles; some mirroring the
traditional norm and others involving partial or complete departure from tradition.
By examining multiple types of coresidence, this dissertation investigates ways in
which the traditional intergenerational living arrangement becomes modified or
transformed in order to accommodate individual needs and desires of aging parents
and their families in a modern social environment. For this purpose, the research in
this dissertation explores the source of diversity in coresidence arrangements across
multi-level social dimensions that entail not only individual conditions but also
broader familial and social contexts.
As suggested by the longitudinal trends in societal attitudes and actual rates
of coresidence, the culture of coresidence in Japan is not a stable and ever-intact
phenomenon; but rather a phenomenon that fluctuates and variegates as society
23
changes. With significant social transformations expected in every level of aging
societies, a better understanding and more accurate projection of diverging
trajectories for the traditional family culture would make a significant contribution to
the establishment of social programs that are both sensitive and responsive to the
diverse needs and cultural expectations of families in society.
1.10 Overview of this Dissertation
As explained above, this dissertation presents three studies that explore
multiple patterns of intergenerational coresidence between older parents and adult
children in contemporary Japan, addressing research questions about the ways in
which family tradition either remains intact or becomes modified to meet the needs
of parents and children. In the next chapter, I will present the methods and analytical
schemes employed in these three studies on coresidence. I will also describe the two
datasets used for each study.
The following three chapters discuss each of the individual studies on
coresidence respectively. The third chapter presents the study that examined the
relationships between the traditional cultural norm embraced by older parents and
their actual coresidence patterns. The fourth chapter describes the study that focused
on the needs and resources of parents and children within their family contexts. The
fifth chapter discusses the study that examined the community contexts within which
coresiding older parents and adult children are embedded, as enabling or disabling
factors of coresidence arrangements. The last chapter will discuss the overall
findings of the three individual studies and present the conclusions drawn from these
24
studies. The chapter concludes this dissertation with implications for future research
as well as for families in modern aging societies at large.
25
CHAPTER II Methods
This dissertation investigates the traditional cultural norm expressed by older
parents and its relationships with several types of coresidence arrangements that are
embedded in multiple levels of social contexts. Specifically, this thesis consists of
three research studies, all of which deal with the outcomes of the likelihood of
traditional and non-traditional styles of coresidence. The first study also investigates
the factors that shape older parents’ normative beliefs in coresidence and its
reciprocal relationship with the actual experience of coresidence. Thus, the research
model of the first study includes an additional outcome of parental normative belief.
Table 1 presents the outcomes of interest that were examined in each study.
Table 1.Outcome of Interest
Study 1 Study 2 Study 3
Study Focus
Relationships
between norm
and coresidence
Family
context
Regional
context
Traditional
Style
Coresidence with
married child
Parent-headed
coresidence
“Life-long”
coresidence
Outcome 1
(Likelihood of
Coresidence)
Non-
Traditional
Style
Coresidence with
unmarried child
Child-headed
coresidence
“Boomerang”
coresidence
Outcome 2
Normative belief
in coresidence
N/A N/A
Two sets of nationally representative data are analyzed in this dissertation.
The first study is based on analyses of the Japanese General Social Survey (JGSS) in
2001 (Osaka University of Commerce, 2003). The JGSS contains a question that
26
specifically asks respondents about their view toward the desirability of coresidence
between older parents and adult children. While the original dataset contains
respondents aged 18 years and older, this study focuses on the sub-group of older
respondents aged 60 years and older.
The second and third studies of multi-level analyses of coresidence are based
on the responses in another national dataset, the Nihon University Japanese
Longitudinal Study of Aging (NUJLSOA) (USC/UCLA Center on Biodemography
and Population Health, 2004). The NUJLSOA is particularly suitable for these
studies because of its rich information about older parents and their families,
including their adult children and children-in-law. The dataset also contains
information about the geographical locations where older respondents reside,
providing ideal data for multi-level analyses. The NUJLSOA exclusively focuses on
the older population in Japan, those aged 65 years and older. The study aims to
investigate the health and social conditions of older people in Japan, and includes
detailed questions about older respondents’ adult children and their spouses, such as
their demographic characteristics and educational and occupational backgrounds.
The following sections discuss each dataset and study samples in more detail. Each
section below also describes the conceptual models developed for the three
individual studies respectively.
2.1 Study 1: Traditional Cultural Norms and Intergenerational Coresidence
This study aims to explicate the reciprocal relationships between older
parents’ normative attitudes toward intergenerational coresidence and actual patterns
27
of living arrangement in relation to the impact of parental needs, resources, and
demographic characteristics. This first study also presents a contrast between
traditional and non-traditional styles of coresidence by differentiating the marital
status of the coresiding children. To examine the non-recursive model in which the
residual variances of the two outcomes (i.e., filial norm and actual coresidence
pattern) are assumed to be correlated with each other, the study employs a two-stage
least square model. The model is equipped with a system of simultaneous equations,
which allows the specification of reciprocal relationships between the two exogenous
predictors (Jöreskog, 2005). Figure 2 depicts the research model of this study.
Figure 2. Conceptual Framework of Study 1
The study is based on the analysis of data from the Japanese General Social
Survey (JGSS) in 2001. The JGSS is a Japanese version of the General Social
Coresidence with
married/unmarried
child
Normative Belief
in Coresidence
Common Predictors: Older
parent’s needs, resources,
and demographic
characteristics
Instrumental variable 1:
unique predictor of
coresidence
Instrumental variable 2:
unique predictor of
normative belief
28
Surveys in the United States, and it has been conducted on a yearly basis as a
repeated cross-sectional study since 1999. The surveys mainly focus on attitudes and
values regarding a variety of social issues, such as gender norms, politics, and family
relationships. The JGSS contains questions that are comparable with the GSS in the
U.S., as well as questions formulated for the issues unique to Japanese society.
The original sample for the JGSS consists of Japanese residents aged 18
years and older. The study in this dissertation uses a sub-group of 909 respondents
whose age is 60 years and older and who have at least one living adult child. Out of
these respondents, 54 older parents (6% of the selected sample) have at least one
missing value in their responses; and they are omitted from the sample, yielding a
final sample size of 855 older parents.
2.2 Study 2: Intergenerational Coresidence and Family Contexts
The second study investigates the ways in which needs and resources are
negotiated for coresidence among older parents, adult children, and children-in-law
(Figure 3). The study also contrasts two different types of coresidence, one headed
by older parents and the other headed by adult children, in relation to the sorts of
needs and resources mobilized for each type of coresidence. The implemented
method for this analysis is multinomial logistic regression with robust standard errors.
The multinomial outcome variable includes two different types of coresidence and a
reference group of adult children who do not live with their parents (i.e., non-
coresidence group). Because of the repeated information of older parents among
29
adult children within the same family, robust standard errors are estimated to adjust
for the clustering effects of the family units.
Figure 3. Conceptual Framework of Study 2
The second study analyzes data from the Nihon University Japanese
Longitudinal Study of Aging (NUJLSOA). This survey originally collected all
information from older parents, including data on their adult children and children-
in-law. In this study, the original data is “reshaped” so that the unit of analysis is
transformed from older parents to individual married adult children reported by
parents. In the reshaped dataset, information of older parents is repeated for each
adult child in the same family.
This study focuses on 3853 self-responded older parents who have at least
one married child. Originally, this sub-group of the NUJLSOA sample had a
significant proportion of older individuals with missing data on relevant
Status difference
with spouse
Adult Child’s
Characteristics
Status difference
with spouse
Adult Child’s
Characteristics
Child-Headed
/Parent-Headed
Coresidence
Older parent’s
Characteristics
30
measurements (about 30 percent of the sample). To retain the original sample with
minimal selection bias, missing values were estimated with the method of multiple
imputation. This imputation method replaced missing data with a specified number
of values (five times in this study) that were estimated in their predictive distribution
(Schafer & Olsen, 1998). In this study, multiple imputation produced five complete
sets of data, all of which were combined for final estimates.
The comparison between the original sample (with missing cases excluded)
and the imputed sample shows that the characteristics of the two samples were
almost identical, except that the imputed sample has a slightly lower level of
education on average. The final analyses are presented with the imputed sample
only, since both groups of samples produce almost the same estimates. The imputed
sample consists of 3858 older parents and 8601 adult children in total.
2.3 Study 3: Intergenerational Coresidence and Social Contexts
The third study examines individual-level conditions and prefecture-level
contextual factors that affect the likelihood of intergenerational coresidence
arrangements for older parents. Another two types of coresidence are contrasted;
one represents the traditional type of “life-long” coresidence and the other
characterizes the non-traditional type of “boomerang” coresidence. Older
individuals’ likelihood for each type of coresidence is analyzed in relation to the
regional-level contextual factors that represent economic well-being, housing space,
cultural climate, and welfare resource utilization. In this study, the regional factors
are measured by the unit of prefecture, which is the sub-national district that divides
31
the country into 47 units. The study tests models in which the likelihood of
coresidence is determined by older parents’ individual characteristics and prefecture-
level contextual factors. Furthermore, the analysis tests a cross-level interaction, in
which the contextual factors modify or alter the extent to which older individuals’
normative attitudes affect their likelihood of coresidence. The overall conceptual
model is summarized in Figure 4.
Figure 4. Conceptual Framework of Study 3
The third study employs a multinomial logistic regression within a two-level
hierarchical structure. In this approach, two types of coresidence (i.e., traditional
“life-long” coresidence and non-traditional “boomerang” coresidence) are contrasted
against a reference group of non-coresidence within a two-level structure, which
locates individuals at the first level and prefecture contexts at the second level (Luke,
2004; Raudenbush & Bryk, 2002).
Social Contexts:
Prefecture
Economic well-being;
Housing space;
Cultural climate;
Welfare resource
Older individual’s
Characteristics
Life-long /
Boomerang
Coresidence
32
The third study also analyzes the initial wave of Nihon University Japanese
Longitudinal Study of Aging (NUJLSOA) collected in 1999. Unlike the second
study, the analysis in this study employs the original unit of analysis, which is based
on the responses of older parents. The data of the four prefecture-level contextual
factors are retrieved from datasets collected by other governmental and research
institutes; and they are integrated with the NUJLSOA data for analyses.
The sample for the third study consists of older individuals who have at least
one living child. Out of 4085 respondents who meet the selection criteria, 520
respondents (12.7% of the selected sub-sample) are excluded, which makes the
effective sample size of 3565 older parents. A comparison between the original
sample and the study sample suggests that the study sample is almost identical to the
original nationally representative sample, except that they are slightly healthier than
the original sample. This is mainly because of the inevitable exclusion of
respondents who have a proxy person answer their survey due to significant health
conditions. Using the two-stage estimator of sample specification error (Heckman,
1979), the preliminary analysis calculates the overall probability for the original
sample to be excluded from the study sample; and tests whether the calculated
probability has a significant impact on the study outcomes. The result suggests that
the sample exclusion did not alter the estimated likelihood of coresidence in
significant ways.
The following three chapters present individual studies based on the analyses
of the data described in this chapter. The detailed sample characteristics will be also
presented in the respective chapters.
33
CHAPTER III
Traditional Cultural Norms and Coresidence
Japan is one of the most rapidly aging societies in the world, with a rate of
aging that is almost three times that of most other developed countries (Maeda &
Ishikawa, 2000). The proportion of those aged 65 and older in the Japanese
population is currently about 19 percent, among the world’s highest (Cabinet Office,
2004). Japanese society is also unique in its cultural orientation toward older adults.
Cultural norms promote familial integration of older people, which is most obviously
seen in the relatively high rate of intergenerational coresidence between elderly
parents and their adult children (Kumagai, 1996; Palmore, 1975; Wada, 1995). In
2002, about 30 percent of those 65 years and older in Japan lived with their children
(Cabinet Office, 2004), while equivalent figures are less than five percent of older
people in the United States and Germany (Maeda & Ishikawa, 2000). Traditionally,
intergenerational coresidence is defined as a living arrangement where older parents
and their married children share a house together, which often leads to the
establishment of three-generation households.
Despite the relatively high proportion of three-generation households, the
prevalence of such traditional living arrangements has precipitously declined
recently. In fact, in just two decades, the percentage in such households dropped
from 50 to 24 percent (Cabinet Office, 2004). If intergenerational coresidence is a
structural manifestation of a cultural norm that prioritizes family over individual
needs (Burr & Mutchler, 1999; Hamon & Blieszner, 1990; John, Resendiz, & De
Vargas, 1997; Kurosu, 1994; Maeda & Ishikawa, 2000; Wallace & Facio, 1992),
34
then the historical decline in coresidence may indicate this norm has fundamentally
weakened or has altered its form of expression in contemporary Japan. In this
investigation, we are concerned with the mutual relationship between traditional
preferences for intergenerational coresidence and several types of residential
circumstances among older people in Japan, and test a reciprocal model of influence
between expressed filial norms concerning coresidence and actual living
arrangements.
3.1 Intergenerational Coresidence and Filial Norms in Japan
Intergenerational coresidence in Japan is legitimated and reinforced by a
family ideology known as ie, “a vertically composite form of nuclear families, one
from each generation” (Kumagai, 1996, p.7). In principle, the ie family form
represents a three generational stem family where elderly parents live with their
eldest son and his family (Kumagai, 1996; Long, 1991; Ochiai, 1997). The
Confucian norm of filial piety, the moral aspect of the ie system, emphasizes
children’s unconditional obedience to their parents and parents-in-law as
reciprocation for the sacrifices that parents made in raising them (Koyano, 1996).
Although intergenerational coresidence is generally interpreted as the
behavioral expression of the belief in traditional cultural norms that emphasize
adherence to filial duty and responsibility (Kurosu, 1994), several factors argue
against interpreting intergenerational coresidence simply as a reflection of cultural
beliefs. First, cultural norms and values are dynamic, particularly in a rapidly
changing social environment such as Japan. Thus, it would be a mistake to assume
35
that Japanese norms of filial piety are static, and, indeed, evidence supports the
premise that they have become less relevant as a value orientation (Ogawa &
Retherford, 1993). Second, if intergenerational coresidence in Japan is driven
increasingly by instrumental concerns of either younger or older generations, then
normative beliefs, regardless of their absolute strength, may have become a relatively
weaker rationale for coresidence. Third, it is possible to argue that multigenerational
coresidence, when experienced due to necessity or convenience, shapes attitudes
toward this type of living arrangement, positively or negatively, as an ideal family
form. We review these three points in more detail below.
In spite of Japan’s unique status as a modernized nation with strong filial
tendencies, it has witnessed a steady weakening in the degree of responsibility that
family members take for supporting and housing their elderly parents. Scholars in
Japan generally consider the decline in social compliance with filial duty to be the
product of rapid economic development facilitated by post-war modernization, and is
evidenced by strengthening preferences for intergenerational autonomy (Sodei,
1995). Ogawa and Retherford (1993), for instance, found a dramatic decline over 30
years in the strength with which Japanese women expressed their obligation to older
parents, and speculated that cultural homogeneity in Japanese society accelerated the
impact of modernization on norms for eldercare responsibility so that once structural
and economic changes took place, cultural changes followed relatively quickly.
Motivations for multigenerational living arrangements have become
increasingly varied and less tied to traditional cultural ideologies. Koyano (1996;
2003) in explaining his finding of weaken normative basis of family functioning in
36
Japanese society, claimed that Japanese families are now transitioning to more
instrumentally oriented relationships (e.g., elderly parent’s needs for assistance).
Several studies have found that strategic factors are becoming more common reasons
for forming multigenerational households in Japan (e.g. Izuhara, 2002; Morgan &
Hirosima, 1983; also reviewed in O'Leary, 1993). This also is exemplified in the
emergence of a new type of arrangement where unmarried adult children live with
their parents long after they graduate from school, arguably called “parasite
singles”(e.g., Lunsing, 2003; Raymo, 2003; Yamada, 1999).
Factors other than norms have been found to influence intergenerational
coresidence in East Asian populations. These can be divided into domains of need,
resource availability, and predisposing factors. Among the strongest predictors of
coresidence is the elderly parent’s health status, with poorer health increasing the
likelihood of coresidence (Brown et al., 2002; Logan & Bian, 1999; Logan et al.,
1998; Raymo & Kaneda, 2003). Widowhood, another indicator of need, is also a
strong predictor of living with an adult child (Martin & Tsuya, 1994; Raymo &
Kaneda, 2003).
In general, parents with advanced age are more likely to live with their
children (Cabinet Office, 2004). However, after controlling relevant factors, studies
have not found consistent results regarding the relationship between parental age and
the likelihood of living with an adult child. Some studies identified a positive
relationship between the two (e.g., Brown et al., 2002; Ogawa & Retherford, 1997),
but others have found them to be negatively related (e.g., Logan & Bian, 1999) or
found no relationship (Ogawa & Retherford, 1997; Raymo & Kaneda, 2003; Tsuya
37
& Martin, 1992). In general, however, older people express stronger preferences for
coresidence than younger people (Cutler et al., 1999).
Elderly parents’ socioeconomic status has been found to enhance the
likelihood of intergenerational coresidence. For instance, parents with greater
housing resources tend to live with their children compared to parents with fewer
such resources, presumably because they are better able to provide such
arrangements to their children (Logan & Bian, 1999; Raymo & Kaneda, 2003).
However, having more education tends to weaken preferences for coresidence, as
educational attainment may expose one to alternative social values (Logan & Bian,
1999; Rindfuss et al., 1994).
As a result of a patrilineal kinship system in Japan, gender is an important
predisposing factor for intergenerational coresidence. The most traditional
intergenerational living arrangement is one in which parents live with their eldest son
and his family (Kumagai, 1996). In terms of parents’ gender, elderly men are more
likely than elderly women to live with their adult child (Tsuya & Martin, 1992).
Other predisposing factors include urban-rural residence and the nature of the
relationship parents maintain with their children. Rates of coresidence tend to be
greatest and normative beliefs in coresidence tend to be stronger in rural areas of
Japan where traditional practices are more established (Raymo & Kaneda, 2003;
Tsuya & Martin, 1992).
The quality of family relationships has the potential to influence coresidence
values and decisions. Children with whom relationships are harmonious may be
viewed as more attractive residential partners, as it has been suggested that affiliation
38
in Japanese families is increasingly based on affective than obligatory ties (Koyano,
1996, 2003). On the other hand, coresidence may compromise the quality of family
relations, lead to intergenerational conflict or call into question the desirability of
such an arrangement, as has been speculated in China (Logan & Bian, 1999) and the
U.S. (Aquilino & Supple, 1991; Coward, Albrecht, & Shapiro, 1996).
Several family studies have formally examined the reciprocal dynamic
between normative ideals about intergenerational relations and actual family
experience. Research on Chinese families by Logan and Bian (1999) found that
preferences for coresidence positively influenced whether older parents lived with
their adult children, but living with adult children negatively influenced older
parents’ preferences for coresidence. Employing a similar reciprocal model, a study
of Israeli families by Spilerman and Elmelech (2003) revealed that normative
expectations concerning the transfer of wealth to younger generations influenced
whether intergenerational transfers were made, but intergenerational transfers did not
influence normative expectations. Both studies support a model where value
orientations motivate behavior toward family members that is consistent with those
values, but not necessarily the reverse; indeed, the results in China suggest that
coresidence with children actually weakens support for this type of living
arrangement.
3.2 Research Questions
The goal of this investigation is to examine the degree to which norms of
intergenerational coresidence guide and are guided by the practice of coresidence.
39
The interdependence between subjective expressions of traditional filial norms and
parent-adult child coresidence has rarely been examined in Japan--the nation in Asia
where social and economic changes have been most acute over the last half-century.
Furthermore, few studies have considered whether coresidence with unmarried adult
children, the result of more recent social change, functions as a new non-traditional
family form.
A better understanding of filial norms and their enactment in diversifying
Japanese families, therefore, requires an examination of the symbiotic relationship
between norms of filial piety and structural family conditions. In this analysis we
use national data collected in 2001 from older people in Japan to explore the
interrelationship between expressed norms toward intergenerational coresidence and
the experience of living in multigenerational households. Specifically, we address
the following questions: To what degree does adherence to traditional norms
regarding the desirability of stem-family households determine whether older people
will live in such households? Does coresidence with unmarried adult children differ
from living with married children in this regard? Does the experience of
intergenerational coresidence shape normative orientations toward multigenerational
living? To address these questions, we rely on a model in which coresidence norms
and practice have the capacity to mutually influence each other (see Figure 5).
40
Figure 5. Research Model
3.3 Method
Data for this investigation derive from the Japanese General Social Survey
(JGSS) conducted in 2001. The JGSS is a Japanese version of the U.S. General
Social Survey and consists of a nationally representative sample of adults 18 and
over in Japan (Osaka University of Commerce, 2003). The sample was identified
through a multistage random sampling procedure that targeted adults living in
households throughout Japan (see details in Osaka University of Commerce &
University of Tokyo, 2002).
For this investigation we focus on 909 respondents to the JGSS who were 60
years and over at the time of the survey with at least one living adult child. Among
Predictors of Coresidence
Only
Coresidence
Common Predictors
Normative Belief
in Coresidence
Predictors of Normative
Belief Only
d1
d2
41
these respondents, we omitted 6% of the sample with missing data on relevant study
variables, yielding an effective sample size of 855. Characteristics of this older
sample are shown in Table 2. More than half (54%) of the sample consisted of
women, about one-quarter (24%) were widowed, and somewhat more than one-third
(36%) had at least a high-school education. The large majority of respondents (90%)
owned their own home, and 71% lived in a suburban or urban neighborhood. The
average age of this sample was 70 years.
42
Table 2. Sample Characteristics (n=855)
Variables Range/Categories Mean/Percentage s.d.
Age 60-89 years 70.1 7.2
Self-rated health 1(very poor) – 5
(excellent)
3.2 1.2
Family-life satisfaction 1 (dissatisfied) –
5(satisfied)
3.6 1.0
Male
45.7 Gender
Female
54.3
Owns home
90.3 Home ownership
Does not own home
9.6
Much lower than average
13.9
Lower than average
34.7
Average
43.9
Income
Higher than average
7.5
Elementary school
31.1
Middle school
32.6
High school
25.8
Education
College and above
10.4
Rural residence
Lives in rural area
29.2
Lives in suburban/urban
area
70.8
Widowhood Widowed 24.0
Not widowed 76.0
Son availability Has at least one son 82.1
Has no son 17.9
43
Table 3 shows the detailed living arrangements of older parents in the
sample. Slightly more than half (53%) was living with an adult child. However,
parents were twice as likely to live with a son (40%) as a daughter (18%). When
considering the marital status of children in conjunction with their gender, the most
common household type consisted of parents living with a married son (20%), while
the least common type consisted of parents living with a married daughter (4%). Of
parents not living with children, four out of five lived with their spouses only.
Table 3. Living Arrangements (n=855)
Living Arrangements Frequency
(%)
Couple-only 311
(36.4)
Living alone 66
(7.7)
Coresidence with child 452
(52.9)
Living with single son(s) 167
(19.5)
Living with married son(s) 178
(20.8)
Living with single daughter(s) 113
(13.2)
Living with married daughter(s) 35
(4.1)
Other 26
(3.0)
Total 855
(100)
44
The analysis consists of two sets of outcomes: coresidence with a child and
normative beliefs about intergenerational coresidence. Based on household
membership rosters reported by each respondent, three categories of coresidence are
derived: living with a married child (25%), living with an unmarried child (28%),
and not living with any children (47%). Normative belief in intergenerational
coresidence, is measured with the following question: “Do you think it desirable for
three generations (older people, their married children, and grandchildren) to share a
home?”, with two response categories “desirable” (66%) and “not desirable” (34%).
Constructing dichotomous coresidence variables from the three household
categories requires that the two most theoretically informative contrasts be
established. Since the normative question specifically triggers a response about the
desirability of three-generation stem-family households, our first variable contrasts
living with married children (=1) with all others (=0). (We note that while a small
minority of parents lived in stem-households with married children who were
childless at the time of the survey, many of these households will potentially
transition into three-generation households). In assessing coresidence with
unmarried children (=1) the contrast group consists of parents who did not live with
any children (=0). Because it is conditional on not living with married children, this
contrast provides a test of whether living with unmarried children can be considered
as a quasi-traditional arrangement (either as a type of household that will eventually
transition into a stem-family, or as an alternative multigenerational household based
primarily on the dependence of one generation on the other). This operationalization
of dummy variable contrasts implicitly treats intergenerational household
45
arrangements as an ordered continuum that goes from the most traditional to the least
traditional arrangement.
Predictors of the two outcome variables were selected based on the findings
of previous studies of intergenerational coresidence in Japan, as well as in other
East-Asian countries. These variables represent factors related to needs of older
parents (health and marital status), socio-economic resources of older parents
(economic status and home ownership), and predisposing characteristics of older
parents and their families (gender, education, urban/rural residence, son availability,
and family relationship quality).
Age is operationalized as a continuous variable ranging from 60 to 99.
Health is rated on a five-point ordinal scale reflecting self-perceptions of health
status, with response categories ranging from very poor (1) to excellent (5).
Economic status is assessed with the following question designed to elicit self-
perceptions of economic rank; “Compared with Japanese families in general, how
would you consider your family income?” Response categories range from much
lower than average (1) to higher than average (4). Parents’ education level is
constructed as a four-point scale for the highest level achieved: elementary school
(1), middle-school (2), high school (3), and college/university and above (4). The
quality of family relationships is assessed with the question “How much satisfaction
do you get from your family life?”, with response categories on a five-point scale
ranging from dissatisfied (1) to satisfied (5). Dichotomous predictors include gender
of parent (1=female, 0=male), marital status (1= widowed, 0= married or other),
home ownership (1= owns home, 0=does not own home), rural context (1= lives in a
46
rural area, 0= lives in urban/suburban area), and son availability (1=has son, 0= does
not have son). The distributions of independent variables are shown in Table 2.
The purpose of this investigation is to explore the reciprocal relationship
between filial norms and intergenerational coresidence, and to examine how each
affected the other relative to needs, resources, and predisposing factors. Models are
estimated separately for each of the two household contrasts because they reflect
different sample bases. In order to estimate the joint determination between
coresidence norms and household structure, our analysis is conducted using two-
stage least squares. In the first stage, we test the effects of all exogenous predictors
on each outcome variable in a recursive model where the relationship between
intergenerational norms and living arrangements is treated only as a residual
correlation (moderately strong at .39). From these models we then chose
instrumental variables that allowed the identification of non-recursive models
capable of specifying the reciprocal causal relationship between the two outcome
variables in a system of simultaneous equations (Jöreskog, 2005)
LISREL 8.54 is used to produce maximum likelihood estimates at each stage
of the analysis. Since all variables except age are either dichotomous or ordinal, a
polychoric correlation matrix is analyzed using weighted least squares. The key
assumption made in using this procedure is that the observed categories of the
dichotomous/ordinal variables represent thresholds for latent propensities that are
normally distributed (Jöreskog, 2005).
47
3.4 Results
We first describe the cross-classification between endorsing stem-family
coresidence as a norm and actual living arrangements. Table 4 shows that the
expression of traditional norms followed a pattern consistent with our theoretical
rank order--weakest for parent who do not live with their children (56%), stronger
for parents living with unmarried children (65%), and strongest for parents living
with married children (88%). Polychoric correlations confirmed this ranking. The
correlation between traditional norms and living arrangements was higher when
contrasting parents in households with married children (.55) than when contrasting
parents in households with unmarried children (.33).
Table 4. Cross-Classification between Living Arrangements and Normative
Belief in Traditional Multigenerational Coresidence
Desirability of Multigenerational Coresidence
Living Arrangements
Desirable Undesirable Total
Not living with child
225
(55.8)
178
(44.2)
403
(100)
Living with unmarried child
155
(64.8)
84
(35.2)
239
(100)
Living with married child
187
(87.8)
26
(12.2)
213
(100)
Total
567
(66.3)
288
(33.7)
855
(100)
Note: Percentages are in parentheses
48
The cross-classification in Table 4 also showed that only 12% of parents
living with married children found a traditional living arrangement to be
undesirable—a fairly low rate of dissatisfaction with current conditions. However,
more than half (55%) of parents not living with children found a traditional living
arrangement to their liking, suggesting many parents may be disappointed or
frustrated over not being able to fulfill their cultural ideal.
We next present multivariate analyses to estimate the mutual relationship
between norms and living arrangements using maximum likelihood two-stage least
squares. We note again that these models entail two different contrasts: (1)
coresidence with married children vs. all other living arrangements (n=855), and (2)
coresidence with unmarried children vs. not living with any children (N= 686).
Based on the results of recursive models we selected instruments that allowed
estimation of two non-recursive models (The results of this part of the analysis are
not presented in this paper but are available upon request). Instruments are variables
that estimate proxy (instrumental) variables that overcome the problem that each
jointly determined exogenous variable is correlated with the error term in its
equation. In order to identify these models, variables serving as instruments in one
equation are restricted to zero in the equation in which they were shown to have no
effect.
Instruments for the equation predicting coresidence with married children
were variables that predicted this type of coresidence but not norms (income and
health), and instruments for the norms equation was a variable that predicted norms
but not this type of coresidence (family-life satisfaction). Estimates for the non-
49
recursive model are presented in Table 5 (homeownership was excluded from this
section of the analysis because it is too highly skewed—98% of parents who lived
with a married child owned their own home). In the equation predicting coresidence
with married children, holding a traditional belief in coresidence increases the
likelihood of coresiding in a stem-family household. In addition, older parents, as
well as those with greater income and who have a son are more likely than their
counterparts to coreside with a married child.
50
Table 5. Maximum Likelihood Estimates of Nonrecursive Model for
Coresidence with Married Children and Normative Belief in
Traditional Multigenerational Coresidence (n=855)
Predictor Coresidence
a
Normative Belief
a
Age .13*
(.06)
-.13
(.07)
Female .06
(.10)
-.13
(.09)
Health -.09
(.05)
n/a
Income .22***
(.07)
n/a
Live in a rural area .15***
(.09)
.11
(.09)
Widowed .14
(.11)
.11
(.11)
Have at least one son .17**
(.07)
-.15*
(.06)
Education -.08
(.09)
-.11***
(.06)
Family-life satisfaction n/a .16
(.06)
Coresident with a child n/a .42*
(.21)
Normative belief in coresidence .52*
(.26)
n/a
R
2
.34 .16
* p <.05 ** p <.01 *** p <.001
a) Error covariance for coresidence and normative belief in traditional
multigenerational coresidence was -.33 (.23).
Note: Standard errors are in parentheses
51
In the equation predicting normative belief, we found that parents who
coresided with a married child were more likely to endorse the traditional norm, as
did those who had no sons, less education, and greater family-life satisfaction.
Notably, the two equations in this table, taken together, identified a reciprocal,
mutually reinforcing, relationship between traditional normative belief and
coresiding with a married child.
Instruments for the equation predicting coresidence with unmarried children
were those variables that predicted this type of coresidence but not norms (gender,
health, home ownership and widowhood), and instruments for the norms equation
were those variables that predicted norms but not this type of coresidence (education
and family-life satisfaction). The estimates from this non-recursive model are
presented in Table 6. Turning to the equation predicting whether parents coresided
with unmarried adult children, we found that normative beliefs in traditional
coresidence did not increase the likelihood of experiencing this type of coresidence.
Younger, male, and widowed parents, as well as those in worse health and who
owned their homes were more likely than their counterparts to coreside with
unmarried children.
52
Table 6. Maximum Likelihood Estimates of Nonrecursive Model for
Coresidence with Unmarried Children and Normative Belief in
Traditional Multigenerational Coresidence (N=686)
Predictor Coresidence
a
Normative Belief
a
Age -.41***
(.07)
.006
(.06)
Female -.29**
(.11)
n/a
Health -.18**
(.07)
n/a
Income .01
(.07)
.003
(.06)
House ownership .38***
(.12)
n/a
Live in a rural area -.12
(.15)
.15
(.08)
Widowed .58***
(.10)
n/a
Have at least one son .07**
(.10)
-.14
(.07)
Education n/a -.20**
(.08)
Family-life satisfaction n/a .17***
(.06)
Coresident with a child n/a .42*
(.20)
Normative belief in coresidence -.005
(.36)
n/a
R
2
.30 .14
* p <.05 ** p <.01 *** p <.001
a) Error covariance for coresidence and normative belief in traditional
multigenerational coresidence was -.23 (.34).
Note: Standard errors are in parentheses
53
The second equation in Table 6 predicting normative beliefs shows that
coresiding with unmarried children increased the likelihood that parents endorse
traditional norms regarding multigenerational living arrangements. This equation
also reveals that parents with greater family-life satisfaction and less education are
more likely to endorse such norms.
3.5 Discussion
The principal goal of this investigation was to study the nature of the
relationship between the ideal and the reality of intergenerational coresidence among
older people in contemporary Japan. Summarizing the main results, we found that
endorsing traditional family arrangements as an ideal promoted traditional stem-
family coresidence but not the less traditional type of coresidence with an unmarried
adult child. Living in a mutigenerational household of any type, however,
strengthened belief in multigenerational living as a social norm.
Our findings present a somewhat different picture depending on the type of
intergenerational arrangement. As for coresidence with an unmarried child, our
results generally support Koyano’s (2003) contention that intergenerational
coresidence in Japan depends more on instrumental contingencies than on traditional
cultural values. Poor health and loss of a spouse are key conditions under which
elderly parents live with their unmarried children, suggesting that the need for
caregiving or companionship may be motivating single children to remain with, or
move back in with their parents.
54
In addition, we find that the possession of property resources enables the
conditions under which coresidence with an unmarried child becomes possible.
These resources can be interpreted as reflections of parents’ ability to accommodate
their single child’s need for housing. This type of coresidence is certainly not trivial,
representing more than one-quarter of elderly households in the sample, slightly
more than the traditional stem-family. Further, though there are more sons than
daughters in households with unmarried children, the ratio is much more even than
in stem-family households where married sons predominate. Thus, the adult children
in these households appear less traditional than their stem-family counterparts in
terms of their gender distribution, but also in their possible functions as providers
and receivers of assistance with respect to their parents. While some Japanese social
critics have pejoratively labeled single children living with their parents as
“parasites,” they also represent the resilience of family life in Japan. Regardless of
whether they or their parents are the “dependent” parties in the household, this
emerging family form is a sign of intergenerational strength and solidarity, albeit that
the arrangement is based more on pragmatic than on ideological reasons.
As expected, holding traditional norms about intergenerational coresidence
appears to enhance the possibility of living in a stem-family household. Yet, our
analysis also highlights a disjunction between traditional cultural norms and actual
family practices in Japanese society. A substantial number of elderly parents in
Japan still express a desire for traditional living arrangements; yet fewer families are
actually able to fit themselves into the “culturally-desirable” image of family (White,
2002). Yet norms for coresidence have weakened over historical time, and this is
55
most acutely seen in younger cohorts who are engaging in life-styles far different
than their parents. The fulfillment of traditional family norms has become
challenging for contemporary young adults who are more economically straddled
than their predecessors, and not even possible for the growing number who are
eschewing marriage and family formation (e.g., Ogawa & Ermisch, 1996; Ogawa &
Retherford, 1997; Retherford, Ogawa, & Sakamoto, 1996). Our findings that older
parents are more likely to live with married children and younger parents are more
likely to live with single children are consistent with the notion that the composition
of multigenerational households has historically changed based on emergent values,
economic circumstances, and marital histories of young adults in Japan.
It is possible to argue from our findings that elderly parents who were
“successful” in accommodating themselves to traditional family arrangements define
themselves as cultural “winners”, while those who were not define themselves as
cultural “losers”. At the same time, older people who live with their children do not
necessarily hold such an arrangement as an ideal. The similar mismatch also exists
among elderly parents who do not live with children. The impact of the mismatch
between norms and residential situations on the well-being of the elderly in Japan is
beyond the scope of this study, but is certainly an important issue to be addressed in
the future.
The experience of coresidence seems to strengthen parent’s support for
traditional norms. This may reflect elderly parent’s reaffirmation of traditional
norms that comes from their actual coresidence experience. Somewhat surprisingly,
the experience of living with an unmarried child also promotes positive attitudes
56
toward traditional family norms. Older parents in this type of living arrangement
seem to use the conventional family norms to justify what they are doing. It is also
possible to interpret that these parents’ support for traditional norms reflects their
expectation for these children to get married and have children of their own.
Normative attitudes toward intergenerational coresidence are shaped by the quality
of family life. These results suggest that traditional family norms are not as
monolithic in Japanese society as they are typically portrayed, but are made
malleable by elderly parent’s preferences and family experiences.
Swidler (1986) proposed that cultural values are closely linked with actions
when there is stability in the cultural life of a nation. However, when rapid social
change alters the cultural landscape, traditional values begin to be questioned and are
no longer treated as “common sense”. Viewed in this context, the rapid rate of
modernization achieved by Japan since World War II should have shaken the
coercive power of core cultural norms regarding traditional family arrangements. In
this technologically advanced society, not far in time from its traditional past,
previously taken-for-granted norms have substantially lost their ability to initiate
actions. Perhaps this indicates that social norms based on instrumental concerns of
older parents and their children will supercede the traditional ones that are in decline,
or that such coresidence decisions will become increasingly based on personal
preferences and individual needs.
In closing this discussion, we note that information about adult children that
may have been consequential for the residential outcomes of their older parents was
not available in the data. For example, the socioeconomic status of adult children
57
can be a key component in the decision to coreside with parents (e.g., Tsuya &
Martin, 1992). Also omitted in the data is information about geographic proximity
between parents and children who do not live together. Proximity may enhance the
availability of intergenerational support in a way that is quite comparable to that
offered by coresidence (e.g., Martin & Tsuya, 1991; Silverstein, 1995). While the
inclusion of children’s characteristics and normative beliefs would likely have
increased the explanatory power of our models, it remains unclear how controlling
for these characteristics would have altered existing findings.
3.6 Conclusion
The ie ideology as an institutionalized norm guiding family arrangements in
Japan remains intricately blended in Japanese society. As a result of the
contradiction between the values associated with ie and current social and economic
realities, many families in Japan are now finding it difficult to comply with
traditional family demands. Indeed, a generational shift in expectations for shared
living arrangements may be exemplified by a cultural lag where the values of older
adults and their adult children are not synchronized with each other. What then will
happen to traditional norms of intergenerational coresidence in Japan in the near and
distant future? How will traditional norms become transformed in order to
legitimate coresidence with single children? Will intergenerational coresidence
merely become a strategy to deal with need (of either generation), rather than a social
dictate? These are important matters to be explored in order to better understand the
role of family life in a rapidly aging society such as Japan.
58
Chapter IV
Coresidence and Family Contexts
Recent statistics show that the prevalence of traditional intergenerational
coresidence between older parents and their children has been steadily declining
among economically developed nations in the world (United Nations, 2005). Japan
is one of such developed countries that has observed the continuous decline of
traditional multigenerational households. In the past two decades, the traditional
three-generational household in Japan has halved from 50 to 24 percent (Cabinet
Office, 2005). Parallel to the decreasing rates of coresidence across time, studies
also indicate that societal attitudes toward the traditional norm of filial
responsibilities are becoming negative (Ogawa & Retherford, 1993).
Another look at the coresidence rates in Japan, however, indicates that
despite the decline of the three-generation household, about half of the older people
live with their children, presumably due to the slight increase in two-generation
households where parents coreside with their adult children only (Cabinet Office,
2005). This suggests that the present intergenerational coresidence in Japan does not
necessarily mirror the traditional culture of three-generation households but entails
different types of arrangements that are newly created by families (Kono, 2000;
Martin & Tsuya, 1994; Takagi, Silverstein, & Crimmins, in press).
The trend of declining intergenerational living arrangements is generally
interpreted from the “common-preference” perspective, which considers the
intergenerational household as an arrangement intended to fulfill the family’s unitary
59
purpose to meet the needs of older parents (Becker, 1991). This approach attributes
the steady decline of multigenerational households to the socio-economic conditions
that have been improved for older individuals in developed societies, subsequently
reducing the practicality and necessity of intergenerational coresidence (Aboderin,
2003; Ogawa & Retherford, 1993).
On the other hand, game theoretic approaches consider intergenerational
coresidence as a family arrangement that accommodates and compromises multiple
needs and resources of each family member that do not necessarily coincide with
each other (Hayashi, 1995; Lundberg & Pollack, 1994; Pezzin & Schone, 1997,
1999). From this perspective, the implied social functions of coresidence is not
solely to attend the need of older parents but also to accommodate demands from
families at various generational positions by exploring a reconciling point among
family members. Thus in this context, residential outcome of intergenerational
coresidence would depend not only on the socio-demographic conditions and
normative orientations of older parents but also on the level of socioeconomic
resources maintained by other family members such as adult children and their
families. In fact, recent studies on coresidence in Japan suggest that negotiations
among family members are increasingly important in creating a living arrangement
that compromises the traditional cultural expectations and imminent needs and
resources exigent in the family (Izuhara, 2000, 2004; Traphagan, 2003).
In this study, we focus on the family contexts where different needs and
resources are negotiated by adult children and their parents in making a coresidence
arrangement in contemporary Japan. In this respect, we regard the coresidence
60
arrangement as a product of family negotiations across multiple family lineages; 1)
the vertical line between older parents and adult children; 2) horizontal lineages
among adult children; and 3) another horizontal line between adult children and their
spouses. As one of the potential outcome products of negotiations for coresidence,
we also consider coresidence as a household arrangement that reflects power
dynamics within the family. We represent the power dynamics in coresidence
arrangements by differentiating the household by headship, whether the household is
headed by adult children or by older parents. The headship of the household
indicates the role and power assumed for parents and children within the coresidence
household (United Nations, 2005). Parent-headed coresidence indicates that the
household management power is retained by the parent and the potential flow of
support is from parents to children, whereas child-headed coresidence suggests older
parents’ relinquishment of their power and resources to their coresiding children,
presumably due to their needs for social and economic support (Mutchler & Burr,
2003; Speare & Avery, 1993). The two types of coresidence also imply a contrast
between the traditional arrangement where older parents retain their power
throughout their life and the non-traditional arrangement in which adult children
accommodate their parents in their house contingent upon parental needs. Using a
nationally representative dataset of older individuals in Japan, we attempt to
delineate divergent forms of coresidence arrangements within the modern nation of
Japan.
61
4.1 Negotiation between Older Parents and Adult Children
Previous studies have identified reasons and motivations for intergenerational
coresidence based on level of needs and resources, normative attitudes, and
demographic characteristics of older parents and adult children. The needs of older
parents predominantly derive from their decline in health, when they generally
require increased assistance from others (Brown et al., 2002; Raymo & Kaneda,
2003; Silverstein, 1995). Older parents’ widowhood status is another indicator of
need for coresidence because of the lack of spousal support (Logan et al., 1998;
Raymo & Kaneda, 2003; Silverstein, 1995). On the other hand, the coresidence
arrangement can benefit some adult children, particularly those who need assistance
for childcare (Morgan & Hirosima, 1983), and those who have a relatively low level
of educational and occupational status, presumably due to their heightened need for
financial assistance. The coresidence arrangement also requires housing resources.
In general, older parents and adult children are more likely to live together if the
home is owned by the family (Logan & Bian, 1999; Raymo & Kaneda, 2003). Also,
parents with limited income may need to live with their children to supplement their
financial needs.
Within the context of Japan, as well as within other East Asian societies,
where the norm of filial piety is rooted in family practice (Koyano, 2003; Logan &
Bian, 1999; Sung, 2000), older parents’ normative attitudes toward filial
responsibility affect their residential decision of coresidence. Studies have shown
that older parents with traditional family values tend to embrace the notion of filial
responsibility for aging parents thus they prefer to live with their children (Logan &
62
Bian, 1999; Takagi & Silverstein, 2006). Normative values also are reflected by the
level of education achieved by older parents and adult children. Because of the
potential exposure to new ideas, parents and children with higher levels of
educational tend to express negative attitudes toward the traditional family norm,
leading to the maintenance of separate households (Logan & Bian, 1999; Rindfuss et
al., 1992; Takagi & Silverstein, 2006).
4.2 Negotiations among Adult Children with their Siblings
The traditional family norm, rooted in the coresidence arrangement in Japan
(as well as in other East Asian societies such as China and Korea), imposes filial
responsibilities primarily on sons in the family; therefore, in general, there is a
stronger expectation for sons to live with their parents than for daughters (Bumpass,
1994; Logan et al., 1998; Takagi & Silverstein, 2006). Rural residency also predicts
the enhanced likelihood of coresidence, presumably due to the relatively traditional
local culture maintained in the area (Raymo & Kaneda, 2003). On the other hand,
there are some potential alternatives to the traditional coresidence arrangement for
married adult children who are expected to live with their parents. For instance,
among adult children, single children are more likely to live with their parents than
married children (Logan & Bian, 1999; Rindfuss & Raley, 1998; Ward et al., 1992),
possibly due to their relatively greater flexibility in accommodating their parents’
needs and/or their own financial and housing needs. In this respect, the presence of a
single child in the family may substitute for the expected coresidence obligation of
married children.
63
Another alternative to coresidence for married children is to have a sibling
living separately but nearby their parents (Logan et al., 1998; Martin & Tsuya, 1991;
Takagi et al., 2006). The “intimacy-at-a-distance”, which tends to be preferred in
other societies such as the U.S.(Silverstein, Burholt, Wenger, & Bengtson, 1998),
has become more acceptable in contemporary Japanese society (Martin & Tsuya,
1994). The presence of a sibling living near their older parents may compromise the
need of coresidence, subsequently reducing the likelihood of married children to
reside with their parents.
In addition, the number of adult children in the family, which indicates the
relative size of the sibling network for supporting older parents, also affects the
likelihood of coresidence for married children (Rindfuss & Raley, 1998). Since
traditional coresidence tends to be formed exclusively with one of the children, a
larger sibling size in the family would mean a smaller possibility for each adult child
in the family to live with their parents.
4.3 Negotiations between Adult Children and their Spouses
The residential decision of coresidence also requires an agreement between
coresiding adult children and their spouses--children-in-law for older parents.
Because of the potential tensions and difficulties that derive from everyday
interactions among families in coresidence households (Aquilino, 1999; Cutler et al.,
1999; Koyano, 2003), the compliance of children-in-law with their spouse’s decision
for household arrangement could play a critical role. In this context, the power
dynamics within the couple may determine adult children’s likelihood of coresidence
64
with their parents. One potential indicator of such power dynamics is the
educational difference within the couple (Elder & Rudolph, 2003). In particular,
yielding to the partner’s decision for intergenerational household arrangement could
be more common when children-in-law hold less power in household management,
reflected by their lower educational attainment relative to their partner.
4.4 Study Purpose and Research Questions
The purpose of this study is to investigate the ways in which normative
beliefs, needs, and resources of older parents and adult children are negotiated within
the family context making possible two different types of intergenerational
coresidence: one headed by adult children and the other headed by older parents. In
examining the likelihood of each type of coresidence for married children, we focus
on family contexts that cut across the three different lineages: 1) between older
parents and adult children; 2) among adult children; and 3) between adult children
and their spouses. In an attempt to differentiate coresidence arrangements by the
power dynamics between older parents and adult children, we examine what
socioeconomic factors within the family motivate or necessitate married adult
children to either accommodate their parents as their cohabiter or live in a household
managed by their parents.
4.5 Method
Data and Sample
For this study, we utilize the data from the initial wave of the Nihon
University Japanese Longitudinal Study of Aging (NUJLSOA) collected in 1999.
65
The NUJLSOA is a national longitudinal study which focuses on the health and
social conditions of the Japanese older population aged 65 years and older
(USC/UCLA Center on Biodemography and Population Health, 2004). The baseline
sample consists of 4997 primary respondents including 606 proxies, with a response
rate of 74.6 percent. The survey asks respondents about up to 10 biological/adaptive
children and children-in-law, if the child is married.
In this study, we focus on the sub-group of self-reported respondents who
have at least one married child. To address the significant portions of respondents
who have missing data on relevant items (about 20% of the sample with the primary
selection criteria), we implement the method of multiple imputation, which estimates
missing values a specified number of times and pools multiple sets of analyses on
imputed data to calculate estimates (Schafer & Olsen, 1998). The imputation
method we utilize in this study is based on the calculation of Chained Equations
formulated by a conditional density of each individual variable with consideration of
all other variables (Royston, 2004). We carry out the five sets of imputation
throughout the analysis; thus, all of our estimates derive from the pooled inferential
statistics based on a set of five individual imputed datasets.
Table 7 and Table 8 present the two sets of sample characteristics of older
parents and adult children; one includes only those parents and adult children with no
missing values (2656 older parents and 5451 adult children) and the other group
represents all parents with their missing values imputed (3853 older parents and
8601 adult children). Overall, the characteristics of the two groups resemble closely
each other. Minor differences exist in their educational status, with the imputed
66
sample being slightly skewed to lower educational levels. Also, a somewhat higher
proportion of children are categorized as those who have a lower level of education
than their spouse. Additionally, the imputed full sample of older parents contains a
slightly higher proportion of parents who have a more traditional view of the family
norm of filial responsibility. However, we note that we ran two separate analyses
using both samples independently and the produced results did not have any
significant differences; thus, we present only the results based on the analyses of our
imputed full sample.
67
Table 7. Older Parents’ Characteristics
Sample without
Imputation (N=2656)
Sample with Imputations (N=3853)
Variables Range/Categories Mean (SD)/Percentage Mean(SD)/Percentage Number of Imputed
Cases
Age 65-99 years 74.7 (6.1) 74.9 (6.2) 0
Number of children 1-11 children 2.6 (1.1) 2.6 (1.2) 0
Income 1 (less than $4160) to 13
(more than $125, 000)
4.7 (2.3) 4.6 (2.3) 805
Gender Female 55.0 57.3 0
Male 45.0 42.7
Home ownership Owns home 90.0 90.4 6
Does not own home 10.0 9.6
Education Junior high school 60.6 63.6 33
High school 30.7 28.8
Vocational school/Junior
college
2.9 2.6
4-year university/Graduate
school
5.8 5.1
68
(Table 7, Continued)
Self-employed in
the past?
Yes 18.5 19.4 116
No 81.5 80.6
Widowhood Widowed 37.0 38.6 0
Not widowed 63.0 61.4
IADL problem Has at least one difficulty 10.7 11.9 11
Has no difficulty at all 89.3 88.1
Yes 47.9 46.4 0 Non-coresidence
child lives in the
same town?
No 52.1 53.6
Single child in the
family?
Yes 30.1 30.0 0
No 69.9 70.0
Family Value Agree 37.6 37.7 192
Somewhat agree 25.8 27.1
Somewhat disagree 17.1 16.8
Disagree 19.5 18.4
69
Table 8. Adult Children’s Characteristics
Sample without
Imputation (N=5451)
Sample with Imputation
(N=8601)
Variables Range/Categories
Mean
(SD)/Percentage
Mean
(SD)/Percentage
Number
of
Imputed
Cases
Age 24-79 years 46.4 (7.2) 46.8 (7.3) 157
Gender Female 50.5 50.7 3
Male 49.5 49.3
Education Junior high school 9.8 12.1 97
High school 50.2 51.7
Vocational school/Junior college 13.6 12.6
4-year university/Graduate school 26.4 23.6
Self-employed? Yes 14.9 15.5 296
No 85.1 84.5
Employed in blue-collar
job?
Yes 25.8 28.9 296
No 74.2 71.1
Employed in clerical job? Yes 29.7 29.6 296
No 70.3 70.4
70
(Table 8, Continued)
Employed in professional
job?
Yes 25.0 24.9 296
No 75.0 75.1
Have a child of pre-school
age?
Yes 14.4 13.3 88
No 85.6 86.3
Have a child of elementary
school age?
Yes 24.5 23.3 90
No 75.5 76.7
Have a child of junior high
school age or older?
Yes 71.8 72.5 89
No 28.3 27.5
Rural residence Lives in rural area 22.3 24.0 135
Lives in urban area 77.7 76.0
Educational difference with
spouse
Same as spouse (reference) 61.5 60.1 922
Has a higher education than spouse 20.0 19.3
Has a lower education than spouse 18.5 20.6
71
Dependent Variable
The outcome variable of this study represents the two different power
dynamics related to coresidence arrangements between older parents and adult
children. The first arrangement is one where the heads of household are the
coresiding adult children; and the second arrangement is one where the heads of
household are the older parents (either the older respondent or his/her spouse). The
headship person in each coresidence arrangement is intended to represent the person
who holds the primary responsibility for attending to household matters (United
Nations, 2005). As a baseline contrast, we establish the reference category as those
adult children who do not live with their parents.
Independent Variables: Adult Children’s Characteristics
Determinants of the likelihood of intergenerational coresidence include
factors that describe the socioeconomic background of adult children and their
family context as represented by their parents’ characteristics. As for predicting
factors relevant to adult children, we assess their demographic characteristics, socio-
economic status, educational difference with their spouse, and presence of their own
children. The demographic background of adult children includes their age
(continuous variable), gender (1=female, 0=male), and rural residency (1=lives in a
rural area, 0=lives in an urban area).
Socio-economic status of adult children is assessed by level of educational
attainment on a four-point scale (1=junior high-school, 2=high-school, 3= vocational
school or junior college, and 4=four-year university and/or graduate school).
Occupational status is also included with a series of dichotomous variables:
72
professional occupation (1=yes, 0=otherwise), blue-collar occupation (1=yes,
0=otherwise), and clerical occupation (1=yes, 0=otherwise). Due to the relatively
high frequency of coresidence among self-employed families (Raymo et al., 2004), a
dichotomous variable of self-employment (1=yes, 0=otherwise) is also introduced
into the model.
The educational difference between adult children and their spouses
represents the status dynamic within the couple, which potentially plays a significant
role in negotiating a living arrangement within the family. In the original survey, the
children-in-law’s educational status is reported by older respondents using the same
measurement as for their own children. We calculate the difference in educational
levels between each reported adult child and his/her spouse; and created a set of
dummy variables: 1) adult children have a higher level of education than their spouse
(1=yes, 0=otherwise) and 2) adult children have a lower level of education than their
spouse (1=yes, 0=otherwise), with the reference group being the same level of
educational attainment within the couple.
Adult children’s own family structure and their potential need for the
assistance with childcare, i.e., whether adult children have their own children, is
assessed with a set of three dichotomous variables, depending on whether adult
children have: 1) a child of the preschool age (1=yes, 0=otherwise); 2) a child of the
junior high school age; and/or 3) a child of the high school age or older. The
reference group includes adult children who have no children.
73
Independent Variables: Family Contexts
Family context is assessed by the overall family composition and older
parents’ characteristics, including their demographic characteristics, socioeconomic
resources and needs, and normative attitudes. The variables of family composition
include the number of adult children in the family as a continuous variable and older
parent’s widowhood as a dichotomous variable (1=widowed, 0=not widowed). In
addition, two dichotomous variables of family compositional factors are introduced
as potential alternatives to the choice of coresidence: the presence of an adult child
living nearby the older parent (1=has at least one child living in the same town/city,
0=otherwise) and the availability of a single child in the family (1= at least one adult
child within the family is not married, 0=otherwise).
The same as with adult children, the demographic characteristics of older
parents are represented by their age and gender. Reflecting the relevant family-level
occupational background for the overall likelihood of coresidence (Raymo et al.,
2004), older parent’s longest occupational background as being self-employed is
included as a dichotomous variable (1=mostly self-employed in the past,
0=otherwise). Educational background of older parents is assessed with the same
scale as adult children described earlier.
Socio-economic resources and needs of older parents are assessed with home
ownership (1=owns home, 0=does not own home) and household income (13-point
scale ranging from 1= less than 500,000 Japanese yen-approximately 4160 US
dollars to 13= more than 15 million Japanese yen-approximately 125,000 US dollars).
As a proxy for level of need for support, parents’ self-reported ability to perform the
74
following seven instrumental daily activities is included: preparing own meals,
leaving home to purchase necessary items, taking care of financial matters, using a
telephone, light housework, taking a bus or train, and taking medication as
prescribed. Due to substantial skewness, the individual seven items, which were
originally assessed with a four-point scale ranging from 1 (not difficult) to 4 (unable
to perform), are combined as a dichotomous variable (1=has a difficulty in
performing at least one item, 0=otherwise).
Older parents’ normative attitudes are represented by their attitudes toward
filial responsibility. The degree of traditionalism in their attitudes is assessed with a
four-point scale, based upon their response to the following statement: “A child
should be expected to support and take care of his or her aged parents, as the child
should feel a sense of gratitude to the parents for raising him/her.” The scale ranged
from disagree (1) to agree (4).
Statistical Model
In this study, we estimate the differential determinants of two different types
of coresidence: child-headed coresidence and parent-headed coresidence from the
perspective of married adult children. Specifically, the selected predictors focus on
the family context shared among adult children in the family, as well as adult
children’s individual socio-economic and family backgrounds. Since the original
data are all collected from older parents, we “reshape” the original dataset to
transform the unit of analysis from older parents to each of their married children.
The reshaped dataset contains adult children’s information with their parent’s (either
father’s or mother’s) data repeated within the family. We adjust the standard errors
75
to handle the clustering effects of the family unit by calculating robust standard
errors. Based on the transformed dataset, we implement a multinomial logistic
regression and estimate the log odds of married children being in either a child-
headed or a parent-headed coresidence household relative to the odds of being in a
reference group of non-coresidence.
4.6 Results
As described in Table 7, the average age of older parents is 75 years, and
slightly less than half of the parents are male (43%) and widowed (39%) respectively.
The majority (90%) own a home and have an average household income between 1.5
million Japanese yen (about 12,500 US dollars) and 3 million Japanese yen (about
25,000 US dollars). Slightly more than one-third of the parents (29%) have up to a
high school education, and slightly over one-fifth of the parents (19%) report having
been self-employed as the longest occupation status in the past. In terms of health
condition, slightly more than one-tenth of the parents (12%) report that they have
some difficulties in performing at least one area of instrumental daily activities.
Slightly less than half of the older respondents (46%) have at least one of their
children living nearby; and one-third (30%) of them have at least one child who is
not married. Regarding normative attitudes, more than half of the older respondents
(65%) agree with a given statement of filial responsibility.
As shown in Table 8, the average age of adult children is 47 years, and sons
and daughters are represented in almost equal proportions. Most of the children have
their own children of junior high school age or older (72%), but a smaller portion of
76
them have either a child of pre-school age (13%) and/or elementary school age
(23%). As for employment, about one-sixth of the children are self-employed (15%),
and each occupation (i.e., blue-collar, clerical, and professional) is represented by
about a quarter to one-third of children respectively. In terms of education, the
majority of the children have at least a high school education (88%). Slightly more
than half of the children (60%) have the same educational level as their spouse, and
the rest is divided into those with a higher educational status than spouse (21%) and
those with a lower educational status than their spouse (19%). One out of four
children in this sample (24%) lives in a rural area.
Table 9 presents the living arrangements of older parents in our study, cross-
classified with gender of adult children. Overall, 11 percent of the adult children
(n=935) reported by older parents are in a child-headed coresidence arrangement,
and 7 percent of the children (n=610) reside with their parents in a parent-headed
coresidence household. As the cross-classification with children’s gender shows,
both types of coresidence are predominantly formed with sons than with daughters,
reflecting the cultural norm of son preference for coresidence in Japan (Koyano,
1996; Takagi & Silverstein, 2006).
77
Table 9. Living Arrangements and Adult Children’s Gender
Living Arrangements Sons Daughters Total
Not living with parent
2978
(42.2)
4078
(57.8)
7056
(100)
Child-headed coresidence
777
(83.1)
158
(16.9)
935
(100)
Parent-headed coresidence
482
(79.0)
128
(21.0)
610
(100)
Total
4237
(49.3)
4364
(50.7)
8601
(100)
Note: Percentages are in parentheses
Table 10 presents the results of the multinomial logistic analysis that shows
the odds ratios of the likelihood of the two types of coresidence relative to their
reference of non-coresidence. Consistent with the cross-tabulation shown earlier,
both types of coresidence are predominantly formed with sons. Also, both
coresidence arrangements are common among married adult children who are older
and live in rural areas. Self-employed children and adult children with a child of
elementary school age are also more likely to live with their parents in both types of
coresidence arrangements.
78
Table 10. Multinomial Logit (N=8601): Odds Ratios to Non-Coresidence Arrangement
Child-Headed Coresidence Parent-Headed Coresidence
Odds Ratio Robust S.E. Odds Ratio Robust S.E.
Adult Children’s Characteristics
Age 1.07*** .01 1.05*** .01
Gender (Male) 22.27*** 6.2 14.62*** 4.52
Education .99 .10 1.16 .12
Gender*Education .66*** .08 .71** .08
Rural residence 5.06*** .45 6.36*** .61
Blue-collar job .86 .11 .87 .13
Clerical job 1.05 .11 .84 .13
Professional job .72* .12 .69* .13
Self-employed 1.48** .21 2.03*** .29
Has a child of pre-school age 1.02 .17 .90 .14
Has a child of elementary school age 1.35* .16 1.32* .14
Has a child of junior high school age 1.26 .16 .93 .12
79
(Table 10, Continued)
Has a higher education than spouse 1.64*** .22 .99 .14
Has a lower education than spouse .82 .10 1.12 .16
Older Parent’s Characteristics
Age 1.00 .01 .96** .01
Gender (Male) .79* .09 1.56*** .18
Widowhood 3.54*** .31 .46*** .07
Home ownership 3.75*** .51 5.27*** 1.90
Household income .93* .02 1.08** .03
IADL problem 1.09 .11 1.09 .17
Education .87 .07 .92 .07
Self-employed in the past 1.11 .11 1.35** .14
Family value 1.03 .04 1.12** .05
Number of children .62*** .02 .62*** .03
Presence of single child .59*** .06 .59*** .08
Non-coresident child in the same city .49*** .04 .50*** .05
*p<.05 **p<.01 ***p<.001
80
On the other hand, several factors originated in parent’s characteristics
clearly differentiate the two types of coresidence. While the parent-headed
coresidence is more common for older fathers, the child-headed coresidence is more
likely to be arranged for older mothers. Also, older parents in a parent-headed
coresidence tend to have a relatively high level of income, whereas those living in a
child-headed coresidence household are likely to report a lower income. In addition,
widowed parents are much more likely to be found in the child-headed coresidence
than in the parent-headed coresidence.
As expected, the more siblings present in a family, the lower the likelihood of
coresidence for each married child, suggesting an exclusive coresidence arrangement
with only one child. In addition, the two alternatives to coresidence tend to have a
significant effect on the likelihood of coresidence for married children. That is, the
presence of a sibling who is single or who lives nearby the parent appears to
substitute for the coresidence arrangement between older parents and married
children in the family.
Furthermore, we examined the interaction between adult children’s gender
and educational status in relation to the likelihood of being in a child-headed or a
parent-headed coresidence arrangement. Figure 6 and Figure 7 present the predicted
probabilities of each type of coresidence arrangement as a function of gender and
education of married adult children. For both types of coresidence arrangements,
sons with higher level of educational status are less likely to live with their parents.
As for the likelihood of being in a child-headed coresidence, daughter’s educational
level does not appear to have a significant impact, staying at very low percentages
81
across educational levels. On the other hand, in case of parent-headed coresidence,
married daughters with a higher level of education are slightly more likely to live
with their parents. Considering that the child’s educational status by itself does not
have a significant impact on the likelihood of coresidence arrangements, the
identified interaction between educational level and gender of married children
suggests a suppressing effect on the independent effect of education on coresidence.
0
0.05
0.1
0.15
0.2
Junior high High school College University
Child's Education
Predicted Probability of
Child-Headed Coresidence
Female Male
Figure 6. Predicted Probability of Child-Headed Coresidence by Adult Children’s
Gender and Education
82
0
0.02
0.04
0.06
0.08
0.1
Junior high High school College University
Child's Education
Predicted Probability of
Parent-Headed Coresidence
Female Male
Figure 7. Predicted Probability of Parent-Headed Coresidence by Adult Children’s
Gender and Education
The analyses show several unique determinants of each type of coresidence
as well. The power dynamics between adult children and their spouse predict only
child-headed coresidence but not parent-headed coresidence. More specifically,
adult children are more likely to live with their parents as a head of the household
when they hold a higher educational status than their spouse. On the other hand, one
of the unique predictors of parent-headed coresidence is the parent’s past self-
employment status. Older parents who were mostly self-employed in the past are
more likely to be found in a parent-headed coresidence arrangement. Also, having
an older parent who holds a relatively more traditional attitude toward filial
83
responsibilities significantly increase the likelihood of married children living in a
parent-headed coresidence; but no relationship with parental normative attitudes is
identified for child-headed coresidence.
4.7 Discussion
This study investigated common and distinctive socioeconomic factors that
determine the likelihood of two different types of coresidence arrangements in
contemporary Japanese society: one represents child-headed coresidence and the
other depicts parent-headed coresidence. We focused our analysis on multiple
family lineages in an attempt to encompass a wide range of factors that are
negotiated in the family contexts to arrange intergenerational coresidence. Overall,
the results support the game theoretic approach to family arrangements, in which the
family deals with divergent needs and resources in order to establish one form of
intergenerational household arrangement (Pezzin & Schone, 1997, 1999). Identified
predictors for coresidence arrangements suggest that the formation of coresidence is
determined not only by the needs and resources of older parents but also by the
socioeconomic and family conditions of adult children and their families. The
comparison between parent-headed coresidence and the child-headed coresidence
suggest both common and unique factors that lead a particular adult child in the
family to share a household with their parents either as a head of the household or as
a coresiding family member in their parent’s household.
Intergenerational coresidence in contemporary Japan still appears to be under
the influence of the cultural norm of son preference. Also, overall, coresidence
84
seems to be more common in rural areas, where the traditional family practice tends
to be relatively intact relative to that in urban areas. Interestingly, results also show
some gender differences in the ways in which married sons and daughters diverge
from the traditional family arrangement. For instance, the significant interaction
between children’s gender and educational status suggests that while married sons
who hold a higher level of educational attainment tend to live separately from their
parents, for married daughters, it is their higher level of educational status that
enables them to live with their own parents. This suggests that, for sons and
daughters, going against the traditional culture of coresidence takes opposite
directions. Since a higher level of educational attainment is generally associated
with more negative attitudes toward the traditional norms (Logan & Bian, 1999;
Takagi & Silverstein, 2006), highly educated sons will be less likely to abide by the
cultural convention of coresidence thus will not live with their parents; whereas for
highly educated daughters, living with their own parents (rather than their parents-in-
law) implies their movement against the tradition.
From adult children’s standpoint, the power dynamics with their spouse
seems to matter particularly for creating a child-headed coresidence arrangement
where the child presumably holds the primary position for household management.
Specifically, adult children who hold a higher level of education than their spouse
are more likely to manage a coresidence arrangement with their own parents,
presumably due to their relatively higher status in power dynamics for family
decision-making (Elder & Rudolph, 2003).
85
Alternatives to coresidence constitute a part of the negotiation process in the
family as well. Findings show that married children’s coresidence with their parents
appears to be either delayed or substituted by siblings who are still single or those
who live separately but nearby their parents. Having a sibling who lives nearby the
parent is likely to be an alternative to coresidence, which has become more accepted
as the arrangement of “intimacy-at-a-distance” (Martin & Tsuya, 1994).
The contrast between the two types of intergenerational coresidence by
household headship shows that the implied function of coresidence varies depending
on which generation plays the primary role in running the household, indicating the
divergent benefits of coresidence arrangements for older parents and adult children.
The results suggest that the coresidence household headed by adult children
primarily serves to fulfill the needs of older parents, particularly for mothers and
those parents who are widowed, as well as for older parents with relatively limited
income. On the other hand, parent-headed coresidence tends to be an arrangement in
which older parents retain relatively rich socioeconomic resources, reflected by their
still having a living spouse and possessing a relatively high income. Older parents in
this type of coresidence also are more likely to embrace the traditional norm of filial
responsibility, whereas such a relationship is absent in the child-headed coresidence.
This may reflect the difference in older parents’ status within the household.
Presumably, the parents who live with their children as head of the household may
have more control in household managements over their children, thus expecting
support from their children in return.
86
Another determinant of coresidence is the occupational background of older
parents and adult children. Adult children’s occupational background reflects the
ways in which the tradition of coresidence arrangements is either contested or
maintained across generations. The discouraging factor is observed with relatively
highly ranked occupational status of adult children (i.e., professional job), which
tends to decrease their likelihood of living with their parents. On the other hand,
self-employed married adult children are more likely to live with their parents, either
in a child-headed arrangement or a parent-headed arrangement. This is consistent
with a past study that found that intergenerational coresidence tends to be more
common among self-employed families (Raymo et al., 2004). However, our results
also show that when the coresidence household is headed by adult children, older
parent’s self-employment status in the past does not play a significant role,
suggesting that the establishment of child-headed coresidence takes place regardless
of the past occupational background of older parents. In other words, this suggests
that while the parent-headed coresidence may be a continuous household
arrangement tied to past arrangements and backgrounds of older parents, child-
headed coresidence may involve more disruptions in older parents’ life-styles
because of its primary purpose to have their imminent needs met by their coresiding
children.
To place the results in a proper context, we note several limitations of our
study. First is the cross-sectional nature of our dataset. Empirical support for our
causal inference of coresidence arrangements will not be complete without a
longitudinal data analysis. The data from the first two waves that were available for
87
analysis did not provide enough numbers of transitions in living arrangements, thus
longitudinal analysis on this matter was impossible. We also note that although most
of the data on adult children we used in our analysis was straightforward,
demographic and occupational characteristics data on adult children in our analysis
was provided by their older parents, not directly from the adult children themselves,
which potentially creates biases in their accuracy (Aquilino, 1999). For instance, the
bias could come from patterns of missing data, in which older parents may fail to
report about their adult children with a relatively more distant relationship than with
their other children. The method of multiple imputation we implemented was
intended to minimize such latent biases that could derive from the missing cases and
maximize the statistical power. Although the preliminary analyses do not show any
significant differences between the imputed and pre-imputation samples, the results
still need to be carefully interpreted.
This study presents intergenerational coresidence in Japan as the result of
negotiations that are based on the needs and resources exigent at multiple levels of
family dynamics. As the bargaining model of household arrangements contends, the
findings show that the establishment of one type of family arrangement requires
compromise and cooperation from multiple family members. Also, the identified
contrast between child-headed coresidence and parent-headed coresidence in our
study reflects the diversified functions of coresidence in contemporary Japanese
society. Our analysis depicts child-headed coresidence as an arrangement that has
departed from the traditional, culturally-prescribed arrangement (Hashimoto, 1996)
to a need-based one that hinges on socioeconomic conditions of older parents and
88
adult children at a particular point in time (Martin & Tsuya, 1994; Takagi et al.,
2006). Somewhat interestingly, the health condition of older parents does not predict
either type of coresidence in this study. One possible interpretation of this is
methodological; in the analysis where the principal sampling unit consists of a group
of adult children whose parents are mostly healthy, there may not be enough
variation in parental health conditions within the sample.
Generally, statistics on coresidence are based on older parents’ perspectives;
thus, for instance, the rate of coresidence among older individuals with their adult
children would be presented as being currently about 50 percent, which seems
substantially high (Cabinet Office, 2004). On the other hand, if the viewpoint is
shifted to adult children’s perspectives, the proportion of married children who
coreside with their parents is only about 20 percent. This suggests that coresiding
children are specifically chosen to live with their parents based on negotiations
among cultural preferences, needs, and resources shared within the family. Japanese
society has recently observed the progressively weakened tie between traditional
cultural norms and actual coresidence arrangements (Ogawa & Retherford, 1993),
creating individual variability in living arrangements of older parents with, or
without their children (Izuhara, 2004; Kono, 2000; O'Leary, 1993). The findings of
this study demonstrate one aspect of such diversity; that implied functions of
coresidence significantly vary depending on the ways in which the coresidence
household is managed by either older parents or adult children. The irrelevancy of
older parents’ normative beliefs in filial responsibilities in the child-headed
coresidence indicates that some intergenerational coresidence arrangements in Japan
89
are managed primarily by adult children and may be governed by a different
normative paradigm – a new normative framework that may be significantly
different from the one traditionally known in Japanese society. The coresidence
management presented in this study also suggests that there is a significant variation
in power dynamics across different types of coresidence households; and such
diversity should certainly be recognized for better understanding of intergenerational
family arrangements in aging modern society such as Japan.
90
CHAPTER V
Coresidence and Social Contexts
Japan is one of the very few developed countries that maintain a relatively
high rate of coresidence between older parents and adult children. As of 2001, 58
percent of older people aged 60 years and older in Japan lived with their children, a
figure three to ten times greater than that found in comparably developed Western
societies such as the United States (17%), Germany (15%), and Sweden (5%)
(Cabinet Office, 2001). The traditional form of coresidence in Japan, where older
parents reside with their eldest son and his family, has been discussed as the
structural manifestation of the cultural norm of filial piety, which assigns children
the responsibilities to take care of aging parents (Koyano, 1996; Palmore & Maeda,
1985; Sung, 1998). At present, however, both the rate of coresidence and social
beliefs in the norm of filial piety are in decline. The percentage of multigenerational
households has dropped from 50 to 24 percent in the past two decades (Cabinet
Office, 2005). Past studies also indicate increasingly negative attitudes toward the
norm of filial piety in Japanese society (e.g., Elliott & Campbell, 1993; Koyano,
1996, 2003; Ogawa & Retherford, 1993).
Scholars interpret the decline of intergenerational coresidence in Japan as the
gradual disappearance of family tradition occurring in most developed societies.
They argue that while coresidence is still partially guided by the norm of filial piety,
the changes accompanying the rapid modernization of Japanese society has also
made the traditional coresidence arrangement impractical and unreasonable for many
91
families (Kono, 2000; O'Leary, 1993; Ogawa & Retherford, 1993; Sodei, 1995). On
the other hand, Japan remains an anomalous case within the community of
economically advanced countries that expects nuclear family households to
predominate (Maeda & Ishikawa, 2000).
Relatively overlooked aspect of intergenerational coresidence in Japan is its
diversity across individuals as well as local communities. At the individual-level,
previous longitudinal studies have presented that the intergenerational coresidence in
Japan has begun to shift from a normatively-driven, “preventive” approach with
advance preparation for expected age-related needs of older parents (Hashimoto,
1996) to a “contingent” approach, in which older parents and adult children become
reunited after the need arises (Brown et al., 2002). With this contingent approach,
families modify, but do not violate, the tradition of filial piety by delaying the timing
of coresidence (Kawabe & Shimizu, 1994; Kono, 2000). The shift as such
presumably creates diversity in current patterns of intergenerational coresidence
arrangements, including both traditional and non-traditional forms.
There is also a variation in the prevalence of coresidence across geographical
areas within the country; the coresidence rate among older individuals tends to be
higher in the north-east region than in the south-west region of Japan (Ministry of
Health Labour and Welfare, 2000; Shimizu, 2004). The regional variation in
coresidence rates indicates the existence of contextual effects on older individual’s
living arrangements, such as the degree of traditionalism in family culture, strengths
of socio-economic and housing conditions, and richness in welfare resources to
support families with older residents in need.
92
In this study, we contend that the overall trend of declining traditional
coresidence in Japan needs to be re-examined for its potentially divergent ways in
which the traditional culture remain influential in the living arrangements of
Japanese older individuals. In particular, our focus is to examine how larger socio-
economic, cultural, and welfare contexts shape coresidence patterns, and to
determine the strength with which filial values are expressed in the practice of
intergenerational coresidence. Potential predictors of intergenerational coresidence
at both the individual- and regional- levels are included in our analysis of a
nationally representative sample of older individuals in Japan. Specifically, we study
the traditional type of life-long coresidence arrangement and a relatively new style of
“boomerang” coresidence, in which a child becomes reunified with a parent in a
shared household later in life in response to family contingencies (Cherlin, 1994;
Martin & Tsuya, 1994). Below, we discuss potential predictors of intergenerational
coresidence in more detail, both at individual- and contextual-levels.
5.1 Diversity of Coresidence across Individuals
Individual-level predictors of coresidence between older parents and adult
children in Japan are commonly identified by studies mainly in East Asian societies,
which include need, kin availability, resources, normative attitudes, and demographic
characteristics. Needs are mainly represented by older parent’s health status, with
poorer health predicting a higher likelihood of coresidence (Brown, 2003; Crimmins
& Ingegneri, 1990; Logan & Bian, 1999; Logan et al., 1998; Raymo & Kaneda,
2003; Silverstein, 1995). Individual level of needs is also reflected as the availability
93
of family members. For instance, lack of spouse through widowhood tends to
increase the likelihood of coresidence (Martin & Tsuya, 1994; Raymo & Kaneda,
2003). Because of the patrilineal kinship system in Japan, availability of sons also
matters; older parents who have at least one son are more likely to live in a
coresidence arrangement than those who do not (Takagi & Silverstein, 2006). One
possible alternative to coresidence is an arrangement where older parents and adult
children live separately but close to each other (Cabinet Office, 2004; Martin &
Tsuya, 1994). Thus, the availability of adult children within close geographical
proximity may decrease the likelihood of coresidence. As for resource availability,
older parents with greater housing resources are more likely to live with their
children than those with fewer resources (Logan & Bian, 1999; Raymo & Kaneda,
2003). Housing resources, such as home ownership and household income,
presumably enable older parents to accommodate housing needs of adult children.
Incidence of coresidence is also determined by older parents’ cultural beliefs
about the desirability of coresidence. Older parents who more strongly embrace
traditional family values are more likely to form a coresidence arrangement than
those with a less traditional view (Kurosu, 1994; Takagi & Silverstein, 2006).
Normative attitudes are also indirectly reflected by the educational attainment of
parents and children. In general, parents and children with higher levels of
educational attainment tend to express a negative attitude toward traditional family
value, and thus less likely to live together (Logan & Bian, 1999; Rindfuss et al.,
1992; Takagi & Silverstein, 2006).
94
In terms of demographic characteristics, parents with advanced age are
generally more likely to live with their children (Cabinet Office, 2004). In Japanese
context, parental gender has not been found as a consistent predictor; but some
studies indicate that older men are more likely to live with their children than older
women (e.g., Tsuya & Martin, 1992). Rural residency is another demographic
characteristic that tends to increase coresidence presumably due to the fact that rural
areas generally maintain a traditional family culture (Raymo & Kaneda, 2003; Tsuya
& Martin, 1992).
5.2 Diversity of Coresidence across Regions
Studies have argued that family arrangements in Japan vary across not only
individual family circumstances but also the social contexts in which they are
embedded, including local housing and economic conditions, cultural climate, and
accessibility of social welfare resources (Izuhara, 2000, 2004; Kumagai, 1996, 2005;
Kurosu, 1994). Very few studies, however, have clarified the role and meaning of
this diversity at the national level.
Observed diversity of intergenerational coresidence across Japanese regions
appears to derive from socio-economic contexts such as economic and housing
conditions. Coresidence arrangements tend to be difficult in the areas with few
economic opportunities because of the scarce work availability for younger
generations (e.g., Thompson, 2003). Insufficiency of financial resources at local
levels could indicate the heightened needs of intergenerational support from older
parents who reside in the area, potentially encouraging their coresidence.
95
Alternatively, the lack of economic resources in the area also means that younger
generations need to move to more prosperous areas to compromise their lack of
financial means. Also, limited housing options and space put significant restrictions
on creating intergenerational living arrangements, which generally require a larger
space for two generations to share the household together (Izuhara, 2000; Logan et
al., 1998; Sodei, 1995).
Regional diversity of coresidence can also derive from the cultural climate
within the area. Generally speaking, the normative environment in the north-east
region tends to embrace intergenerational coresidence as an ideal arrangement to
take care of older parents, whereas the northern island (Hokkaid) and the south-
west region maintain a culture that encourages independent living arrangements for
older parents and adult children (Shimizu, 2004). These regional differences tend to
exist independently from the rural-urban contrast in prevalence of coresidence,
suggesting the tenacity of regional culture that either embraces or discourages
coresidence arrangement, rather than the local economic conditions that tend to
differentiate rural areas from urban areas (Raymo & Kaneda, 2003).
Finally, the community-level availability of social welfare services for older
people is another potential contextual factor that influences intergenerational living
arrangements in Japan. In response to the declining social belief in filial
responsibilities, there has been a steady increase of societal demands for public
welfare resources for Japanese older people (Eto, 2001; Peng, 2003). In the past few
decades, Japanese government has put forth their political initiatives to increase the
level of welfare supplies for the older population, in an attempt to achieve a better
96
balance between familial responsibilities and social welfare responsibilities
(Ishikawa & Maeda, 2000). One of the major political shifts accompanied by this
recent welfare trend is to give local community more responsibilities and authorities
for service management (Masuzoe, 2000; Peng, 2003). As a result, the present social
welfare service for the Japanese older population entails significant variability across
the sub-national local units.
5.3 Study Purpose and Research Questions
The purpose of our study is in three fold. First is to examine the differential
determinants of two different forms of coresidence and delineate the contrastive
functions of intergenerational living arrangements, between the one predominantly
governed by traditional cultural norms and the one made in response to exigent
personal needs. Reflecting the emerging contingent arrangement of coresidence
(Brown et al., 2002), we contrast the traditional type of “life-long” coresidence
where older parents and adult children live together since the birth of the child and
the non-traditional type of “boomerang” coresidence where parents and children start
living together again after some time passed. Second is to investigate variation in
coresidence arrangements across regions of Japan that vary by economic
opportunities afforded the working population, housing space, utilization of public
services for older adults, and the extent of traditionalism in the cultural milieu. Third
is to study whether the socio-economic constraints, cultural milieu, public welfare
service, and housing availability modify the relationship between older individual’s
expectation of filial piety and coresidence with a child.
97
We hypothesize that the traditional type of life-long coresidence tends to be
primarily associated with the cultural norm embraced by older parents, whereas the
non-traditional type of boomerang coresidence is more specifically related to the
needs of older parents. In terms of the observed regional variation of coresidence
rates in Japan, we introduce the sub-national districts of prefecture as a geographical
unit; and examine the relationship between the prefecture-level socio-economic and
cultural contexts and individual-level living arrangement of intergenerational
coresidence. We expect that the likelihood of coresidence tends to be higher in a
prefecture that provides relatively rich economic and housing resources for
coresiding parents and children. We also hypothesize that a traditional cultural
climate that prevails within the prefecture would promote coresidence arrangements.
On the other hand, we consider the local welfare resources for older residents as a
potential replacement of the family support of coresidence. Furthermore, because of
its potential linkage between the cultural norm and modernization of social structure
(Ogawa & Retherford, 1993), we speculate cross-level effects between older
individual’s normative attitudes and nesting contextual factors, where the extent to
which the coresidence arrangement is supported by older parent’s normative belief
differs depending on the nesting socio-economic and cultural contexts.
5.4 Methods
Data
This study analyzes the initial wave of the Nihon University Japanese
Longitudinal Study of Aging (NUJLSOA) collected in 1999. The NUJLSOA is a
98
national longitudinal study conducted by the researchers from Nihon University, and
the survey specifically focuses on the health and social issues of the older population
aged 65 years and older in Japan (USC/UCLA Center on Biodemography and
Population Health, 2004). The first wave of data collection had a response rate of
74.6 percent, which included 4997 participants: 4391 self-respondents and 606
persons whose responses were provided by a proxy.
Our study focuses on the sub-group of the NUJLSOA sample whose
information was provided in self-reports and who had at least one living child. Out
of 4085 respondents who satisfied these selection criteria, we further excluded 520
respondents (12.7% of the selected sub-sample) who had missing values for our
study variables that could not be imputed, yielding an effective sample size of 3565.
We note that the majority of respondents excluded from our sample were those who
provided their answers through proxies due to their significant health conditions that
precluded them from answering by themselves. We decided to exclude them from
our sample because of their complete missing on our focal question of cultural value.
As a result, the preliminary comparison between our analytic sample to the full
sample revealed that there were no differences except with regard to health status,
where we found that our sample tended to have a better health than the original
sample. Thus, our results are mostly generalizable to the older population with
relatively less significant health problems.
Dependent Variable
The outcome variable of this study is a multinomial variable representing
three different types of intergenerational living arrangements. The first is the
99
reference group of non-coresidence where older parents live separately from their
children. The second category is labeled as life-long coresidence, where older
parents and adult children have been living together since the child was born. This
type of coresidence was considered to reflect the traditional family culture in East
Asian societies that promotes intergenerational dependency in shared living
arrangements to prepare for expected needs in the future (e.g., Bumpass, 1994;
Hashimoto, 1996; Hirschman & Minh, 2002; Kurosu, 1994; Ting & Chiu, 2002).
The third category is called “boomerang” coresidence, where older parents and adult
children started living together later in their lives. The label of “boomerang”
indicates the reunification of parents and children. This arrangement represents the
relatively new type of intergenerational living arrangement that responds to the
family needs that arise at a particular point in time (Cherlin, 1994; Martin & Tsuya,
1994).
Although our data does not provide the distinctions, the “boomerang”
coresidence could include both older parents who had moved into their adult
children’s home and those who had accommodated their children moving into their
house. We also note that a small number of parental households (n=21) had both a
child who never left home and at least one who previously lived independently and
returned. Since only 21 parents are in the hybrid category, their assignment to one
category or the other is of little consequence to our findings. Nevertheless, we
classify this hybrid form of coresidence as a “boomerang” type of household because
we believe that having at least one boomerang adult child makes the household
arrangement in its totality non-traditional from the parent’s point of view.
100
Individual-Level Independent Variables: Older Parents’ Characteristics
Individual-level predictors of coresidence reflect older parents’ demographic
characteristics, socioeconomic resources and needs, family composition and
availability, and normative attitudes. Older parent’s age is operationalized as a
continuous variable ranging from 65 to 99. Dichotomous variables are created for
gender of older parents (1=female, 0=male) and rural residency (1=lives in a rural
area, 0=lives in an urban area).
As for socioeconomic resources and needs of older parents, home ownership
is coded as a dichotomous variable (1=owns home, 0=does not own home) and
income is assessed with a 13-point ordinal scale ranging from 1 (less than 500,000
Japanese yen-approximately 4160 US dollars) to 13 (more than 15 million Japanese
yen-approximately 125,000 US dollars). Due to the substantial proportion of
missing values (about 20% of the original dataset), we imputed the household
income variable within the stratum of parental age and gender using the hotdeck
method, which replaces missing values based upon other cases with no missing data.
We caution that the single imputation of hotdeck generally decreases the estimates of
standard errors, thus increases the risk of Type I error. Older parents’ health status is
assessed as their self-reported ability to perform seven instrumental daily activities:
preparing own meals, leaving home to purchase necessary items or medication,
taking care of financial matters, using a telephone, dusting, cleaning up and other
light housework, taking bus or train, and taking medication as prescribed. These are
measured with a four-point scale ranging from 1 (not difficult) to 4 (unable to
perform). Due to skewness in this variable, we combined all seven items and created
101
a dichotomous variable (1=has a difficulty in performing at least one item,
0=otherwise).
Family composition and child availability are estimated by three dichotomous
variables of widowhood (1=widowed, 0=not widowed), son availability (1=has at
least one son, 0=has no son), and presence of child living nearby (1=has at least one
child living in the same town/city, 0=otherwise). We note that because of the
minimal divorce rate of older people in Japan (1.7% of our study sample), we
grouped these married and divorced/separated parents in one group against those
who were widowed. In addition, the number of children is included as a continuous
variable.
Education of both older parents and adult children is measured on a four-
point scale: junior high-school (1), high-school (2), vocational school or junior
college (3), and four-year university and/or graduate school (4). As the proxy of
attitudinal climate of adult children, we also include the averaged educational level
of all children about whom older parents reports having children, including both
those who coresided with parents and who lived separately from parents, within each
family. Older parents’ normative attitudes toward filial responsibilities is measured
with the following statement: “A child should be expected to support and take care
of his or her aged parents, as the child should feel a sense of gratitude to the parents
for raising him/her.” Response was provided on a four-point scale ranging from
disagree (1) to agree (4).
102
Prefecture-Level Independent Variables: Socio-Economic and Cultural Contexts
We use characteristics of prefectures to examine the impact of contextual
factors on residential arrangements. Prefectures are sub-national jurisdictions led by
elected governors. Japan consists of 47 prefectures, each containing smaller
governmental units of cities, towns, and villages; and the unit of prefecture serves as
a geographical unit that represents socioeconomic contexts commonly shared by the
residents. We investigate four characteristics of prefectures that are expected to
influence the context in which intergenerational coresidence takes place: economic
well-being, housing space, cultural milieu, and public welfare service utilization by
the aged. All indexes are converted to z-scores in our analysis to maintain a
comparability of scales in our analysis.
We utilize two socioeconomic indexes as prefecture characteristics. First one
indicates the financial state of the prefecture, called financial power index (zaisei-
ryoku shisu in Japanese). This index, calculated as the ratio of total financial income
to total financial demand (Development Bank of Japan, 2006), is produced every
year by the government to assess the financial independence of each prefecture. The
calculation considers the amount of tax income allocated to the prefecture, which for
instance includes income and property taxes paid by local industries and individual
consumers. Thus, the larger value would indirectly represent the strength and
vitality of local economy, as well as the richness in employment opportunities for
working age populations.
In general, coresidence arrangement also requires a larger space to
accommodate multiple generations in one household (Logan & Bian, 1999; Sodei,
103
1995). Thus as an estimate of the prefecture-level residential condition for
coresidence, we include the variable of housing space, which was indicated by the
average floor area in square-meters per household in each prefecture (one square
meter = 10.76 square-feet) as retrieved from the 2000 Japan Census (Statistical
Bureau at Ministry of Internal Affairs and Communications, 2006).
We represent the normative climate toward intergenerational coresidence in
each prefecture by calculating the average prefecture-wide response to the following
statement; “Children ought to live with their parents when their elderly parents can
no longer live by themselves,” retrieved from the National Family Research of Japan
in 1998 (NFRJ98). Responses are scored on a four-point scale, ranging from 1
(disagree) to 4 (agree). The NFRJ is a national family survey conducted by the
research group called National Family Research of Japan, and the survey contains
the data of 6,985 residents of Japan aged between 28 and 77 years old (National
Family Research of Japan, 2004). Since intergenerational coresidence is an
arrangement formed by two different generations, we retrieve data on prefecture
attitudes from this different national dataset, because we intend to capture prefecture-
level attitudes that span all age groups, not just older residents. As mentioned earlier,
the degree of traditionalism in cultural milieu tends to have a variation at the regional
level, in which multiple prefectures are grouped together. We consider this regional-
level factor to be only relevant to the cultural climate but not to other socioeconomic
contextual factors; thus, we represent all the contextual factors with the unit of
prefecture.
104
To measure public welfare service utilization by older residents, we use an
index that represents the availability of old-age welfare services at the prefecture-
level. The measurement, called local caregiving power index (chiiki-kaigo-ryoku in
Japanese), provides an estimate for the strength and richness of local welfare support
for older residents in the community (Sumitomo Life Insurance Research Institute,
1998, 2001). This index is a standardized score based on the average days of
utilization for the three types of elder care services (home care service, daycare
service, and short-stay service) per 100 older individuals aged 65 years and older in
each prefecture.
Statistical Model
This study aims to investigate the relationship between individual-level and
prefecture-level characteristics and intergenerational coresidence. For this purpose,
we implement a multinomial logistic regression within a hierarchical modeling
approach. This method is also called hierarchical generalized linear model (HGLM),
which transforms the outcome variable without a normal distribution by using a logit
function in a multi-level structure (Luke, 2004; Raudenbush & Bryk, 2002). In this
method, the likelihoods of two types of coresidence are separately estimated at two
levels: the first level of individuals and the second level of prefecture. We estimate
three models by sequentially including individual-level predictors, prefecture-level
predictors, and finally cross-level interactions. Since the focus of our analysis is to
estimate individual inferences embedded in prefectural contexts rather than their
“fit” to their prefectural averages, we select the method of grand-mean centering for
all the predictor variables (Luke, 2004). This centering method is implemented to
105
establish meaningful intercepts in the model. The equation at the first level is
specified as
mij = ß0j(m) + ß1(m) * (IC
1
) …+ ßz(m) * (ICz)
where mij represents the relative logits of category m (i.e., life-long coresidence or
“boomerang” coresidence) to the reference group (i.e., non-coresidence) for the
person i in prefecture j, ß0j(m) measures the random intercepts of the overall adjusted
log odds of type m coresidence within the j prefecture, and ß1 to ßz are the fixed-
effect coefficient of Z individual characteristics (IC).
The second model adds a level-2 equation in which the overall incidence of
each type of coresidence is estimated as a function of prefecture-level characteristics.
The equation at the second level is thus specified as follows.
ß0j(m) = G00(m) + G01(m) * (PC
1
)+… G0k(m) * (PC
k
) + u0j(m)
where for each contrast with the type m coresidence, G00(m) represents the constant,
G01(m) to G0k(m) are the effects of K prefecture characteristics (PC), and u0j(m)
represents the residual effects in each contrast of m.
The third model tests a cross-level interaction, in which we allow level-1
effects of filial responsibility to vary across prefectures along with the intercepts
(other covariates were estimated as fixed effects). Thus, at level 2, we estimate both
equations above and the following equation.
ß1j(m) = G10(m) + G11(m) * (PC
1
)+… G1k(m) * (PC
k
) + u1j(m)
where ß1j(m) are prefecture-level contextual effects of older individual’s normative
attitudes on each of the m residential contrasts, G10(m) represents the constant, G11 to
G1k are the contextual effects of prefecture characteristics, u1j represents the residual
106
term. In this equation, our interest is in identifying how prefecture attributes modify
the influence of personal attitudes on coresidence outcomes.
5.4 Results
First, we present the characteristics of the sample (Table 11). Slightly more
than half (57%) were women, somewhat more than one-third (38%) were widowed,
and the 30% had at least a high-school education. Most of the respondents (90%)
owned their home, had a household income between 1.5 million Japanese yen (about
12,500 US dollars) and 3 million Japanese yen (about 25,000 US dollars) on average,
and about 60% lived in an urban area. The average age of the sample was 75 years
old. The respondents had about three children on average (2.6), the majority (78%)
had at least one son, and slightly less than half of them had at least one child living in
the same town or city. Adult children’s average level of educational attainment
corresponded to a value signifying more than high-school but less than a college
graduate.
107
Table 11. Sample Characteristics
Variables Range/Categories Mean/Percentage SD
Living Arrangements Does not live with children 42.6
Life-long coresidence 44.7
Boomerang coresidence 12.7
Age 65-99 years 74.7 6.1
Number of children 1-11 children 2.6 1.2
Income 1 (less than $4160) to 13
(more than $125, 000)
4.6 2.3
Gender Female 56.8
Male 43.2
Home ownership Owns home 90.1
Does not own home 9.9
Education Junior high school 62.5
High school 29.4
Vocational school/Junior
college
2.8
4-year university/Graduate
school
5.3
Junior high school (1-1.5) 7.5
High school (1.5-2.5) 44.3
Vocational school/Junior
college (2.5-3.5)
24.6
Educational level of
children (averaged by
parent)
4-year university/Graduate
school (3.5-4.0)
23.6
Rural residence Lives in rural area 40.7
Live in urban area 59.3
Widowhood Widowed 37.8
Not widowed 62.2
IADL problem Has at least one difficulty 11.4
Has no difficulty at all 88.6
Son availability Has at least one son 78.5
Has no son 21.5
Yes 46.1 Non-coresident child
lives in the same town? No 53.9
Agree 37.7
Somewhat agree 26.6
Somewhat disagree 16.8
Filial responsibility of
children
Disagree 18.9
108
Table 12 presents means, standard deviations, and ranges of the four
prefecture characteristics in their original metrics. Average sample size for each
prefecture was 77, ranging from 21 to 287. The range of value for each variable
indicates a significant amount of variability among prefectures in socioeconomic and
cultural characteristics. Our preliminary analysis of these contextual characteristics
and demographical compositions at the prefecture level (results not shown) identified
a positive correlation between the stronger economic climate and the proportion of
the working population aged between 15 and 64 years old in the prefecture,
confirming the critical role played by the younger population for maintaining the
financial vitality at the prefecture-level. We note that individual data were available
from all 47 prefectures, thus all prefectures were represented in our study.
Table 12. Contextual Characteristics of 47 Prefectures in Japan
Table 13 presents logit coefficients and converted odds ratios from the
multinomial logistic regression that includes only individual characteristics. These
Contextual
Variables
Measurement Minimum Maximum Mean SD
Local welfare
availability
Per capita use of
services in elderly
population
40.8 57.3 50.2 4.6
Housing space Square meters 63.0 146.4 103.8 18.3
Economic well-
being
Financial power
ratio
.22 1.08 .46 .20
Traditional
normative climate
Attitudes toward
coresidence
2.4 3.0 2.6 .14
109
results show that older mothers were 20% more likely than older fathers to live with
a child in a life-long coresident arrangement. In terms of need, widowed parents
were almost twice as likely to live in either type of household, but IADL limitation
was only associated with boomerang coresidence. Resources also played a role in
the likelihood of coresidence. Older parents who owned a home were three to five
times more likely than renters to find themselves in coresidential arrangements of
either type. Income played no role in the likelihood of coresidence.
110
Table 13. Multinomial Hierarchical Linear Model with Individual Predictors
(N=3565).
Living Arrangements
Life-Long Coresidence
a
Boomerang
Coresidence
a
Individual characteristics Logit OR
b
Logit OR
b
Age -.009 .99 .011 1.01
Female .19 1.21* .20 1.22
Widowhood .64 1.90*** .77 2.16***
Rural residence .34 1.40* -.26 .77
Home ownership 1.59 4.90*** 1.13 3.10***
Household income .012 1.01 .013 1.01
IADL problem .043 1.04 .53 1.70***
Education -.16 .85* -.098 .91
Filial responsibility .20 1.22*** .064 1.07
Number of children .15 1.16*** .27 1.31**
Children’s education -.49 .61*** -.32 .73***
Son availability .56 1.75*** .14 1.15
Non-coresident child in same city -.86 .42*** -1.06 .35***
Intercept -.072 .93 -1.27 .28***
Level-2 Random Effects
Variance component (Intercept) .47*** .12*
*P<.05; **p<.01; ***p<.001
a
Reference = Non-Coresidence
b
OR = Odds Ratios
111
The cultural variables significantly affected the likelihood of life-long
coresidence. Parents who expressed more traditional view towards the filial
responsibilities were more likely to live in this traditional style of coresidence than
those with less traditional attitudes. Similarly, older parents with a lower level of
educational background were more likely to have the traditional style of life-long
coresidence than those higher educated respondents. Rural residents, typically more
traditional than their urban counterparts, were 40% more likely to live in life-long
arrangements with their children.
Child characteristics also affected the likelihood of coresidence. Parents with
more children, less educated children, and those who had no child living
independently in the same town or city tended to coreside with children in a life-long
or a boomerang arrangement. Parents with at least one son were 75% more likely
than those parents without a son to live in a life-long arrangement.
We introduce prefecture variables to the models in Table 14. Results for the
main effects are shown in the first two equations. We see in Table 14 that with the
introduction of contextual variables, the effects for individual-level variables (as
level-2 intercepts) change little from the previous table because the prefecture-level
variables, as standard scores, are mean-centered. Examining the influence of four
contextual variables, we observe that elderly residents of prefectures with more
spacious housing on average are more likely to coreside with children on a life-long
basis (67% more for each standard unit change in housing space). Older residents
living in prefectures with stronger economies were more likely than those in weaker
economic contexts to engage in both types of coresidence. The effect is substantially
112
higher for life-long coresidence, with a 58% increase for every standard unit increase
in financial power index. The availability of local welfare resources tended to
reduce the likelihood of both types of coresidence. Living in a more service-rich
prefecture caused life-long coresidence and boomerang coresidence to decline by
16% and 13% respectively, for a standard score increase in service utilization. The
normative climate of a prefecture had no influence on the overall likelihood of either
type of coresidence.
113
Table 14. Multinomial Hierarchical Linear Models with Individual and Prefecture Predictors (N=3565)
Main Effects Model Cross-level Effects Model
Life-Long
Coresidence
a
Boomerang
Coresidence
a
Life-Long
Coresidence
a
Boomerang
Coresidence
a
Individual characteristics Logit OR
b
Logit OR
b
Logit OR
b
Logit OR
b
Age -.010 .99 .011 1.01 -.010 .99 .012 1.01
Female .18 1.20* .20 1.22 .18 1.20* .20 1.22
Widowhood .65 1.92*** .77 2.16*** .65 1.92*** .78 2.18***
Rural residence .34 1.40* -.24 .79 .34 1.40* -.24 .79
Home ownership 1.59 4.90*** 1.14 3.13*** 1.59 4.95*** 1.15 3.16***
Household income .011 1.01 .013 1.01 .010 1.01 1.01
IADL problem .054 1.06 .55 1.73*** .048 1.05 .55 1.73***
Education -.15 .86* -.094 .91 -.15 .86* -.095 .91
Filial responsibility .20 1.22*** .064 1.07 .22 1.23*** .11 1.10*
Number of children .23 1.26*** .27 1.31** .22 1.25*** .26 1.30**
Children’s education -.50 .61*** -.32 .73*** -.49 .61*** -.32 .73***
Son availability .56 1.75*** .14 1.15 .57 1.77*** .14 1.15
Non-coresident child in same city -.87 .42*** -1.06 .35*** -.87 .42*** -1.07 .34***
114
(Table 14, continued)
Level 2: Prefecture Characteristics
Housing .52 1.68*** .10 1.10 .52 1.68*** .12 1.13
Local finance .46 1.58*** .16 1.17* .47 1.60*** .18 1.20*
Normative climate .065 1.07 .095 1.10 .063 1.01 .088 1.09
Public service availability -.18 .84* -.14 .87* -.18 .84* -.14 .87*
Cross-level interactions
Filial responsibility by housing .027 1.03 -.061 .94
Filial responsibility by local finance -.018 .98 -.057 .94
Filial responsibility by normative climate .050 1.03 .10 1.11*
Filial responsibility by public service
availability
.028 1.05 .11 1.12*
Intercept -.096 .91 -1.30 .27*** -.098 .91 -1.31 .27***
Level-2 Random Effects
Variance Component (Intercept) .17*** .08** .17*** .07*
Variance Component (Filial responsibility) .013 .007
*p<.05 **p<.01 ***p<.001
a
Reference = Non-Coresidence; OR
b
= Odds Ratios
115
Cross-level interactions between values of filial responsibility and the four
contextual variables are introduced in the last two equations of Table 14. Two such
interactions are found, both with respect to boomerang coresidence. As described in
the Figure 8, living in a more traditional normative climate increased the strength
with which older parent’s values toward the filial responsibility increased the
likelihood of boomerang coresidence. In relatively non-traditional prefectures, the
predicted probabilities of boomerang coresidence tended to stay the same level
among older individuals with varying degree of normative attitudes. On the other
hand, in the more traditionally-oriented prefectures, older parents with the most
traditional attitudes were much more likely to live with their children than those
parents with the least traditional attitudes.
10.0%
15.0%
20.0%
25.0%
30.0%
12 3 4
Older Parent's Normative Attitudes
Highly traditional prefecture (more than one SD above the mean)
Moderately traditional prefecture (mean)
Non-traditional prefecture (more than one-SD below the mean)
Figure 8. Predicted Probability of Boomerang Coresidence by Older Parent’s
Normative Attitude and Cultural Climate of Prefecture
116
The second interaction is with respect to public welfare service utilization.
The relationship between traditional filial values and living with a boomerang child
was increasingly more positive for older parents residing in prefectures with greater
availability of welfare services (Figure 9). The graphs suggest that in the prefectures
that were low in welfare service utilization, the likelihood of boomerang coresidence
stayed the same level regardless of older parents’ normative attitudes, maintaining
the overall higher likelihood of this type of coresidence comparing with those in the
prefectures with relatively high levels of welfare resources. On the other hand,
within the prefectures of high levels of welfare service utilization, the probability of
boomerang coresidence became progressively higher as the older parent’s normative
attitudes toward filial responsibilities became more traditional.
10.0%
15.0%
20.0%
25.0%
12 34
Older Parent's Normative Attitudes
Prefecture with high welfare availability (more than one SD above the mean)
Prefecture with modest welfare availability (mean)
Prefecture with low welfare availability (more than one SD below the mean)
Figure 9. Predicted Probability of Boomerang Coresidence by Older Parent’s
Normative Attitude and Welfare Resource Availability of Prefecture
117
The results of our conditional models highlight differences and
commonalities between the two types of coresidence. The unique predictors for the
life-long coresidence included older parents’ being a female, living in a rural area,
having a lower level of education, and having at least one son. Holding a traditional
belief in adult children’s filial obligations was a unique predictor for the life-long
coresidence before the interaction terms were introduced, but it became a significant
predictor for both types of coresidence in our final model. On the other hand, the
predictor that was significant only for the “boomerang” coresidence was older
parents’ having a difficulty with their IADL abilities.
The variance component of the level-2 random effect represents the residual
or unexplained variation at the prefecture level (unlike hierarchical linear models
there is no level-1 residual variance). We note that in Table 13, before we added
prefecture-level variables, residual variance in the two contrasts were .47 and .12.
These figures reduced to .17 and .08 respectively when contextual characteristics
were added in Table 14. Although these figures do not necessarily provide a
straightforward measurement of residual variance reduction (Snijders & Bosker,
1994), they still suggest the fairly strong explanatory power of our prefecture-level
variables.
5.5 Discussion
This study examined the extent to which intergenerational coresidence of
older people in Japan is guided by norms, needs, and resources. In general, we
conclude (1) that functions of intergenerational coresidence in Japan entail
118
significant variability, which hinges on the ways in which the residential
arrangement were formulated by parents and children, and (2) that filial piety is less
a dichotomy (present or absent, strong or weak), than it is a malleable cultural script
that has different meanings and manifestations depending on individual and
contextual characteristics.
Our findings suggest that the traditional type of life-long coresidence tends to
be both a value-driven and resource-driven arrangement. The value-driven nature of
the life-long coresidence was clearly indicated by the result that this type of living
arrangement tends to be guided by older parent’s embracement of family norms. In
contrast, the non-traditional style of boomerang household appeared to be more
common with parents who have poor health, suggesting its implied function of
providing support on as-needed basis. The contrast between the two types of
coresidence arrangements implies that the etiology of coresidence is important to
consider in contemporary Japanese society, where family behavior is increasingly
motivated by the point-in-time emergence of need, and less by prescriptive
obligations regarding life-style (e.g., Koyano, 2003; Pimentel & Liu, 2005).
There were similarities between the two arrangements as well. For instance,
both types of coresidence tended to accommodate widowed parents. In boomerang
coresidence, the parental widowhood may trigger another sign of needs of support
from parents. On the other hand, in the traditional life-long coresidence, the loss of a
spouse by the coresiding parent may impose an extra pressure for coresiding children
to remain in their shared household. In addition, both living arrangements were
more common among parents with a relatively large number of children, but they
119
also had a tendency to be replaced with a separate living arrangement if the parent
had a proximate child living nearby.
Results also suggest that the sustenance of the traditional cultural practices
requires a sufficient level of socioeconomic resources. This was particularly
reflected by housing conditions. As the past studies suggested (Logan & Bian, 1999;
Raymo & Kaneda, 2003; Takagi & Silverstein, 2006), older parent’s home
ownership seems to be a critical condition for coresidence arrangements. At the
contextual level, larger amounts of housing space enabled families to live together on
a life-long basis—an arrangement more likely to entail the possibility of having a
three-generations in the same household. It is possible that healthier economies
produce work opportunities for younger families that keep them (or allow them to
be) coresident with their parents. Consistent with modified-modernization theory
(Inglehart & Baker, 2000), our results suggest that socioeconomic development
provides some younger generations the resources that enable them to fulfill their
perceived cultural duties.
On the other hand, more generous local welfare policies at the prefecture
level produced a reduction in coresidence, which suggests their potential role of
replacing families for supporting older parents in need. However, at the same time,
our analysis also shows that traditional values at the individual-level are more likely
to be expressed in boomerang coresidence in prefectures with greater welfare
provisions for older adults. This suggests that formal services and traditional
orientations are complementary with respect to coresidence decisions, which is
reflected in the principles underlying the “Japanese-style” welfare system that
120
emphasize the importance of partnerships between families and public welfare sector
in support of older individuals (Goodman & Peng, 1996). In this respect, local
welfare resources in Japanese communities may implicitly reinforce the contribution
of family support for needy older individuals.
There are several limitations to our investigation that merit discussion. First
of all, the cross-sectional nature of our dataset puts some limitations in our causal
inferences. The dataset on which our study is based offers the possibility of
longitudinal analysis. However, relatively few respondents changed their living
arrangements over the two years between the panels. Therefore, we decided to focus
only on the first wave of the data, which entailed the largest number of sample
respondents, before attrition produced selection biases. Confirmation of causal
mechanisms needs to await longer-term longitudinal data to become available.
We also note that detailed variables describing characteristics about co-
residing adult children, such as their gender, marital statuses, and health conditions,
were omitted from our analyses because they were not distributed well enough to
produce convergence in our multilevel models. We also did not have available direct
measurements of the values, economic needs, and health conditions of the adult
children themselves. In our analysis, however, for both types of coresidence, higher
educational attainment of adult children tends to decrease the overall likelihood of
coresidence. Although children’s educational level does not directly measure their
normative attitudes, this suggests the potential effect of children’s normative
orientations in practice of coresidence arrangements. While it is unclear how these
omissions of adult children’s characteristics affected our results, it seems likely that
121
additional variance will be explained with the introduction of these variables to our
model.
Finally, we note that our contextual variable of economic well-being entails
factors that may merit another independent examination in future research. For
instance, we originally considered the variable of housing costs to be another
important contextual factor (Mutchler & Burr, 2003); however, the average cost of
housing was highly correlated with other prefecture variables in our analysis,
particularly the indicator of economic well-being (r = .9)., thus we excluded it from
our final analysis.
Population aging and structural changes led by modernization demand
societies a reassessment of the roles of family for older generations, which involves
modifications of traditional family values and actual family arrangements, as well as
reconfigurations of the contracts between families and the social welfare state
(Hashimoto, 1997). Most of the previous research on aging and families have
focused on only a part of this complex mechanism among societal modernization,
intergenerational family tradition, and the welfare state in modern aging societies;
and they particularly lack consideration for the broader socio-cultural contexts where
families are embedded (Hagestad & Dannefer, 2001; Marshall, Matthews, &
Rosenthal, 1993). In this study, we attempted a distinctive approach to encompass a
multi-level factors to explain multiple and diverse trajectories that the family
traditions may follow in modern society. Our study demonstrated that
intergenerational family arrangements hinge on not only family circumstances and
individual cultural beliefs but also on socioeconomic and welfare resources at the
122
community-level. The family tradition embedded in a multi-level structure of
modern society thus follows complex and diverse trajectories to elastically remain
for older people and their families as well as for service providers.
123
CHAPTER VI Discussion
This dissertation examined the current status of the traditional culture of
intergenerational coresidence between older parents and adult children in
contemporary Japan, where the prevalence of intergenerational coresidence is
steadily declining and the influential effect of the cultural norm of filial piety is
dissipating. A number of studies have argued that the postwar modernization that
significantly transformed the structural environment of Japanese families has
reduced the feasibility and practicality of coresidence arrangements in Japan (Ogawa
& Retherford, 1993; Sodei, 1995). The decreasing rates of three-generation
coresidence are often viewed as the gradual disappearance of the traditional family
culture in modern society. However, if the definition of coresidence is extended to
include two-generation households, statistics suggest that coresidence is still a fairly
common living arrangement for Japanese older individuals – consisting of about half
of the households, indicating the ubiquity and the potentially overlooked diversity of
this type of living arrangement.
The purpose of this dissertation was to study the variability in coresidence
arrangements between Japanese older parents and their children and to investigate
the elastic nature of the cultural norm still implicitly existing in many of the
intergenerational living arrangements in Japan. The results based on nationally
representative data suggest that the culture of coresidence in Japan remains in
divergent forms, including both traditional styles that mirror the cultural norm of
filial piety and non-traditional styles that modify the normatively expected
124
arrangement to accommodate both the needs of the families and perceived normative
expectations. The analyses implemented in this study revealed that the seemingly
one type of living arrangement of coresidence significantly varies in its normative
orientations and functions, depending on the individual backgrounds, familial
conditions, and socioeconomic contexts where the coresiding older parents and adult
children are nested.
This chapter discusses the major findings of this study from the following
three points of view. First, cultural norms and values are not a stable entity; rather,
the norms are dynamic and fluid, containing a significant variability based on the
living experiences of older parents and their children. Second, intergenerational
coresidence in Japan, which used to be predominantly a value-laden living
arrangement, now entails divergent styles, some of which exist without the influence
of the traditional cultural norm. Third, sustaining the culture of coresidence requires
not only the embracement of the cultural norms of coresidence but also sufficient
levels of socioeconomic resources available to the family across multi-level social
dimensions, such as the socioeconomic conditions of individual older parents and
adult children, needs and resources negotiated among family members, as well as the
socioeconomic climate where coresiding parents and children are embedded.
The rest of this chapter discusses the findings from these three perspectives.
The discussion of findings concludes with the overarching theme of divergence in
the culture of family coresidence and its implications for the future studies on family
coresidence in aging societies.
125
6.1 Elasticity of Coresidence Culture
The first study of this dissertation examined the reciprocal relationship
between older parents’ normative beliefs in coresidence and actual coresidence
arrangements. The analysis revealed that the traditional style of coresidence with
married children tends to be promoted by older parents’ normative beliefs in
coresidence, but not by the non-traditional style of coresidence with unmarried
children. At the same time, older parents’ actual experience of living with adult
children, whether in a traditional form or in a non-traditional form, appeared to
strengthen older parents’ normative beliefs in multigenerational coresidence.
The analyses also revealed that parents’ normative beliefs depend on the
evaluation of their own family-life and on their educational background. Older
parents’ relatively high level of satisfaction with their family-life strengthened their
traditional attitudes toward coresidence. On the other hand, parents’ high level of
educational attainment promoted negative attitudes toward traditional coresidence
arrangements. These findings suggest that the sustenance of cultural norms is not
monolithic but malleable, depending on older parents’ family experiences and
relationships, and the extent to which they have been exposed to new ideas through
education.
The study also showed that significant portions of older adults in Japan
experience a gap between what they consider to be a normatively desirable living
arrangement and what they actually have as their family arrangement. The
disjunction between the expressed desire of coresidence and actual living
arrangement indicates the challenge of fulfilling the traditional family expectation in
126
modern Japanese society (White, 2002). In this context, the reproduction of cultural
norms in Japan may be carried out by a specific group of families, possibly by those
“cultural winners” who were successful in satisfying the social expectation for
intergenerational family arrangements. At the same time, there exists a space for
other “cultural losers” in accommodating the traditional expectation in somewhat
modified styles, which would subsequently create divergent family arrangements
(White, 2002). Findings suggest that the traditional norms and values of coresidence
elastically survive through diverse channels, creating a range of family arrangements
that vary by the extent to which the arrangement “successfully” accommodates the
normative principles of coresidence.
6.2 Traditional and Non-Traditional Styles of Coresidence
The three studies implemented in this dissertation presented different ways of
contrasting traditional styles of coresidence and non-traditional forms of coresidence.
Specifically, the studies in this dissertation differentiated coresidence arrangements
by: 1) the marital status of coresiding adult children; 2) household headship; and 3)
the timing of coresidence. Across the three studies, the traditional form of
coresidence was delineated as the living arrangement that reflects the traditional
cultural norm of filial piety, where older parents reside with their married children
and play a primary role in household management, and where parents and children
share a household on a life-long basis. On the other hand, non-traditional
coresidence was contrasted as an arrangement in which the traditionally expected
arrangement was modified, such that parents reside with unmarried children, adult
127
children manage the household instead of older parents; and parents and children live
together on a contingent basis.
Each contrast between traditional and non-traditional styles of coresidence
suggested that some portion of coresidence arrangements are traditional in the sense
that they are primarily guided by the cultural norm of coresidence and filial
responsibility, while significant portions of coresidence are non-traditional because
of their loosened link with cultural norms and their closer connection with the
imminent needs and resources of the family. For instance, the first study (Chapter 3)
suggested that older parents’ endorsement of coresidence as ideal tends to have a
substantial influence on the heightened likelihood of the traditional form of
coresidence with married children; however, the non-traditional form of coresidence
with unmarried children did not appear to be affected by older parents’ normative
attitudes. Also, the second study (Chapter 4) indicated that the traditional style of
parent-headed coresidence tends to be more common among families with older
parents who hold more traditional attitudes toward filial responsibilities. On the
other hand, the counterpart of child-headed coresidence did not have any significant
influences from older parents’ normative attitudes. Similarly, the third study
observed a greater influence of older parents’ normative attitudes on the traditional
style of life-long coresidence than on the non-traditional type of boomerang
coresidence.
These contrasts suggest that the embracement of the cultural norm still exerts
an influential effect on the sustenance of actual coresidence with adult children,
particularly those traditional styles of coresidence. On the other hand, there seems to
128
exist significant proportions of intergenerational coresidence that have been arranged
regardless of older parent’s attitudes toward the cultural norm of filial piety. Instead,
non-traditional styles of coresidence are generally driven by the needs of older
parents, particularly those parents with poor health and those with no living spouse
available.
As discussed by some scholars before (Koyano, 1999; Martin & Tsuya, 1994),
the differences between the traditional and non-traditional styles of coresidence
indicate the potential shift of intergenerational living arrangements in Japan from
value-laden, preventive living arrangements for the expected needs of older parents
to the contingent arrangement in which coresidence is considered as the need for
support arises in the family. The cross-sectional nature of the studies in this
dissertation are limited in proving whether the identified contrasts between the
traditional and the non-traditional forms are the reflections of the substantial shift in
the functions and formations of coresidence in Japan. However, the snap-shot of the
variation in coresidence arrangements presented in this dissertation shows that within
one form of coresidential living arrangement between older parents and adult
children, different styles and functions are present, some of which have already
departed from the normatively expected living arrangement.
6.3 Multi-Level Determinants of Coresidence Culture
One of the goals of this dissertation was to examine the ways in which family
and social contexts affect older individuals’ living arrangements with adult children.
For this purpose, a multi-level approach was implemented across the two separate
129
studies; one addressed the family contexts that potentially influence married adult
children’s likelihood of coresidence and the other investigated the community
contexts that either enable or disable older parents to have a coresidence arrangement
with children. The results of these studies suggest that the sustenance of the
coresidence culture, whether it is in a traditional form or in a non-traditional style,
hinges on the level of socioeconomic resources available to the family. In other
words, in contemporary Japanese society, the embracement of the traditional
coresidence culture itself is not enough to maintain the intergenerational coresidence
practice.
The analysis of family contexts showed that across multiple types of
coresidence, the process of deciding which child will coreside with the parent
involves both conventional culture and the socioeconomic background of adult
children. For instance, across the studies, findings consistently suggested that
Japanese coresidence entails the conventional norm of son preference for the
coresiding child, which is quite common in East Asian societies (Logan et al., 1998;
Raymo & Kaneda, 2003). Coresidence in Japan also tends to be more prevalent
among self-employed families, indicating the existence of another sub-group within
the Japanese society that may hold a unique culture of intergenerational living
arrangements (Raymo et al., 2004).
On the other hand, the study also observed negotiating factors within the
family context that are potentially considered as alternatives to coresidence.
Particularly, the presence of single children and the availability of adult children who
live near to their parents appeared to substitute for some coresidence arrangements of
130
married children. The rise of alternatives suggest that the etiology of coresidence
needs to be understood in terms of the entire family context which often contains
subsidiary persons providing support to older parents.
Riley and Riley (1993) previously introduced the concept of the latent kin
matrix, in which family members located in diverse positions are called upon for
assistance when a substantial need becomes realized by the family. The analysis of
family context and coresidence in this study indicates the existence of such a latent
kin matrix in Japanese intergenerational living arrangements as well, where family
members from different generational positions become involved in negotiating a
coresidence arrangement. The diminished power of traditional norms and increased
space for individual negotiations in intergenerational living arrangements in Japanese
society indicates the heightened importance of other family members who did not
used to play a primary role in creating a coresidence arrangement.
The third study of coresidence and regional variation examined another level
of determinants that potentially create diversity in coresidence arrangements. This
analysis utilized the sub-national unit of prefecture as the geographical unit and
investigated the sources of regional variation in the likelihood of coresidence
arrangements for older parents. The results revealed that the overall likelihood of
coresidence tends to be higher in the prefecture with a relatively strong economic
climate, which indicates that the sustenance of coresidence requires sufficient levels
of socio-economic resources afforded by the community, particularly for younger
generations who need to maintain their own finances while living with their parents.
In contrast, the findings showed that residing in a prefecture that is relatively rich in
131
welfare resources for older residents tends to decrease the likelihood of
intergenerational coresidence, suggesting that the enriched welfare services in the
community may serve as a substitute for families who would otherwise have chosen
to have a coresidence arrangement for their older parents (Goodman & Peng, 1996;
Kono, 2000).
At the same time, the third study also discovered that the extent to which
older parents’ normative attitudes affect the likelihood of non-traditional boomerang
coresidence varies, depending upon the cultural milieu and welfare environment of
the prefecture where older parents and adult children reside. According to the
analyses, living in a relatively traditional cultural environment, as well as residing in
a prefecture with rich welfare services, tends to reinforce older parents’ normative
inclinations for reunification with adult children in a shared household. The
identified cross-level interaction in this part of the study suggests that the modified
tradition of coresidence in Japan is arranged in tandem with the cultural and welfare
contexts in which families are embedded. In Japanese contexts, decision making for
parents and children to form a coresidence arrangement later in their life is still
influenced by the traditional culture that is either implicitly exigent in the community
atmosphere or subtly implied in the current “Japanese-style” welfare services.
6.4 Elastic and Divergent Family Culture
The findings of this study have delineated two overarching themes: elasticity
of the traditional family culture and diversity of intergenerational family
arrangements. The declining prevalence of traditional multigenerational households
132
has been interpreted as the diminished, and possibly disappearing, effect of the
underlining cultural norm of filial piety in modern Japanese society (Ogawa &
Retherford, 1993). Although it is indisputable that the influence of the traditional
family norm in actual living arrangements has become weakened in contemporary
Japanese society, the analyses in this study suggested that the significant numbers of
coresidence arrangements are still linked with the normative values embraced by the
family. However, the ways in which the cultural norm remains effective in actual
living arrangements vary depending on a variety of factors, such as the type of
coresidence arrangement parents and children end up with, the family and
educational experiences of older individuals and their children, and the familial and
community contexts in which the families are nested. This indicates that the
traditional cultural norm remains in society in elastic styles, even in a highly
modernized society like Japan where social structures often conflict with the survival
of culturally expected family arrangements.
The elasticity of cultural norms in modern Japanese society also echoes the
notion raised by the modified modernization theory (Inglehart & Baker, 2000),
particularly in its thesis that the traditional cultural values do not simply disappear as
the society becomes modernized; rather, the ways in which the cultural values either
remain intact, modified, or dismissed, depends upon the socio-historical background
of the society. In case of the norm of filial piety in Japan, possibly because of its
long-term ubiquity in virtually every aspect of intergenerational family arrangements
(Koyano, 1996), extensive efforts have been made by many Japanese families to
fulfill the cultural expectations prescribed by the norm (White, 2002). Whether the
133
modifications of the cultural norm take a significant toll on the families in Japan is
beyond the scope of this dissertation. However, the disjunction between the desire
for coresidence expressed by older parents and actual living arrangements suggest
that some older individuals may be in a dilemma about not being able to achieve a
normatively expected living arrangement with their children.
Another theme that came out of this research is diversity in family
arrangements. Diversity in family formations and relationships is one of the
emerging themes in recent literature on the family (Allen, Blieszner, Roberto,
Fransworth, & Wilcox, 1999; Coontz, 2000; Johnson, 1996; Lowenstein & Bengtson,
2003). This literature argues that the ways in which a family accommodates their
relationships and household arrangements significantly varies with multiple social
dimensions, such as the cultural values shared by the family and the nesting
socioeconomic environment that either encourages or restricts the availability and
necessity of family support. The findings from this study indicate the substantial
diversity in intergenerational family arrangements even within a relatively
homogeneous cultural environment of Japan, where there is relatively little
variability in ethnic composition and normative orientations. The implications of
these findings for more ethnically diverse national contexts such as the United States
should be considerable because of their substantive importance in acknowledging
and understanding the tremendous amount of ethnic and cultural diversity in family
arrangements. As this study clearly presented, there tends to be a significant
relationship between the availability of community resources and family living
arrangements, suggesting that a better grasp of diversity in family formations across
134
generations is particularly important for establishing social services that are capable
of attending to the needs and demands that different groups of families may be
dealing with in their own ways.
6.5 Implications for Future Studies
The findings of this study entail mainly four implications and agendas for
future research. First, future studies need to overcome the limitations imposed by the
cross-sectional nature of the datasets. This study has delineated diverse and
distinctive patterns that currently exist among coresidence households arranged by
older parents and adult children in Japan; it successfully provided information about
the ways in which some older individuals and their families in Japan have already
diverged themselves from the traditional, value-laden living arrangement. In fact,
the shift in coresidence functions speculated about in this study correspond with the
findings of previous longitudinal studies (e.g., Brown et al., 2002) that have
empirically shown that the poor health of older parents serve as an initiating factor of
coresidence arrangements. It should be noted here that estimates for need-based
aspects of coresidence in this study are generally underestimated because the cross-
sectional nature of the analyses were unable to detect either past or future transitions
to coresidence that may have occurred with some of the older respondents currently
categorized in the non-coresidence group. In other words, the estimated impact of
older parents’ health conditions on the likelihood of coresidence could be more
substantial than presented in this study. More elaborate longitudinal analyses need to
wait for the longer periods of longitudinal data in order to empirically observe
135
whether there is a substantial shift from the traditional type to the non-traditional
type of coresidence in Japanese society.
Second, for a more accurate comprehension of normative beliefs and family
living arrangements, future studies need to incorporate the normative attitudes of
both older parents and adult children in the same family. Although a part of this
study included measurements of children’s educational levels as proxies for their
normative attitudes, the potential bias that might have occurred as a result needs to
be examined by using more direct and comparable measurements. For instance, as
previous studies have demonstrated, evaluation and assessments of family
relationships tend to differ depending on the kinship position (Aquilino, 1999;
Giarrusso, Stallings, & Bengtson, 1995; Raley & Rindfuss, 1999). The generational
disagreement between older parents and adult children as such may also be
applicable to the preference and evaluation of coresidence arrangements (Cutler et al.,
1999; Eto, 2001; Ting & Chiu, 2002). Future studies thus need to include normative
attitudes from both generations and address questions about how generational
differences in normative attitudes within families affect the actual intergenerational
household arrangements.
Third, future studies need to incorporate more detailed data on the help and
support actually exchanged between older parents and adult children in a coresidence
household. This study depicted non-traditional styles of coresidence as a need-
driven arrangement; yet, it is not clear about what sorts of support are actually
exchanged between older parents and adult children in coresidence settings. The
patterns of support exchanged may significantly vary depending on the types of
136
coresidence identified in this study. Future studies that address more specific aspects
of the social functions of each type of coresidence arrangement would greatly
enhance the understanding of the benefits and needs of this type of living
arrangement for older parents and adult children.
Finally, to place the results in a proper context, the results presented in this
study need to be examined in a cross-national framework. For instance, the ways in
which the traditional norm of filial piety is reflected in actual coresidence
arrangements may differ significantly in nations such as China and Korea that
embrace a similar cultural norm yet maintain distinctively different socio-political
environments (Eun, 2003; Lee, 2000; Logan et al., 1998; Pimentel & Liu, 2005). On
the other hand, in societies where coresidence is similarly prevalent yet families do
not share the same kind of cultural norm as Japanese families, such as southern
European countries (Daatland & Herlofson, 2003; Glaser & Tomassini, 2000),
individual motivations for coresidence may be different from what this study has
observed. To determine the generalizability of the results in this study to other
cultural and national contexts, comparable cross-national analyses are required.
6.6 Implications for Families in Aging Societies
Population aging in modern nations accompanied by socio-structural changes
facilitated by modernization processes requires a constant reevaluation of traditional
family culture. The findings of this study have provided insights about the elastic
nature of the cultural norms that are reflected in divergent styles of intergenerational
living arrangements of coresidence. Placing the results in larger context, this study
137
concludes with three implications for families in aging societies. The first
implication has to do with redesigning cultural values implied in social and welfare
policies for older populations. Hashimoto (1997) argues that aging populations in
society demand not only the restructuring of socio-political structures but also the
redefining of societal values to diversify its capacity to accommodate a variety of
needs and resources dealt with by both families and service providers. Cultural
values and beliefs do not stay intact across time; instead, they become modified and
transformed as the society changes its structural environment (Swidler, 1986). For
instance, in Japanese contexts currently only about 10 percent of older individuals
expect that they would depend on their children in future; this is a dramatic decline
from being almost 70 percent about five decades ago (Ogawa & Retherford, 1993).
While the results depicted significant departures from the traditional cultural
norm in some non-traditional styles of coresidence, this study also observed the
Japanese welfare contexts that tend to reinforce the prescription of traditional values
in the non-traditional type of boomerang coresidence. This indicates that in some
cases, the current welfare context in Japan may be putting extra pressure on families
to fulfill their traditionally expected responsibilities for older parents when such
norms are not well accepted by them anymore. In this context, some of the present
political and welfare systems may need to depart themselves from the traditionally
embraced norm of filial responsibility and seek a better understanding and
accommodation of new support styles, where both older parents and adult children
can comfortably involve themselves. This does not negate the importance of family
138
in lives of older parents; rather, it suggests a redefinition of the balance between
family and public sectors in modern Japanese society.
The second implication has to do with the roles expected of single adult
children for older parents, which may replace some of the responsibilities
traditionally assigned to married children. Throughout the analyses in this study,
unmarried children were depicted as either a child living in a non-traditional
coresidence for older parents in need or as an alternative figure to coresidence with
married children. In Japan, the number and prevalence of adult children who either
delay marriage or never marry has been gradually increasing (Raymo, 2003). The
younger cohorts of single adult children are arguably called “parasite singles” –
children who live in their parents’ house and use their parents’ financial resources
long after they finish their education (Lunsing, 2003; Raymo, 2003; Yamada, 1999).
Most of the adult children examined in this study belong to a somewhat older cohort
than the cohort of parasite singles. Thus, compared with the cohort of parasite
singles, unmarried adult children in this study may already have passed the age that
is typically expected for marriage and have maintained their single status in more
stable terms. Their relatively fewer family responsibilities may make these
unmarried children more readily available for aging parents who need a family
member with whom to share a household. In this respect, the gradually increasing
population of single adult children in Japanese society may create a new family
dynamic in the near future, in which single children play a significant role in
intergenerational household arrangements.
139
The third implication of this study is the potential emergence of new types of
cultural norms in intergenerational family arrangements. The results of this study
showed the loosened linkage between the traditional normative principles and non-
traditional types of coresidence. This suggests that some coresidence arrangements
may already be under the new social norm, which could be significantly different
from the traditional norm of filial piety. Obviously, societal values and norms do not
change overnight; rather, they gradually change as the cultural norm becomes poorly
fit with the individual desires and capacities of modern society (Swidler, 1986). The
gradual transformation and diversification of traditional cultural norms, which may
already be present in some of the living arrangements examined in this study, offers
an informative insight for policy makers and service providers about the emergence
of a potentially new social norm that guides a significantly large sub-group of older
individuals and their families in modern aging society. Socio-political and welfare
policies for older people and their families should be sensitive to such diversifying
social values that constitute the multiple styles of family arrangements that older
parents and children currently consider desirable and ideal.
140
Bibliography
Aboderin, I. (2003). "Modernization" and economic strain: The impact of social
change on material family support for older people in Ghana. In V. Bengtson,
L. & A. Lowenstein (Eds.), Global aging and its challenge to families (pp.
284-302). New York: Aldine De Gruyter.
Allen, K. R., Blieszner, R., Roberto, K. A., Fransworth, E. B., & Wilcox, K. L.
(1999). Older adults and their children: Family patterns of structural diversity.
Family Relations, 48, 151-157.
Aquilino, W. S. (1990). The likelihood of parent-adult child coresidence: Effect of
family structure and parental characteristics. Journal of Marriage and the
Family, 52, 405-419.
Aquilino, W. S. (1999). Two views of one relationship: Comparing parents' and
young adult children's reports of the quality of intergenerational relations.
Journal of Marriage and the Family, 61, 858-870.
Aquilino, W. S., & Supple, K. R. (1991). Parent-child relations and parent's
satisfaction with living arrangements when adult children live at home.
Journal of Marriage and the Family, 53, 13-27.
Bawin-Legros, B., & Stassen, J. F. (2002). Intergenerational solidarity: Between the
family and the state. Current Sociology, 50, 243-262.
Becker, G. S. (1991). A Treatise on the Family (2nd ed.): Harvard University Press.
Bengtson, V. L., Rosenthal, C. J., & Burton, L. M. (1995). Paradoxes of families and
aging. In R. H. Binstock & L. K. George (Eds.), Handbook of aging and the
social sciences (4th ed., pp. 253-288). San Diego, CA: Academic Press.
Brown, J., Winchester, Liang, J., Krause, N., Akiyama, H., Sugisawa, H., & Fukaya,
T. (2002). Transitions in living arrangements among elders in Japan: Does
health make a difference? Journal of Gerontology: Social Sciences, 57B,
s209-s220.
Brown, N. (2003). Under one roof: The evolving story of three generation housing in
Japan. In J. W. Traphagan & J. Knight (Eds.), Demographic change and the
family in Japan's aging society (pp. 53-71). New York: State University of
New York Press.
141
Bumpass, L., L. (1994). A comparative analysis of coresidence and contact with
parents in Japan and the United States. In L.-J. Cho & M. Yada (Eds.),
Tradition and change in the Asian family (pp. 221-246). Honolulu, Hawaii:
University of Hawaii Press.
Burr, J. A., & Mutchler, J. E. (1999). Race and ethnic variation in norms of filial
responsibility among older persons. Journal of Marriage and the Family, 61,
674-687.
Burr, J. A., & Mutchler, J. E. (2003). English language skills, ethnic concentration,
and household composition: Older Mexican immigrants. Journal of
Gerontology: Social Sciences, 58B, s83-s92.
Cabinet Office. (2001). Koreisha no seikatsu to ishiki ni kansuru kokusai hikaku
chosa (International comparison of life and opinions of older people).
Retrieved August 8, 2006, from the World Wide Web:
http://www8.cao.go.jp/kourei/ishiki/h12_kiso/html/0-1.html
Cabinet Office. (2004). Korei Shakai Hakusho (Annual Report on the Aging Society).
Tokyo, Japan: Gyosei.
Cabinet Office. (2005). Korei-shakai Hakusho (Annual Report on the Aging Society
2005) (in Japanese). Retrieved October 14, 2005, from the World Wide Web:
http://www8.cao.go.jp/kourei/whitepaper/w-2005/zenbun/17index.html
Cabinet Office. (2006). Korei-shakai Hakusho (Annual Report on the Aging Society
2006) (in Japanese). Retrieved January 31, 2007, from the World Wide Web:
http://www8.cao.go.jp/kourei/whitepaper/w-2006/zenbun/18index.html
Cherlin, A. (1994). The Japanese family in comparative perspective. In L. J. Cho &
M. Yada (Eds.), Tradition and change in the Asian family (pp. 421-433).
Honolulu, Hawaii: University of Hawaii Press.
Coontz, S. (2000). Historical perspectives on family studies. Journal of Marriage
and the Family, 62, 283-297.
Coward, R. T., Albrecht, S. L., & Shapiro, A. (1996). The perceptions of elderly
parents about the possibility of discontinuing their coresidence with adult
children. Research on Aging, 18, 325-348.
Crimmins, E., M., & Ingegneri, D., G. (1990). Interaction and living arrangement of
older parents and their children: Past trends, present determinants, future
implications. Research on Aging, 12, 3-35.
142
Cutler, L. J., Cho, J., Yust, B. L., & Franklin, E. M. (1999). Generational differences
in orientations to coresidence in Chongju, South Korea. Journal of Family
and Economic Issues, 20, 367-386.
Daatland, S. O., & Herlofson, K. (2003). 'Lost solidarity' or 'changed solidarity': a
comparative European view of normative family solidarity. Ageing and
Society, 23, 537-560.
Development Bank of Japan. (2006). Chiiki deta beisu (Local data base) (in
Japanese). Retrieved February 10, 2006, from the World Wide Web:
http://www.dbj.go.jp/japanese/download/database/main.html
Elder, H. W., & Rudolph, P. M. (2003). Who makes the financial decisions in the
households of older Americans? Financial Services Review, 12, 293-308.
Elliott, K. S., & Campbell, R. (1993). Changing ideas about family care for the
elderly in Japan. Journal of Cross-Cultural Gerontology, 8, 119-135.
Estes, C., L. (2003). Theoretical perspectives on old age policy: A critique and a
proposal. In S. Biggs & A. Lowenstein & J. Hendricks (Eds.), The need for
theory: Critical approaches to social gerontology (pp. 219-243). New York:
Baywood Publishing Company, Inc.
Eto, M. (2001). Public involvement in social policy reform: Seen from the
perspective of Japan's elderly-care insurance scheme. Journal of Social
Policy, 30, 17-36.
Eun, K. (2003). Changing roles of the family and state for elderly care: A Confucian
perspective. In V. L. Bengtson & A. Lowenstein (Eds.), Global aging and its
challenge to families (pp. 253-271). New York: Aldine De Gruyter.
Giarrusso, R., Stallings, M., & Bengtson, V. L. (1995). The "intergenerational stake"
hypothesis revisited: Parent-child differences in perceptions of relationships
20 years later. In V. L. Bengtson & K. W. Schaie & L. M. Burton (Eds.),
Adult intergenerational relations: Effects of societal change (pp. 227-263).
New York: Springer Publishing Company.
Glaser, K., & Tomassini, C. (2000). Proximity of older women to their children: A
comparison of Britain and Italy. The Gerontologist, 40, 729-737.
Goodman, R., & Peng, I. (1996). The East Asian welfare states: Peripatetic learning,
adaptive change, and nation-building. In G. Esping-Andersen (Ed.), Welfare
states in transition: National adaptations in global economies (pp. 192-224).
Thousand Oaks, California: Sage Publications.
143
Hagestad, G. O., & Dannefer, D. (2001). Concepts and theories of aging: Beyond
microfication in social science approaches. In R. H. Binstock & L. K. George
(Eds.), Handbook of aging and social sciences (5th ed., pp. 3-21). San Diego:
Academic Press.
Hamon, R. R., & Blieszner, R. (1990). Filial responsibility expectations among adult
child-older parent pairs. Journal of Gerontology: Psychological Sciences, 45,
p110-p112.
Hashimoto, A. (1996). The gift of generations: Japanese and American perspectives
on aging and the social contract. New York: Cambridge University Press.
Hashimoto, A. (1997). Designing family values: Cultural assumptions of an aging
society. Japan Quarterly, 44(4), 59-65.
Hayashi, F. (1995). Is the Japanese extended family altruistically linked? A test
based on engel curves. The Journal of Political Economy, 103, 661-674.
Heckman, J. J. (1979). Sample selection bias as a specification error. Econometrica,
47, 153-162.
Hirschman, C., & Minh, N. H. (2002). Tradition and change in Vietnamese family
structure in the Red River Delta. Journal of Marriage and the Family, 64,
1063-1079.
Inglehart, R., & Baker, W., E. (2000). Modernization, cultural change, and the
persistence of traditional values. American Sociological Review, 65, 19-51.
Ishikawa, H., & Maeda, D. (2000). Development of long-term care for elderly people
in Japan. In D. R. Phillips (Ed.), Aging in the Asia-pacific region: Issues,
policies, and future trends (pp. 133-157). New York: Routledge.
Izuhara, M. (2000). Changing family tradition: Housing choices and constraints for
older people in Japan. Housing Studies, 15, 89-110.
Izuhara, M. (2002). Care and inheritance: Japanese and English perspectives on the
'generational contract'. Ageing and Society, 22, 61-77.
Izuhara, M. (2004). Negotiating family support? The 'generational contract' between
long-term care and inheritance. Journal of Social Policy, 33, 649-665.
Jenike, B., Robb. (1997). Gender and duty in Japan's aged society: The experience of
family caregivers. In J. Sokolovsky (Ed.), The cultural context of aging:
Worldwide perspectives (pp. 218-238). Westport, Connecticut: Bergin &
Garvey.
144
John, R., Resendiz, R., & De Vargas, L. W. (1997). Beyond familism? Familism as
explicit motive for eldercare among Mexican American caregivers. Journal
of Cross-Cultural Gerontology, 12, 145-162.
Johnson, C., L. (1996). Cultural diversity in the late-life family. In R. Blieszner & V.
H. Bedford (Eds.), Aging and the family: Theory and research (pp. 307-331).
Westport, Connecticut: Praeger Publishers.
Jöreskog, K. G. (2005). Structural Equation Modeling with Ordinal Variables using
LISREL. Retrieved January 31, 2007, from the World Wide Web:
http://www.ssicentral.com/lisrel/techdocs/ordinal.pdf
Kawabe, H., & Shimizu, H. (1994). Japanese perceptions of the family and living
arrangements: The trend toward nuclearlization. In L.-J. Cho & M. Yada
(Eds.), Tradition and change in the Asian family (pp. 121-133). Honolulu,
Hawaii: University of Hawaii Press.
Kono, M. (2000). The impact of modernization and social policy on family care for
older people in Japan. Journal of Social Policy, 29, 181-203.
Koyano, W. (1996). Filial piety and intergenerational solidarity in Japan. Australian
Journal on Ageing, 15(2), 51-56.
Koyano, W. (1999). Population aging, changes in living arrangement, and the new
long-term care system in Japan. Journal of Sociology and Social Welfare, 26,
155-167.
Koyano, W. (2003). Intergenerational relationships of Japanese seniors: Changing
patterns. In V. L. Bengtson & A. Lowenstein (Eds.), Global aging and its
challenge to families (pp. 272-283). New York: Aldine De Gruyter.
Kumagai, F. (1996). Unmasking Japan today: The impact of traditional value on
modern Japanese society. Westport, Connecticut: Praeger.
Kumagai, F. (2005). Co-Residency living arrangement as a factor to facilitate labor
force participation of Japanese women: Community level analysis of
Japanese families. International Journal of Sociology and Social Policy,
25(3), 42-62.
Kurosu, S. (1994). Who lives in the extended family and why? The case of Japan. In
L.-J. Cho & M. Yada (Eds.), Tradition and change in the Asian family (pp.
177-198). Honolulu, Hawaii: University of Hawaii Press.
145
Lee, Y. J. (2000). Support between rural parents and migrant children in a rapidly
industrializing society: South Korea. In A. Mason & G. Tapinos (Eds.),
Sharing the wealth: Demographic change and economic transfers between
generations (pp. 282-305). New York: Oxford University Press.
Logan, J. R., & Bian, F. (1999). Family values and coresidence with married
children in urban China. Social Forces, 77, 1253-1282.
Logan, J. R., Bian, F., & Bian, Y. (1998). Tradition and change in the urban Chinese
family: the case of living arrangements. Social Forces, 76, 851-882.
Long, S., O. (1991). Family change and the life course in Japan. New York: East
Asia Program Cornell University.
Lowenstein, A., & Bengtson, V. L. (2003). Challenges of global aging to families in
the twenty-first century. In V. L. Bengtson & A. Lowenstein (Eds.), Global
aging and its challenge to families (pp. 371-377). New York: Aldine De
Gruyter.
Luke, D. A. (2004). Multilevel Modeling. Thousand Oaks, California: Sage
Publications.
Lundberg, S., & Pollack, R. A. (1994). Noncooperative Bargaining Models of
Marriage. American Economic Review, 84, 132-137.
Lunsing, W. (2003). 'Parasite' and non-parasite' singles: Japanese journalists and
scholars taking positions. Social Science Japan Journal, 6, 261-265.
Maeda, D., & Ishikawa, H. (2000). Ageing in Japan: Retirement, daily lives,
pensions and social security. In D. R. Phillips (Ed.), Aging in the Asia-pacific
region: Issues, policies, and future trends (pp. 113-132). New York:
Routledge.
Marshall, V. W., Matthews, S., & Rosenthal, C. (1993). Elusiveness of family life: A
challenge for sociology of aging. In G. Maddox & M. Lawton (Eds.),
Kinships, aging, and social change: Annual review of gerontology and
geriatrics (Vol. 13, pp. 39-72). New York: Springer Publishing Company.
Martin, L., G., & Tsuya, N., O. (1991). Interactions of middle-aged Japanese with
their parents. Population Studies, 45, 299-311.
Martin, L., G., & Tsuya, N., O. (1994). Middle-aged Japanese and their parents:
Coresidence and contact. In L.-J. Cho & M. Yada (Eds.), Tradition and
change in the Asian family (pp. 153-178). Honolulu, Hawaii: University of
Hawaii Press.
146
Masuzoe, Y. (2000). Socializing care for the elderly. Japan Echo, 27(3), 25-27.
Ministry of Education; Culture; Sports; Science; and Technology. (2006). Statistical
Abstract 2006 Edition:School Education. Retrieved January 17, 2007, from
the World Wide Web: http://www.mext.go.jp/english/statist/index.htm
Ministry of Health Labour and Welfare. (2000). Kosei Hakusho (White Paper:
Annual Report on Health and Welfare). Tokyo: Gyosei.
Morgan, S. P., & Hirosima, K. (1983). The persistence of extended family residence
in Japan: Anachronism or alternative strategy? American Sociological Review,
48, 269-281.
Mutchler, J. E., & Burr, J. A. (2003). Living arrangements among older persons: A
multilevel analysis of housing market effects. Research on Aging, 25, 531-
558.
National Family Research of Japan. (2004). Profile of NFRJ98 (in Japanese).
Retrieved September 18, 2004, from the World Wide Web:
http://www.waseda.jp/assoc-nfroffice/NFRJ98_profile.htm
National Institute of Population and Social Security Research. (2004). Ippan Jinko
Toukei (General Demographic Statistics) (in Japanese). Retrieved September
24, 2004, from the World Wide Web:
http://www1.ipss.go.jp/tohkei/Popular/Popular_f.html
O'Leary, J., S. (1993). A new look at Japan's honorable elders. Journal of Aging
Studies, 7, 1-24.
Ochiai, E. (1997). The Japanese family system in transition: a sociological analysis
of family change in postwar Japan. Tokyo, Japan: LTCB International
Library Foundation.
Ogawa, N., & Ermisch, J., F. (1996). Family structure, home time demands, and the
employment patterns of Japanese married women. Journal of Labor
Economics, 14, 677-702.
Ogawa, N., & Retherford, R., D. (1993). Care of the elderly in Japan: Changing
norms and expectations. Journal of Marriage and the Family, 55, 585-597.
Ogawa, N., & Retherford, R., D. (1997). Shifting costs of caring for the elderly back
to families in Japan: Will it work? Population and Development Review, 23,
59-94.
147
Osaka University of Commerce. (2003). JGSS Menu (English). Retrieved May 4,
2003, from the World Wide Web:
http://jgss.daishodai.ac.jp/english/index.html
Osaka University of Commerce, & University of Tokyo. (2002). Japanese social
surveys: JGSS-2001 Codebook. Tokyo, Japan: Institute of social science at
the University of Tokyo.
Palmore, E. B. (1975). What can the USA learn from Japan about aging? The
Gerontologist, 15, 64-67.
Palmore, E. B., & Maeda, D. (1985). The honorable elders revisited: A revised
cross-cultural analysis of aging in Japan. Durham: Duke University Press.
Peng, I. (2003). Pushing for Social Care Expansion: Demography, Gender, and the
New Politics of the Welfare State in Japan. American Asian Review, 21(2),
25-55.
Pezzin, L. E., & Schone, B. S. (1997). The allocation of resources in
intergenerational households: Adult children and their elderly parents. The
American Economic Review, 87, 460-464.
Pezzin, L. E., & Schone, B. S. (1999). Intergenerational household formation, female
labor supply and informal caregiving: A bargaining approach. The Journal of
Human Resources, 34, 475-503.
Pimentel, E. E., & Liu, J. (2005). Patrilineal coresidence in urban China: A life
course perspective. The Intergenerational Journal of Sociology and Social
Policy, 25(3), 63-91.
Raley, R. K., & Rindfuss, R., R. (1999). The effect of respondent's kinship position
on reported levels of integrational contact: USA and Japan. Population
Research and Policy Review, 18, 279-298.
Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical Linear Models: Applications
and Data Analysis Methods (2nd ed.). Thousand Oaks, California: Sage
Publications.
Raymo, J., M., Liang, J., Sugisawa, H., Kobayashi, E., & Sugihara, Y. (2004). Work
at older ages in Japan: Variation by gender and employment status. Journal
of Gerontology: Social Sciences, 59B(3), s154-s163.
Raymo, J. M. (2003). Premarital living arrangements and the transition to first
marriage in Japan. Journal of Marriage and the Family, 65, 302-315.
148
Raymo, J. M., & Kaneda, T. (2003). Changes in the living arrangements of Japanese
elderly: The role of demographic factors. In J. Traphagan, W. & J. Knight
(Eds.), Demographic change and the family in Japan's aging society (pp. 27-
52). New York: State University of New York Press.
Retherford, R., D., Ogawa, N., & Sakamoto, S. (1996). Values and fertility change in
Japan. Population Studies, 50, 5-25.
Riley, M. W., & Riley, J. W., Jr. (1993). Connections: Kin and cohort. In V.
Bengtson, L. & A. Achenbaum, W. (Eds.), The changing contract across
generations (pp. 169-189). New York: Walter de Guyer.
Rindfuss, R., R., & Raley, R. K. (1998). Intergenerational contact in the United
States and Japan. In K. O. Mason & N. O. Tsuya & M. K. Choe (Eds.), The
changing family in comparative perspective: Asia and the United States (pp.
193-213). Honolulu, Hawaii: University of Hawai'i Press.
Rindfuss, R. R., Liao, T. F., & Tsuya, N. O. (1992). Contact with parents in Japan:
Effects on opinions toward gender and intergenerational roles. Journal of
Marriage and the Family, 54, 812-822.
Rindfuss, R. R., Liao, T. F., & Tsuya, N. O. (1994). Contact with parents in Japan:
Effects on attitudes toward gender and intergenerational roles. In L.-J. Cho &
M. Yada (Eds.), Tradition and change in the Asian family (pp. 199-219).
Honolulu, Hawaii: University of Hawaii Press.
Royston, P. (2004). Multiple Imputation of Missing Values. Stata Journal, 4, 227-
241.
Schafer, J. L., & Olsen, M. K. (1998). Multiple imputation for multivariate missing-
data problems: A data analyst's perspective. Multivariate Behavioral
Research, 33, 545-571.
Shimizu, H. (2004). Family and support, Kazoku Kakumei (Family Revolution) (in
Japanese) (pp. 200-206). Tokyo, Japan: Kobundo.
Silverstein, M. (1995). Stability and change in temporal distance between the elderly
and their children. Demography, 32, 29-45.
Silverstein, M., Burholt, V., Wenger, G. C., & Bengtson, V. L. (1998). Parent-child
relations among very old parents in Wales and the United States: A test of
modernization theory. Journal of Aging Studies, 12, 387-409.
Snijders, T. A. B., & Bosker, R. J. (1994). Modeled Variance in Two-Level Models.
Sociological Methods & Research, 22, 342-363.
149
Sodei, T. (1995). Care of the elderly: A women's issues. In K. Fujimura-Fanselow &
A. Kameda (Eds.), Japanese women: New feminist perspectives on the past,
present, and future (pp. 213-228). New York: The Feminist Press.
Speare, A., Jr., & Avery, R. (1993). Who helps whom in older parent-child families?
Journal of Gerontology, 44(1), s64-s73.
Spilerman, S., & Elmelech, Y. (2003). Israeli attitudes about inter vivos transfers. In
V. L. Bengtson & A. Lowenstein (Eds.), Global aging and its challenge to
families (pp. 175-195). New York: Aldine De Gruyter.
Statistical Bureau at Ministry of Internal Affairs and Communications. (2006).
Population Census. Retrieved August 11, 2006, from the World Wide Web:
http://www.stat.go.jp/english/data/kokusei/2000/kihon1/00/13.htm
Sumitomo Life Insurance Research Institute. (1998). Chiiki kaigoryoku (Local
caregiving power) (in Japanese). Tokyo, Japan: Chuo Houki.
Sumitomo Life Insurance Research Institute. (2001). Chiiki kaigoryoku deta bukku
(Databook of Local Caregiving Power) (in Japanese). Tokyo, Japan: Chuo
Houki.
Sung, K. (1998). An exploration of actions of filial piety. Journal of Aging Studies,
12, 369-386.
Sung, K. (2000). An Asian perspective on aging East and West: Filial piety and
changing families. In V. L. Bengtson & K.-D. Kim & G. C. Myers & K.-S.
Eun (Eds.), Aging in East and West: Families, states, and the elderly (pp. 41-
56). New York: Springer Publishing Company.
Swidler, A. (1986). Culture in action: Symbols and strategies. American Sociological
Review, 51, 273-286.
Takagi, E., & Silverstein, M. (2006). Intergenerational Coresidence of the Japanese
Elderly: Are Cultural Norms Proactive or Reactive? Research on Aging, 28,
473-492.
Takagi, E., Silverstein, M., & Crimmins, E., M. (in press). Intergenerational
Coresidence of Older Adults in Japan: Conditions for Cultural Plasticity.
Journal of Gerontology: Social Sciences.
Thompson, C. S. (2003). Depopulation in rural Japan: "Population politics" in Towa-
cho. In J. Traphagan, W. & J. Knight (Eds.), Demographic change and the
family in Japan's aging society (pp. 89-106). New York: State University of
New York Press.
150
Ting, K. F., & Chiu, S. W. K. (2002). Leaving the parental home: Chinese culture in
an urban context. Journal of Marriage and the Family, 64, 614-626.
Traphagan, J. W. (2003). Contesting coresidence: Women, in-laws, and health care
in rural Japan. In J. W. Traphagan & J. Knight (Eds.), Demographic change
and the family in Japan's aging society (pp. 203-226). New York: State
University of New York Press.
Tsuya, N., O., & Martin, L., G. (1992). Living arrangements of elderly Japanese and
attitudes toward inheritance. Journal of Gerontology: Social Sciences, 47,
s45-s54.
United Nations. (2005). Living arrangements of older persons around the world.
New York: United Nations Publication.
USC/UCLA Center on Biodemography and Population Health. (2004). Nihon
University Japanese Longitudinal Study of Aging. Retrieved December 11,
2004, from the World Wide Web:
http://www.usc.edu/dept/gero/CBPH/nujlsoa/index.htm
Wada, S. (1995). The status and image of the elderly in Japan: Understanding the
paternalistic ideology. In M. Featherstone & A. Wernick (Eds.), Images of
aging: Cultural representations of later life (pp. 48-60). New York:
Routledge.
Walker, A. (1996). Intergenerational relations and the provision of welfare. In A.
Walker (Ed.), The new generational contract: Intergenerational relations,
old age and welfare (pp. 10-35). London: UCL Press.
Wallace, S. P., & Facio, E. L. (1992). Moving beyond familism: Potential
contributions of gerontological theory to studies of Chicano/Latino aging. In
J. F. Gubrium & K. Charmaz (Eds.), Aging, self, and community: A collection
of readings (pp. 207-224). Greenwich, Connecticut: JAI Press.
Ward, R., Logan, J., R., & Spitze, G. (1992). The influence of parent and child needs
on coresidence in middle and later life. Journal of Marriage and the Family,
54, 209-221.
White, M., I. (2002). Perfectly Japanese: Making families in and era of upheaval.
Berkeley and Los Angeles, California: University of California Press, Ltd.
Yamada, M. (1999). The age of parasite single. Tokyo, Japan: Chikuma Shobo (In
Japanese).
Abstract (if available)
Abstract
Intergenerational coresidence is one of the distinct types of living arrangements of older adults that reflects their socioeconomic conditions and the cultural values shared by their family and society at large. Although the prevalence of coresidence is decreasing in most of the developed nations in the world, Japan remains an anomalous case because of its relatively high prevalence of coresidence and the intricate connections between coresidence and the traditional cultural norm of filial piety.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Transmission and transitions: young adults' beliefs, values and life course transitions in familial context
PDF
The measurement, life course patterns, and outcomes of intergenerational ambivalence among parent-adult child dyads
PDF
Children 's migration and the financial, social, and psychological well-being of older adults in rural China
PDF
Normative obligations and parental care in social context
PDF
The relationship of expectation and satisfaction of filial piety with stroke-related outcomes of Chinese older adults in stroke rehabilitation
PDF
Internet communication use, psychological functioning and social connectedness at older ages
PDF
Intergenerational support between grandparents and grandchildren in rural China and its effect on the psychological well-being of older adults
PDF
The effects of time and space on health status of older adults in China
PDF
A study of depression disclosure among Latino older adults in Los Angeles
PDF
Diverging fertility patterns? Racial and educational differences in fertility behaviors and their implications for socioeconomic mobility
Asset Metadata
Creator
Takagi, Emiko
(author)
Core Title
Intergenerational coresidence of older adults in contemporary Japan: traditional cultural norms in divergent styles
School
College of Letters, Arts and Sciences
Degree
Doctor of Philosophy
Degree Program
Sociology
Publication Date
06/29/2007
Defense Date
02/27/2007
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
filial piety,intergenerational coresidence,Japanese elderly,living arrangement,OAI-PMH Harvest,socio-cultural context
Place Name
Japan
(countries)
Language
English
Advisor
Silverstein, Merril (
committee chair
), Crimmins, Eileen M. (
committee member
), Giarrusso, Roseann (
committee member
)
Creator Email
takagi@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-m568
Unique identifier
UC1233659
Identifier
etd-Takagi-20070629 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-507581 (legacy record id),usctheses-m568 (legacy record id)
Legacy Identifier
etd-Takagi-20070629.pdf
Dmrecord
507581
Document Type
Dissertation
Rights
Takagi, Emiko
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
Tags
intergenerational coresidence
Japanese elderly
living arrangement
socio-cultural context