Multigenerational family structure in Japanese society: impacts on stress and health behaviors among women and men

Abstract

Rapid population aging in Japan has led to rising demands for informal care giving. Traditionally, care giving for aging parents has fallen disproportionately on the shoulders of women living in multigenerational households. However, rising labor force participation by Japanese women, declining marriage and fertility rates, and women's changing expectations have combined to produce unprecedented strains on traditional multigenerational households where care giving to elders traditionally takes place. In this paper, we explored gender-specific relationships between family structure, stress and worries, and health behaviors, using linked data from two national surveys conducted in Japan: the 1995 Comprehensive Survey of the Living Conditions of People on Health and Welfare, and the 1995 National Nutrition Survey. We found that women in multigenerational households reported more care-giving worries, and also less future health and financial worries. Living with parents was associated with protective health behaviors (less smoking, less heavy drinking), but also more sedentary behavior among women, while men in “sandwich” families (i.e., living with both parents and children) reported heavier smoking. The association of family structure and health behavior was not mediated by worries. Living alone was associated with worse health for women. These findings suggest gender-specific patterns of worries and health behaviors that reflect both the health-protecting and health-damaging effects of living in multigenerational households.

Introduction

Industrialized countries throughout the world have witnessed a rapid growth in their older populations, but nowhere more quickly than in Japan. Among OECD countries in 1978, Japan had the lowest population share of people aged 65 years and older (just 8%), but by 2000, the country had the second highest share (17%) behind Italy (United Nations, 2002). Over the next half century, the old-age share of the Japanese population is projected to double, from 17% in 2000 to 33% in 2050 (Endo & Katayama, 1998, Chapter 5). No other nation in history has experienced a comparably rapid pace of population aging.

Unlike countries such as the United States, the aging of the population in Japan has little to do with the maturing of the “baby boom generation”, but is due principally to the precipitous decline in the fertility of Japanese women, combined with spectacular increases in longevity since the post-War period (Horlacher, 2002). In turn, the rapid decline in fertility among Japanese women is not because of falling marital fertility, but because of a combination of rising age at marriage, increasing proportion of women who never marry, and the rise in divorce rates. The population divorce rate in Japan more than doubled over a 35-year period, from about 80 per 1000 in 1960 to over 200 by 1995 (Retherford, Ogawa, & Matsukura, 2001). Reflecting these trends, the proportion of unmarried women during the peak childbearing years (25–29) rose from less than 10% in 1930 to 54% in 2000 (Kojima, 1995).

Meanwhile, the elderly population (aged 65 years and above) in Japan is also growing older due to improved survival rates. For the oldest of the old segment of the population, Manton and Vaupel (1995) have estimated that an 80-year old Japanese woman and man can expect to live for an additional 8.5 and 6.9 years, respectively. These figures contrast dramatically with life expectancy in 1935, when Japanese men and women could expect to live only to 47 and 50 years, respectively. The norm in pre-industrial Japanese society was for parents to die before they became frail and in need of care. Indeed, famous Noh plays and novels (such as “The Ballad of Narayama” by Shichiro Fukazawa) have depicted the tragedy of customs in some impoverished rural areas of Japan which demanded that elders who lived beyond a certain age be carried away by their sons to a remote mountain top and left to die. By contrast, the increasing longevity of seniors in Japan today implies long years of dependency and need for care. In 1987, the average length of time that families nursed an aging relative was 4 years. By 1997, it had grown to 7.4 years (Efron, 2001).

The pattern of care giving in Japan is highly gendered. Japanese tradition requires the wife of the oldest son to care for her husband's parents as well as disabled relatives until their death. Until recently, children who placed their parents into nursing homes often faced shame and community ostracism. Currently, 85% of those who care for elderly relatives are women. More than half of caregivers are themselves over 60 years. In short, there is a growing mismatch between the demand for long-term care, and its supply. Rapid population aging combined with changes in family structure (declining fertility, rising divorce rates, more women living alone) pose a unique set of challenges to Japanese society in terms of its future burden of dependency (Horlacher, 2002). Ogawa and Retherford (1997) have calculated the so-called familial support ratio, defined as the female population aged 40–59 divided by the population of both sexes aged 65–84. In Japan, this ratio was about 1.75 until 1960, but fell to 1.09 in 1995, and projected to decline further to 0.6 by 2025. By that time, there will be only one potential female middle-aged caregiver for every two elderly persons. Such projections, moreover, assume that Japanese society (and government) can continue to count on women to shoulder the burden of care giving to the elderly. Increasingly, this assumption is unlikely to be sustained.

Some observers have noted a dramatic change in Japanese citizens’ attitudes toward gender equality in recent decades. According to the latest nationwide opinion surveys conducted by the government in 2002, only 47% of respondents agreed with the statement, “a husband's place is at work, while a wife's place is in the home”. This contrasts with 57.8% who agreed with the view in 1997, and more than 70% in 1992 (Cabinet Office, Japan, 1992, 1997, 2002). On the other hand, the actual reality within households has lagged behind public opinion. The person who does the lion's share of domestic chores and care giving in Japanese households remains the woman (whether or not she is also in the paid workforce). The 2002 governmental surveys mentioned above also reported that more than 80% of wives carry out the majority of household tasks, and that this proportion has hardly budged during the past decade. For example, 84.5% of respondents answered that house cleaning was mainly done by the wife in the 2002 survey (82.4% in 1992 survey). The corresponding figures were 90.2% for washing (88.1% in 1992), 90% for preparing meals (87.3% in 1992), and 83% for cleaning after meals (80.9% in 1992) (Cabinet Office, Japan, 1992, 1997, 2002).

Some have noted a direct connection between the aging of Japanese society and its plummeting birth rate. Continuing societal expectations for women to provide the bulk of care giving to household members (whether as daughters-in-law, spouses, or parents of children) seems linked to their increasing preference to put off marrying at an early age, or perhaps not marrying at all. As Horlacher (2002, p. 50) has noted:

When Japanese women consider whether or not to marry, they realize that if they do, in middle age they may be forced to give up full time employment to care for an elderly parent. Over the next 35 years, the number of elderly in Japan who are frail, demented, or bedridden will increase by a factor of three. At the same time, there will be fewer middle-aged women to be caregivers. So the probability that a married woman may have to give up full-time employment to care for an elderly parent-in-law is increasing. Surely this must give pause to single women who are contemplating marriage.

Reflecting their rebellion against traditional norms, the traditional multigenerational household is declining in Japan (Strom, 2001). The proportion of eldest sons who resided with their parents fell from 65% in 1957 to 25% in 1998. According to the population census, multigenerational families declined by half, from 16.1% in 1970, to 13.9% in 1985, and 8.5% in 2000. Married couples co-residing with their parents also continued to fall from 18.1% in 1970, to 16.1% in 1985, and 10.8% in 2000. Meanwhile, the percentage of the households with persons living alone steadily climbed from 20.8% in 1985, to 25.6% in 1995, and 27.6% in 2000. The percentage of married couples living without other family members also nearly doubled over the same 30-year period, from 9.8% in 1970, to 13.7% in 1985, and 18.9% in 2000 (Ministry of Public Management, Home Affairs, Posts and Telecommunications, Japan, 1970–2000).

It is worth remarking at the outset that residing in a multigenerational household can be both a source of stress and reward. For women in the paid workforce, the presence of an elderly woman living in the same household is an important stress buffer in terms of the help that the latter often provide in performing domestic duties including child care (Yashiro et al., 1997). A recent nationwide survey found that more than 90% of wives under the age of 40 years had a living mother and mother-in-law, whether they lived with them or not (National Institute of Population and Social Security Research (1993), National Institute of Population and Social Security Research (1998)).

What are the psychological and physical health consequences for women (and men) confronting a rapidly aging society? As an extreme example of what many other Western societies will face in the coming century, Japan offers an interesting case study. Yet, aside from anecdotal evidence (such as the travails of nursing hell depicted in popular novels, such as “The Hateful Age” by Yasushi Inoue), there has been very little survey evidence published on the impacts of Japan's changing family structure. Some survey data provide information on the changing prevalence of co-residence with parents. Among women aged under 50 years, the rate of co-residing with parents has been decreasing (30.0% in 1993, 25.2% in 1998), although the same figures for women aged 45–49 years increased slightly (29.5% in 1993, 30.4% in 1998) (National Institute of Population and Social Security Research (1993), National Institute of Population and Social Security Research (1998)). Middle-aged women therefore seem to be at highest risk of juggling the demands of returning to the workplace (following years of child-rearing) and meeting expectations of providing care to aging parents.

The purpose of our study was to present evidence from a nationally representative survey on the health impacts of Japanese multigenerational families, for both women and men. Specifically, we sought to examine the psychological and physical health profiles of Japanese women and men living under different family circumstances. By family circumstance, we mean the diverse family structures, ranging from single women and men living alone, married couples with no children, married couples with children (the so-called nuclear family), to childless couples living with their parents, and so-called “sandwich” families consisting of multiple generations co-residing within a single dwelling.

The aims of this study were three-fold: (i) to examine the gender-specific prevalence of daily worries (e.g., about future health, care giving, and financial needs) associated with different kinds of family structure; (ii) to examine the gender-specific associations between various types of family structure and health-related behaviors, such as cigarette smoking, alcohol drinking, physical activity, and regular health check-ups; and (iii) to examine the associations between daily worries and health-related behaviors. We hypothesized that women who reside in a multigeneration family with her parents-in-law would be particularly vulnerable (through stress and worries) to deleterious health behaviors. We also hypothesized that both women and men residing in multigeneration families would exhibit higher worry levels (about health, finances, future care-giving needs), compared with families consisting of married couples alone, or single women and men living alone.

To explore these issues, we linked data from two major nationwide surveys separately conducted in Japan, i.e., the 1995 Comprehensive Survey of the Living Conditions of People on Health and Welfare (CSLCPHW) and the 1995 National Nutrition Survey (NNS). The questionnaires of these surveys provided information on family relationships within each household, self-reported stress situation as well as socioeconomic indicators at the household level (Hashimoto et al., 2 (1995), Ministry of Health, Labor and Welfare, Japan (1995)).