This Article Abstract Full Text (PDF) Services Download to citation manager Citing Articles Citing Articles via HighWire PubMed PubMed Citation

Work at Older Ages in Japan: Variation by Gender and Employment Status

¹ Department of Sociology, University of Wisconsin–Madison.
² School of Public Health, University of Michigan, Ann Arbor.
³ Obirin University, Tokyo, Japan.
⁴ Tokyo Metropolitan Institute of Gerontology, Japan.

Address correspondence to Dr. James M. Raymo, Department of Sociology, University of Wisconsin, 1180 Observatory Dr., Madison, WI 53706. E-mail: jraymo{at}ssc.wisc.edu

	Abstract

TOP Abstract Limitations of Existing... Institutional and Social Context... Methods Results Discussion References

 
Objectives. This study describes the correlates of labor force participation among Japanese men and women aged 60–85 and examines differences by gender and employment status.

Methods. Using four waves of data collected from a national sample of older Japanese between 1990 and 1999, we estimate multinomial logistic regression models for three measures of labor force participation (current labor force status, labor force exit, and labor force re-entry) as a function of individual and family characteristics measured 3 years earlier.

Results. Labor force participation is significantly associated with socioeconomic status, longest occupation, and family structure. The strength and nature of these relationships differ markedly for men and women and for wage employment and self-employment.

Discussion. The emphasis on life course experiences and work–family interdependence characterizing recent research on retirement in the United States is clearly relevant in Japan as well. To better understand later-life labor force participation in Japan, subsequent research should incorporate more direct measures of life course experiences and family relationships and attempt to make explicit cross-national comparisons of these relationships.

RAPID population aging, in conjunction with the trend toward earlier retirement among older men (Kohli, Rein, Guillemard, & van Gustern, 1991; Quinn & Burkhauser, 1995), is expected to threaten the viability of existing pay-as-you-go social security systems in most industrialized countries (Gruber & Wise, 1999). In response, many countries have shifted away from longstanding policy efforts to encourage earlier retirement in favor of efforts to encourage continued economic activity at older ages (Kohli et al., 1991; O'Rand & Henretta, 1999). This policy shift has stimulated comparative research on the economic, political, and social forces underlying regional differences in levels of labor force participation at older ages (e.g., National Research Council, 2001).

For several reasons, comparative studies of later-life work often include Japan (Gruber & Wise, 1999; Hurd & Yashiro, 1997; Schulz, Borowski, & Crown, 1991). First, the economic and policy challenges facing Japan, as one of the most rapidly aging industrial societies, are more immediate and more serious than in countries whose populations are aging less rapidly (Bass, 1996a). Second, levels of later-life labor force participation are substantially higher in Japan than in other industrialized countries (e.g., Bass, 1996b). For example, the labor force participation rate of men aged 65 and over in 1996 was.36 in Japan but only.17 in the United States,.05 in Germany, and.02 in France (Kinsella & Velkoff, 2001). Third, the institutional and social context of later-life work in Japan differs markedly from that in most other industrialized countries. The high prevalence of self-employment, mandatory retirement at relatively young ages, relatively weak pension disincentives to work, and distinctive cultural and family characteristics have all been discussed as reasons for the particularly strong labor force attachment of older Japanese. Fourth, Japan is the only industrialized country with a history of policy efforts to improve labor market opportunities and increase labor force attachment among older people (Bass, 1996b; Campbell, 1992; Kimura & Oka, 2001; Rix, 1996). Projected labor force shortages and public pension shortfalls associated with extremely rapid population aging in Japan have intensified policy efforts to increase the already high levels of labor force participation among older men and women (e.g., Kimura & Oka, 2001). Japan's experience with efforts to promote labor force attachment at older ages by reducing earnings tests on early pension benefit receipt, increasing the age of mandatory retirement, reducing pension benefit levels, increasing the age of eligibility for benefits, and promoting retraining and re-employment programs (Campell, 1992; Kimura & Oka, 2001) is of substantial relevance to the discussion of similar measures in other rapidly aging industrialized countries.

Previous research provides rich description of the institutional and social context of later-life work in Japan. Unfortunately, however, similarly detailed empirical studies of the individual-level correlates of labor force behavior are not currently available. This gap in the literature is due largely to data limitations. Previous analyses have relied upon cross-sectional surveys containing a relatively limited range of information. As a consequence, very little is known about labor force transitions, about the importance of noneconomic correlates of labor force participation, and about differences between self-employment and market employment at older ages. Furthermore, the labor force participation of older Japanese women has been largely ignored in previous research. The primary goal of this article is to address these limitations by using a nationally representative longitudinal survey of older Japanese men and women to describe the correlates of labor force participation beyond the age of 60.

	LIMITATIONS OF EXISTING EMPIRICAL RESEARCH ON LABOR FORCE PARTICIPATION AT OLDER AGES IN JAPAN

TOP Abstract Limitations of Existing... Institutional and Social Context... Methods Results Discussion References

 
Previous research on the individual-level correlates of later-life labor force participation in Japan has relied primarily on data from the Employment Status Survey of the Elderly (ESSE). Because the ESSE is a cross-sectional survey with limited retrospective information, existing empirical evidence on the correlates of labor force participation is limited to age-specific associations between labor force participation and a narrow range of concurrently measured characteristics. In the absence of longitudinal data, researchers have struggled to make causal inferences regarding the correlates of labor force status (e.g., Nagase, 1999; Ogawa, 2000; Seike, 1993) and have been unable to examine the correlates of labor force transitions.

Because previous studies of Japan have focused almost exclusively on estimating the "effects" of income and pension characteristics, they provide little information with which to evaluate the generalizability of many interesting findings from recent research on retirement in the United States. For example, recent analyses convincingly demonstrate that, net of individual economic characteristics, patterns of later-life labor force participation in the United States are closely linked to life histories, the interdependence of life spheres, and the mutual influences of family members (e.g., Blau, 1998; Hayward, Friedman, & Chen, 1998; Szinovacz, DeViney, & Davey, 2001). Almost nothing is known about these relationships in Japan.

Furthermore, very little is known about how correlates of later-life work differ by gender or employment status, that is, wage employment and self-employment. The limited attention to older Japanese women presumably reflects the fact that rates of labor force participation are much higher among older men. However, it is clear that the labor force participation of middle-aged women has increased over time (Ogawa & Retherford, 1997; Yashiro & Oshio, 1999) and that the encouragement of older women's work will be an increasingly important policy objective as the Japanese labor force contracts and ages in the years to come. It is more difficult to understand why many previous empirical studies have not distinguished self-employment from other forms of employment. Given the prevalence of self-employment among older Japanese and the well-documented relationship between self-employment and retirement age in the United States (e.g., Hayward et al., 1998), it is not clear how results based on data aggregated over the two groups should be interpreted.

In this article, we begin to address these limitations by (a) using longitudinal data to model labor force status at a given time as a function of individual characteristics measured at an earlier point in time, (b) providing the first analysis of labor force transitions among older Japanese, (c) moving beyond the narrow focus on pension characteristics to explicitly consider the role of respondents' work history and family characteristics, and (d) focusing on how correlates of later-life labor force participation in Japan differ by gender and by employment status.

	INSTITUTIONAL AND SOCIAL CONTEXT OF LATER-LIFE WORK IN JAPAN

TOP Abstract Limitations of Existing... Institutional and Social Context... Methods Results Discussion References

 
To provide a framework in which to interpret the results of these analyses, we summarize several key features of the institutional and social context of later-life work in Japan. In particular, we focus on self-employment, mandatory retirement and re-employment, public pension provisions, family structure, and gender relations.

Self-Employment
Self-employment is very common among older Japanese. Table 1 presents data from the 2000 Population Census of Japan, indicating that 44% of male and 53% of female labor force participants aged 60 and over were self-employed. Among men and women younger than 60 years, only 12% of labor force participants were self-employed. This difference primarily reflects a long-term decline in self-employment (Yashiro & Oshio, 1999), with current cohorts of older Japanese far more likely than subsequent cohorts to have worked in agriculture, small family businesses (e.g., retail shops, restaurants), and other forms of self-employment. Table 1 indicates that among the self-employed, 61% of men and 70% of women were working in either the primary sector (row 1) or retail occupations (row 5). The potential salience of work–family interdependence among older self-employed Japanese is suggested by the fact that two thirds of the self-employed women are classified as "family workers" (i.e.,.36 ÷.53 =.68). This group comprises primarily women working in businesses or farms operated by a household member but also includes those doing piece work at home. Greater flexibility and proximity to home presumably make self-employment more compatible than wage employment with the family responsibilities of older women. Because self-employed workers are eligible only for the relatively limited benefits provided by the basic public pension plan (kokumin nenkin) (Yashiro & Oshio, 1999), economic necessity is presumably another important reason for higher levels of labor force attachment among the self-employed.

Pathways into postretirement jobs depend, to a large extent, on characteristics of the career job. For those retiring from larger companies with lifetime employment policies and re-employment programs, it is not uncommon to continue working for the same company or to be transferred (shukk) to an affiliated company (Kimura & Oka, 2001; Rix, 1996; Schulz et al., 1991). Others may find employment in different types of work, perhaps in self-employment or facilitated through work agencies such as the Silver Human Resource Centers, described by Bass (1996b), Campbell (1992), and Schulz and colleagues (1991). Despite the prevalence of postretirement re-employment, it is likely that the experience of mandatory retirement is associated with a relatively lower likelihood of labor force participation. Receipt of severance pay, eligibility for relatively generous wage-indexed public pension benefits (ksei nenkin) (Yashiro & Oshio, 1999), and the fact that postretirement jobs are typically characterized by substantially lower pay, poorer working conditions, and lower job security than career jobs (Kimura & Oka, 2001; Schulz et al., 1991) are disincentives to continued work for those who have experienced mandatory retirement.

Gender Roles and Family Structure
Gender differences in the correlates of later-life work in the United States (Blau, 1998; Pienta, Burr, & Mutchler, 1994; Ruhm, 1996) have been interpreted as the cumulative reflection of gender-asymmetric roles across the life course, with the work decisions of older men remaining contingent on the presence of economically dependent family members and those of women remaining contingent on the presence of family members requiring instrumental support (e.g., elder care, child care). The highly asymmetric gender division of labor in Japanese families (Tsuya & Mason, 1995), in combination with high levels of intergenerational coresidence (Rein & Salzman, 1995) and the associated exchange of support among family members (Ogawa & Retherford, 1997), suggest that similar gender differences may be particularly relevant in Japan. For example, the presence of coresident children may limit the labor force participation of older women who are primarily responsible for the provision of domestic services while promoting work among older men who are more likely to remain in the provider role. Alternatively, it is possible that economies of scale and immediate access to economic support from children may contribute to lower levels of labor force attachment among older men and women coresiding with adult children.

	METHODS

TOP Abstract Limitations of Existing... Institutional and Social Context... Methods Results Discussion References

 
Data
The previous section suggests several ways in which the context of later-life work in Japan may contribute to gender and employment status differences in the relationships between labor force participation and correlates such as work history and family structure. To examine these relationships, we use data from waves 2–5 of the National Survey of the Japanese Elderly (NSJE). We do not use data from wave 1 because the initial survey did not distinguish wage employment from self-employment. Wave 2, conducted in 1990, is a nationally representative sample of 2,037 noninstitutionalized men and women aged 60 and over. The response rate at wave 2 was 78%, and among surviving wave 2 respondents, response rates were 88% in 1993, 77% in 1996, and 76% in 1999. Because we are able to use information on labor force status from proxy interviews, effective response rates for waves 3–5 are about 7 percentage points higher. Following respondents to the wave 2 survey across three subsequent waves allows us to establish appropriate temporal ordering of labor force outcomes and their correlates while also generating a sample size large enough to reliably estimate coefficients in the models described below. To impute data missing at wave t (t = 2, 3, 4, 5), we employed the following procedures. If consistent responses were provided in waves prior and subsequent to wave t (including wave 1), the value of these responses was used. For four of our independent variables (joint spousal income, longest occupation, current occupation, and whether longest occupation was self-employed), we created a category to include missing data. After dropping observations for respondents over the age of 85, observations with missing data on the dependent variable, and observations with missing data on independent variables that could not be readily imputed, the sample upon which we base our analyses consists of 5,276 person–period observations (2,267 men and 3,009 women) of the 1993 60- to 85-year-old respondents to the wave 2 survey.

Compared with the data analyzed in previous studies, the most obvious advantage of these NSJE data is the longitudinal information necessary for the analysis of labor force transitions. Given our interest in relationships between family structure and labor force participation, the detailed household rosters collected in the NSJE are another important advantage over the ESSE and other previously examined data.

Variables
Labor force participation
The dependent variable is constructed from questions about current work status and self-employment. Using a standard definition of labor force participation, we first collapse responses to the question on current work status into two categories: "in the labor force" (working now, working as an unpaid family employee, temporarily not working for reasons such as illness, unemployed and looking for work) and "not in the labor force" (retired, not working because of disability, engaged in housework, other). We then separate respondents who are "in the labor force" into two groups: wage employees and self-employed. Assuming that unemployed respondents (<1% of the sample) are seeking work in wage labor rather than self-employment, we combine these observations with the wage employees.

Work history
Measures of work history include a five-category representation of longest occupation and an indicator of whether the longest occupation was self-employed. Models also include a dichotomous indicator of whether the respondent retired from a job between the ages of 50 and 60 as a proxy for the experience of mandatory retirement. This is a crude proxy in that it includes some respondents who retired between ages 50 and 60 for reasons other than mandatory retirement and presumably excludes others who experienced mandatory retirement after age 60.

Socioeconomic characteristics
Socioeconomic characteristics include categorical indicators of spouses' combined income, current occupation, whether current occupation is self-employed, and whether the respondent lives in an urban or rural area. Educational attainment was included in preliminary analyses but was insignificant and thus excluded from the models presented. An obvious omission from this set of socioeconomic characteristics is a direct measure of pension eligibility and receipt: the primary focus of previous research. The absence of direct information on pensions and the relatively limited information on other economic characteristics of respondents and their families (e.g., wealth, financial transfers across generations) are the main comparative weaknesses of the NSJE data. Following other studies based on similarly limited data (e.g., Szinovacz et al., 2001), we seek to minimize potential omitted variable bias by proxying pension eligibility and receipt with correlates such as the characteristics of respondents' longest occupation and the experience of mandatory retirement. Given their age, nearly all NSJE respondents are eligible to receive public pension benefits at either full or reduced levels. Our primary concern is thus the extent to which these variables adequately proxy benefit levels. This concern is mitigated somewhat by evidence that the retirement-promoting effect of pension eligibility and benefit levels in Japan is relatively weak (Nagase, 1999; Ogawa, 2000; Oishi & Oshio, 2000; Seike, 1993; Shimono & Tachibanaki, 1984).

Age and health
All models include a linear term for age and a dichotomous measure of functional status. Respondents who reported some difficulty performing at least one of three activities of daily living (bathing, climbing two or three flights of stairs, walking a few blocks) were coded as 1 and those with no difficulties were coded as 0.

Family structure
All models include one measure of marital status and two measures of living arrangements. Because the large majority of older men and women in the sample are either currently married or widowed, we operationalize marital status as a dichotomous indicator (0 = spouse present, 1 = spouse not present). Two dichotomous measures of living arrangements indicate whether or not respondents are coresiding with married child[ren] or with unmarried child[ren] (0 = not coresiding with child[ren], 1 = coresiding with child[ren]). Distinguishing the marital status of coresident children is motivated by assumptions that married children are more likely to have children themselves and that unmarried children are more likely to be economically dependent upon their parents. Coresidence with married children may thus have a stronger negative relationship with the labor force participation of women, whereas coresidence with unmarried children may have a stronger positive relationship with that of men.

Analytic Strategy
To examine the relationship between labor force participation and these individual and family characteristics, we estimate three models based on three different subsamples. In model 1, we use the full sample to model current labor force status at year t (t = 1993, 1996, 1999) as a function of respondents' characteristics at year t – 3 (i.e., at the previous wave). This approach is similar to existing cross-sectional analyses but has the advantage of appropriate temporal ordering of labor force status and its correlates. In model 2, we restrict the sample to those respondents who were in the labor force at year t – 3 to examine the correlates of labor force exit. In model 3, we restrict the sample to those respondents who were not in the labor force at year t – 3 to examine the correlates of labor force (re)entry. All three models are estimated separately for men and women.

Recognizing that nonrandom attrition due to either death or nonresponse may result in biased estimates of the correlates of labor force participation (Hayward, Grady, Hardy, & Sommers, 1989), we treat loss to follow-up or death between waves as a competing risk. All models are estimated using multinomial logistic regression:

where p_ijt is the probability that the ith respondent was observed in labor force status j (j = not in the labor force, wage employee, self-employed, dead/lost to follow-up) at year t (t = 1993, 1996, 1999) and p_i0t is the probability that the ith respondent was observed in labor force status 0 at year t. In models 1 and 3, the reference outcome (i.e., labor force status 0) is "not in the labor force." In model 2, the reference outcome is "in the labor force." X_it–3 is a vector consisting of the independent variables described above. As indicated by the subscripts, all independent variables are measured 3 years prior to the dependent variable. Because we are pooling multiple waves of data, all models include a dummy variable for survey year to control for temporal influences. We also estimate models with robust standard errors to account for nonindependence across repeated observations. We use these standard errors to calculate the significance of gender differences in the estimated coefficients.

	RESULTS

TOP Abstract Limitations of Existing... Institutional and Social Context... Methods Results Discussion References

 
Table 2 presents a cross-tabulation of labor force status at years t – 3 and t. Row marginals, representing the distribution of work status at year t – 3, indicate that 45% of men and 21% of women were in the labor force. Among those in the labor force, 59% of men and 68% of women were self-employed. Summing the diagonal cells in the first three rows and columns shows that later-life work status is quite stable, with 66% of men and 75% of women remaining in the same status across adjacent waves.

View this table: [in this window] [in a new window]   Table 2. Transitions in Labor Force Status, Year t – 3 Year t.

Corresponding results for women are presented in Table 5. Coefficients in bold are significantly different (at p <.10) from the corresponding coefficients in the models for men. Model 1 clearly shows that the correlates of older women's labor force participation are different from those for men. Particularly notable are the large gender differences in the relationships between family characteristics and wage employment. Married women and women coresiding with children (either married or unmarried) are significantly less likely than otherwise similar women to be employed in wage labor. These coefficients are insignificant in the corresponding contrast for men. Other notable gender differences include (a) the relatively weak relationship between joint spousal income and labor force participation among women, (b) the relatively low likelihood of wage employment among women whose longest occupation was professional/managerial or in sales/service/clerical, and (c) the weaker relationship between functional limitations and labor force participation among women. Gender differences in the correlates of self-employment are generally smaller than those for wage employment, presumably reflecting greater gender similarity in work opportunities, working conditions, and the interdependence of work and family responsibilities among the self-employed elderly.

	DISCUSSION

TOP Abstract Limitations of Existing... Institutional and Social Context... Methods Results Discussion References

 
Using data from a nationally representative longitudinal survey of men and women aged 60 and over, we have extended existing research on the individual correlates of later-life labor force participation in Japan in several ways. Not only have we provided the first analysis of labor force participation based on longitudinal data, we have documented the salience of characteristics largely ignored in previous work. Importantly, we have also shown that the correlates of labor force participation differ by employment status and by gender.

Net of variables typically included in previous studies (i.e., age, income, and health), it is clear that labor force participation is influenced by work history and family characteristics. The experience of mandatory retirement substantially reduces the likelihood of labor force participation, whereas longest occupation in agriculture and other forms of self-employment are strong predictors of continued self-employment at older ages. These findings suggest that low public pension benefit levels and job flexibility associated with self-employment contribute to the high rates of labor force participation among older Japanese. The relevance of job flexibility is further suggested by observed relationships between family structure and women's labor force participation. Evidence that the presence of a spouse and coresiding children reduces the likelihood of wage employment, but not self-employment, is consistent with a primarily domestic role for older women and suggests the incompatibility of relatively inflexible wage labor and family responsibilities. These findings also suggest that wage employment among older women is primarily driven by economic necessity. That is, older women without immediate access to family-provided financial support may need to work to support themselves.

Taken as a whole, our findings strongly suggest the need for care in interpreting the results of existing analyses of labor force participation among older Japanese. First, it is not clear to what extent the relationships between labor force participation and the economic variables central to previous studies are independent of the work history and family characteristics included in this study. Second, evidence that the correlates of wage employment are, in many cases, different from those of self-employment makes it difficult to interpret results of studies in which data are aggregated across employment types. Third, significant gender differences in the effects of income and family characteristics clearly indicate that studies of older men provide only a partial picture of the correlates of later-life work. Finally, our results confirm the need for caution in using estimated relationships between labor force status and its correlates to make inferences about the correlates of labor force transitions. Comparing the results of model 1 with those of models 2 and 3 indicates that causal inferences drawn from cross-sectional data would be correct in some cases and incorrect in others. Because the results of model 1 reflect labor force exit and re-entry as well as the likelihood of labor force participation at baseline, labor force transitions can be understood only by focusing on those who are at risk of experiencing them.

Results also suggest that, all else (e.g., pension policy) equal, changes in the characteristics of future cohorts of older Japanese may influence levels of labor force participation. For example, a simplistic extrapolation of the estimated relationship between longest occupation and labor force participation suggests a potential decline in the labor force attachment of subsequent cohorts of older Japanese who are much less likely to have been self-employed or to have worked in agriculture. Similar reasoning suggests that projected declines in coresidence with married children (Hirosima, 1997) may contribute to increased labor force participation among subsequent cohorts of older Japanese women. The value of simplistic extrapolations such as these is obviously not to make accurate projections but rather to emphasize the relevance of changing life course experiences of older men and women. Attention to work histories and family characteristics will enhance the effectiveness of efforts to promote later-life participation in the rapidly aging and contracting Japanese labor force.

Results also provide some evidence with which to evaluate the generalizability of findings from recent studies emphasizing life course influences on later-life work in the United States. Despite substantial contextual differences, relationships between labor force participation and the occupational history and family structure of older Japanese men and women are generally similar to those observed among their U.S. counterparts. For example, in both countries, labor force participation is significantly higher among men and women who worked in agriculture and other forms of self-employment. In contrast to the United States (e.g., Hayward, Hardy, & Liu, 1994), however, higher education and careers in higher status occupations do not appear to be associated with higher rates of labor force participation among older Japanese. Work–family relationships among older Japanese are also generally similar to those of their U.S. counterparts and largely consistent with expectations reflecting gender-asymmetric family roles and the importance of intergenerational exchange of resources and services. Two important differences between our findings for Japan and the general pattern of results for the United States are (a) the very large gender differences in the relationship between marital status and wage employment and (b) the importance of coresidence with children as a correlate of labor force participation.

We view these results as an important extension of previous empirical analyses of later-life work in Japan but recognize several important limitations that should be addressed in subsequent research. First, coefficients for our measures of work history and family structure are largely in the hypothesized direction, but few are estimated precisely. This is not surprising, given the small number of respondents in some categories (e.g., men without a spouse present, women in wage employment) and the stability of later-life life labor force participation. Analyzing larger samples that cover a broader range of ages may allow for more precise estimation of these relationships. Second, the NSJE data contain very limited measures of economic status. A better understanding of the ways in which economic characteristics interact with work history, family characteristics, and health to influence patterns of labor force participation is essential for speculation regarding future patterns of labor force participation in Japan and for the discussion of related policy measures. Finally, it is important to recognize that the measures of family structure we have examined are indirect representations of the theoretical concepts of ultimate interest, that is, family relations and intergenerational exchanges. Direct measures of financial exchanges, child care and elder care provision, and time use would facilitate the evaluation of hypothesized work-limiting effects of family responsibilities for older women and work-promoting effects of economic responsibilities for older men. Direct measures of spouse's health, spouse's work status, and marital quality would also allow for examination of joint retirement and spousal influences on work emphasized in recent studies of older Americans. Addressing these limitations will allow for a better contextual understanding of later-life work in Japan and, more importantly, contribute to the foundation of subsequent studies of Japan in a comparative perspective.

	Acknowledgments

 
This research was supported by Grant R37-AG154124 (Jersey Liang, principal investigator) from the National Institute on Aging. Additional support was provided through a grant from the Japanese Ministry of Health, Labor, and Welfare Longevity Foundation, by the Tokyo Metropolitan Institute of Gerontology, and by a fellowship from the Abe Foundation.

	Footnotes

 
Decision Editor: Charles F. Longino, Jr., PhD

Received for publication September 15, 2003. Accepted for publication December 30, 2003.

	References

TOP Abstract Limitations of Existing... Institutional and Social Context... Methods Results Discussion References

 

• Bass, S. A. 1996a. Introduction: Japan's aging society. In S.A. Bass, R. Morris, & M. Oka (Eds.), Public policy and the old age revolution in Japan (pp. 1–12). New York: Haworth Press.
• Bass, S. A. 1996b. An overview of work, retirement, and pensions in Japan. In S.A. Bass, R. Morris, & M. Oka (Eds.), Public policy and the old age revolution in Japan (pp. 57–78). New York: Haworth Press.
• Blau, D. M. 1998. Labor force dynamics of older married couples. Journal of Labor Economics, 16,595-629.
• Campbell, J. C. 1992. How policies change. Princeton, NJ: Princeton University Press.
• Gruber, J., Wise, D. A., (Eds.). 1999. Social security and retirement around the world. Chicago: University of Chicago Press.
• Hayward, M. D., Friedman, S., Chen, H. 1998. Career trajectories and older men's retirement. Journal of Gerontology: Social Sciences, 53B,S91-S103.
• Hayward, M. D., Grady, W. R., Hardy, M. A., Sommers, D. 1989. Occupational influences on retirement, disability, and death. Demography, 26,393-409.[Medline]
• Hayward, M. D., Hardy, M. A., Liu, M.-C. 1994. Work after retirement: The experiences of older men in the U.S. Social Science Research, 23,82-107.
• Hirosima, K. 1997. Projection of living arrangements of the elderly in Japan, 1990–2010. Genus, 53,79-111.
• Hurd, M. D., Yashiro, N. 1997. The economic effects of aging in the United States and Japan. Chicago: University of Chicago Press.
• Kimura, T., Oka, M. 2001. Japan's current policy focus on longer employment for older people. In V. W. Marshall, W. R. Heinz, H. Kruger, & A. Verma (Eds.), Restructuring work and the life course (pp. 348–359). Toronto: University of Toronto Press.
• Kinsella, K., Velkoff, V. A. 2001. An aging world: 2001. Washington, DC: U.S. Bureau of the Census.
• Kohli, M., Rein, M., Guillemard, A.-M., van Gustern, H., (Eds.). 1991. Time for retirement: Comparative studies of early exit from the labor force. Cambridge, U.K.: Cambridge University Press.
• Nagase, N. 1999. The public pension and the labor supply of older women in Japan. Review of Population and Social Policy, 8,27-48.
• National Research Council. 2001. Preparing for an aging world: The case for cross-national research. Washington, DC: National Academy Press.
• Ogawa, H. 2000. The effect of household structure on the employment behavior of elderly male workers. Review of Population and Social Policy, 9,25-46.
• 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.
• Oishi, A., Oshio, T. 2000. Social security wealth and retirement behavior. Social Security Research, 35,405-419 Japanese.
• O'Rand, A. M., Henretta, J. C. 1999. Age and inequality: Diverse pathways through later life. Boulder, CO: Westview Press.
• Pienta, A. M., Burr, J. A., Mutchler, J. E. 1994. Women's labor force participation in later life: The effects of early work and family experiences. Journal of Gerontology: Social Sciences, 49B,S231-S239.
• Quinn, J. F., Burkhauser, R. V. 1995. Retirement and labor force behavior of the elderly. In L. G. Martin & S. H. Preston (Eds.), Demography of aging (pp. 50–101). Washington, DC: National Academy Press.
• Rein, M., Salzman, H. 1995. Social integration, participation, and exchange in five industrial countries. In S. A. Bass (Ed.), Older and active: How Americans over 55 are contributing to society (pp. 237–262). New Haven, CT: Yale University Press.
• Rix, S. E. 1996. The challenge of an aging work force: Keeping older workers employed and employable. In S. A. Bass, R. Morris, & M. Oka (Eds.), Public policy and the old age revolution in Japan (pp. 79–96). New York: Haworth Press.
• Ruhm, C. J. 1996. Gender differences in employment behavior during late middle age. Journal of Gerontology: Social Sciences, 51B,S11-S17.
• Schulz, J. H., Borowski, A., Crown, W. H. 1991. Economics of population aging: The "graying" of Australia, Japan, and the United States. New York: Auburn House.
• Seike, A. 1993. Labor markets in aging societies. Tokyo: Toyo Keizai Shimposha. Japanese.
• Seike, A. 1997. Labor market implications of social security: Company pension plans, public pensions, and retirement behavior of the elderly in Japan. In M. D. Hurd & N. Yashiro (Eds.), The economic effects of aging in the United States and Japan (pp. 295–315). Chicago: University of Chicago Press.
• Shimono, K., Tachibanaki, T. 1984. Analysis of the employment behavior of older men and women. Social Security Research, 19,398-413 Japanese.
• Szinovacz, M. E., DeViney, S., Davey, A. 2001. Influences of family obligations and relationships on retirement: Variations by gender, race, and marital status. Journal of Gerontology: Social Sciences, 56B,S20-S27.
• Tsuya, N. O., Mason, K. O. 1995. Changing gender roles and below replacement fertility in Japan. In K. O. Mason & A.-M. Jensen (Eds.), Gender and family change in industrialized countries (pp. 139–167). Oxford, U.K.: Clarendon Press.
• Yashiro, N., Oshio, T. 1999. Social security and retirement in Japan. In J. Gruber & D. A. Wise (Eds.), Social security and retirement around the world (pp. 239–267). Chicago: University of Chicago Press.

This article has been cited by other articles:

				J. M. Raymo, J. Liang, E. Kobayashi, Y. Sugihara, and T. Fukaya Work, Health, and Family at Older Ages in Japan Research on Aging, March 1, 2009; 31(2): 180 - 206. [Abstract] [PDF]

This Article Abstract Full Text (PDF) Services Download to citation manager Citing Articles Citing Articles via HighWire PubMed PubMed Citation