This study assesses the effect of spouse and marital relationship characteristics on labor force withdrawal. The authors further explore differences between husbands’ and wives’ retirement. The analyses are based on two waves of the National Survey of Families and Households. Marital characteristics influence retirement decisions in several ways and contribute significantly to model fit. Husbands adjust their retirement in terms of wives’ benefit eligibility, whereas wives’ retirement is contingent on the couple’s income. Husbands also tend to leave the labor force when their wives are ill. In addition, retirement decisions seem to reflect considerations about postretirement marital quality and husband’s status in the marriage. Gender differences prevail with regard to the impact of work and marital history on retirement decisions.

Our data suggest the importance of family obligations and relationships in retirement decisions and demonstrate considerable diversity in these processes. Models of retirement should pay greater attention to the interdependence of work and family spheres and to the diversity of retirement processes among various population groups.

The findings suggest that effects of surrogate parenting differ by gender, and that they are partially contingent on grandparents' vulnerabilities (marital status, education, and presence of childless dependent children in the household) before grandchildren join the household. Mediating effects of other life changes are relatively small.

The scope of explanation for retirement behavior could be greatly enlarged by acknowledging workers' extended engagement with the question before the event-its "givenness" in their future, and their course of action toward it. This study provides evidence for such extended involvement among workers aged 51 to 61 in the 1992 Health and Retirement Study (HRS) who saw themselves being within 15 years of retiring. To the extent that workers foresaw less time left at work, they reported more frequent thinking and talking about retirement. This pattern was quite general. The consideration given to retirement was, as expected, more frequent in circumstances that might orient workers toward the future. However, even absent these circumstances, subjective proximity still predicted the topicality of retirement. Widely held, albeit individual, timetables for retirement demonstrate its embeddedness in the subjective life course of older workers.We suppose retirement to be a course of action. In this article, we demonstrate a basic feature of that course. Working people are engaged in the idea of retirement years in advance, and this engagement has a general and regular pattern. We have argued elsewhere that workers' decision making for retirement extends over a considerable period (Ekerdt 1998;Ekerdt, DeViney, and Kosloski 1996). The extended stream of decision making for retirement can be observed in the

The authors used data from the first wave of the Health and Retirement Study ( F. Juster and R. Suzman 1995) to evaluate whether certain job-related gratifications might reduce retirement planning. Three definitions of retirement planning were evaluated and then regressed separately on a set of variables that included 3 types of job-related satisfactions (intrinsic gratification, positive social relations, and ascendance in the workplace) and 7 covariates: education, age, sex, health, marital status, race, and pension eligibility. Findings indicated that jobs high in ascendance were related to an increase in certain types of retirement planning, but jobs high in intrinsic rewards and positive social relations were related to less planning, regardless of how planning was defined. The findings suggest that information about work-related rewards may be useful in targeting individuals who might benefit from retirement planning programs, in developing planning programs to help workers realize more complex retirement plans, and in assisting employers who hope to retain older workers.