Retirement transition is a major life event in later adult life. Its timing is important for older individuals for economic, personal and family reasons, as well as for aging societies contemplating a comprehensive plan for population changes, including sustainability of the labor force, pension system, and welfare services such as eldercare. This thesis explores retirement timing in contemporary Sweden, which serves as an interesting case study because of its aging population, high labor force participation of men and women, universal pension system and generous welfare services. The overarching aim of the thesis is to investigate how relationships in the private sphere associate with retirement timing by focusing on leisure engagement, family relations and intergenerational ties.

The thesis consists of an introductory chapter and four empirical studies. The purpose of the introductory chapter is to place the four studies in context by focusing on the Swedish population structure, labor force participation and pension system and by highlighting some of the central theories and empirical findings related to retirement transition.

Study I addresses leisure engagement before retirement and retirement timing, and how engagement in leisure changes after retirement. The study finds that retirement timing varies by both the type of preretirement activity domain and the level of engagement. For instance, occasional or frequent engagement in dance and music postponed retirement compared to no engagement in these activities. The study also finds that patterns of leisure engagement after transition into retirement tend to be a continuation of the corresponding preretirement patterns.

Study II investigates the association between grandparenthood and retirement timing. The results show that grandparents at different life stages are more likely to retire compared to non-grandparents, but there is also variation among grandparents, and the more complex the family situation, the more likely grandparents are to retire.

In Study III, the focus shifts to the relationship between survival of elderly parents and retirement timing. The study finds that parental survival is positively linked to retirement timing and that the effects are stronger and more consistent for women thanfor men, in particular when only one parent is still alive. Additionally, women have a higher propensity of retiring in the immediate period after parental death, especially when the father is widowed. In contrast, men have a higher propensity of retiring when either the mother or father has been widowed for some years.

Study IV examines married couples’ propensity to coordinate retirement. The study finds that the likelihood that spouses will coordinate their retirement decreases as their age difference increases but that age differences have a similar effect on retirement coordination for couples with a larger age difference. The study also finds that coordination is largely gender neutral in opposite-sex couples with age differences, regardless of whether the male is the older spouse.

The thesis shows that, compared to wealth or health predictors of retirement, factors concerning the private sphere are also most relevant in non-trivial ways to large shares of retirees in Sweden. Increased knowledge of these relationships is important both for individuals’ retirement planning and for decision-makers’ and policy-makers’ planning and organization.