Conference Agenda

Ageing aligns with changes in living arrangements, social networks, and companionship patterns that may increase time spent alone. But social isolation not necessarily implies loneliness, the subjective feeling of being alone. We jointly test the determinants of loneliness: living alone, spending time alone, and the existence of a network in case of need, together with an objective measure of daily pattern of time spent alone and with others.

Using the most recent Italian Time Use survey data (2013-14) we focus on individuals aged over 60 and model a measure of loneliness, the unmet desire for company measured through the following question: Would you like to spend more time in the company of other people? We use Sequence Analysis and clustering techniques to identify relevant daily patterns of time spent alone and with others, at home and outside. Regression models test how those time-use profiles are linked with loneliness and help identify high-risk groups. Results provide evidence that the patterns of daily activities contribute to explaining the perception of loneliness more than solo living or the lack of social networks. Those who spend time alone and at home faring well signal that being alone and feeling lonely are separate dimensions.

Studies have proposed that grandparental investment can improve grandchildren’s educational success. Grandparental death causes the loss of this investment. When preceded by disabling illness and support needs, grandparents and grandchildren may compete for the resources of the middle generation or the other grandparent. Thus, the timing and cause of grandparental death may influence grandchildren’s educational success. To examine this, this study used Finnish Census Panel data to compare grade point average (GPA) at the end of compulsory schooling for siblings who had experienced grandparental death at different ages. Using sibling fixed effects models, a modest, but robust, gradient was found based on the timing of grandparental death. The younger the grandchild at the time of death, the lower the grandchild’s GPA. This gradient was most pronounced when the maternal grandmother died. However, if a grandfather died from dementia – a particularly disabling illness – the gradient was reversed – the older the grandchild at the time of death, the lower the GPA. This suggests that burdensome causes of death may lead to competition over grandmaternal resources during school years. These results suggest that a complete loss of or substantial competition over grandmaternal investments contributes to poorer educational outcomes for grandchildren.

This study focuses on the relationship between daily time allocation and life-satisfaction (LS) in old age in Italy, through an innovative approach. Building on the Active Aging framework, it investigates how various activities – such as productive work, active leisure, and passive leisure – differently contribute to LS. The idea is that, given the 24-hour constraint, dedicating time to one activity reduces time available for others, prompting significant trade-offs, which can potentially influence LS. Rather than viewing each activity in isolation, therefore, we assess how the balance among daily activities shape LS for older adults. To address the interdependent nature of time-use diary data, where activities sums to 24 hours, thus generating collinearity issues, this study uses a compositional data analysis approach to measure each activity’s relative influence on LS. Additionally, using isotemporal substitution models, it explores the potential shifts in LS when reallocating time from one activity to another. Findings show that reallocating time from productive activities generally reduces LS, except when shifted to active leisure, which has no significant effect. Conversely, increasing time in productive activities at the expense of active leisure decreases LS, emphasizing the importance of balancing personal enjoyment and social contribution.

Few studies have examined longitudinal patterns of loneliness and associated health and social network resources. Using data from six waves of the English Longitudinal Study of Ageing (2008/09 to 2018/19, N=4740) and group-based trajectory modelling (GBTM), we identified five groups of loneliness trajectories: three with stable levels (37% “stable low”; 26% “stable medium”; 9% “stable high”) and two with time-varying scores of loneliness (8% “increasing”; 20% “decreasing”). Multinomial regression models and GBTM were used to examine baseline and time-varying factors associated with these trajectories, with a focus on health and social network characteristics (size, perceived social support, closeness, and frequency of contact) with partners, children, immediate family, and friends. Results showed respondents in poor health, depressed, and whose health deteriorated over time were more likely to have stable high or increasing levels of loneliness. Although lack of family and friends was associated with higher levels of loneliness, social network resources were also important determinants of loneliness. Respondents with high social network resources and those who maintained or improved them over time reported lower levels of loneliness. The study highlights the need to investigate loneliness from a life course perspective and account for the presence of social networks and their characteristics.