The Long-Term Returns to Childbearing: Adult Off-Springs and the Wellbeing Of Older Parents

Arnstein Aassve and Chiara Pronzato with Steve Pudney

Arnstein Aassve (Dondena Centre) and Chiara Pronzato (Università di Torino), with Steve Pudney (University of Essex), presented The Long-Term Returns to Childbearing: Adult Off-Springs and the Wellbeing Of Older Parents at Alp-Pop 2012. The Alpine Population Conference (La Thuile, 15-18 January 2012).

Abstract: Our focus concerns the role of children’s life course events and their impact on the older parents in terms of mental wellbeing. This is a different focus on what one typically finds in the social science – most analyses are concerned with the impact of parents’ behaviour on the children’s outcomes. This line of research is extensive and includes for instance parental investment in time and money, but also on demographic events such as divorce, on a range of child outcomes. Such an imbalance in the literature is not entirely unexpected given that there is ample evidence to suggest that net transfers between parents and children follow a downward path. Still, there is a huge literature that concerns health and mental wellbeing of older individuals, but rarely is the role of the children taken into account. However, recently social scientists have begun looking into the role of children for the wellbeing of the older parents. In our mind, this issue has not been looked into sufficiently, and the aim of this analysis is to help filling this gap.

There are several good reasons to consider this relationship. First, from an economics point of view, an important motivation for childbearing in the first place is that prospective parents will derive utility or enjoyment from their children. However, the focus tends to be on the enjoyment of “young” parents from having children. In developed countries, there is a well established trade-off between quantity and quality of children, meaning that as societies develop, parents prefer a smaller number of children, but at the same time invest more in them. In developing societies, the argument is that children provide cheap labour and is therefore an important resource for the household. Moreover, with lacking welfare provision, children is considered as an insurance against old age disadvantage. Whereas this argument is not generally put forward for developed societies, one can nevertheless imagine that old people find considerable comfort in having children around. Despite having access to welfare services, children will provide emotional support that certainly might matter for their mental wellbeing. In fact, there is evidence to suggest that childless old people, suffer more from mental and psychological distress. As people grow older, it is hard to see how institutional welfare can substitute the emotional support that children can provide. Despite the obvious reversed causality since mental distress may be negatively linked to partner formation and childbearing, the argument does make intuitive sense.

This paper studies the determinants of mental health of people over 50 in several European countries and, in particular, focuses on children’s life course outcomes.

Research on inter-generational issues makes great demands on data resources. We need to observe

children who are no longer resident in the parental home, good indicators of mental wellbeing, retrospective and panel information. We use SHARE (Survey on Health Aging Retirement in Europe) for the years 2004, 2006, and 2008. It comprises questions relating to (a maximum of four) non-resident children, has reliable mental health variables measured by the 13-item EURO-D instrument, has retrospective (SHARELIFE, 2008) and panel information (SHARE, 2004 and 2006). Finally, it is available for 11 European countries.

We use the 13 indicators of mental health as a latent continuous variable which represents the level of depression, and which is explained by personal and family characteristics, by children’s experiences and life events. We apply the item response theory to study the relationships between a set of categorial observed variables and a continuous latent variable, and we estimate the model using Mplus.

Concerning the personal characteristics, we find a strong positive effect of unemployment and bad health on old people mental health, while a negative one of years of schooling. Being in a couple decreases depression but having experienced separation and widowhood increases it. Having experienced the death of a child, the divorce of a child, and unemployment of a male child increase depression while having a child who achieved good results at school decreases it.

By exploiting cross-country variability, we also find that in countries where divorce rates are higher, parents suffer less from children’s divorce. In the same way, in countries where the unemployment rate is higher, son’s unemployment is less worrying.