At a Glance
Linked Lives: The Importance of the Family for Socioeconomic Attainment, Health, and Mortality
Kieron Barclay, Mikko Myrskylä, Pekka Martikainen (MPIDR / University of Helsinki, Finland)
In this research project, we focus on how the attributes of one’s family members, as well as demographic events that affect family members, might influence a range of outcomes for the index person, from educational and socioeconomic attainment to health and mortality. Literature is vast on this broad topic and addresses questions such as: What influence does widowhood have on the mortality of the surviving spouse? To what extent is marital status associated with health and mortality? What impact do the resources of one’s partner have on health and mortality? Does the death of a parent affect long-term outcomes for children? Does the status attained by children have any effect on the parents?
A puzzle common to all of these questions is the extent to which the patterns observed are "causal" and the extent to which they are explained by selection. Complex endogenous processes, genetic relatedness, and homophilous patterns of relationship formation mean that identifying the causal effects is particularly challenging in the context of the family. Nevertheless, it is important to address these questions given that the family unit, broadly defined, is one of the most salient group identities for humans, and given that the relationships of individuals with their parents, children, and romantic partner or partners are often key defining features of one’s life.
To examine how characteristics of family members and demographic events within the family affect outcomes for the index person, we use a range of data sources, including Nordic population data from Sweden and Finland and data from the Utah Population Database (UPDB). The UPDB contains information on more than eleven million individuals, representing birth cohorts that date back to the period around 1760. We apply a range of statistical tools and research-design strategies, including event history analysis, within-family fixed-effects regression, and antemortem socioeconomic trajectory-matching techniques in order to try and adjust for confounding in terms of factors that predict the timing of death as well as socioeconomic attainment and health outcomes.
Some key findings from this project are that selection does play an important role in driving associations between deaths within the family and later outcomes for the index person, but that there is likely to be a direct causal effect as well. For example, we have found that individuals who will later become widowed begin to see an increase in hospitalization rates before the death of the partner, indicating that shared patterns of health deterioration explain part of the increase in the mortality of the surviving spouse after experiencing widowhood. Likewise, when studying how the death of a parent impacts upon the educational and socioeconomic attainment of children over time, we find that most of the socioeconomic variation in this association (e.g., that it is worse to lose a parent with a higher socioeconomic status (SES) than with a lower SES) disappears after matching on age, education, and crucially, the earnings trajectories of those who did or did not die.
Aging, Mortality and Longevity, Intergenerational Relationships
Journal of Marriage and Family, 1–24. accepted. (2021)
Demography 58:3, 1011–1037. (2021)
Demography 57:6, 2169–2198. (2020)
MPIDR Working Paper WP-2019-010. (2019)
MPIDR Working Paper WP-2019-008. (2019)
American Journal of Epidemiology 188:1, 110–118. (2019)
MPIDR Working Paper WP-2019-007. (2019)
SSM-Population Health 4, 271–279. (2018)