Understanding the Spousal Concordance in Mortality Using Kin-Linked Data

Sven E. Wilson, Brigham Young University
Joseph P. Price, Brigham Young University
Hannah Hammond, Brigham Young University
Kate Wilson, Brigham Young University

Spouses may share similar lifespans for a variety of factors. First, assortative mating, driven both by market forces and by social norms and networks, leads to interspousal correlations in health at the time of marriage. Second, married couples share environmental risks, economic resources, and health-related behaviors. Third, extensive research has shown that the death of a spouse raises mortality risk (the “broken heart syndrome”), thereby shortening the lifespan of the surviving spouse and driving it closer into concordance with the deceased spouse. In this study, we exploit an exciting new data source. The Census Tree collection, which combines machine learning algorithms with crowd-sourced genealogical data allows us to draw 3.8 million couples drawn from the 1940 US Census who have complete lifespan data. We compare the difference in spousal lifespans for the actual couples with a synthetic set of couples put together by random assignment and see a clear bulge in age differences around zero for the actual couples, compared to the synthetic couples, indicating the widowhood effect found in other studies. From the Census Tree, we also obtain lifespans for many of the male and female siblings of the couples, and we use spousal sibling lifespan as an instrument for spousal lifespan. Our identification strategy exploits the fact that sibling correlation in lifespan is strong, while the correlation of one’s lifespan with the spousal sibling’s lifespan is low. Our simple IV model finds a statistically significant causal relationship between the lifespan of both husbands and wives on the lifespan of their respective spouses that are roughly twice as big as the OLS estimates.

No extended abstract or paper available

 Presented in Session 105. Kinship Dynamics and Effects: Historical Insights