Inter Vivos Health Transfers : Final Days of Japanese Elderly Parents

The empirical evidence of the effect of intergenerational coresidence by elderly parents and their adult children on parental health remains fairly inconclusive. The aim of this paper is twofold: (1) to estimate a reliable coresidence effect and (2) to investigate why the coresidence effect can be negative. I argue that coresidence may worsen parental health, with coresidence burdens on children creating disincentives for parents to invest in their health. Studying Japanese data reveals: (i) an insignificant, negative average treatment effect of coresidence; (ii) a significant, negative treatment effect on the treated; and (iii) that parents with high care needs and limited resources, typically widowed mothers, are in coresidence experiencing the largest negative impact. Results are consistent with the theory predictions. In particular, what leads to the negative coresidence effect is not the base period health status but coresidence burdens, such as disability and dementia