The causal effects of caregiving on the spousal carer's health, life satisfaction, and employment

In light of the development of the Long-term Care Insurance Policy in China, an emerging issue of concern is whether to compensate the families of those who need long-term care and who are cared for by family carers, and if so, by how much. It is therefore important and urgent to understand the effects of caregiving on the family carer's health and life satisfaction, as well as the financial costs of family caregiving. Despite expanding research investigating the association between caregiving and the family carer's health and life satisfaction, both internationally and within the specific context of China, a consistent conclusion has not been reached, potentially because of the existence of endogeneity (which refers to a problem that an independent variable [x] is correlated with the error term [e] in a regression model) related to the family effects (i.e. having a family member in poor health may directly affect an individual's health) and selection effects (i.e. individuals who are healthier are more likely to provide care). Furthermore, little is known about the effects of caregiving on older spousal carers' employment or work hours. Consequently, my research aimed to examine the causal effects of caregiving on older spousal carers' health, life satisfaction, work participation, and work hours in China.
My research expanded our understanding of the source of endogeneity in the relationship between caregiving and the spousal carer's health and life satisfaction, and further developed the use of a fixed effects triple difference model in order to address endogeneity related to family effects and selection effects.
Based on two separate empirical analyses, using data from waves 2015 and 2018 of the China Health and Retirement Longitudinal Study, my research found that:
(1) Individuals aged 60+ years who were willing and able to provide care to their spouse had better health than those who were unwilling or unable to provide care, after controlling for the caregiving status and other variables (the family effects, demographic, and socioeconomic characteristics) in the model. This evidence indicates the existence of selection effects.
(2) Individuals aged 60+ years whose spouse reported needing care had worse health than those whose spouse did not report needing care, after controlling for the caregiving status and other variables (the family effects, demographic, and socioeconomic characteristics) in the model. This evidence shows the existence of family effects.
(3) Female carers aged 60+ years were less likely to report difficulties in performing instrumental activities of daily living than female non-carers. Furthermore, female carers were more likely to report poor self-reported health than female non-carers, controlling for the family effects, selection effects, demographic, and socioeconomic characteristics in the model.
(4) Male carers aged 60+ years were 15.4% less likely to be in paid employment than male non-carers.
These results have important policy implications, particularly with the intersecting policy agendas of encouraging individuals to work for longer in the context of increasing longevity (i.e. extending working life agenda), and supporting informal carers to continue providing care to their spouses. The results suggest that policy makers need to pay more attention to older spousal carers' health and seek ways to delay the onset of those older carers' own need for long-term care. Moreover, although stopping work may be associated with an individual's personal motivation to provide care, the results of the empirical analysis support the provision of financial compensation to family carers in order to replace the wages forgone by those who cease paid employment in order to take up unpaid family care responsibilities.