Shi Jiaxin

Jiaxin Shi

Jiaxin Shi, Christian Dudel, Christiaan Monden, and Alyson van Raalte

There are persistent and substantial socioeconomic disparities in remaining life expectancy at older ages. Yet how socioeconomic status plays a role in retirement lifespan has not been fully understood. This study aims to assess gender and educational differences in retirement expectancy and retirement lifespan variation, taking into account individual labor-force exit and re-entry dynamics. We analyzed the 1996–2016 Health and Retirement Study (HRS). Participants were 32,228 individuals aged 50 and above at their first interviews. Discrete-time Markov Chain models were used to estimate multistate life tables, based on transition probability matrices estimated from multinomial logistic regression models. Retirement expectancy, and absolute and relative Gini indices were calculated. We found that among females there is a persistent educational gradient in retirement expectancy over the whole period studied; among males, the relationship between education and retirement expectancy is different across periods, which may have been caused by changes in economic conditions. Females and the lower- educated have higher absolute Gini in retirement lifespan than males and the higher- educated—yet these relationships are reversed when examined by relative Gini. When considering policies to mitigate various challenges in an aging society, policymakers may consider gender and socioeconomic specific retirement expectancy as well as the variation of retirement lifespan to better distribute resources.