Endogenous retirement due to education capital or health capital of the elderly

Elderly people retire because their disutility from working increases with their age. If workers’ physical condition is crucial for their tasks, the disutility of working is determined by their stock of health capital. However, the further the economy develops the more important workers’ productivity becomes. If this is the case, individuals would retire when they felt it was difficult to perform their tasks due to their stock of education capital. In this study, we develop a model of the joint determination of education, health, and retirement decisions, in which the two reasons for the elderly to retire are examined. The problem of a three-period-lived agent in our overlapping generations model is solved in two stages. In the first stage, the agent optimally chooses her desired levels of investments in education and health. In the second stage, given the optimal decisions, the agent compares lifetime utilities derived for the two scenarios, to continue to work or to retire, and chooses the one with a higher utility. If people retire due to poor health conditions, our model predicts a positive effect of life expectancy on their labor participation rate. Whereas, once low productivity of elderly workers becomes the main reason for their retirement, the effect of rising life expectancy disappears. Thus, in the latter case, the complementarity between education and health capital is lost. The model is calibrated to fit the Japanese economy to examine effects of an increase in the Pay-As-You-Go (PAYG) social security contribution rate on economic growth and life expectancy. We find that, if distortional effects inherent in the PAYG system are removed, raising the contribution rate makes the Japanese economy grow faster in the case of retirement due to health capital although it has no effect on growth in the case of retirement due to education capital. The raise lowers life expectancy of the economy in both cases.