Does Retirement Change Lifestyle Habits?

Abstract

This paper studies the effect of retirement on lifestyle habits, including drinking, smoking, exercise, and sleeping, by using panel data from the Japanese Study of Aging and Retirement (JSTAR). Rich information in JSTAR enables us to use an interesting instrumental variable to account for endogeneity. We have three contributions in this paper. This is the first paper that focuses on and investigates the mechanism of the relation between retirement and health, namely, lifestyle habits. Second, new results show that people reduce drinking after retirement and increase sleeping time on weekdays although smoking, frequency of exercise, and sleeping time on holidays seem to be unchanged. Third, controlling important factors also allows us to inspect the detailed channels between retirement and lifestyle habits. Our estimation suggests that the peer effect in the workplace may be influential mainly on drinking habits.

Changes in Smoking During Retirement Transition: A Longitudinal Cohort Study

“…We found a subgroup of smokers who reduced their smoking intensity during retirement transition. This reduction in smoking intensity, which seems to persist for a long period of time, is in agreement with two previous Japanese studies demonstrating reduction in smoking intensity during a shorter, two [14] to four year [11] follow-up period. In addition, we observed two subgroups of people who sustained their smoking intensity, either at high or low level, throughout the retirement transition.…”

Publisher's note

“…We found that elderly changed their smoking habits in three of seven countries; decreasing after retirement in the U.S., Denmark and Japan. Our finding that in no country did retirees increase smoking is consistent with previous research, including Insler (2014), Eibich 2015, Motegi et al (2016) and Zhao et al (2017). The magnitude of the marginal effect of smoking is small in the U.S. and Japan (-0.012 and -0.026), which is evaluated for 6.7% and 13.6% at mean level for non-retirees for each country.…”

“…16 For example, Eibich (2015) restricted 16 e.g. Motegi et al (2016); over 50, Insler (2014); over 50, Eibich (2015); 55-70, Celidoni and Rebba (2017); 45-85, the sample to those aged 55-70 because he finds that the probability of retirement increases sharply at 60 to 65 years of age in Germany. It is better to narrow age range for precise estimation.…”

“…Another key to understanding these differences is a better grasp of the mechanisms through which retirement influences health outcomes and health investment behaviors (i.e., health lifestyle), and attempts to analyze these have increased recently following the Eibich (2015) study, the first to clearly identify and investigate the mechanisms rigorously, using German data. 2 Subsequent examples include Motegi et al (2016); Kämpfen and Maurer (2016); Ayyagari (2016); Celidoni and Rebba (2017); Zhao et al (2017); Bertoni et al (2018); and Kesavayuth et al (2018). 3 However, as each paper focused only on one country and applied different estimation methods, their findings cannot be directly compared or generalized.…”

What Explains the Difference in the Effect of Retirement on Health? Evidence From Global Aging Data

“…(); Insler (); Eibich (); Ayyagari (); Motegi et al . ()). Eibich () is the first study to clearly point out the importance of the mechanism to explain the difference in the effect of retirement on health.…”