How long retirement lasts matters: the effect of retirement duration on frailty in Europe.

Many OECD countries have implemented retirement policies aimed at increasing the minimum retirement age, with the dual objectives of improving older workers' employment rates and ensuring the financial sustainability of pension systems. However, the long-term effects of retirement on health - and particularly on the onset of frailty - remain underexplored. In this paper, we examine the impact of retirement duration on the probability of being frail using an instrumental variable approach and assess heterogeneity by gender and occupation. Our analysis draws on longitudinal data from seven waves of the Survey of Health, Ageing and Retirement in Europe (SHARE), spanning 2004-2019, and restricted to individuals aged 50-75. We instrument retirement duration using both the number of years since the early retirement age and the number of years since the statutory retirement age. The results reveal a non-linear relationship between retirement duration and frailty: in the short and medium term, retirement reduces the likelihood of being frail, whereas in the long term, it increases this likelihood. The marginal effect of retirement duration becomes positive after approximately 4.4 years of retirement, indicating that beyond this point, each additional year of retirement is associated with a higher probability of being frail. These findings suggest that while retirement may initially benefit health, prolonged retirement can impact health negatively, highlighting the need for preventive programs targeting individuals exiting the labor market early.