The study explores the possible unintended health effects of reforms aimed at making eligibility criteria for occupational retirement more severe. The causal link between retirement age and hospitalization for cardiovascular diseases is investigated in a large sample of male Italian retirees (N=94,521). Endogeneity is addressed by an Instrumental Variable identification strategy, in a quasi-natural experiment set-up. The instrument exploits the variation in pension age determined by the standardization of the labour market transitions, which induce workers born during the first months of the year to retire at an older age. The analysis is performed on a longitudinal dataset that combines several Italian administrative archives on pensions, working histories and hospitalization.
Results show a significant health detrimental effect of extended working life. A one-year delay in retirement increases the incidence of hospitalization for cardiovascular diseases (CVD) at 68-70 years old by 2.4 percentage points (p-value<0.01). Retirees who, during their careers, were lower income earners, mainly employed in the secondary sector and in manual occupations, are the groups paying the highest price for staying longer at work, as for them the impact of pension age on CVD is even higher. Sensitivity analyses show that results are robust to different model specifications; to the inclusion of career controls and to seasonality.