T his paper investigates the causal im pact of retirem ent on late life depression, a grow ing concern for public health as m ajor depressive disorders are the second leading cause of disability. W e shed light on the role of econom ic conditions in shaping the effect of retirem ent on m ental health by exploiting the tim e and regional variation in the severity of the econom ic crisis across ten E uropean countries over the 2004 -2013 period. W e use data from four w aves of the Survey of H ealth, A geing and R etirem ent in E urope (SH A R E ) and address the potential endogeneity of retirem ent decision to m ental health by applying a fixed-effect instrum ental variable approach. R esults indicate that retirem ent im proves m ental health of m en, but not of w om en. T his effect is stronger for those m en w orking in regions that are severely hit by the econom ic crisis and in blue -collar jobs. T hese findings m ay be explained by the w orsening of w orking conditions and the rise in job insecurity stem m ing from the econom ic dow nturn: In these circum stances, the exit from the labor force is perceived as a relief. Abstract This paper investigates the causal impact of retirement on late life depression, a growing concern for public health as major depressive disorders are the second leading cause of disability. We shed light on the role of economic conditions in shaping the effect of retirement on mental health by exploiting the time and regional variation in the severity of the economic crisis across ten European countries over the 2004-2013 period. We use data from four waves of the Survey of Health, Ageing and Retirement in Europe (SHARE) and address the potential endogeneity of retirement decision to mental health by applying a fixed-effect instrumental variable approach. Results indicate that retirement improves mental health of men, but not of women. This effect is stronger for those men working in regions that are severely hit by the economic crisis and in blue-collar jobs. These findings may be explained by the worsening of working conditions and the rise in job insecurity stemming from the economic downturn: In these circumstances, the exit from the labor force is perceived as a relief. All the authors state that there are no financial nor personal relationships between themselves and others that might lead to a conflict of interests. Keywords

This study exploits a new dataset in order to quantify the effect of financial incentives on retirement choices. This dataset containsfor the first time in Italy -information on seniority. In accordance with the general finding in Gruber and Wise (2004), we find that financial incentives have an effect on retirement. The effect goes in the expected direction; when employees become eligible for pension benefits the change in financial incentives they experience is so high that their retirement probability increases in a sizable way.We also find that the procedure to impute seniority used in previous studies leads to a sizable measurement error. Due to this measurement error, the key parameters of the model are inconsistently estimated. Our sensitivity analysis suggests that the lack of appropriate information on seniority is an important reason for the unclear evidence so far obtained in retirement studies for Italy.

This study exploits a new dataset in order to quantify the effect of financial incentives on retirement choices. This dataset containsfor the first time in Italy -information on seniority. In accordance with the general finding in Gruber and Wise (2004), we find that financial incentives have an effect on retirement. The effect goes in the expected direction; when employees become eligible for pension benefits the change in financial incentives they experience is so high that their retirement probability increases in a sizable way.We also find that the procedure to impute seniority used in previous studies leads to a sizable measurement error. Due to this measurement error, the key parameters of the model are inconsistently estimated. Our sensitivity analysis suggests that the lack of appropriate information on seniority is an important reason for the unclear evidence so far obtained in retirement studies for Italy.

No abstract