Saved by retirement: Beyond the mean effect on mental health.

We analyze the causal effect of retirement on mental health, exploiting differences in retirement eligibility ages across countries and over time using data from the Survey of Health, Ageing and Retirement in Europe. We estimate not only average effects, but also use distributional regression to examine whether these effects are unequally distributed across the mental health distribution. We find unequally distributed protective effects of retirement on mental health. These gains are larger among those just below and above the clinically defined threshold of being at risk of depression. The preserving effects are larger for women and blue collar workers. Our results suggest that the magnitude of the protective effect is independent of the availability of family support.