This study evaluates the effect of delaying retirement on mortality. Despite the policy importance of the effect of working longer on health, the literature has reached no consensus. This study exploits a natural experiment in Soviet Russia. A 1964 reform allowed pensioners to keep a greater share of their pension if they continued working. In response, workers substantially delayed retirement (Malkova 2019).
I construct a novel data-set by collecting and digitizing archival data on mortality from the Soviet archives. I exploit a difference-in-differences strategy by comparing cohorts who reached pensionable age (60 for men and 55 for women) before the reform, representing my comparison group, to cohorts who reached pensionable age after the reform, representing my treatment group.
I compare the evolution of mortality rates among these treatment and comparison groups four years before and ten years after they reached pensionable age. I find that working longer increased the mortality rates among men both in the short-run and in the long-run. However, I find no effects of working longer on mortality among women. This heterogeneity may be a result of men working longer in more difficult work conditions.