The introduction of social pensions and elderly mortality: Evidence 1870-1939

The strong association between income and mortality raises the question whether more generous social security systems could improve poor people's health outcomes. Thus, in this paper, I analyze whether a major social security innovation, the introduction of social pensions targeted at poor elderly people in the late 19th-early 20th century, has reduced mortality rates of senior citizens. Therefore, I use a cross-country dataset spanning from 1870 to 1939 consisting of 13 countries of which 9 eventually implemented social pensions before World War II. Applying a difference-in-difference-in-difference as well as a regression discontinuity design, I find no evidence for a decline in elderly mortality due to the introduction of social pensions. Based on aggregate census data, I argue that social pensions have reduced elderly labor supply. The reduction is much smaller than social pension recipiency rates, though. These findings suggest that social pensions have raised elderly incomes which, however, did not translate into lower mortality.