Does retirement (really) lead to worse health among European men and women across all educational levels?

Soc Sci Med. 2016 Feb:151:19-26. doi: 10.1016/j.socscimed.2015.12.018. Epub 2015 Dec 15.


Objectives: Whereas the view that retirement would have negative effects on health is widespread, many existing studies may be biased because they do not sufficiently take into account the issue of reverse causation. Using a large longitudinal dataset for twelve Western European countries, this study uses an instrumental variables approach to assess effects of retirement on health.

Methods: Longitudinal data for the years 2009-2012 for 75,722 men and 63,911 women from twelve Western European countries are derived from the European Union Statistics on Income and Living Conditions. Health outcomes used in this study are self-rated health, activity limitations and chronic conditions. Country- and sex-specific early- and full-pension ages are used as instruments. Models were stratified by sex as well as educational levels to assess potential effect heterogeneity.

Results: Results obtained from conventional random-effects models suggest that retired men and women have higher chances of reporting bad self-rated health, activity limitations as well as chronic conditions. However, using an instrumental variables (IV) approach the results suggest that retirement can lead to health improvements in self-reported health as well as activity limitations among men and women. The health improvements associated with retirement among men and women exist across all educational levels.

Discussion: Contrary to several previous studies, the results suggest that retirement may have health preserving effects. The positive effects of retirement and health exist for low as well as high educated men and women.

Keywords: Education; Europe; Health; Instrumental variables; Retirement.

MeSH terms

  • Aged
  • Aged, 80 and over
  • Europe
  • Female
  • Health Status*
  • Humans
  • Male
  • Pensions / statistics & numerical data
  • Retirement*
  • Self Report*