With many OECD countries raising the statutory pension age to cope with increasing fiscal demands from increased life expectancy, the question of whether this extension of working life is good or bad for health is a hot research topic. One of the key problems of research on this topic is that it is hard to identify a causal effect of increasing retirement age on health and mortality. We cannot directly compare the illness and mortality rates of people who retire at different ages, as the reasons why they retire may depend on their health and related factors. So it is hard to separate out the effect of retirement age on health and mortality, when retirement itself is affected by a person’s health.
One of the standard econometric approaches to teasing apart such causal effects is to use changes in policies that affect retirement age. For example, between 1989 and 1998, Norway progressively introduced an early retirement scheme between for some employees, allowing them to retire as early as at age 62, while for other employees, the retirement age remained at 67. Crucially, this difference in the ability to retire at different ages was not related to a person’s own health, but was driven by policy changes that occurred regardless of a person’s health. The study was able to calculate the difference in retirement age induced by such policy changes in Norway, and use this difference to estimate a causal effect of retirement age on survival up to the age of 77. The results suggest that there is no effect of early retirement on mortality. Employees who retired earlier than the age of 67 as a result of changes in Norwegian employment policies had similar mortality rates as other employees who retired at 67.
The results of the paper are similar to other econometric analyses that use policy changes around retirement age to estimate a causal effect on health and survival. These studies have also found little effect of early retirement on health. However, the study cannot tell us anything about the effects of extending working life on health and survival, even though the authors frame their research around the trend of increasing retirement age in many OECD countries and the need to know the consequences for life expectancy of such policies. The authors also acknowledge that the results may not be generalisable to other countries with less generous welfare provision compared to Norway. Employees with poor health in Norway are able to retire earlier on more generous pension arrangements than their peers in other countries. The policy implications of the research are that early retirement does not lead to lower life expectancy in Norway. The question of whether later retirement leads to improving or worsening health remains unanswered.
Professor Tarani Chandola, University of Manchester
Erik Hernaes et al. Journal of Health Economics, May 2013, Pages 586–598