1985-2011: Elderly West Germans were much less satisfied throughout their last 5 years of life & spent a larger proportion of their remaining lifetime in dissatisfaction; mechanisms that contributed were health & social isolation

Increasing longevity and life satisfaction: is there a catch to living longer? Janina Nemitz. Journal of Population Economics, May 3 2021. https://rd.springer.com/article/10.1007/s00148-021-00836-3

Abstract: Human longevity is rising rapidly all over the world, but are longer lives more satisfied lives? This study suggests that the answer might be no. Despite a substantial increase in months of satisfying life, people’s overall life satisfaction declined between 1985 and 2011 in West Germany due to substantial losses of life satisfaction in old age. When compared to 1985, in 2011, elderly West Germans were, on average, much less satisfied throughout their last five years of life. Moreover, they spent a larger proportion of their remaining lifetime in states of dissatisfaction, on average. Two important mechanisms that contributed to this satisfaction decline were health and social isolation. Using a broad variety of sensitivity tests, I show that these results are robust to a large set of alternative explanations.

Conclusion

Given the rapid increase in human life expectancy throughout the last decades, this study asks: Are longer lives more satisfied lives? Using data from the German Socio-Economic Panel, this study suggests that the answer might be no. Although expected satisfied lifetime increased for West Germans at the age of 60 by two years between 1985 and 2010, this increase likely did not compensate for the substantial losses of life satisfaction that occurred at the end of people’s lives. In 2010, average life satisfaction scores in the last five years of life were roughly one-third to half a standard deviation lower than in 1985. Moreover, the period of terminal satisfaction decline was substantially longer. With, on average, slightly more than two additional dissatisfied life years, 60-year-old survivors in 2010 were expected to spend 10% more of their remaining lifetime in states of dissatisfaction, which suggests a drop in this study’s measure of overall quality of life. Nevertheless, in 2010, the proportion of expected satisfied lifetime to expected total lifetime at age 60 was still relatively high, with an average level of about 65% to 70%.

To better understand what contributed to the decline in terminal life satisfaction in West Germany, I explored the role of two potential mechanisms: health and social isolation. Several health indicators (e.g., severe disability, number of hospitalizations) indicated a deterioration of the end-of-life health status over time and, thus, provided support for the expansion of morbidity hypothesis (Gruenberg 1977; Olshansky et al. 1991). Among all health indicators, the increase in legally attested disability had the most detrimental impact on terminal life satisfaction. All measures of social isolation contributed to the decline in terminal satisfaction over time, but individual-level life satisfaction regressions indicated that increased isolation mainly worked through the health channel. These results are in line with studies that show that the onset of disability relates to a lasting well-being decline (Lucas 2007; Oswald and Powdthavee 2008), and that a socially active life is associated with higher late-life well-being, less pronounced late-life decline, and a later onset of terminal satisfaction decline (Gerstorf et al.2016).

One likely explanation for the findings of this study is the decline in sudden death. Over the last three decades, age-standardized mortality from ischaemic heart disease has fallen by more than half in high income countries (Finegold et al. 2013; Hartley et al. 2016). So while in earlier times, when smoking was still very common, quite healthy and satisfied people suddenly dropped dead, nowadays new medical technologies (e.g., drug-eluting stents) allow the medical profession to extend people’s lives even with disease. As a consequence, people are much more likely to experience novel types of diseases as well as an increased burden and complexity of multimorbidity (WHO 2010). Moreover, thinking of slowly progressing diseases such as Alzheimer’s disease or dementia that come with a progressive decline in memory and cognitive function and eventually lead to severe disability (Alzheimer’s Association 2016), it is very plausible that people are much less satisfied throughout their final period of life nowadays. Overall, higher dissatisfaction levels may result from the burden associated with disease, including increased social isolation, as well as the fact that people know that there is no cure or modifying treatment for a disease (Daviglus et al.2010).

The final conclusion that the overall quality of life decreased with improved longevity between 1985 and 2011 rests on a very strong assumption, namely that at the age of 60, people value a high proportion of satisfied to total lifetime more strongly than the actual number of satisfied life years. Although consistent with the literature (Pennington et al. 2015; Ahlert et al. 2016; Fischer et al. 2018), this assumption may not hold. Another related issue is that of acceptable satisfaction levels. Many people would argue that satisfaction levels above the neutral (i.e., 5 on the 0 to 10 scale) are still quite satisfying and, thus, the elderly in our sample, though more dissatisfied in the final period of life, were still quite satisfied in 2011. However, there is a well-documented issue of over-reporting satisfaction scores in surveys with face-to-face interviews (see Diener et al. 2013for review). Even satisfaction scores of about four (on the 0 to 10 scale) may be predictive of suicide (Koivumaa-Honkanen et al. 2001). If people nevertheless believe that life is still satisfying at very low satisfaction scores (i.e., three and lower), then the final conclusion of this study will no longer hold.

Should people and policymakers further invest in life extensions? This study shows that it is important to complement investments that extend the length of human life with investments that improve the quality of life in old age. Under-investments in the latter result in declining satisfaction levels at the end of people’s lives. Quality-of-life-improving policies may have a more positive effect on increasing overall life satisfaction. This is because they would increase satisfaction during a person’s lifetime, and furthermore, may also extend the length of life itself since more satisfied people tend to live longer (Veenhoven 2008; Steptoe et al. 2015).

Which quality-of-life-improving policies should be targeted? As suggested by the analysis of potential mechanisms, potential candidates would be policies that aim to prevent noncommunicable diseases (e.g., via reduced tobacco use, healthy diets, or physical activity) and policies that aim to achieve a better integration of the elderly in today’s societies (e.g., via better provision of public transportation in remote areas). Further research on these and other potential mechanisms is required to decide upon the policies that are most promising. Moreover, future research needs to explore potential heterogeneity in order to better target policies to groups of recipients.