Health Investments and Retirement – University of Copenhagen

Forward this page to a friend Resize Print Bookmark and Share

CHEP > Current research projects > Health Investments and...

29 April 2015

Health Investments and Retirement

The world population is aging. Within the next few years UN projections suggest that there will be more people on the planet above the age of 65 than below the age of five (Kinsella and He, 2009). Probably for the first time in human history. Whereas declining fertility, mortality and rising urbanization were the major demographic trends of the 20th century, "aging" is sure to be one of the key trends for the 21st.

An aging society is likely to have important economic implications, most obviously perhaps in ways of the sustainability of the Welfare state. In this context key issues relate to the likely path of health expenditures and years in retirement. But beyond issues pertaining to fiscal sustainability: what might be the implications of aging for average consumption and savings behavior? For learning and human capital accumulation? In order to think about issues such as these it seems necessary to allow for economic models where agents age in a physiological meaningful way. That is, models where aging is defined as "the intrinsic, cumulative, progressive, and deleterious loss of function that eventually culminates in death".

Mette Ejrnæs, Carl-Johan Dalgaard, and Nick Fabrin Nielsen from UCPH collaborate with Holger Strulik from University of Göttingen on this project financed by The Danish Council for Independent Research:

"In this project we plan on examining the influence of retirement on health investments as well as frailty. Various theoretical predictions have been made regarding the impact from retirement. Wolfe, for instance, suggests that retirement should be associated with a discontinuous upward "jump" in health investments, in contrast to the predictions in Dalgaard and Strulik where no such jump ought to occur. In any event, we wish to look into what the empirical facts are in this regard.

Previous research has examined the link between retirement and measures for the stock of "health", which we can supplement with new results pertaining to the frailty index. Moreover, we are not aware of studies that examine the link between retirement and health investments. We plan on proxying health investments by medical expenses, in keeping with the existing literature.

The empirical part of the project analyzes the impact of retirement on health investments. There exists a huge literature on the impact of health on retirement and a smaller but growing literature on the impact of retirement on mortality or health status. Most of these papers find a negative correlation between retirement and health or mortality, suggesting that those who go on early retirement have a worse health or die earlier. However, the retirement decision is likely to be endogenous to the health status and when accounting the endogeneity most of the papers find that the impact of the retirement either has no effect or a small positive effect on health.

We will analyze the impact of retirement on health investment using Danish register data. To deal with the endogeneity of the retirement decision, we will use two major reforms of the retirement system in Denmark that took place within our sample period. The first reform was in 1999 and decreased the age for old-age-pension from 67 to 65. Furthermore, it made it more attractive to postpone the entry into an early retirement scheme starting from the age of 60. The second reform in 2006 increased the old-age-pension to 67 again and also made it less attractive to go on early retirement. By using the exogenous variation in the retirement age imposed by the reforms, we can identify the causal effect on health investments in a reduced form model

The contribution of this project will be threefold. First we will examine how health investments are affected by retirement. This will provide more knowledge on the mechanisms that drive the results of the effect of retirement on health status and mortality. Second, this study links the empirical findings to the theoretical models. This is done by comparing the theoretical implications with the empirical facts. Finally, we will estimate the parameters of theoretical model lay-out in this project."

Contact: Mette Ejrnæs
Project period: 2013 - 2017