Art der Publikation: Beitrag in Zeitschrift

Ageing and Health Care Costs

Autor(en):

Karlsson, Martin; Iversen, Tor; Øien, Henning

Titel der Zeitschrift:

Oxford Encyclopedia of Economics and Finance

Veröffentlichung:

2018

Digital Object Identifier (DOI):

doi:10.1093/acrefore/9780190625979.013.24

Zitation:

Download BibTeX

Kurzfassung

There has been a rise in recent research regarding the economics on the consequences of population ageing for health care expenditure. The focus of this research is on expenditures in the last years of life and the implications for the age profile of health care spending. One open issue in the literature is whether health care expenditure is so concentrated in the last years before death that the age profiles in spending will change when longevity increases. The previous literature also finds that the age profile of health expenditures, after controlling for time to death, is dependent on the type of health care expenditure that is being predicted. Some studies find that the age is more powerful in predicting long-term care compared to health care expenditures. However, this result appears to depend on the empirical model used.

There are conflicting findings that can be reconciled by carefully reviewing and validating the methods that have been used in the literature. In particular, there are advantages of using a two-part model for health care expenditure, in which health care utilization is allowed to have different determinants on the extensive and the intensive margin. This can be seen when this standard modeling approach is compared to alternatives that require weaker assumptions at the cost of possibly inferior fit to the data. Different proxies are evaluated for morbidity affect the estimated end-of-life costs and age profiles in spending. Using a high-quality health insurance dataset from Germany, which includes all relevant determinants of health and long-term care expenditure, the accuracy of popular methods in predictions of future health and long-term care spending are assessed. The basis for this validation exercise is the prediction accuracy within the same dataset. The endogeneity of death is an issue that has not been addressed satisfactorily in the literature, and there is a solution to the problem based on standard approaches to causal inference.