Nursing home residents make a difference – The overestimation of saving rates at older ages

While life-cycle theory makes the clear prediction that people dissave at old-age, this prediction is not at all borne out by the data from many countries. Various suggestions have been made to explain this discrepancy. This paper sheds more light on the effect of the exclusion of institutionalized individuals in estimating saving rates over old-age, a conceptual aspect often mentioned but never investigated. Particularly this group is expected to decumulate wealth since nursing home expenses net of private (and public) insurance exceed disposable income on average. This paper uses the Health and Retirement Study (HRS) for the USA and the Income and Expenditure Survey (EVS) for Germany to show that there is an increasing overestimation of saving rates from age 75 on if institutionalized households are not included. In the USA, the overestimation of the mean (median) saving rates is 3.3 percentage points (4.3pp) at age 80, 5.4pp (9.4pp) at age 90 and even more for age 90+. The overestimation of the German mean saving rate increases to almost 6pp at age 90. This strong overestimation is based on the fact that nursing home residents strongly reduce their wealth holdings. Referring to the USA, the representative median single nursing home resident reduces wealth holdings by 90% over a two-year period; the representative mean single nursing home resident diminishes total net wealth by 19%. The dissaving is less strong for couples. The ongoing aging of industrialized populations and the connected increase in the fraction of the nursing home population will strengthen the importance of including the nursing home population to estimate aggregate saving rates in micro empirical studies.