Life Expectancy, Income and Long-Term Care: The Preston Curve Reexamined

The Preston Curve - the increasing relation between income per capita and life expectancy - cannot be observed in countries where old-age dependency is widespread (that is, where long-term care (LTC) spending per capita is high). The absence of the Preston Curve in countries with high old-age dependency can be related to two other stylized facts: (1) the inverted-U relation between LTC spending and life expectancy; (2) the inverted-U relation between LTC spending and preventive health investments. This paper develops a two-period OLG model where survival to the old age depends on preventive health spending chosen by individuals while anticipating (fixed) old-age LTC costs. In that model, anticipated LTC costs are shown to have a non-monotonic effect on preventive health investment, thus rationalizing stylized facts (1) and (2). This framework is shown to provide an explanation for the absence of the Preston Curve in countries where old-age dependency is more acute.