Impact of personal income on mortality: Decomposition into biological

Humans have two facets, biological and social. Whereas it is a common idea that poor social conditions affect the biological facet for instance by reducing life expectancy, there are few known cases where an economic effect is affected by the biological facet in a way which can be predicted. The purpose of the paper is to present such a case. In other words, we are going to decompose an economic phenomenon into its biological and social components, a step which provides a marked conceptual simplification. The economic phenomenon that we consider here is one of the most basic that one can think of, namely the relationship between personal income and life expectancy. Intuitively, one is not really surprised that wealthy people live longer than poor people. Here, however, we show an effect which is far less obvious, namely the fact that this relationship does not hold at both extremities of the lifespan interval. The disconnection between income and neonatal (i.e., in the first 28 days after birth) mortality is quite impressive. This observation is particularly significant on account of the fact that the infant mortality (i.e., in the first 365 days after birth) is often taken by economists as a proxy of development when no reliable income data are available. This indicator may be valid for very poor countries where the burden of death due to infectious diseases remains very high, but it is not valid in developed countries such as those considered in this paper. More specifically, we explore the influence of income on mortality by age, separately in France, the United States and South Korea. The same pattern appears for the curves of the (income-mortality) correlation as a function of age. We conjecture that this pattern will be observed in any developed country where the test can be performed. For the test to be possible the main requirement is the availability of income and age-specific mortality data at regional level