http://www.econ.ucdavis.edu/faculty/dlmiller/workingpapers/millerpaxson.pdf

The central question addressed in this paper is whether mortality is affected by relative income, where relative income is defined as an individual's own income relative to the incomes of those who live in the individual's geographical area. The idea that there is a gradient in mortality, so that rich people live longer than do poor is not new, and there is ample empirical evidence of a gradient. What is more controversial is whether richer individuals live longer because they are relatively rich -so that, for example, if everyone's income doubled, mortality rates might remain unchanged- or whether life expectancy is a monotonically increasing function of individual's income levels regardless of the incomes of others. (Au)