Citation

Deaton A, Paxson C. NBER Work. Pap. Ser. 1999; 1999(online): w7140.

Copyright

(Copyright © 1999, National Bureau of Economic Research)

DOI

unavailable

PMID

unavailable

Abstract

People whose family income was less than $5,000 in 1980 could expect to live about 25 percent fewer years than people whose family income was greater than $50,000. We explore this finding using both individual data and a panel of aggregate birth cohorts observed from 1975 to 1995. We assume that health status is determined by social status, defined as income relative to the mean income of a reference group. When reference groups are not observed, health is a function of income whose slope (the gradient) depends on the ratio of within to between-group inequality. We derive results on how this relationship changes at different levels of aggregation. Our results on individuals show that income reduces the risk of death, and does so even controlling for education. Only some of the effect of income can plausibly be attributed to the reduction in earnings of those about to die. The panel of cohorts also shows a strongly protective effect of income, but there is evidence that cyclical increases in income may raise mortality, even when the long-run effects of income are in the opposite direction. There is no evidence that recent increases in inequality raised mortality beyond what it would otherwise have been.