et al., 2004). Two examples of mortality differentials in the United States that reflect largely socioeconomic differences are the substantially higher mortality among the black relative to the white population and the large geographic inequalities in health (Arias, 2010; Murray et al., 2006; Williams and Collins, 1995).

This chapter describes the gradient in mortality by socioeconomic status in the United States and other countries. It examines whether the relative size of the mortality disparity by socioeconomic status in some of this study’s comparison countries could account for the current disparities in life expectancy. It also examines whether the magnitude of the mortality gradient with socioeconomic status has changed over time and how these changes could contribute to disparities in survival.


The association between indicators of socioeconomic status, such as income and education, and mortality implies that the distribution of socioeconomic status within a country could affect mortality—in particular, that two countries with the same average income or education could have differences in health and mortality if income or education were differentially distributed. For instance, a country with greater income inequality—with more wealthy but also more poor people—may have worse average health and greater average mortality because the health benefits to the wealthy from their extra income are outweighed by the health deficits experienced by the poor. This is possible because the marginal benefits of additional income are greater for the poor than for the wealthy—an extra $10,000 per year can make a much greater difference to the health of a person earning $20,000 a year than to that of someone earning $200,000 a year. Thus when inequality is great, the decrease in life expectancy among those of lower socioeconomic status can outweigh the increase in life expectancy among those of higher socioeconomic status, leading to a life expectancy below that likely to be seen in a country with the same average level of the social indicator but less inequality (Preston, 1975; Rodgers, 1979).

Over time, a number of hypotheses have been offered concerning the precise relationship between inequality and health. The most straightforward of these, the absolute level hypothesis, holds that inequality plays no role beyond the simple one described in the previous paragraph. That is, an individual’s health is affected by his or her own socioeconomic status, but is not further affected by how the status of everyone else in the society is distributed (Kawachi et al., 2010; Lynch et al., 2004). Others have suggested that the presence of inequality itself may lead to poorer health and increased mortality for at least some of the population. Most of the pro

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement