Understanding Racial Disparities in Health: The Income-Wealth Paradox

We examine the ways in which racial differences in health vary over the income-wealth distribution, comparing the self-reported health status of non-Hispanic whites with those of individuals of other races and ethnicities. Paradoxically, we find that although the largest unadjusted racial differences in health are between poor whites and poor nonwhites, after adjusting for income, wealth, and other demographic characteristics, health differences between nonwhites and whites are only significant among those in the upper half of the income-wealth distribution. The results suggest that unexplained racial differences in reported health status increase with socioeconomic status among individuals aged 25–54. Understanding Racial Disparities in Health: The Income-Wealth Paradox INTRODUCTION There are large and well-documented differences between the health of non-Hispanic whites and the health of other racial and ethnic groups. In the United States, African Americans have particularly poor health outcomes compared to whites. African Americans have shorter life expectancies and are more likely to live with a chronic condition than whites. Like African Americans, Hispanics and Native Americans have higher rates of diabetes and infectious diseases, such as tuberculosis and syphilis, than white individuals (Keppel, Pearcy, and Wagener 2002). A significant portion of these differences, however, may be due to differences in socioeconomic status (SES). On average, nonwhites have lower incomes, lower levels of education, and less wealth than non-Hispanic whites. A relatively large literature has explored the role of SES in explaining average racial differences in health. However, given the nonlinear relationship between income, wealth, and health, it is plausible that the role of SES in explaining the health gap differs over the income-wealth distribution. This possible distributional variation is explored below by assessing unadjusted and adjusted racial differences in self-reported health status at various levels of wealth and income. We also investigate what may account for apparent health differences. Many researchers have found that income, wealth, and education can explain a large portion, if not all, of the mean racial disparities in health regardless of the measure of health, sample, or methodology employed. (See Hayward, Crimmins, Miles, and Yang 2000; Smith and Kington 1997; and Williams and Collins 1995, for reviews.) There is variation, however, regarding the degree to which SES accounts for all racial differences, and whether any significant differences remain. Kington and Smith (1997), for example, find that while racial disparities in functional limitations (among those with chronic illnesses) can be fully explained by income and wealth differences, unequal prevalence of chronic conditions remains to some extent. 2 The portion of the race gap that can be explained by SES also depends on how SES is defined. Relatively few studies examining the racial gap in health have incorporated wealth into their analysis. Those that have, such as Kington and Smith (1997), Bond Huie, Krueger, Rogers, and Hummer (2003), and Schoenbaum and Waidmann (1997) have found that wealth helps explain a significant portion of racial differences in both morbidity and mortality. Such results are expected, as the racial gap in wealth is much larger than the racial gap in income (Scholz and Levine 2004), and wealth has been shown to be a significant correlate of health over and above income and education, particularly among non-Hispanic whites (Robert and House 1996; Wenzlow, Mullahy, Robert, and Wolfe 2004). Decompositions of racial differences in this literature typically analyze differences at the mean. That is, such research answers the question, What portion of the difference between black and white average health outcomes can be explained by SES? Although conceptually useful, this approach is limiting in that the relationship between health and SES greatly differs over the distribution of income and wealth, and decomposing averages may mask important variation. The concave character of the health gradient in income, reflecting a declining marginal product of income in producing health, is well known. Particularly large health disparities among those with little wealth have also been documented—among the working-aged population (Wenzlow et al. 2004) as well as among individuals nearing retirement (Kington and Smith 1997). This suggests that income and wealth may play a different role in explaining racial health disparities among the poor than among individuals who are better off. The value of understanding distributional differences is that such knowledge can assist in identifying populations that have particularly unequal health outcomes or are more susceptible to having poor health, and may give insight to why racial disparities exist and suggest policy interventions. To our knowledge, the only research that addresses health and the distribution of SES has been on infant birth outcomes. For example, Kleinman and Kessel’s (1987) research on low-weight births found the largest unexplained racial gap in low-weight births among highly educated women, and the smallest differences among those not completing high school. This unexpected pattern of racial disparities by 3 education has been replicated in a number of studies in the context of infant birth weight (Krieger, Rowley, Hermann, Avery, and Phillips 1993). Since education is highly correlated with income and wealth, this suggests that a similar pattern may be observed between health and economic variables. Our objective is to evaluate whether the racial gap in health varies by SES among individuals aged 25–54. Are unexplained racial differences in health larger among the rich or the poor? Our findings suggest that in the United States, although the largest unadjusted health differences between races are among the poor, most of these differences can be explained by income, wealth, and education. Unexplained health differences between poor whites and poor nonwhites are relatively small. Instead, among richer Americans, racial differences persist even after holding income and wealth fixed. This pattern of results can be accounted for by a stronger estimated relationship between income, wealth, and health among whites compared to nonwhites, and by the important role played by wealth as a determinant of health differences among the poor.