Population Composition as an Intervening Mechanism in the Relationship between Income Inequality and Mortality in U.S. States

Elayne J. Heisler, Duke University

This paper addresses the ongoing debate about whether there is a fundamental relationship between income inequality and mortality at the state-level. An extensive body of research has addressed whether greater income inequality is associated with higher mortality rates. Although many researchers working with different populations and measures of inequality have found a strong association between income inequality and mortality, others have found weak or nonexistent relationships (see Kawachi, Kennedy and Wilkinson 1999 for review). Other researchers have also found that after controlling for key variables, specifically population composition (Deaton and Lubotsky 2002) the relationship disappears. Still much debate remains, these debates address whether the relationship exists and what mechanisms may explain the association if it does exist. This paper examines both whether the relationship exists and whether population composition is an important intervening or explanatory mechanism. The basic premise of the debate is that in developed nations more egalitarian countries will have higher levels of overall population health compared to countries like the United States where both the standard of living and level of inequality are high (Rogers 1979; Wilkinson 1992, 1996). This explanation is often used to explain why the United States lags behind less developed countries in fundamental indicators of health like infant mortality (Kawachi, Kennedy and Wilkinson 1999). However, while many have found strong evidence for the relationship others have found concluded that no relationship exists or that it is a statistical artifact (Fiscella and Frank 1997, Gravelle 1998). Even amongst those that believe that income inequality and mortality are related few conclusions have been drawn about the mechanisms that cause the association. In the absence of a coherent explanation for the relationship between income inequality and health, some have debated whether there is actually a link between income distribution and mortality. While Lynch and his colleagues (1998) find many instances where the relationship between income inequality and mortality holds both in the United States and in other developed countries, others have found that when the relationship is examined at a city, state, or census tract level the results are less robust and disappear after relevant controls are added. Deaton and Lubotsky (2002) found that the relationship between inequality and mortality is confounded by the racial composition of cities and states. They conclude that once the population composition of the area is statistically controlled for, there is no relationship. They believe that since African Americans have higher mortality rates and are more likely to be economically disadvantaged, areas with large numbers of African Americans will have higher mortality rates and higher inequality because of population composition differences and not because of a direct link between inequality and mortality. Despite this potential explanation, recent research (Ross et al. 2000) has controlled for population composition, and still finds that the link between inequality and mortality exists. Thus several long-standing debates remain in this area of research. First, does a fundamental relationship between income inequality and mortality exist, and if so, at what geographic aggregate does this relationship hold? If the relationship does exist, does controlling for population composition “explain?the relationship? Further, population composition is a multifaceted construct. Previous research has focused on racial composition (Deaton and Lubotsky 2002) or health behaviors (Kennedy, Kawachi and Prothrow-Stith 1996) but other facets of an area’s population composition may also matter. In particular, the age structure of an area may be important for several reasons. Older adults typically have universal healthcare through the Medicare program, which may reduce the linkage between income and health in older adults. Second, older adults are also less healthy and have higher mortality rates than the general population does, which may confound the relationship between income and health in areas with a larger older population. Third, social spending is a critical factor that should be examined as an intervening mechanism in the relationship between income inequality and health. Public spending may improve the plight of the poor, thus could influence mortality rates, but whether or not public spending reduces inequality will depend on which programs are invested in and if the programs are targeted at specific age groups or spread more generally in the population. Previous research has shown the importance of percent aged for determining levels of social spending and has linked that to mortality rates (Pampel and Williamson 1989). This paper will attempt to address some of these questions through examining the relationship between income inequality and health at the US state-level. This paper draws on data compiled from a variety of government sources over a 19 year time period (1980-1998) to examine the relationship between income inequality and health in the United States. The data contains specific information about the state’s population, economic climate, income distribution, degree of unionization, political context, healthcare system and other attributes. Unlike previous work that has typically focused on all-cause mortality this paper examines specific causes of death and their relationship to inequality. Specifically, infant mortality, cardiovascular diseases and diabetes, all obtained for the Center for Disease Control Compressed Mortality File (CDC 1998), will be examined because these causes of death are related to socioeconomic status through healthcare, lifestyle and education (Rogers, Nam and Hummer 1999). The analysis will employ geographic information system technology (GIS) to graphically represent the relationship between the various aspects of population composition and mortality and to document changes over time. The second part of the analysis will attempt to examine the causal mechanism in the relationship through using pooled cross sectional time series analysis to examine the influence of population composition, income inequality and mortality over time. The base models will examine the fundamental relationship between income inequality and health at the state-level. Subsequent models will examine the influences of the percentage of the population that is children, working age, and older adult on the relationship between income inequality and health. To address the specific question of the importance of universal healthcare coverage for reducing the relationship between income inequality and health in areas where there are larger percentages of older adults the percentage elderly will be disaggregated into percentage over age 65, 75 and 85. Additionally, later modeling will take into account social spending at the state-level, the state’s fiscal health, and the state’s political and social context as context is an important determinant of spending. Research on the linkages between income inequality and mortality has many implications for both health policy and economic policy. If the linkages between health and income inequality exist then policy mechanisms of income redistribution could be used to address public health problems. However, if the reality is that population composition or differences in social class explains this relationship than policy solutions would involve reorganizing society, which is a more complicated and potential less feasible policy option. This paper attempts to shed light on this underlying relationship in order to speculate about possible solutions. References: Center for Disease Control. 1998. "CDC Wonder" [Web Page]. Accessed 20 Feb 2002. Available at http://wonder.cdc.gov/. Deaton, Angus and and Darren Lubotsky. 2002. "Mortality, Inequality and Race in American Cities and States." Social Science and Medicine In Press. Fiscella, K and P Franks. 1997. "Poverty or Income Inequality As Predictor of Mortality: Longitudinal Cohort Study." British Medical Journal 314:1724-28. Gravelle, Hugh. 1998. “How Much of the Relationship Between Population Mortality and Inequality is a Statistical Artifact??British Medical Journal 316: 382-85. Kaplan, George A., Elsie R. Pamuk , John W. Lynch, Richard D. Cohen , and Jennifer L. Balfour. 1996. "Inequality in Income in the United States: Analysis of Mortality and Potential Pathways." British Medical Journal 312:999-1003. Kawachi, Ichiro, Bruce P. Kennedy, and Richard G. Wilkinson. 1999. The Society and Population Health Reader Volume I: Income Inequality and Health. New York : New Press. Kennedy, BP, I Kawachi, and D Prothrow-Stith. 1996. "Income Distribution and Mortality: Cross-Sectional Ecological Study of the Robin Hood Index in the United States." British Medical Journal 312:1004-7. Lynch, John W., George A. Kaplan, and Elsie R. Pamuk. 1998. "Income Inequality and Mortality in Metropolitan Areas of the United States." American Journal of Public Health 88:1074-80. Pampel, Fred C. and John M. Williamson. 1989. Age, Class, Politics and the Welfare State. Canada: Cambridge University Press. Rogers, GB. 1979. "Income and Inequality As Determinants of Mortality: An International Cross-Section Analysis." Population Studies 33:343-51. Rogers, Richard G., Robert A. Hummer, and Charles B. Nam . 2000. Living and Dying in the USA: Behavioral, Health and Social Differentials of Adult Mortality. New York: Academic Press. Ross, NA, Wolfson MC, Dunn JF, Berthelot JM, Kaplan GA, and Lynch JW. 2000. "The Relation Between Income Inequality and Mortality in Canada and in the United States: Cross Sectional Assessment Using Census Data and Vital Statistics." British Medical Journal 320:898-902. Wilkinson, Richard G. 1992. "Income Distribution and Life Expectancy." British Medical Journal 304:165-68. ____ 1996. Unhealthy Societies: The Afflictions of Inequality. London: Routeledge