Economic Status

Summary prepared by Judith Stewart in collaboration with the Social Environment working group. Last revised, December, 2002.

Chapter Contents
a. Background
b. Measurement approaches
c. Comments
d. Bibliography

Background
Education, occupational status and income are the most widely used indicators of socioeconomic status (SES). Though moderately correlated, each of these measures can capture distinctive aspects of social position, and they are not interchangeable. Income has been used widely as a measure of SES, with the most typical income-based measure being a household's total cash income, measured over some period of time such as a month calendar year, or the 12-month period preceding measurement. Some researchers suggest that income is perhaps the strongest and most robust predictor of health (McDonough, Duncan, Williams & House, 1997; Lantz,  House, Lepkowski, Williams, Mero & Chen, 1998) because to some degree the impacts of other SES variables are mediated through it (House & Williams, 2000). Others would disagree, since a strong case can be made that education alters health-related behavior and some psychosocial factors, such as master/control, and that these influence health independent of eduation's effect on income.

In assessing socioeconomic status, and more particularly economic status, measuring variables other than household income may be useful, for example assets such as inherited wealth, savings, employment benefits, or ownership of homes or motor vehicles (Berkman & Macintrye, 1997). While income represents a flow of resources over some period of time, wealth captures the stock of assets at a given point in time, and thus economic reserves. Wealth is a source of economic security providing an index of a household's ability to meet emergencies or absorb economic shocks such as unemployment. However the importance of wealth as a source of economic security may vary among societies (e.g., the vast majority of people in Sweden have relatively little wealth, but the social welfare system provides the resources to absorb economic shocks). Income and wealth are positively correlated, but they are not interchangeable, as shown by the example of an elderly person with a modest fixed income but substantial accumulated wealth.

Evidence of association between income and adult mortality.

It is a common finding that mortality has a strong inverse association with income. Rogot, Sorlie, Johson, & Schmitt (1992), using data from the National Longitudinal Mortality Study (NLMS), showed that people whose reported family incomes in 1980 were less than $5,000 in 1980 prices are estimated to have a life expectancy around 25 percent lower than those whose family incomes were over $50,000. Some investigators suggest that the relationship between socioeconomic status and health is best characterized as a linear gradient of risk, with even those in relatively high socioeconomic groups having better health than those just below them in the social hierarchy (Adler, Boyce, Chesney, Cohen,  Folkman,  Kahn & Syme, 1994; Marmot, Smith, Stanfsfeld, Patel,  North, Head, White,  Brunner & Feeney,1991). Many studies however have indicated that the relationship of socioeconomic position to health, especially when indexed by income, is monotonic, but not linear. Backlund, Sorlie & Johnson (1996) showed that small differences in income are associated with much larger changes in health status among low income as compared to high income families.  House and Williams (2000) report that a number of other studies have shown diminishing or even non-existent relationships of income with mortality (Wolfson, Rowe, Gentleman & Tomiak, 1993; Chapman & Hariharan, 1996; McDonough et al., 1997) and morbidity (House, Kessler, Herzog, Mero, Kinney & Breslow, 1990; Mirowsky & Hu, 1996) at higher levels of income. House, Lepkowski, Kinney, Mero, Kessler & Herzog (1994) suggest that there is a "ceiling effect"; that people throughout the upper socioeconomic strata maintain overall good health until quite late in life, leaving little opportunity for further improvements in average health among those who are especially wealthy. Backlund, Sorlie & Johnson (1999) shed additional light on the nonlinear functional form of income's relationship with mortality. Using data from the National Longitudinal Mortality Study (NLMS) they found that a two-sloped function better described the association between income and mortality than did a linear function for both men and women. The decrease in mortality associated with a US $1,000 increase in income was shown to be much greater at incomes below US $22,500 than at incomes above US $22,500. Ecob & Davey Smith (1999) using the Health and Lifestyle Survey (a national sample survey of adults in England, Wales and Scotland, 1984-85) demonstrate that indices of morbidity are approximately linearly related to the logarithm of income, in all except very high and low incomes. They found that throughout the middle 80% of the income distribution (i.e., from the 10th to the 90th percentiles) a doubling of income is associated with a similar positive effect on health.  These studies taken together argue that assuming a constant effect per unit change in income, or using income as a simple continuous linear variable, may be inappropriate (Krieger, Williams & Moss, 1997).

The impact of fluctations in income on mortality risk has received relatively little attention but is an important topic. McDonough et al (1997), using the U.S. longitudinal Panel Study of Income Dynamics (PSID) of adults 45 years and older, found that persistent low income was a particularly strong determinant of mortality. They also showed however that income instability was an important predictor of mortality particularly among middle-income adults. They suggest that income fluctuations may be more normative at lower incomes, or may be ameliorated by community support or public aid, while at higher levels of income, individuals may have accumulated assets, that is economic reserves, that can compensate for lost income. Middle-income adults, on the other hand, are less likely to have either of these resource types available to help bridge times of income instability. (See K. Newman's (1988) "Falling from grace: The experience of downward mobility in the American middle class" for a qualitative treatment of this topic.)

Evidence of association between assets and adult mortality.
Wealth, in the form of assets such as inherited wealth, savings, stocks and bonds, and employment benefits, has been less frequently used as a measure of  household economic status, largely because it is more difficult to collect these data, and investigations of the relationship between health and wealth (as opposed to income) are a relatively recent phenomenon in the U.S. and elsewhere (Krieger et al.1997). Wealth is often an indicator of income over the life course, and thus may be a better indicator of overall socioeconomic status than is contemporaneous income. Households with comparable lifetime incomes may differ on sources of wealth (e.g., inherited wealth, patterns of savings, and differential rates of return on savings), and these sources may vary by age, race/ethnicity and gender. Retired and elderly individuals may have low pension or social security incomes but substantial accumulated wealth. Krieger et al. (1997), reporting on data from SIPP, note that in 1991 the median net worth of US white households was 9.6 times that of black households, and 8.3 times that of Hispanic households; that of married households was 4.1 times that of female-headed households. These racial/ethnic inequalities were more evident among households in the lowest income quintile (e.g., median net worth of white households equaled $10,257, as compared to only $1 for black households and $645 for Hispanic households). [For more details see: Eller, T.J. (1994). Household wealth and asset ownership: 1991. US Bureau of the Census. Current Population Reports, Ser. P70-34. Washington DC: US GPO.]

Kington & Smith (1997), in their study of socioeconomic status and racial/ethnic differences in functional status associated with chronic diseases, emphasize that household income and household wealth have sizable independent relationships with both the likelihood of experiencing a chronic condition and the number of functional limitations for those with these conditions. In addition, the relationships of income and wealth with these health outcomes are highly nonlinear, with the greatest influence of these SES factors shown in the poverty and near-poverty population. They found that income and wealth disparities associated with the presence/absence of a chronic condition are much larger among women than among men and also appear to be larger among African Americans than among Whites (based on a cross-sectional analysis of a national sample of men and women aged 51 through 61 from the 1992 Health and Retirement Survey).  Other studies that have used wealth or permanent income in health research in the U.S. and Canada include Robert and House (1996), Schoenbaum and Waidmann (1997), and Wolfson et al. (1993).

Wealth can also be assessed by classifying people according to household assets such as whether the family home is owned or rented, and whether there is a car or garden. In Britain, markers of low available income, such as not being a home owner or having access to a car, are strongly associated with increased mortality risk. Among studies that have used such measures are Arber & Ginn (1993) and Marmot, Ryff, Bumpass, Shipley & Marks (1997).

The above mentioned finding by McDonough et al (1997) that income instability has its greatest impact on the health of those at middle income levels is consistent with the buffering effect that accumulated assets can provide.

Non-simplicity of the relationship between economic status and health.
Knowledge of household income may not be predictive of family purchasing power or the income available to individual household members. Studies have shown that goods and services available in lower income neighborhoods and African-American neighborhoods are often poorer in quality and costlier than those available in higher income and white neighborhoods (Kaplan, 1996; Troutt, 1993). In a study of two socially contrasting localities in Glasgow City, Sooman & MacIntyre (1993) reported that in the more deprived neighborhoods the price of healthy foods is higher, the availability and quality of fruits and vegetables are lower, and the price differential of a "healthy food basket" and an "unhealthy food basket" is greater.

Gender may also affect the availability of income within the household. Studies have shown that poor and working class family mothers may skimp on using available money for their own needs to provide first for the needs of their children and husband (Krieger et al., 1997).

The effect of health on income (reverse causality, selection, drift) although probably a minor contributor to the overall association of economic status and health,can have important consequences for some people (Smith, 1999). Economists and others have documented the effect of poor health on earnings in some contexts. For example, disability is a major cause of low income and poverty, and ill health is not infrequently the proximate cause of retirement  (Angus Deaton, 2002).

How might the flow of resources in the form of income and economic reserves in the form of assets benefit health?
Income and wealth represent material resources, potential access to different lifestyles, a sense of security, and contribute to a sense of power and control. [See Mirowsky, Ross & Reynolds (2000) for a discussion of the resource substitution which occurs between education and economic status, in which they propose that "education apparently acts as an alternative to income" (pg. 59).] Economic resources support the discretionary use of time for health-promoting and leisure activities (e.g., having sufficient funds to hire assistance in household upkeep).  They widen the range of options available to cope with unexpected stressors (e.g., household and automobile repairs), and increase the individual's ability to integrate key multiple roles in mutually accommodating ways (e.g., the ability to satisfy the demands of a job and the demands associated with the care of young children (Pearlin, 1999)). For those who experience a chronic health condition economic resources may allow them to alter their environment to reduce the impact of changes in physical functioning, and to moderate environmental exacerbation of such conditions. Wealth in the form of accumulated assets, which is not impacted as is income by reductions in employability due to chronic conditions, would be of particular importance in this case.

Income and economic reserves probably impact access to primary, secondary and tertiary care. Income/wealth is probably associated with obtaining routine screening for blood pressure, cholesterol, mammography, prostate screening, having routine physicals and receiving vaccinations (e.g., children's vaccines, plus tetanus, hepatitis, flu vaccines in adults). Those with substantial income/wealth also have greater access to expensive treatments, premier medical experts, and care in premier institutions.

Measurement approaches
Income
Income can be used as a quantitative variable or grouped into categories. The categorical approach is more common since individuals tend to be reticent about providing exact income information or don't know it, but are less uncomfortable indicating their placement in categories. Despite the use of the categorical approach to income responses, refusal rates are higher than for the other two commonly used indicators (i.e., education and occupation). Liberatos, Link & Kelsey (1988) report that data from the General Social Survey conducted by the National Opinion Research Center between 1972 and 1987 show a refusal rate of 9 percent for income items.

Categories are often determined by the expected range of incomes of participants in the sample under study. This fact reduces comparability across studies since the ranges of income levels are affected by the geographic area of the study, the characteristics of the study respondents, and the time period under study. For purposes of analysis, income categories are usually recoded to their midpoints and often are transformed to logarithms.

An important consideration in the construction of survey items is the scope of the income sources the respondent should consider when determining "household income". Questions about income received from jobs, social security, retirement annuities, unemployment benefits, and public assistance are fairly standard, and these sources would probably occur to a respondent even if not specifically probed. Income sources such as interest dividends, income from rental properties, child support and alimony might less frequently be spontaneously considered in a calculation of household income. In addition, household income may also include income earned from the "informal economy" (e.g., jobs that pay cash but have no benefits or job security) particularly in communities of recent immigrants and minorities, as well as informal transfers (e.g., of goods and services). These latter two income sources may be ones that respondents do not wish to disclose or for which they would have difficulty determining a monetary value.

Household incomes cannot be compared without knowledge of the size of the household. The impact of a given income is significantly dependent on family size and composition. A total household income of $30,000 would mean something quite different to a family of two and a family of eight. It also means something different depending on whether one breadwinner earns all/most of the income while the other is able to attend to other household responsibilities versus if two adults have to work full-time to earn this income. While some researchers ignore the issue of family size and composition, others divide the total household income by the number of household members to produce a per capita income. This tends to overcompensate because the costs of maintaining a given standard of living do not increase propostionately (there are "economies of scale"). Other researchers, such as Tim Smeeding, suggest an intermediate adjustment, dividing the family income by the square root of the family size. This approach suggests that a family of four needs about double the income of a single person to have the comparable standard of living. [For further discussion of equivalence scales see: Buhmann, B., Rainwater, L., Schmaus, G. & Smeeding, T. (1988). Equivalence scales, well-being, inequality and poverty: Sensitivity estimates across ten countries using the Luxembourg Income Study database. The Review of Income and Wealth; vol. 34(2), 115-142.]

Wealth
The measurement of wealth in the form of assets such as inherited wealth, savings and benefits is much less frequent than household income. Wealth can also be assessed by classifying people according to household assets such as whether the family home is owned or rented, and whether there is a car or garden. Wealth in the form of assets may be offset by accumulated debt, thus suggesting that getting a sense of the balance of assets to debt is important. Some people's wealth derives from their ability to borrow, or find investors, very large sums of money to invest; and in these less frequent cases they may be "making a living" off borrowed wealth. Examples of questions to assess income and wealth appear in Table 1.

Comments
Each socioeconomic indicator has its own set of advantages and limitations (Krieger, Williams & Moss, 1997; Berkman & Macintyre, 1997). The advantages of using income include:

• Analyses using income are likely to be open to reverse causation arguments
• Income is a more unstable measure of SES than education or occupation, and is sensitive to changes in life circumstances (thus the advantage of using, for example, 5-year income)
• Level of current income is age dependent, tending to increase up to age 65 (or retirement)
• Income information is especially sensitive for some people, resulting in greater errors in reporting and non-response for income questions than for some other SES indicators
• Measuring income well can be costly and time consuming
• Income varies within occupations and is only moderately correlated with education
• Income measures fail to include income earned from the "informal economy", informal transfers, and assets (e.g., inherited wealth, savings, benefits, or ownership particularly of homes and motor vehicles)

The advantages of including a wealth measure when determining socioeconomic status are:

• Wealth may be more strongly linked to social class position than earned income
• Wealth may be associated with health independent of other SES indicators

The limitations of including a wealth measure parallel some of those associated with the indexing of income:

• Given the multiple categories that may contribute to wealth assessment this may be a difficult calculation for respondents
• Wealth information is especially sensitive for some people, resulting in greater errors in reporting and nonresponse for wealth questions than for some other SES indicators
• Measuring wealth well can be costly and time consuming

Bibliography

Adler, N.E., Boyce, T., Chesney, M.A., Cohen, S., Folkman, S., Kahn, R.L. & Syme, S.L. (1994). Socioeconomic status and health: The challenge of the gradient. American Psychologist 49:15-24.

Arber, S. & Ginn, J. (1993). Gender and inequalities in health in later life. Social Sciences and Medicine, 36:33-46.

Backlund, D., Sorlie, P.D. & Johnson, N.J. (1996). The shape of the relaitonship between income and mortality in the United States: evidence from the National Longitudinal Mortality Study. Annuals of Epidemiology:6:12-23.

Backland, D., Sorlie, P.D. & Johnson, N.J. (1999). A comparison of the relationships of education and income with mortality: the national longitudinal mortality study. Social Science & Medicine, 49:1373-1384.

Berkman, L.F. & Macintyre, S. (1997). The measurement of social class in health studies: old measures and new formulations. In: Kogevinas, M., Pearce, N., Susser, M. & Boffetta, P. (eds) Social Inequalities and Cancer. Lyon:IARC Scientific Publications No. 138, International Agency for Research on Cancer.

Buhmann, B., Rainwater, L., Schmaus, G. & Smeeding, T. (1988). Equivalence scales, well-being, inequality and poverty: Sensitivity estimates across ten countries using the Luxembourg Income Study database. The Review of Income and Wealth; vol. 34(2), 115-142.

Chapman, K.S. & Hariharan, G. (1996). Do poor people have a stronger relationship betwee income and mortality than the rich? Implications of panel data for health analysis. Journal of Risk and Uncertainty, 12:51-63.

Ecob, R. & Davey Smith, G. (1999). Income and health: What is the nature of the relationship? Social Science & Medicine, 48:693-705.

House, J.S., Kessler, R.C., Herzog, A.R., Mero, R.P., Kinney, A.M. & Breslow, M.J. (1990). Age, socioeconomic status, and health. Milbank Quarterly, 68:383-411.

House, J.S., Lepkowski, J.M., Kinney, A.M., Mero, R.P., Kessler, R.C. & Herzog, A.R. (1994). The social stratification of aging and health. Journal of Health and Social Behavior, 35:213-234.

House, J.S. & Williams, D.R. (2000). Understanding and reducing socioeconomic and racial/ethnic disparities in health. In: Smedly, B.D. & Syme, S.L. (eds) Promoting Health: Intervention Strategies from Social and Behavioral. Washington D.C.:National Academy Press, 82-124.

Kaplan, G. (1996). People and places: contrasting perspectives on the association between social class and health. International Journal of Health Services, 26:507-19.

Kington, R.S. & Smith, J.P. (1997). Socioeconomic status and racial and ethnic differences in functional status associated with chronic diseases. American Journal of Public Health,87:805-810.

Krieger, N., Williams, D.R. & Moss, N.E. (1997). Measuring social class in US public health research: Concepts, mehtodologies, and guidelines. Annual Review of Public Health, 18:341-78.

Libertos, P., Link, B.G. & Kelsey, J.L. (1988). The measurement of social class in epidemiology. Epidemiologic Reviews, 10:87-121.

McDonough, P., Duncan, G. J., Williams, D. & House, J. (1997). Income dynamics and adult mortality in the United States, 1972 through 1989. American Jounal of Public Health, 87(9), 1476-1483.

Marmot, M.G., Ryff, C.D., Bumpass, L.L., Shipley, M. & Marks, N. (1997). Social inequalities in health: Next questions and converging evidence. Social Science and Medicine, 44:901-10.

Marmot, M.G., Smith, G.D.., Stanfsfeld, S.., Patel, C., North, F., Head, J.., White, I., Brunner, E. & Feeney, A. (1991). Health inequalities among British civil servants: The Whitehall II study. Lancet 337:1387-93.

Mirowsky, J. & Hu, P.N. (1996). Physical impairment and the diminishing effects of income. Social Forces, 74: 1073-96.

Mirowsky, J., Ross, C.E. & Reynolds, J. (2000). Links beteween social status and health status. In C. Bird, P.Conrad & A. Fremont (eds.) Handbook of Medical Sociology. London:Sage Publishing Company.

Newman, K. (1988). Falling from Grace: The Experience of Downward Mobility in the American Middle Class. New York, NY:Free Press.

Pearlin, L.I. (1999). The stress process revisited. In Aneshensel, C.S. & Phelan, J.C. (eds.) Handbook of the Sociology of Mental Health. New York:Kluwer Academic/Plenum Publishers.

Robert, S.A. & House, J.S. (1996). SES differentials in health by age and alternative indicators of SES. Journal of Aging and Health, 8:359-88.

Robert, S.A. & House, J.S.(2000). Socioeconomic inequalities in health: An enduring sociological problem. In C. Bird, P.Conrad & A. Fremont (eds.) Handbook of Medical Sociology. London:Sage Publishing Company.

Rogot, E., Sorlie, P.D., Johson, N.J. & Schnitt, C. (1992). A mortality study of 1.3 million persons by demographic, social, and economic factors: 1979-1985.  Bethesda, Md.:NIH.

Schoenbaum, M. & Waidmann, T. (1997). Race, socioeconomic status, and health: Accounting for race differences in health. The Journal of Gerontology, 52B:61-73.

Smith, J. (1999). Healthy bodies and thick wallets: The dual relationship between health and socioeconomnic status. Journal of Economic Perspectives, 145-166.

Sooman, A. & MacIntyre, S. (1995). Health and perceptions of the local environment in socially contrasting neighbourhoods in Glasglow. Health Place, 1:15-26.

Troutt, D.D. (1993). The Thin Red Line: How the Poor Still Pay More. Oakland, CA: West Coast Regional Officeof   Consumers Union.

Wolfson, M., Rowe, G., Gentleman, J.F. & Tomiak, M. (1993). Career earnings and death: A longitudinal analysis of older Canadian men. Journal of Gerontology, 48: S167-79.