Educational status

John D. and Catherine T. MacArthur
Research Network on Socioeconomic Status and Health

Summary prepared by Judith Stewart in collaboration with the Social Environment working group. Last revised August, 2001.

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. Each of these measures can capture distinctive aspects of social position but they are not interchangeable, nor are they immune to interactions with such variables as race/ethnicity and gender. Education has become one of the most widely used indicators of socioeconomic position used for mortality and health studies in both demography and epidemiology in the United States (Liberatos, Link & Kelsey, 1988). Elo & Preston (1996) suggest that there are two basic analytic reasons for preferring educational attainment to other common markers of social standing such as occupation or income. First, educational level can be determined for all individuals, whereas not everyone has an occupation or an income (e.g., retired persons, homemakers). Second, health impairments that emerge in adulthood rarely affect educational attainment, since educational attainment is normally complete by the early adult years, around age 25, before decrements to health that accompany aging occur. In contrast both occupation and income can be importantly affected by health impairments that develop in adulthood. Thus using educational attainment as a SES indicator avoids the potential contamination of reverse causation inherent in the other two standard measures.

Evidence of association between educational attainment and adult mortality. There is considerable evidence demonstrating that an individual’s educational status is an important predictor of mortality and morbidity. Kitagawa & Hauser (1973) did one of the most complete studies of mortality differentials done in the U.S.; they matched a sample of death certificates for persons 25 years of age and older who died from May through August 1960 with census records on the same persons enumerated in the 1960 census on April 1, 1960. They found that in 1960 higher SES groups exhibited lower rates of all-cause mortality than did lower SES groups. One of their principal measures of socioeconomic status was educational attainment as assessed by years of school completed. Mortality varied inversely with educational attainment. The range of mortality differentials was larger among individuals 25-64 years of age than among older individuals, and greater among women than men.

Using data on 2380 participants from a five city project, Winkleby, Jatulis, Frank & Fortmann (1992) examined the independent contribution of education, income, and occupation to a set of cardiovascular disease risk factors, including cigarette smoking, systolic and diastolic blood pressure, and total and high-density lipoprotein cholesterol. Their results showed that the relationship between SES measures and risk factors was strongest and most consistent for education. Higher risk was associated with lower levels of education.

Trends in educational attainment differentials in mortality. A number of studies in the years since the Kitagawa & Hauser work have investigated trends in educational attainment differentials in mortality. Feldman, Makuc, Kleinman & Cornoni-Huntley (1989) examined mortality differentials for men and women age 45 years and over using the First National Health and Nutrition Examination Survey (NHANES I) and the NHANES I Epidemiologic Follow-up Study (NHEFS). They found that educational differentials in mortality had increased among middle-aged men and were present among older men (whereas in 1960 no inverse relationship was observed in the older men). For women, the inverse relationship between educational attainment and morality remained at about the same magnitude as in 1960. Educational differentials remained even after adjusting for current cigarette smoking, high blood pressure, high cholesterol, and obesity at baseline.

Pappas, Queen, Hadden & Fisher (1993) used data from the 1986 National Mortality Followback Survey (NMFS) to replicate the analysis done by Kitagawa & Hauser. The NMFS is a nationally representative sample of individuals aged 25 years or more who died in 1986. Information from death certificates was linked with data from questionnaires filled out by next of kin. The 1986 National Health Interview Survey (NHIS) sample was used to estimate populations at risk. Findings showed that although death rates declined between 1960 and 1986, the decline was steeper for men and women with higher educational attainment. Among men aged 25 to 64 years the direct age-adjusted death rate declined by 50%, whereas among men with low educational attainment the decline was 15%. The differences were smaller for women.

Preston & Elo (1995) used the National Longitudinal Mortality Survey (NLMS) to explore the educational differentials question, comparing and contrasting in some detail their work with that of Feldman et al and Pappas et al (e.g., characteristics of data sources, analytic approaches). They conclude that inequalities have widened for males but contracted for working-age females (25-64). In their analyses educational inequality trends are more adverse for older persons (those aged 65+) with the largest increase in inequality occurring for men aged 65-74.

Non-simplicity of the relationship between educational attainment and health. The relationship between educational attainment and health may not be uniform across the social hierarchy. In reviewing a number of large studies in the United States, Krieger & Fee (1994) noted that education seems to have little relation to health status among persons in households with incomes below poverty thresholds. This finding suggests that the effects of absolute material deprivation cannot be successfully buffered or mediated by the resources represented by education. Further, Krieger, Williams and Moss (1997) point out that economic returns for a given level of education may vary depending on gender and race/ethnicity. For example, they report on findings from the US Bureau of the Census, that in 1989 among persons in the U.S. who had completed high school and who were over age 18 and working full-time, average annual earnings of white men ($26,526) were $5,000 more than those of black and Hispanic men, and white women earned about $8,000 less per year than white men, but about $1,000 more than black and Hispanic women.

Valkonen (1989) used education to compare socioeconomic differences in all cause mortality in different European countries in the 1970s. The relation between number of years of education and mortality rate is remarkably similar for England and Wales, the Scandinavian countries and Hungary for men. Men with the most educational attainment have 40-60% lower death rates compared to those with the least education. The general relationship is similar among women, but there are exceptions to the general decline of mortality with rising educational attainment. For a woman, material income and wealth may be determined not only by her own characteristics, but also those of her husband, and the degree to which this is the case may vary among countries, thus to some degree explaining the variation in these associations for women.

How might educational attainment benefit health? Yen & Moss (1999) suggest that there are skills and social benefits which come with increasing educational levels. "Skills may include: 1) ability to process certain kinds of information or critical thinking and 2) ability to interact with bureaucracies, institutions, and health practitioners. Social benefits may include: 1) credentials and the economic access they provide, 2) social networks and extension of cultural capital, 3) socialization to adopt health-promoting behaviors; and 4) enhanced expectations for the future leading to helpfulness, planning, self-efficacy, and a sense of control." The relative importance of these effects may be period and cohort specific, and affected by race/ethnicity and gender.

Ross and Wu (1995) propose three explanations for the association between education and health (work and economic conditions, social and psychological resources, and health lifestyle). They present data from two national samples showing that a large part of the association between education and health is accounted for by these explanatory variables, but a significant direct effect of education remains.

Measurement approaches
Educational status is typically assessed either by years of education attained, or in terms of credentials earned. Examples of such items appear in Table 1.

Table 1. Measuring educational attainment.

a) Years of educational attainment.

What is the highest grade (or year) of regular school you have completed?
(Check one.)
Elementary School	High School	College	Graduate School
01_____	09_____	13_____	17_____
02_____	10_____	14_____	18_____
03_____	11_____	15_____	19_____
04_____	12_____	16_____	20+____
05_____
06_____
07_____
08_____

b) Credentials earned.

What is the highest degree you earned?

____High school diploma or equivalency (GED)

____Associate degree (junior college)

____Bachelor's degree

____Master's degree

____Doctorate

____Professional (MD, JD, DDS, etc.)

____Other specify

____None of the above (less than high school)

Although both approaches are common, it may be more meaningful to measure educational level in terms of credentials rather than simply years of education. A one-year difference between completing 9^th versus 10^th grade is not the same as the one-year difference between completing 11^th and 12^th grade, since only a person with a 12^th grade education is certified as a high school graduate. Using data from the 1977 General Social Survey of the National Opinion Research Center, Faia (1981) found that certification was more influential in determining occupational prestige than years of schooling. Possession of a certificate of high school graduation has important implications for employment prospects (Krieger & Fee, 1994). In addition there are discontinuities when years of education is used as a continuous variable.

Generational effects need to be taken into account when using an individual level SES variable. The significance of educational achievements such as a high school or college degree will vary between birth cohorts, due to the shifting population patterns of educational attainment. Many more persons have completed high school and college in the last 20 years than in any previous period. Liberatos et al (1988) illustrate this point with 1980 census data showing that in 1940 24.5% of the U.S. population held a high school degree or more, while by 1980 68.6% fell in this category. In other words, the significance of holding a college degree for someone now 80 years old and someone now 25 years old will differ markedly since a much higher proportion of the latter's age cohort will have obtained a college-level education. The social meaning of a level of educational achievement will be quite dissimilar for these two individuals.

Comments

Each socioeconomic indicator has its own set of advantages and disadvantages (Krieger, Williams and Moss, 1997; Williams & Collins, 1995). The advantages of using educational attainment include:

Education is fairly stable beyond early adulthood
Its measurement is practical and convenient in many contexts
It is one of the socioeconomic indicators especially likely to capture aspects of lifestyle and behavior

But education used as an indicator of SES also has several limitations:

Educational attainment varies by age cohort

Relative to income, there is decreasing variability in years of education

In at least some national data, inequalities in health associated with income are larger than those associated with education, such that using education as a measure of SES may minimize estimates of social inequalities in health

The lack of volatility in education for most adults precludes the opportunity to assess how health status is affected by changes in SES

Many studies that use education as an indicator of SES are individualistic in approach and do not incorporate information about the education level of other members of the household

The economic return for a given level of education varies importantly by race and gender
The economic and social return for a given level of education may vary importantly by the prestige of the educational institution attended

Bibliography

Elo IT & Preston SH (1996). Educational differentials in mortality: United States, 1979-85. Social Science& Medicine, 42(1):47-57.

Faia MA. (1981). Selection by certification: a neglected variable in stratification research. American Journal of Sociology,86(5):1093-1111.

Feldman JJ, Makuc DM, Kleinman JC & Cornoni-Huntley J (1989). National trends in educational differences in mortality. American Journal of Epidemiology, 129:919-1033.

Kitagawa EM & Hauser PM (1973). Differential Mortality in the United States: A study in socioeconomic epidemiology. Cambridge, MA: Harvard University Press.

Krieger N & Fee E (1994). Social class: The missing link in U.S. health data. International Journal of Health Services, 26:391-418.

Krieger N, Williams DR & Moss N (1997). Measuring social class in U.S. public health research: concepts, methodologies, and guidelines. Annual Review of Public Health, 18:341-378.

Liberatos P, Link BG & Kelsey JL (1988). The measurement of social class in epidemiology. Epidemiologic Reviews 10, 87-121.

Pappas G, Queen S., Hadden W. & Fisher G. (1993). The increasing disparity in mortality between socioeconomic groups in the United States, 1960 and 1986. New England Journal of Medicine, 329: 103-109.

Preston SH & Elo IT (1995). Are educational differentials in adult mortality increasing in the United States? Journal of Aging and Health, 7(4); 476-496.

Ross CE & Wu C (1995). The links between education and health. American Sociological Review, 60:719-745.

Valkonen T (1989). Adult mortality and level of education: a comparison of six countries. In: Fox J (ed.), Health Inequalities in European Countries. Aldershot: Gower Publishing; 142-162.

Wickleby MA, Jatulis DE, Frank E & Fortmann SP (1992). Socioeconomic status and health: How education, income, and occupation contribute to risk factors for cardiovascular disease. American Journal of Public Health, 82(6): 816-820.

Yen IH & Moss N (1999). Unbundling education: A critical discussion of what education confers and how it lowers risk for disease and death. In Adler, N.E., Marmot, M., McEwen, B.S. & Stewart, J. (Eds.) (1999). Socioeconomic Status and Health in Industrial Nations: Social, Psychological and Biological Pathways. Ann NY Acad Sci. Vol # 896, 350-351.