| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||
TOPIC 3. RACIAL AND ETHNIC INEQUALITIES IN HEALTH |
Department of Sociology, Indiana University, Bloomington.
Address correspondence to Pamela Braboy Jackson, PhD, Department of Sociology, Indiana University, Ballantine Hall 744, Bloomington, IN 47405-7103. E-mail: pjackson{at}indiana.edu
Abstract
There are several challenges facing scholars studying health inequalities among minority populations. Primary among these challenges are developing adequate measures of social inequality and introducing appropriate strategies for eliminating health disparities. More research is sorely needed on both of these fronts as evidenced by the health paradox facing black, middle class men and women. This effort, however, can best move the study of health inequalities forward when juxtaposed against theoretical paradigms that embrace the complexity of the intersection of race, class, and gender.
THE distribution of health is a prominent topic on many health policy agendas. Some are concerned with the way in which lifestyle choices, social factors, and globalization of the market economy continue to generate disparate health outcomes for different groups. Given the increase in average health inequality across the world (Kawachi, Marshall, & Pearce, 1991
; Kunst, Geurts, & van den Berg, 1995; Kunst, Groenhof, Mackenbach, & Health, 1998; Mackenbach & Kunst, 1997; Marmot et al., 1991; World Bank, 2002), these issues will remain at the forefront of health research.
The purpose of this commentary is to draw attention to two challenges facing those interested in health inequalities among minority populations: (a) measuring inequality and (b) tackling health inequality. I argue that the article by David Williams (this volume) speaks well to both issues; however, a series of questions posed by health scholars provide even more food for thought that may stimulate efforts to link social scientific research to social policy initiatives. Here, they ask, "Exactly what would we like to be equally distributed in the population? Would we consider perfect equality to be when all individuals live the same number of years? When they enjoy the same level of health? When they have exactly the same health status at all points in their lives?" (Gakidou, Murray, & Frenk, 2000, p. 1). In the discussion of measuring inequality, I draw attention to the fact that the very concept of social inequality (often assessed in terms of socioeconomic status [SES]) may need to be revised as health status is often associated with the subjective (e.g., differential association) and objective (e.g., actual income) conditions of people's lives. I then describe an ongoing debate among a group of health scholars regarding appropriate assessments of health inequality that is not described in the Williams piece, although it has relevance for the study of health inequality.
The approach by Williams can also be extended to incorporate a theoretically based model of inequality that may prove useful to those investigators who are sensitive to the historical and current circumstances that generate/perpetuate health inequalities: the intersectionality paradigm (Mullings & Wali, 2001; Weber & Parra-Medina, 2003). Here, I argue that even if the issue of measuring inequality is adequately addressed, tackling certain health problems will remain challenging for those interested in reducing health inequality. I end this discussion with examples provided in the case of specific health problems faced by some members of the Black middle class.
MEASURING INEQUALITY
Race, gender, and socioeconomic status (SES) lead to the differential distribution of health risks and thus to variations in health (Krieger, Rowley, Herman, Avery, & Phillips, 1993). SES is usually measured by education, income, occupational status, and/or work conditions. These objective indicators of SES are subsequently explored as predictors of health (see Borg & Kristensen, 2000; Williams, 1990; Williams & Collins, 1995). A growing body of research suggests that the working definitions of social inequality as assessed by traditional indictors of SES can be expanded to improve our understanding of health inequality.
There has been renewed interest in the SES–health relationship, perhaps as a result of rising economic and/or health disparities around the world. This explosion is most clearly demonstrated in an American Journal of Public Health commentary by Kaplan and Lynch (1997), who show the exponential growth in the number of citations on this topic in MedLine. While some of this work is being conducted by sociologists, SES has been "discovered" by scholars in multiple disciplines. Nonetheless, it is important to consider how multiple indicators of social and economic conditions are related to particular health disparities (Lynch, Smith, Harper, & Hillemeier, 2004) as much scholarly work demonstrates the variety of pathways through which separate indicators of SES operate to affect health (for review, see Kessler, 1982; Krieger, Williams, & Moss, 1997).
Measures of Social Inequality
One scale that expands on current measures of social position is the MacArthur Scale of Subjective Social Status (Adler, Epel, Castellzaao, & Ickovics, 2000). This approach is in line with an older body of research by social psychologists (see Centers, 1949; Jackman & Jackman, 1973; Rosenberg, 1953). Like these earlier measures, the MacArthur Scale of Subjective Social Status was developed to capture the common sense of social status across the SES indicators. In an easy pictorial format, it presents a "social ladder" with 10 steps and asks individuals to place an "X" on the rung on which they feel they stand. There are two versions of the ladder: one linked to one's standing in the general population based on traditional SES indicators (national ladder) and the second linked to standing in one's community (community ladder).
The relationship between objective SES (as assessed by measurement of education, individual and family income, occupation, and wealth) and subjective social status indicates that the ladder rankings are not redundant. Subjective social status appears to measure something other than what is being measured by the objective indicators. Preliminary data suggest that individuals' perceptions of their place in the hierarchy as assessed by the ladders are associated with health. The SES–health relationship may be mediated by the subjective feeling of being lower or higher on the ranking. This specific line of research begs the question of the appropriate comparison group. More importantly, however, it speaks to the issues of network composition and who are the significant others within the network, as well as the social comparison processes used by socially disadvantaged populations.
In terms of research on race and health, the difference between these two ladders may be of particular interest in poorer communities in which individuals may not be high on the SES ladder in terms of income, occupation, or education but may have high standing within social groups such as a religious or local community. Insofar as social standing has beneficial effects on biological processes related to health, standing on the community ladder may be as important as standing on the SES ladder. The nonequivalence of SES indicators across racial groups in terms of their material implications has been recognized (William & Collins, 1995), but further consideration may need to be given to fully capturing the social meaning of class across racial groups. The following measure of social class brings this topic directly to the forefront of much health research.
Another recent innovation in social class assessment is represented by the Cambridge Social Interaction and Stratification Scale (Bergman, Lambert, Prandy, & Joye, 2002) where the focus is on differential association as a feature of social stratification arrangements. The Cambridge Scale is an index of social advantage derived empirically from knowledge of which particular pairs of occupations are the substance of friendships and marriages. This scale reflects differences in generalized advantage and lifestyle that seem to be quite sensitive to women's position in the social structure. This scale seems to be quite flexible in its measurement of "occupational groups" by broadly defining this concept to include, in addition to occupation as normally understood, differences in status in employment (self-employed). However, it is possible (and usual) for the scale to be gender specific and for other bases for groupings, such as ethnicity and education or qualifications (for those not in paid employment), to be incorporated. In fact, its originators argue that the scale measures social and material advantage, two concepts that are indivisible. This scale attempts to reflect a social reality of finer gradations more clearly than traditional class schemas.
In a comparison study of health inequality among men and women in Britain, Sacker, Firth, Fitzpatrick, Lynch, and Bartley (2000) found that in women, social class based on individual employment relations and job conditions showed only a weak association to mortality, whereas larger differences in risk of mortality in women were found when social position was measured according to the Cambridge Scale. As such, conventional measures of SES based on characteristics of employment considerably underestimate the role played by such conditions. Here, then, is an opportunity for researchers to think more broadly about the way in which social structure impinges on the lives of minority women, in particular.
Measures of Health Inequality
Williams (this issue) focuses on disparities in health in lieu of the complex interactions among the social categories of race, gender, and social class. There is an ongoing debate, however, among some health investigators regarding the ideal measure of health inequality—a debate not acknowledged in the Williams piece. For example, some argue that an individual approach is best, where health inequality should be assessed across individuals in a population (Gakidou et al., 2000). The World Health Organization most recently adopted the individual approach. Others believe that health inequality should continue to be measured across groups, the more standard approach utilized by most health investigators. These scholars believe that no measure is value-free as it is inevitably used to rank how well national health systems are performing (Asada & Hedemann, 2002). A more recent call has been made for the integration of these approaches leading to a measure of total health inequality because, when used separately, these indicators may very well underestimate total health inequality (Gakidou & King, 2002). Besides these issues, some emphasize other methodological concerns related to measurement of inequalities in health (Wolfson & Rowe, 2001).
Williams's focus on between-group as well as within-group differences would be consistent with this more holistic view of health inequality. By emphasizing the fact that there are differences within racial/ethnic groups, no individual variation in health status is ignored. This is an important issue as it pertains, especially, to public health policy because a total health inequality measure would allow policymakers to attempt to reduce inequalities across individuals in addition to reducing disparities in average health status across groups in society.
TACKLING HEALTH INEQUALITY
Assuming, then, that one had adequate measures of inequality, the next step is tackling health inequality. The questions raised by Gakidou and colleagues (2000) restated in the beginning of this article are relevant for the description of health disparities outlined by Williams, especially in the context of his discussion of understanding how health risks change over the life course. There is a body of work describing profiles of health risk that rely on health survivorship functions. The fact that ethnic minorities are more likely to experience death from heart disease, cancer, diabetes, homicide, and accidents on the job underscores the importance of taking multiple factors into account when assessing health inequality (Lynch et al., 2004). Of course, the precursors to such risk factors may themselves be linked to other forms of inequality in society. Exposure to poverty in childhood, for example, has been linked to a variety of health outcomes across the life course (Hayward, Miles, Crimmins, & Yang, 2000; Lynch, Kaplan, & Salonen, 1997; McLeod & Shanahan, 1996). So, then, some individuals face higher risks of ill health and mortality at every age, and others face much lower risks. As such, there is a need for more discussion regarding the likelihood of achieving health equality given the structural conditions that continue to impinge on the life chances of many adults.
One of the critical issues raised by much of Williams's work that cannot be overstated is the role that institutional and individual-level discrimination plays in these on-going patterns of health. There is consistent evidence that African-Americans with higher educational, occupational, and income levels are more likely than their African-American counterparts to report racial discrimination. On the one hand, African-American men are more likely to report discrimination than African-American women (Forman, Williams, & Jackson, 1997; Ren, Amick, & Williams, 1997). On the other, African-American women are "doubly disadvantaged" by their race and gender status (Collins, 1990). In attempts to tackle health disparities, then, the topic of discrimination should not be understated. In any discussion of reducing health inequalities, one must consider the individual-level as well as group-level factors that put individuals/groups at higher risk of that health outcome. Two case examples that I bring to your attention are the high infant mortality rate among highly educated Black women and hypertension rates among middle-class Black men.
A glaring disparity in health outcomes is the race/ethnic discrepancy in the incidence of infant mortality. African American women, in particular, are over two times more likely to suffer the loss of an infant than their non-Hispanic White counterparts. This black/white differential exists regardless of education or income (National Center for Education in Maternal and Child Health, 2000). The relative gap between the infant mortality rates of non-Hispanic White and African American mothers who have 16 or more years of schooling is greater than the one between those with less than 12 years of education.
The Black/White difference in infant mortality has been linked to a complex web of medical (e.g., genetic heritage), socioeconomic (e.g., access to neonatal technology), and behavioral (e.g., diet) factors. The majority of neonatal deaths in the United States are due to low birth weight and prematurity or preterm delivery. Although only 17% of all babies are born to Black families, these babies account for 33% of all low birth weight babies and 38% of all very low birth weight babies (Shiono & Behrman, 1995).
A more recent emphasis has been placed on the role played by psychosocial stressors, especially racism and a lack of support systems available to middle-class African American women. These concerns are crystallized when we consider the way in which race/ethnicity, gender, and SES converge to create the paradox facing those interested in the alarming rate of black infant mortality: Highly educated African American women have a higher infant mortality rate than non-Hispanic white, Asian, and Latino women who have not completed high school (Pamuk, Makuk, Heck, & Reuben, 1998).
Middle-class African American men are an understudied group in much of health research despite their increased risk for some of the most pressing health conditions (Williams, 2003). Heart disease, for example, remains the leading cause of death among most Americans, with hypertension serving as a major risk factor. The association between SES and hypertension, however, varies by race and gender. For African American women, there is an inverse association between income and hypertension. Among African American men, however, income is unrelated to hypertension (Pamuk et al., 1998). This association is not evidenced among non-Hispanic White adults (Hypertension Detection and Follow-Up Program Cooperative Group, 1977). A recent study of a predominantly African American population in Harlem similarly found that men with a college degree had higher levels of hypertension than high school graduates. In contrast, hypertension risk declined with each higher level of income and education for women (Diez-Roux, Northridge, Morabia, Bassett, & Shea, 1999). The elevated level of psychosocial stress reported among middle-class African American men may contribute to their increased hypertension risk (James, 1994).
The intersectionality paradigm, evidenced in feminist scholarship (Collins, 1990) and more recent work of other scholars (see Mullings & Wali, 2001; Weber & Parra-Medina, 2003), implores health researchers to move beyond the linear approaches to multiple minority statuses. These authors argue, in fact, that this framework would allow scholars to move beyond the cumulative adversity perspective to illuminate some of the paradoxes evidenced in the health literature, including the examples provided above on the Black middle class.
A linear approach would not expect to find, for example, middle-class African Americans to report poorer health outcomes than their lower-class peers (regardless of ethnicity). A cumulative adversity perspective would not expect Black men to reap fewer health benefits from their income status than Black women. As such, investigators must be sensitive to the oppressive conditions faced by each group that speak to their unique position in the social structure. Given the unique historical experiences underlying the social positions of race, gender, and social class, considerable evidence is suggesting that at their intersection is formed a completely new status that is more than the sum of its individual parts.
Studies of health disparities among minority populations should be sensitive to the crude measures of socioeconomic position that are often adopted in large-scale studies of the population. These indicators do not, necessarily, take into consideration the subjective dimensions of social position, including daily activities, social relationships, or relative standing in a community. These factors may very well dictate beliefs, values, and behaviors that are embraced among various groups. For example, the most prevalent structural arrangement impinging on the lives of the majority of African Americans is residential segregation (Williams & Collins, 2001), where regardless of socioeconomic position, African Americans live in poorer neighborhood conditions than their White peers (Sampson & Wilson, 1995). The neighborhood often serves as the backdrop for the exchange of social capital, which includes a sense of mutual trust and shared expectations (Sampson, Raudenbush, & Earls, 1997). In general, "individuals residing in neighborhoods with high social capital are more likely to have effective channels of communication, to have reciprocal relationships providing mutual support, and to be influenced by positive social norms, all increasing the likelihood that healthy behaviors will be adopted" (Reagan & Salsberry, 2005, p. 2219). Thus, besides the added burden of the race "tax" that culminates in perceived discrimination among the Black middle class (Jackson & Stewart, 2003; Williams, 1998) and the cumulative or "weathering" effect of extended exposure to stressful life conditions (Geronimus, 2001), middle-class African Americans must contend with a constricted social network that may not afford them the social capital to address these initial and ongoing situations (Sampson et al., 1997). Perhaps when the complexity of social inequality is taken into account, we will be in a better position to understand why groups that appear advantaged on one dimension of stratification (e.g., education, income) continue to be disadvantaged on other dimensions of stratification (e.g., health).
In essence, the Williams piece on "The Health of U.S. Racial and Ethnic Populations" raises important issues of measurement, policy initiatives aimed at tackling health inequalities, and health paradoxes that cannot go unnoticed. Although examining differences in average health levels across groups is an important endeavor, attention should also be paid to developing and improving current measures of health inequality. These empirical exercises should be juxtaposed against applicable and challenging theoretical paradigms that embrace, rather than avoid, the complexity of the intersection of race, class, and gender.
References
| ||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
