Cover Image

PAPERBACK
$51.75



View/Hide Left Panel

SECTION V
TWO INTERNATIONAL COMPARISONS



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 675
Critical Perspectives on Racial and Ethnic Differences in Health in Late Life SECTION V TWO INTERNATIONAL COMPARISONS

OCR for page 675
Critical Perspectives on Racial and Ethnic Differences in Health in Late Life This page intentionally left blank.

OCR for page 675
Critical Perspectives on Racial and Ethnic Differences in Health in Late Life 18 Ethnic Disparities in Aging Health: What Can We Learn from the United Kingdom? James Y. Nazroo Ethnic inequalities in health have been a major concern in the United Kingdom for several decades. The collection of evidence on the nature, extent, and causes of these has been a focus of numerous studies, both of immigrant mortality (Balarajan and Bulusu, 1990; Harding and Maxwell, 1997; Marmot, Adelstein, Bulusu, and Office of Population Censuses and Surveys, 1984) and difference in morbidity across ethnic groups (Erens, Primatesta, and Prior, 2001; Nazroo, 1997, 2001). Over this time we have seen several paradigm shifts in the focus of studies, with an initial emphasis on genetic and cultural differences, to an emphasis on socioeconomic inequalities, to a more recent emphasis on racism and a revisiting of culture as ethnic identity (Nazroo, 1998; Smaje, 1996). Throughout this work there has been little emphasis on the issues of age and aging, perhaps because of the recency of migration to the United Kingdom for many of the ethnic minority groups studied (and their consequent relatively young age profiles). Nevertheless, given both a concern about the policy implications of ethnic inequalities in health, and the academic interest in using the additional diversity in experience provided by ethnic comparisons to help understand causes, the experiences of older ethnic minority people are very important. Studies of ethnic inequalities in health among older people have the potential to address issues of age, generation, and cohort, which are of fundamental importance to understanding processes and causes. The following specific questions arise in three categories—age, generation, and cohort:

OCR for page 675
Critical Perspectives on Racial and Ethnic Differences in Health in Late Life Age: Is the emergence of ethnic inequalities in health intimately linked to the aging process? If so, how do the factors that lead to ethnic inequalities play out over the life course? Generation: Are ethnic inequalities in health a consequence of migration experiences? Are they a product of the unique experiences of a migrant generation? Do they “transmit” to second and subsequent generations and, if so, why? Cohort: How far are ethnic inequalities in health related to the specific historical context of a new migrant population? Can we anticipate that the health experiences of younger ethnic minority people will be similar to those of middle-aged and older ethnic minority people? Or have sufficient shifts occurred in the cultural and economic contexts of their lives to make their experiences of aging different? The form that migration to the United Kingdom has taken potentially allows us to address these questions, and this chapter sets out to begin to map out elements of the U.K. data on these issues. MIGRATION OF ETHNIC MINORITY GROUPS TO THE UNITED KINGDOM Although some black people settled in the United Kingdom prior to World War II (mainly in London and the ports on the west coast of the United Kingdom—Bristol, Cardiff, Liverpool, Glasgow—and primarily related to the slave trade), most of the nonwhite migration to Britain occurred after World War II. This was driven by the postwar economic boom and consequent need for labor, a need that could be filled from British Commonwealth countries—primarily countries in the Caribbean and the Indian subcontinent. This “economic” migration was followed by migration of spouses and children and, sometimes, older relatives, in a climate when the legislation regulating entry into the United Kingdom became increasingly restrictive. Migration from these countries was not evenly spread over time: immigration from the Caribbean and India occurred throughout the 1950s and 1960s, peaking in the early 1960s; from Pakistan, largely in the 1970s; from Bangladesh, mainly in the late 1970s and early 1980s; and from Hong Kong, in the 1980s and 1990s. In addition, there was a notable flow of immigrants from East Africa in the late 1960s and early 1970s, made up of

OCR for page 675
Critical Perspectives on Racial and Ethnic Differences in Health in Late Life migrants from India to East Africa who were subsequently expelled. Over the past 10 years, migration to the United Kingdom has taken a very different form, including mostly refugees. This pattern of migration means that the vast majority of older, nonwhite ethnic minorities in the United Kingdom are first generation migrants, making the situation different from that in other non-European countries. However, alongside this “visible” migration, there has been a long history of migration to England from Ireland, which continued during the active recruitment of labor from the Caribbean and the Indian subcontinent. The history of Irish migration to England, as to the United States, holds important lessons on the circumstances of economic migrants and their descendants, and how far skin color is a demarcating factor. The collection of data on ethnicity in the U.K. Census has happened only twice, for 1991 and 2001. Data from the 2001 Census, which included a fairly comprehensive assessment of ethnicity, have only recently become available. The 1991 Census asked respondents to indicate which ethnic group they belonged to from a range of fixed choices that encompassed both skin color and country of origin, but it did not identify white minority groups. Responses to this question are shown in Table 18-1, along with the percentage in each group who were born in the United Kingdom. The table TABLE 18-1 Ethnic Composition of United Kingdom Population Ethnic Group Number/1,000 Percent Percent born in United Kingdom White 51,844 94.5 95.8 All ethnic minorities 3,007 5.5 46.8 All black 885 1.6 55.7 Black-Caribbean 499 0.9 53.7 Black-African 208 0.4 36.4 Black-othera 179 0.3 84.4 All South Asian 1,477 2.7 44.1 Indian 841 1.5 42.0 Pakistani 476 0.9 50.5 Bangladeshi 160 0.3 36.7 Chinese and others 644 1.2 40.6 Chinese 158 0.3 28.4 Other-Asian 197 0.4 21.9 Other-otherb 290 0.5 59.8 aThe “black-other” group contains people recorded as “black” with no further details, those identifying themselves as “black British,” and people with ethnic origins classified as mixed black/white and black/other ethnic group. Most of the “black-other” group members seem to have had Caribbean family origins, but were born in Britain. bThe “other-other” group contains North Africans, Arabs, and Iranians, together with people of mixed Asian/white, mixed black/white, and “other” mixed categories. SOURCE: 1991 Census (Nazroo, 1999).

OCR for page 675
Critical Perspectives on Racial and Ethnic Differences in Health in Late Life shows that at the 1991 Census, 5.5 percent of the U.K. population (just over 3 million people) identified themselves as belonging to one of the nonwhite ethnic minority groups. Table 18-1 also shows that in 1991, just under half of the nonwhite ethnic minority population was born in the United Kingdom, though this varies across specific groups reflecting both period of migration and patterns of fertility. Children formed a third of the ethnic minority population, compared with less than a fifth of the white population. In contrast, while 16 percent of the population as a whole was aged over 65, only just over 3 percent of the ethnic minority population fell within this age group. Differences in age profiles also varied across ethnic minority groups, with the Caribbean, Indian, and Chinese groups having a slightly older profile than the Pakistani and Bangladeshi groups. SOCIAL AND ECONOMIC CIRCUMSTANCES OF OLDER ETHNIC MINORITIES IN THE UNITED KINGDOM Ethnic minority and white people live in markedly different areas of the United Kingdom. Analysis of the 1991 Census (Owen, 1992, 1994) has shown that the nonwhite ethnic minority population is largely concentrated in England, mainly in the most populous areas. Key findings are as follows: More than half of the ethnic minority population lives in southeast England, where less than a third of the white population lives. Greater London contains 44.8 percent of the ethnic minority population and only 10.3 percent of the white population. Elsewhere, the West Midlands, West Yorkshire, and Greater Manchester display the highest relative concentrations of ethnic minority people. Nearly 70 percent of ethnic minorities live in Greater London, the West Midlands, West Yorkshire, and Greater Manchester, compared with just over 25 percent of whites. There are even greater differences when smaller areas, census enumeration districts (equivalent to census tracts in the United States), are considered; more than half of ethnic minorities live in areas where the total ethnic minority population exceeds 44 percent, compared with the 5.5 percent national average. There are also large differences in household structure among ethnic groups. Analysis of the 1991 Census (Coleman and Salt, 1996) and the Fourth National Survey of Ethnic Minorities (FNS)1 (Modood et al., 1997) showed that white, Caribbean, Indian, and Chinese families had similar numbers of children, while Pakistani and Bangladeshi families had many

OCR for page 675
Critical Perspectives on Racial and Ethnic Differences in Health in Late Life more children. South Asian households were also larger because of the number of adults they contained. Half of Pakistani and Bangladeshi households had three or more adults, compared with two-fifths of Indian and Chinese households and less than one-fifth of white and Caribbean households. How this plays out in the household composition of older people has been explored by Evandrou’s (2000) analysis of households containing one or more people aged 60 or older using General Household Survey data. This analysis showed that white British, white Irish, and, to a lesser extent, Caribbean people aged 60 or older were more likely to be living alone or as a couple (about 40 percent of the two white groups and 28 percent of the Caribbean group) than Indian (12 percent) or Pakistani and Bangladeshi (6 percent) people. Nearly half of Indians, Pakistanis, and Bangladeshis aged 60 or older were found to be living in large households (three or more adults plus children), compared with about a quarter of white British and white Irish people and 30 percent of Caribbeans. One consequence of large households is overcrowded accommodation, as shown by analysis of data drawn from a recent Health Survey of England (HSE),2 which focused on ethnic minority groups, including white minority groups (who have been combined into one group in analyses presented here that encompasses migrants from Ireland, about two-thirds of this group; Scotland; Wales; and mainland Europe). Focusing on respondents aged 50 or older and using an occupancy rate of 1.5 or more people per bedroom as the threshold to define overcrowding, shows the extent of overcrowding in the Bangladeshi and Pakistani groups (63 and 44 percent respectively of those aged 50 or older were in this category), which contrasts with much lower rates in the Caribbean, white minority, and white English groups (all 4 to 5 percent). The Indian group sits in between (20 percent) and, given the similarity in sizes of households among the Indian, Pakistani, and Bangladeshi groups described in the previous paragraph, this suggests that household size is not the only determinant of overcrowding. Another determinant is, of course, economic position. Tables 18-2 and 18-3, based on HSE data, explore some dimensions of this. The first part of Table 18-2 focuses on men aged 65 or younger (the age of receipt of a state pension for men is 65 in the United Kingdom), showing rates of paid employment for three age groups. Concentrating on the oldest group first, for the white English group just over a third of men aged 50 to 65 were not in paid employment. Figures are higher for all of the ethnic minority groups, except for the Chinese group, with particularly high rates in the Pakistani group (70 percent not in paid employment) and Bangladeshi group (with only one in seven in paid employment). Similar, though smaller, ethnic differences in participation in paid employment can be seen for men aged 30 to 49, with Bangladeshi men again having particularly high rates of not being in paid employment (nearly one in two). Comparing rates for those

OCR for page 675
Critical Perspectives on Racial and Ethnic Differences in Health in Late Life TABLE 18-2 Economic Activity and Occupational Class (percent)   Caribbean Indian Pakistani Bangladeshi Chinese White minority White English Employed Men aged 50-65 42 57 31 16 62 47 63 Men aged 30-49 74 86 78 55 88 84 88 Men aged 16-29 48 49 47 49 39 74 67 Registrar General’s class, men aged 50-65   I/II 12 37 26 9 49 43 39 IIInm 8 9 12 4 1 6 8 IIIm 53 27 29 39 35 30 36 IV/V 27 27 33 48 15 20 17 SOURCE: 1999 HSE (Erens, Primatesta, and Prior, 2001).

OCR for page 675
Critical Perspectives on Racial and Ethnic Differences in Health in Late Life aged 50 to 65 with those aged 30 to 49 shows that the fall in participation rates is greater in all but one of the minority groups (Chinese men) compared with the white English group, and is particularly large for Pakistani and Bangladeshi men, for whom rates drop by around two-thirds. Rates of participation in paid employment are low for the youngest group, in part reflecting a large proportion in school, although again rates for white English men are higher than for most other ethnic groups (white minority men are the exception here). The second part of the table shows occupational class for men aged 50 to 65. The data suggest that the profiles of white English, white minority, and Indian men in this age group are similar, with Chinese men better off and Caribbean, Pakistani, and particularly Bangladeshi men worse off. The striking difference among women in these ethnic groups (not shown in the table) is the level of participation in paid employment. Analysis of the 1999 HSE shows that among women of working age, approximately a quarter of Caribbean, white minority, and white English women are economically inactive, compared with just over a third of Indian and Chinese women and about four-fifths of Pakistani and Bangladeshi women. These figures increase for all groups if women aged 50 to 60 are considered (the age of receipt of a state pension for women is 60 in the United Kingdom), but the broad pattern remains the same. The most stark finding is that only 2 percent of Bangladeshi women in this age group are in paid employment compared with about 10 percent of Pakistani women, just over a third of Indian women, and nearly two-thirds of Caribbean, white minority, and white English women (there were too few Chinese women in this category in the sample to provide an estimate for them). Table 18-3 shows household income from all sources for households with a respondent aged 50 or older split into tertiles, that is, three bands that reflect the general population income distribution, with a third of the general population (of all ages) in each band. The first part of the table (which is not equivalized to account for variations in household size) suggests that the white groups and the Indian group have similar levels of income, with the Caribbean and Pakistani groups worse off and the Bangladeshi group much worse off—90 percent were in the bottom tertile. The second half of the table is equivalized to account for variations in household size using the McClemens scoring system (Erens et al., 2001). In comparison with the nonequivalized data, the white groups appear to be better off (because older people in these groups live in smaller households) and the Pakistani and Caribbean groups appear a little worse off. The ethnic comparison changes a little for the equivalized data. Again they suggest that the two white groups are equivalent, but with the Indian as well as the Caribbean group worse off, and the Pakistani as well as the Bangladeshi group much worse—three quarters of the Pakistani group and

OCR for page 675
Critical Perspectives on Racial and Ethnic Differences in Health in Late Life TABLE 18-3 Household Income: Households with One Person or More Aged 50 or Older (percent)   Caribbean Indian Pakistani Bangladeshi Chinese White minority White English Not equivalized Bottom tertile 72 53 65 90 — 53 57 Middle tertile 20 29 23 7 — 30 25 Top tertile 8 18 12 2 — 17 18 Equivalized Bottom tertile 54 55 77 93 59 36 36 Middle tertile 36 29 19 5 12 35 37 Top tertile 10 17 5 2 29 29 27 SOURCE: 1999 HSE (Erens, Primatesta, and Prior, 2001).

OCR for page 675
Critical Perspectives on Racial and Ethnic Differences in Health in Late Life more than 90 percent of the Bangladeshi group were in the bottom tertile. In terms of the top income tertile, the Chinese group is equivalent to the two white groups, but it also has substantially more households in this age group in the bottom tertile. One of the pervading experiences of ethnic minority people in the United Kingdom (both white and nonwhite) is racial harassment and discrimination. There are no data that allow an adequate exploration of how experiences of racism and discrimination vary by age, but for the adult nonwhite population, this issue was investigated in some depth by the FNS (Modood et al., 1997). This suggested that more than one in eight ethnic minority people had experienced some form of racial harassment in the past year. Although most of these incidents involved racial insults, many of the respondents reported repeated victimization, and a quarter of the ethnic minority respondents reported being fearful of racial harassment. The FNS also showed that among ethnic minority respondents, there was a common belief that employers discriminated against ethnic minority applicants for jobs and widespread experience of such discrimination (Modood et al., 1997). Indeed, when white respondents to the survey were asked about their own racial prejudice, 26 percent admitted to being prejudiced against South Asians, 20 percent to being prejudiced against Caribbeans, and 8 percent to being prejudiced against Chinese. A study by the Commission for Racial Equality has suggested that white minority groups, such as the Irish, also face extensive racial harassment (Hickman and Walters, 1997). PATTERNING OF ETHNIC INEQUALITIES IN HEALTH IN THE UNITED KINGDOM The recording of country of birth at the Census and on death certificates has allowed fairly extensive analysis of the patterning of death rates and cause of death by country of birth (Balarajan and Bulusu, 1990; Harding and Maxwell, 1997; Marmot et al., 1984). This is obviously inadequate for understanding the experiences of ethnic minority people born in the United Kingdom and, in the U.K. context, runs the difficulty of conflating migrants returning back to the United Kingdom from British Commonwealth countries and new immigration from those countries. Table 18-4 provides a summary of findings from the most recent analysis of immigrant mortality around the 1991 Census, showing age standardized mortality ratios (SMRs) by country of birth for all causes and four specific causes of death (chosen for illustrative purposes). There are fewer analyses of morbidity, with the key national sources in the United Kingdom being the 1999 HSE (Erens et al., 2001) and the FNS (Nazroo, 1997, 2001). Table 18-5 provides a summary of data on morbidity for the adult population, drawn from the FNS, showing the relative risk for ethnic minority

OCR for page 675
Critical Perspectives on Racial and Ethnic Differences in Health in Late Life The second half of Table 18-6 shows differences in alcohol consumption. For both men and women, rates of drinking alcohol were highest for the two white groups, with equivalent rates for Caribbean men; slightly lower rates for Caribbean women, Indian men, and Chinese men and women; and low rates for Indian women and for Pakistani and Bangladeshi men and women (who are predominantly Muslim). This pattern is reflected in the measure of drinking more than the recommended weekly limit. Given the overall pattern shown in Table 18-6, it is unlikely that the poorer health experience of ethnic minorities in the United Kingdom can be explained by differences in health behaviors. EXPLAINING THE RELATIONSHIP BETWEEN AGE AND ETHNIC INEQUALITIES IN HEALTH A Migration Effect? One possible explanation for the patterning of ethnic inequalities in health across age groups that was shown in Figures 18-2 and 18-3 is that this is a consequence of differences between migrants and nonmigrants; that either experiences prior to migration, or factors related to migration, have led to poorer health for migrants, which consequently appears as the emergence of ethnic inequalities in health at older ages. This would also explain the appearance of differences at a younger age for the most recent migrant groups, Bangladeshi and Pakistani. The facts that the period of significant migration for many of these groups was relatively narrow (more or less a decade for most) and that migration generally involved specific age groups (young adults) make the separation of migration and age effects difficult. One possibility is to plot levels of health by age separately for migrant and nonmigrant groups; another is to focus on a group that has had a longer period of migration. Figure 18-4 does both of these, looking at age and reported fair or bad health for migrants and nonmigrants in the white minority group, the only group with a sufficient overlap in age between migrants and nonmigrants to make a graphic representation useful. The figure is striking in that it suggests that the health profile of the two groups is remarkably similar. Although confounding with age cannot be easily adjusted for when the age profiles of the compared groups do not overlap to any great degree, a regression analysis can begin to unpack separate effects. In this case, regression analysis was used to explore whether age on migration and years since migration contributed to health risk independently of age for each of the migrant groups included in the 1999 HSE (i.e., Caribbean, Indian, Pakistani, Bangladeshi, Chinese, and white minority). No significant, nor large, relationship between age on migration or years since migration and reported fair or bad health was found for four of these groups—Caribbean, Indian, Bangladeshi, and Chinese. For the Irish and Pakistani groups, there

OCR for page 675
Critical Perspectives on Racial and Ethnic Differences in Health in Late Life FIGURE 18-4 Fair/bad health by age and migration: White minority. SOURCE: 1999 HSE (Erens, Primatesta, and Prior, 2001). was a small but significantly increased risk of fair or bad health with increasing years since migration, which would be consistent with a health selection effect wearing off over time. Although these findings are clear, their interpretation is not entirely straightforward. Those who migrate are a selected subgroup of the population from which migration occurs, and health may be significant in this selection. Those who are healthier may be more likely to have migrated in certain contexts (long distances and active recruitment into jobs), while those who are less healthy may be more likely to have migrated in other circumstances (short distances). If the latter were the case, we would expect the health of migrants to be poorer than that of nonmigrants. If the former were the case, we would expect the health of migrants to be better than that of nonmigrants. Furthermore, we might expect any health selection effect to wear off over time, with health becoming poorer as time since migration increases. Thus, the evidence of a deterioration in health with time since migration for Irish and Pakistani groups lends some support to the suggestion of positive health selection for them. However, previous evidence has suggested that Irish migrants were not positively health selected (Marmot et al., 1984), and it is not clear why Pakistani migrants would be positively health selected, but not Indian and Bangladeshi migrants. Therefore these findings should be interpreted with care. An alternative interpretation is that if the process of migration itself, and experiences postmigration, led to a deterioration in health, we might expect the effects of positive health selection to be attenuated, and a pattern similar

OCR for page 675
Critical Perspectives on Racial and Ethnic Differences in Health in Late Life to that shown graphically in Figure 18-4 to appear; that is, the lack of difference shown in this figure, and in the regression analysis, might be a consequence of the competing drives of positive health selection and the adverse health consequences of migration. Thus the positive health selection effect becomes suppressed. This would then explain why evidence for positive health selection is only present for two of the six migrant groups studied. Generation Differences Despite the apparent lack of difference in health between migrants and nonmigrants, the process of migration to the United Kingdom has not been neutral. Ethnic identity is one dimension we might expect to be influenced by the process of migration, the experiences of second generation people, and the globalization of media. Work on ethnic identity, based on the FNS and using factor analysis, identified a number of potential underlying dimensions of ethnic identity, including one reflecting a traditional identity (Karlsen and Nazroo, 2002a). This included dimensions such as: wearing traditional clothes; speaking traditional languages; thinking of oneself as a member of an ethnic minority group; not thinking of oneself as British; and believing that close relatives should marry a member of the same ethnic group. This dimension of identity correlated strongly with a number of demographic factors, including age at migration. However, after the inclusion of age, gender, and occupational class in a regression model, it did not correlate with health outcomes (Karlsen and Nazroo, 2002a). The implication is that a change in the strength of traditional ethnic identities is not an explanation for ethnic inequalities in health in general, and does not contribute to the emergence of ethnic inequalities in health at older ages. However, health behaviors appear to be strongly correlated with generation. For example, in the United Kingdom, first generation South Asian migrants have much lower rates of smoking than second generation South Asian migrants (Nazroo, 1998). There are also economic differences between migrant and nonmigrant ethnic minorities in the United Kingdom. Evidence indicates significant downward social mobility for most postwar migrant groups (Heath and Ridge, 1983; Smith, 1977), and such downward mobility may have impacted on health. There is also evidence of a correction of some of this downward mobility for second generation ethnic minority people. This is illustrated in Table 18-7, which shows the proportion of men who are in a manual rather than nonmanual occupation, focusing on those of working age (16 to 65) and split between migrants and nonmigrants. It shows that

OCR for page 675
Critical Perspectives on Racial and Ethnic Differences in Health in Late Life TABLE 18-7 Occupational Class by Country of Birth: Men Aged 16 to 65 (percent)   In a Manual Occupational Class Caribbean Indian Pakistani Bangladeshi Chinese White Minority Not born in England 73 53 64 77 50 43 Born in England 56 52 43 35 32 53   SOURCE: 1999 HSE (Erens, Primatesta, and Prior, 2001). although there is no difference for the Indian group, for the white minority group migrants are more likely to be in nonmanual jobs, and for all of the other groups second generation men are less likely to be in a manual job than first generation men. It is worth noting that while the inclusion of men from all ages was necessary because of small samples with more restricted age groups, this leads to underestimation of the relative advantage of second generation men because younger men are more likely to be second generation and at the beginning of an occupational career. Such differences may be very important. There is an extensive literature on socioeconomic inequalities in health and how these might relate to ethnic inequalities in health (Davey Smith, Wentworth, Neaton, Stamler, and Stamler, 1996; Lillie-Blanton and Laveist, 1996; Navarro, 1990; Nazroo, 1998, 2001; Rogers, 1992). However, most of this work has been applied across the population as a whole. The next section explores socioeconomic effects further. Socioeconomic Effects The process of standardizing for socioeconomic position when making comparisons across groups, particularly ethnic groups, is not straightforward. As Kaufman and colleagues (Kaufman, Cooper, and McGee, 1997; Kaufman, Long, Liao, Cooper, and McGee, 1998) point out, the process of standardization is effectively an attempt to deal with the nonrandom nature of samples used in cross-sectional studies—controlling for all relevant “extraneous” explanatory factors introduces the appearance of randomization. But, attempting to introduce randomization into cross-sectional studies by adding “controls” has a number of problems, neatly summarized by Kaufman et al. (1998, p. 147) in the following way: When considering socioeconomic exposures and making comparisons between racial/ethnic groups … the material, behavioral, and psychological circumstances of diverse socioeconomic and racial/ethnic groups are distinct on so many dimensions that no realistic adjustment can plausibly simulate randomization.

OCR for page 675
Critical Perspectives on Racial and Ethnic Differences in Health in Late Life Indeed, an analysis of ethnic differences in income within class groups in the FNS emphasizes this point. Table 6.11 in Nazroo (2001) showed that while total household income adjusted for household size followed the class gradient for each ethnic group, within each class group, ethnic minorities had a smaller income than whites. Indeed, for the poorest groups—Pakistani and Bangladeshi—differences were twofold and equivalent in size to the difference between the top- and bottom-class groups in the white population. A similar pattern existed for other indicators of socioeconomic position. One way of beginning to deal with this is to enter several indicators of socioeconomic position into the analysis at the same time. Figure 18-5 presents 1999 HSE data where this has been done for the total population. It shows odds ratios of reporting fair or bad health in comparison with the white English group, not adjusted and adjusted for several indicators of socioeconomic position. The natural logarithm of the odds ratio is used, so that the visible size of the difference is more meaningful. Adjustment for socioeconomic indicators produces a reduction in the difference between the ethnic minority group and the white English group for all except the white minority and Indian groups. For the Caribbean, Pakistani, and Bangladeshi groups, the reduction is large. These findings are consistent with other explorations of the contribution of socioeconomic position to ethnic inequalities in health in the United Kingdom, which have suggested that across ethnic groups and across health outcomes, socioeconomic in- FIGURE 18-5 Odds ratio for reported fair or bad health compared with white English: All ages. SOURCE: 1999 HSE (Erens, Primatesta, and Prior, 2001).

OCR for page 675
Critical Perspectives on Racial and Ethnic Differences in Health in Late Life FIGURE 18-6 Odds ratio for reported fair or bad health compared with white English: Aged 50 or older. SOURCE: 1999 HSE (Erens, Primatesta, and Prior, 2001). equalities make a major contribution to ethnic inequalities in health, particularly for the poorer groups (Nazroo, 1997, 2001). Figure 18-6 repeats the analysis shown in Figure 18-5, but for respondents aged 50 or older. Although the overall pattern is similar, the reduction in effects following adjustment is smaller for most groups. Racial Harassment and Discrimination Not all of the social disadvantage faced by ethnic minority groups can be summarized with economic variables. As described earlier, experiences of racism and discrimination are commonplace for ethnic minority groups in the United Kingdom. Furthermore, such experiences appear to be related to health in the United Kingdom, as well as the United States (Krieger, Rowley, Herman, Avery, and Phillips, 1993; Krieger and Sidney, 1996). Table 18-8, drawn from analysis of the FNS (Karlsen and Nazroo, 2002b), shows that reporting experiences of racial harassment and perceiving employers to discriminate against ethnic minorities are independently related to likelihood of reporting fair or poor health, and that this relationship is independent of socioeconomic effects. This may represent three dimensions of inequality operating simultaneously: economic disadvantage; a sense of being a member of a devalued, low-status group (British employers dis-

OCR for page 675
Critical Perspectives on Racial and Ethnic Differences in Health in Late Life TABLE 18-8 Racial Harassment or Racial Discrimination and Risk of Fair or Poor Health   All Ethnic Minority Groups Odds Ratio* 95% Confidence Intervals Experience of racial harassment     No attack 1.00 — Verbal abuse 1.54 1.07-2.21 Physical attack 2.07 1.14-3.76 Perception of discrimination     Fewer than half of employers discriminate 1.00 — Most employers discriminate 1.39 1.10-1.76 *Adjusted for gender, age, and occupational class. SOURCE: Fourth National Survey (Karlsen and Nazroo, 2002b). criminate); and the personal insult and stress of being a victim of racial harassment. Age, Generation, and Cohort How far the ethnic patterning of health described in previous sections can be attributed to an aging process, the generation-specific impact of migration, or contextual effects that vary across age cohorts is not clear, partly because of the cross-sectional nature of available data. What is clear is that ethnic inequalities in health in the United Kingdom widen with age, but are present during early childhood. Such a pattern may be a consequence of an accumulation of risk across the life course, or the playing out of childhood differences across the life course, with early “exposures” having an impact in early childhood that reappears in adulthood. The fact that socioeconomic effects appear to be greater when all ages are considered, rather than just older people, might be a consequence of this—contemporary measures of economic position may have less predictive value for outcomes that are a consequence of early or accumulated socioeconomic effects. So it may be that early life experiences are crucial, which in turn may mean that it will take a few generations of upward social mobility for ethnic inequalities in health at older ages to diminish. The evidence presented suggested that the experience of migration, either in terms of the event itself or longer term consequences, was not an explanation for the emergence of ethnic inequalities in health at older ages. However, this may be a premature conclusion to draw, particularly as effects may have occurred alongside health selection into a migrant group, with the subsequent suppression of any relationships at a statistical level. Related to this, it may be that current context is important; that the contemporary experiences of particular cohorts will determine differences

OCR for page 675
Critical Perspectives on Racial and Ethnic Differences in Health in Late Life in health outcomes. Here changes in the economic position of second generation people in comparison with a migrant generation, which occur alongside other shifts, such as those in ethnic identity, may be important. It may be that we cannot predict future health experiences for younger cohorts on the basis of the experiences of older cohorts. However, one context that does not appear to be changing dramatically in the United Kingdom is experiences of racial harassment and discrimination. The impact of this on health was briefly illustrated earlier, and shown to be a potentially very important determinant of ethnic inequalities in health. CONCLUSION AND RECOMMENDATIONS This chapter has raised, but been unable to come to clear conclusions about, a number of issues. How far is the growth of ethnic inequalities in health with age a consequence of aging? The implication of such a consequence is that the inequality is a product of prenatal events or those occurring in childhood (as the observed ethnic inequalities in health in childhood might themselves be), or the product of the accumulation of risk over the life course. In contrast, it might be a consequence of a specific generational effect—the act of migration and its consequences—although the evidence presented suggested this might not be the case. Of course, the experiences of migrants and second generation ethnic minorities are different, with the second generation having less traditional identities and being economically more successful. Such differences between cohorts might suggest different future experiences for younger ethnic minorities compared with the contemporary experiences of middle-aged and older ethnic minorities. The historical pattern of migration to the United Kingdom means there is a close correlation between age, generation, and period of migration within ethnic minority groups, making it impossible to come to clear conclusions with cross-sectional data such as those reported here, even though we can use them to begin to explore possible explanations. Similar difficulties are likely to exist in other countries, and such difficulties require investment in panel data. While not a panacea, panel data will enable us to begin to sort out age, generation, and cohort effects, and the relevance of these for an understanding of population differences in health is great. The additional diversity in experience offered by studies of ethnicity, and the relationships among age, ethnicity, and health, will greatly strengthen our ability to understand the social mechanisms underlying inequalities in health. However, it is important to recognize the overriding importance of national and historical context, and how this influences the lives of ethnic minority and migrant populations. Racism can be considered to be fundamentally involved in the structuring of economic, social, and health opportunities for ethnic minorities. At an individual level, experiences and per-

OCR for page 675
Critical Perspectives on Racial and Ethnic Differences in Health in Late Life ception of discrimination and harassment appear to be strongly related to health. At an institutional level, discrimination clearly influences the economic opportunities that people have, as well as the quality of health and social services that they receive. In addition, historically racism has fundamentally structured the construction of ethnic minority groups and patterns of migration at both an international level (when, why, and where to) and a national level (which locations, which industries). For example, Fenton (1999), building on the work of Eriksen (1993), has distinguished five types of ethnic-making or migration situations: Urban minorities, who are often migrant worker populations. Proto-nations or ethnonational groups, who are peoples who have and make a claim to be nations and to some form of self-governance. Ethnic groups in plural societies, who are the descendants of populations who have typically migrated as coerced, voluntary, and semivoluntary workers. Indigenous minorities, those dispossessed by colonial settlement. Postslavery minorities: the descendants of (African) people formerly enslaved in the “New World.” Athough this typology might not be comprehensive, it does point to the different contexts within which ethnicity or “race” become mobilized to form distinct groupings. Implicit in the typology is that the differing processes listed will lead to different forms of racialization, of subsequent disadvantage, and to different historical trajectories for the groups involved. Understanding the process and the context within which ethnic groups are “made” should aid in understanding disadvantage and how disadvantage might develop. The value in international studies is that such processes have occurred differently across countries. Understanding how they are related to future trajectories for ethnic minority groups (across generations and cohorts) and individuals (over time) is important to any understanding of ethnic difference, including ethnic inequalities in health. ACKNOWLEDGMENTS Work for this chapter was supported by a grant from the U.K. Economic and Social Research Council under the Growing Older Programme, Grant Number: L480254020. ENDNOTES 1.   FNS (Modood et al., 1997) was the fourth in a series of studies on the lives of ethnic minorities in Britain, conducted by the Policy Studies Institute. It was a representative survey of the main ethnic minority groups living in Britain, together with a comparison sample of

OCR for page 675
Critical Perspectives on Racial and Ethnic Differences in Health in Late Life     whites. Topics covered included economic position, education, housing, health, ethnic identity, and experiences of racial harassment and discrimination. 2.   HSE data will be used extensively in this chapter. The HSE is an annual survey conducted jointly by the National Centre for Social Research and University College London on behalf of the Department of Health. Interviews are administered to a nationally representative sample identified using a stratified sampling design. A follow-up biomedical assessment is usually carried out on the sample, which involves measurement of height, weight, blood pressure, etc., and taking a blood sample. The focus of the HSE shifts from year to year. In some years it takes a specific disease focus (e.g., cardiovascular disease), and in other years it focuses on particular population groups (e.g., children). In 1999 the HSE focused on ethnic minorities, providing much of the data shown in this chapter. REFERENCES Balarajan, R., and Bulusu, L. (1990). Mortality among immigrants in England and Wales, 1979-83. In M. Britton (Ed.), Mortality and geography: A review in the mid-1980s, England and Wales (pp. 104-121). London, England: Her Majesty’s Stationery Office. Battle, R.M., Pathak, D., Humble, C.G., Key, C.R., Vanatta, P.R., Hill, R.B., and Anderson, R.E. (1987). Factors influencing discrepancies between premortem and postmortem diagnoses. Journal of the American Medical Association, 258(3), 339-344. Bloor, M.J., Robertson, C., and Samphier, M.L. (1989). Occupational status variations in disagreements on the diagnosis of cause of death. Human Pathology, 30, 144-148. Coleman, D., and Salt, J. (1996). Ethnicity in the 1991 Census: Volume 1: Demographic characteristics of the ethnic minority populations. London, England: Her Majesty’s Stationery Office. Davey Smith, G., Wentworth, D., Neaton, J., Stamler, R., and Stamler, J. (1996). Socioeconomic differentials in mortality risk among men screened for the Multiple Risk Factor Intervention Trial: II. Black men. American Journal of Public Health, 86(4), 497-504. Erens, B., Primatesta, P., and Prior, G. (2001). Health survey for England 1999: The health of minority ethnic groups. London, England: Her Majesty’s Stationery Office. Eriksen, T.H. (1993). Ethnicity and nationalism: Anthropological perspectives. London, England: Pluto Press. Evandrou, M. (2000). Social inequalities in later life: The socio-economic position of older people from ethnic minority groups in Britain. Population Trends, 101, 11-18. Fenton, S. (1999). Ethnicity: Racism, class and culture. Basingstoke, England: MacMillan Press. Harding, S., and Maxwell, R. (1997). Differences in the mortality of migrants. In F. Drever and M. Whitehead (Eds.), Health inequalities: Decennial supplement Series DS no. 15 (pp. 108-121). London, England: Her Majesty’s Stationery Office. Heath, A., and Ridge, J. (1983). Social mobility of ethnic minorities. Journal of Biosocial Science, 8(Suppl.), 169-184. Hickman, M., and Walters, B. (1997). Disability and the Irish community in Britain: Report of research undertaken for the CRE. London, England: Commission for Racial Equality. Karlsen, S., and Nazroo, J.Y. (2002a). Agency and structure: The impact of ethnic identity and racism on the health of ethnic minority people. Sociology of Health and Illness, 24(1), 1-20. Karlsen, S., and Nazroo, J.Y. (2002b). The relationship between racial discrimination, social class and health among ethnic minority groups. American Journal of Public Health, 92(4), 624-631.

OCR for page 675
Critical Perspectives on Racial and Ethnic Differences in Health in Late Life Kaufman, J.S., Cooper, R.S., and McGee, D.L. (1997). Socioeconomic status and health in blacks and whites: The problem of residual confounding and the resiliency of race. Epidemiology, 8, 621-628. Kaufman, J.S., Long, A.E., Liao, Y., Cooper, R.S., and McGee, D.L. (1998). The relation between income and mortality in U.S. blacks and whites. Epidemiology, 9(2), 147-155. Krieger, N., and Sidney, S. (1996). Racial discrimination and blood pressure: The CARDIA study of young black and white adults. American Journal of Public Health, 86(10), 1370-1378. Krieger, N., Rowley, D.L., Herman, A.A., Avery, B., and Philips, M.T. (1993). Racism, sexism, and social class: Implications for studies of health, disease, and well-being. American Journal of Preventive Medicine, 9(Suppl.), 82-122. Lillie-Blanton, M., and Laveist, T. (1996). Race/ethnicity, the social environment, and health. Social Science and Medicine, 43(1), 83-91. Marmot, M.G., Adelstein, A.M., Bulusu, L., and Office of Population Censuses and Surveys (OPCS). (1984). Immigrant mortality in England and Wales 1970-78: Causes of death by country of birth. London, England: Her Majesty’s Stationery Office. Modood, T., Berthoud, R., Lakey, J., Nazroo, J., Smith, P., Virdee, S., and Beishon, S. (1997). Ethnic minorities in Britain: Diversity and disadvantage. London, England: Policy Studies Institute. Navarro, V. (1990, November 17). Race or class versus race and class: Mortality differentials in the United States. Lancet, 1238-1240. Nazroo, J. (1999). Ethnic inequalities in health. In D. Gordon, M. Shaw, D. Dorling, and G. Davey Smith (Eds.), Inequalities in health: The evidence presented to the Independent Inquiry into Inequalities in Health, chaired by Sir Donald Acheson (pp. 155-169). Bristol, England: Policy Press. Nazroo, J.Y. (1997). The health of Britain’s ethnic minorities: Findings from a national survey. London, England: Policy Studies Institute. Nazroo, J.Y. (1998). Genetic, cultural or socio-economic vulnerability? Explaining ethnic inequalities in health. Sociology of Health and Illness, 20(5), 710-730. Nazroo, J.Y. (2001). Ethnicity, class and health. London, England: Policy Studies Institute. Owen, D. (1992). Ethnic minorities in Great Britain: Settlement patterns. (National Ethnic Minority Data Archive 1991 Census Statistical Paper No. 1). University of Warwick, England: Centre for Research in Ethnic Relations. Owen, D. (1994). Spatial variations in ethnic minority groups populations in Great Britain. Population Trends, 78, 23-33. Rogers, R.G. (1992). Living and dying in the USA: Socio-demographic determinants of death among blacks and whites. Demography, 29, 287-303. Smaje, C. (1996). The ethnic patterning of health: New directions for theory and research. Sociology of Health and Illness, 18(2), 139-171. Smith, D. (1977). Racial disadvantage in Britain. Harmondsworth, England: Penguin.