Health selection: the role of inter- and intra-generational mobility on social inequalities in health
Introduction
Potential explanations for social inequalities in health include differences in material circumstances, health behaviour and health selection, as discussed in the Black Report (Townsend & Davidson, 1992). The basis of health selection is that health exerts a strong effect on the attainment of social position resulting in a pattern of social mobility through which unhealthy individuals drift down the social scale and the healthy move up. The literature on health and social mobility suggests that, in general, health status influences subsequent social mobility (Illsley, 1955; Wadsworth, 1986; West, 1991), but evidence is patchy and not entirely consistent across different life stages. Also, there has been limited and inconclusive evidence on the effect that this could have on health gradients (Blane, Davey-Smith, & Bartley, 1993; Bartley & Plewis, 1997). Recently, it was proposed that health-related social mobility does not widen health inequalities, but rather it reduces them, because the upwardly mobile have poorer health status, and the downwardly mobile have better health, than those in the class of destination (Bartley & Plewis, 1997). Again, the evidence is inconsistent with some studies suggesting that health selection acts to reduce the magnitude of inequalities (Power, Stansfeld, Matthews, Manor, & Hope, 2002a; Blane, Harding, & Rosato, 1999a; Blane, Davey-smith, & Hart, 1999b), whereas others do not (Elstad, 2001). Whilst, health selection is not regarded as the predominant explanation for health inequalities (Fox, 1985; Blane et al., 1993; Davey Smith & Morris, 1994; Macintyre, 1997; Marmot, Ryff, Bumpass, Shipley, & Marks, 1997), more insight is needed on whether it acts to widen or narrow inequalities across different health measures and stages of the life course.
Several approaches have been used to establish the role and magnitude of health selection on the social gradient. One approach focuses on the effect of social mobility, that is all social mobility and not just that related to health status, on health or health gradients (Power, Matthews, & Manor, 1996; Rahkonen, Arber, & Lahelma, 1997). A second approach focuses on the effect of health status at an earlier life stage in relation to health gradients later on (Power, Manor, Fox, & Fogelman, 1990; van de Mheen, Stronks, Looman & Mackenbach, 1998). Both of these approaches overstate the effect of health selection. In the first instance, the entire effect of social mobility is captured, particularly that associated with education, rather than mobility specific to prior ill-health. In the second instance, the contribution of health selection is again overstated, because health status at earlier life stages is itself likely to be socially patterned. By allowing for the effect of health status at an earlier life stage, we simultaneously incorporate the social inequality accompanying prior health status. Social gradients in health at earlier life stages have themselves been influenced by material circumstances and health behaviour, as well as health selection. Thus, a third approach has been suggested to overcome these difficulties by focusing on both prior health status and social mobility (Lundberg, 1991; Rodgers & Mann, 1993). This can be achieved using statistical models, which examine how prior health status influence the direction of subsequent mobility, allowing for class of origin.
It has been argued that health selection would have a stronger effect around the time of labour market entry, that is inter-generationally, when the likelihood of social mobility is greatest (West, 1991; Blane et al., 1993). Thus, it is important to distinguish between inter- and intra-generational health selection, although few studies are available that examine selection at both stages of mobility. In previous work, we examined mobility and health status at labour market entry in the 1958 British birth cohort, using the first two approaches mentioned above (Power, Manor, Fox, & Fogelman (1990), Power, Matthews, & Manor (1996); Power, Manor, & Fox, 1991). As we have already argued, these approaches have limitations in accurately assessing the contribution of health selection. The present paper extends previous work on selection and social inequalities in health in that it uses the third approach (mentioned above) to estimate the contribution of selection due to ill-health. This is distinct from selection related to other characteristics, such as education and health-related behaviour, which is summarised by the overall pattern of social mobility, examined in detail elsewhere (Power et al., 1996). Here we consider both inter and intra-generational health selection within the 1958 birth cohort. In order to fully examine the contribution of health selection on health gradients, we examine: (a) the extent and pattern of social mobility in the cohort; (b) the influence of health status on subsequent social mobility; and (c) the effect of health selection on health inequalities.
Section snippets
Study sample
The 1958 birth cohort includes all children born in England, Wales and Scotland during the 3–9 March 1958 (Centre for Longitudinal Studies, 1994). The study has been described in detail elsewhere (Ferri, 1993). In brief, information was collected on 98% of births totalling 17,414. Subjects have been followed from birth through to age 41, with 11,405 subjects (73% of the target population in 1991) responding at age 33. Those remaining in the study were found to be generally representative of the
Social mobility from birth to age 33
Table 1 shows that at birth around 70% of fathers were in manual classes, declining to 62% when the cohort member was aged 16. By age 23 when cohort members were classified by their own occupation, there was a slight decrease for men and a large decrease for women in the percentage in manual occupations. The main change between ages 23 and 33, was the increasing percentage of subjects in classes I & II. The size of the professional and managerial class had almost doubled comparing the two
Discussion
Studies of health and social position show a continuing interest in the role of health selection (West, 1991). Traditionally, much of the focus has been on health selection as an explanation for inequalities in health and the findings of our paper are particularly relevant to this continuing debate on health selection effects. However, there is now growing appreciation that factors at different life stages are important in relation to adult health outcomes (Kuh & Ben-Shlomo, 1997). This
Acknowledgements
This research was supported by a grant from the (UK) Economic and Social Research Council under the Health Variations Programme (L128251021) to Chris Power, Sharon Matthews, Stephen Stansfeld and Orly Manor. CP is a Fellow with the Canadian Institute for Advanced Research and is grateful to them for personal support.
References (42)
- et al.
Adolescence and health inequalitiesExtensions to Macintrye and West
Social Science & Medicine
(1992) The Black report and beyond what are the issues?
Social Science & Medicine
(1997)- et al.
Social inequalities in healthNext questions and converging evidence
Social Science & Medicine
(1997) - et al.
Childhood and adulthood risk factors for socio-economic differentials in psychological distress
Social Science & Medicine
(2002) - et al.
The influence of adult ill health on occupational class mobility and mobility out of and into employment in The Netherlands
Social Science & Medicine
(1999) Rethinking the health selection explanation for health inequalities
Social Science & Medicine
(1991)Categorical data analysis
(1990)- et al.
Socio-economic mobility among patients with Schizophrenia or major affective disorder
British Journal of Psychiatry
(1995) - et al.
Does health-selective mobility account for socioeconomic differences in health? Evidence from England and Wales, 1971–1991
Journal of Health and Social Behavior
(1997) - et al.
Social selection what does it contribute to social class differences in health?
Sociology of Health and Illness
(1993)