Health, Medicine and Society

Key Theories, Future Agendas

Michael Calnan, Jonathan Gabe, Simon J. Williams

• Publisher: Routledge
• Year: 2002
• Print ISBN: 9780415221351
• eBook ISBN: 9781134598250
• Topic: Social Sciences
• Subtopic: Health Care Delivery
• Index: Social Sciences

Taking as its point of departure recent developments in health and social theory Health, Medicine and Society brings together a range of eminent, international scholars to reflect upon key issues at the turn of the century.
Contributors draw upon a range of contemporary theories, both modernist and postmodernist, to look at the following themes:
*health and social structure
*the contested nature of the body
*the salience of consumption and risk
*the challenge of emotions
Health, Medicine and Society provides a 'state-of-the-art' assessment of health related issues at the millennium and a cogent set of arguments for the centrality of health to contemporary social theory. Written in a clear, accessible style it will be ideal reading for students and researchers in health studies, public health, medical sociology, medicine and nursing.

Part I
Rethinking social structure and health

Chapter 1

Class, time and biography

Mildred Blaxter

Social class has always been a fundamental concept in medical sociology, demonstrating its empirical value for the understanding of ‘health chances’ for the individual ever since the early years of this century when Stevenson constructed a classification based on father’s occupation for the purpose of analysing infant mortality in England and Wales (Stevenson 1925). In the past, however, medical sociologists have been criticised for an atheoretical use of class. Registrar General’s Social Class (RGSC) was undoubtedly useful. In many decades of national statistics, and in countless studies of health outcomes, experiences, behaviour and attitudes, linear trends by RGSC have been the norm. But, increasingly, not only is this timehonoured instrument beginning to falter in certain circumstances, but the processes which lie behind such a classification are coming under scrutiny.

Medical sociology, and especially the ‘inequality in health’ debate, have thus been criticised as being isolated from developments in wider sociology. The theme of this chapter, however, is to document how this is changing. It is argued that, currently, medical sociology is both taking note of contemporary theory of class and contributing to it.

This is occurring largely through an attempt to incorporate the concept of time. Health is a characteristic where time cannot be ignored: the sociology of health is concerned with birth and death, ageing and the lifecourse, becoming ill and getting better, moving through both personal and historical trajectories. Health is neither simply a characteristic of the individual nor an event, but their meeting as they come together in biography. Thus health is a topic which adds in a special way to both structure and action as they are conceived of in the theory of class.

The questions addressed here (and illustrated in an inevitably selected way by reference to a variety of bodies of research) are:

• in what ways has medical sociology articulated with contemporary debates about the concept of class?
• how is time being incorporated, both theoretically and empirically?
• in what ways does biography represent the synthesis of class and time?

The journey is in part from ‘class and health’ to ‘biography and health’. This is a journey from an area which is stereotypically, though not invariably, quantitative, cross-sectional, static, depending on measures of health and of class which are as precise as possible, to a field of work which is probably, though not necessarily, qualitative, encompassing change and the constructed nature of both health and social structure. This journey is mapped in more detail throughout this volume.

Class

In ‘inequality’ studies particularly, social class has always been a key concept. For most of this century RG Social Class has played a major role in the monitoring of trends in mortality and morbidity. The principal question of recent decades has been how to explain the observed linear relationship between health and occupational class. This general pattern is seen throughout industrialised societies and across most measures of health, and remains relatively unaffected by social policies and by generally improving health and lengthening expectation of life. Specific diseases may have specific causes, but cutting across these there is a vulnerability which is clearly related to social structure. Thus the importance of ‘class’ remains, and RG Social Class is still commonly used in analysis on the grounds that it permits comparison with data over a long period of the past, and that it is still a useful predictor of ill health.

In the wider sociological arena, however, the way in which medical sociology has used the system has been criticised for an unclear theoretical basis, and it is argued that contemporary discussion of the meaning of class has been ignored. What Holton and Turner (1994) called the ‘debate and pseudo-debate’ about the ‘future’ of class analysis (Goldthorpe and Marshall 1992), or its ‘death’ (Clark and Lipset 1991), ‘attenuation’ (Morris and Scott 1996) and ‘fragmentation’ (Compton 1996), cannot be rehearsed here. In the practical empirical terms which were perhaps first seen as relevant in medical sociology, the basis of the mounting criticism of class analysis was that large and growing numbers of any population are routinely omitted from the standard classification: the retired, welfare recipients, women engaged in household duties, those who have never been employed. In particular, the use of a system designed for male occupations and lifestyles was increasingly found to be inappropriate for women.

The wider debate on class involved more than simply pointing to the problems of detail in a system which might be outmoded, however. Though changes in the social standing of particular occupations and shifts in the occupational structure have led to modifications in the Registrar General system at successive censuses, there are more fundamental criticisms. Among these is that the class structure of modern industrialised societies, and indeed the very meaning of class, have changed: this is not simply historical change in the relative positions of occupations, but fundamental changes in the significance of occupation. There have been extensive changes in the world of production, with the decline in manufacturing industry. The middle classes have not only increased in size, in both absolute and relative terms, but have also become more differentiated. There has been a shrinkage of the wage labour society, through extended education, earlier retirement, shorter hours, and the development of part-time, shared, and contract work. The boundaries between work and nonwork become more fluid, with flexible forms of employment and domestic and wage labour less clearly separated. There is a shortening of the proportion of the lifespan spent in work. Rising living standards, a decline in the influence of traditional institutions, and the erosion of traditional status orders, have all been implicated in the changing meaning of class.

These practical problems of applying RG Social Class, and doubts about the continuing validity of the system, have caused increasing unease about using class as an explanatory variable in health. In the field of inequality of health, for instance, class continues, despite all the problems noted above, to be a useful descriptive variable, but it offers little to explanation, to the identification of the factors which cause social variation. There is no clarity about what RG Social Class actually measures, or with what accuracy. The basis is officially described as level of occupational skill, implicitly presumed to be associated with both a material, economic dimension and a status dimension. The conflation has been criticised by Weberians and Marxists alike. In fact, rather little attention has been paid by theoretical sociologists to mapping either changing rewards or shifting prestige in RG classes over time, since in the wider sociological arena it is preferred to dismiss the simple RGSC I–V altogether. It is only medical sociology which has remained to some extent tied to the system because of its use for census and mortality data.

The elaboration of class in medical sociology

Thus it is in medical sociology, particularly, that a large body of work has developed in the elaboration of RG Social Class, seeking associations and explanations for socially patterned health in terms of the possible components of class—education, income, occupation, work conditions, lifestyles. This work was, certainly at its beginning, empirically rather than theoretically driven. It does, however, feed back into the concept of class by trying to ‘unpack’ its dimensions.

The use of, for instance, house tenure or car ownership can be seen as an early approach to the replacement of occupational categories by consumption patterns (see e.g. Goldblatt 1990; Davey Smith et al. 1990). Again, recognising that income and living conditions vary widely within social class groupings, research workers have constructed indicators which combine social class with living conditions or financial difficulties (Carstairs and Morris 1989; Bartley et al. 1994; Power et al. 1996). Whilst the UK has continued to emphasise occupationally-based concepts of class, other European countries have tended to use educational qualifications either together with, or in place of, occupation (Rahkonen and Lahelma 1992; Lahelma et al. 1994; Kunst and Mackenbach 1994). Dahl (1994) looked at the joint effects of income, occupation, and education in Norway, concluding that in this study, as in others, occupational class remained the most consistent and important predictor of health. Other work sees class as predicting other measures, such as income or education, but something which ought to be kept separate. Townsend, for instance, has argued for the importance of keeping social class out of his area-based deprivation measure, on the grounds that to include it would confuse the measure of deprivation with its causes (Townsend et al. 1987).

The particular study of groups to which RG Social Class is less easily applicable has made special contributions. Various elaborated measures have been used to analyse health and class in adolescence (Macintyre and West 1991), for older people (Arber and Ginn 1991, 1993; Martelin 1994), or to test alternative classifications for women (Moser et al. 1988; Pugh et al. 1991).

The work on women can be instanced as a particular example of this. Just as, in the past, comparison of the health of men in certain occupations with that of their wives was a central tactic of classical epidemiology, so the ‘new’ social epidemiology is illuminated by considering the meaning of social class for women and for men. Traditionally, the individualistic approach to socio-economic variation in women’s health, using married women’s own occupations rather than their husband’s class, produces narrower class differentials for women than for men, seeming to show that, for married women, ‘own’ occupation is not so clearly an indicator of the household’s material position. For many years alternative ways of classifying have used a combined husband and wife class measure, or have used both the partner’s occupation and own occupation separately as indicators for women’s health (Britten and Heath 1983; Martikainen 1995) Arber (1997) suggests that the increase in employment rates among married women and the greater fluidity in marital status may mean that in future the individualistic approach may be favoured. In an analysis of a large sample from the British General Household Survey, she demonstrated that the usefulness of different approaches may depend on what outcome measures are being used. Women’s ‘limiting long-standing illness’ was associated with their own labour market characteristics, whereas self-assessed health was better predicted by a range of variables including husband’s class and the material conditions of the household. It was concluded that several indicators of social class, each depicting distinct aspects of socio-economic status, should be used.

Class, as a dimension, cross cuts with other social statuses. Arber (1991) pointed out that while women entered into the British debate on class differences in health somewhat belatedly, an American tradition had long been dominated by role analysis, with women’s health considered primarily in terms of marital, parental and employment roles. Using, again, the General Household Survey, she demonstrates how both traditions can be reformulated and integrated. The ways in which women’s roles are associated with health status is determined by material circumstances, but these cannot be captured by occupational class alone. Much other research has similarly explored the ways in which women’s roles and health have to be seen within a structural context, and the interactions between employment status and other variables. Except for those with young children, exclusion from the labour market is clearly associated, for women, with poorer health.

New occupation-based classifications

Another important trend is the work which is beginning to make comparisons of different ways in which occupational class might be defined in explicitly theoretical terms, seeking not just to ‘unpack’ RG Social Class but to explore other systems. Occupation-based classifications used for other areas in sociology such as mobility studies have, for instance, been associated with the names of Goldthorpe and colleagues (Goldthorpe and Hope 1974). The Erikson-Goldthorpe schema, an eleven-category validated measure based on an explicit theory of occupational groupings (Erikson and Goldthorpe 1992) is currently being used for health studies. ‘Classes’ are distinguished in terms of such dimensions of the work setting as conditions of employment, occupational security and promotion prospects. This system has been adopted for a large international comparative study (Kunst and Mackenbach 1994) and has been used in Britain by Bartley et al. (1996a). Using a 1971 and a 1981 cohort from the OPCS Longitudinal Study, these authors found similar magnitudes of class difference to those represented by RGSC, and they comment

It is of considerable significance that substantial and persistent differences in mortality between social groups…have been identified by a schema designed explicitly to group occupations with similar employment relations and with no reference to health data.

(p. 467)

Another example, this time considering morbidity rather than mortality, is the analysis of Wolfarth (1997), who compared classification systems using both conventional measures of socio-economic status (education, occupational prestige) and operationalisations of what was called a neo-Marxist concept of class in terms of control of production (ownership of the means of production, control over labour and investment and control over own work). The latter classification distinguished eight classes, such as bourgeosie, decision-makers, workers, semi-autonomous employees, etc. It was suggested that socio-economic status describes a gradual, quantitative difference between strata, while these neo-Marxist classes are clear entities qualitatively different from one another. In the empirical test, which was of various measures of psychiatric morbidity in a large Israeli sample, both classifications provided predictors for morbidity, but the overlap between them was small. Each appeared to have a unique relationship to psychiatric outcome variables, with ‘class’ adding significantly to the prediction provided by ‘socio-economic status’. The conclusion is that they are distinguishable both theoretically and empirically, and conceptualising social inequality in different ways can enhance the understanding of—in this case— psychiatric morbidity.

Elaborated and differentiated measures of class position were also used by Pierret (1993), but in this case for the study of health-related concepts and discourses rather than health outcomes. In a sample of ‘residents of an old quarter of Paris’, ‘residents of a new city’, and ‘farmers from a rural commune’, the traditional occupational classification proved inadequate in the search for correlates of discourses about health. A classification based on ‘positions in the production system’ proved more illuminating. Five groups were formed:

• small farmers
• unskilled or semiskilled workers and persons with unstable jobs
• middle-level employees in the public sector
• middle-level employees in the private sector
• school teachers.

Pierret asked ‘Might discourses about health (and illness) be organised …on the basis of “constants” such as a person’s sense of time, relations to the state, or feelings of security?’ These groups did indeed provide distinct discourses. For instance, for the groups with manual occupations, bodies were tools, or instruments for work. There were, however, differences between farmers and workers: for farmers, health fitted into a relatively homogenous world view based on a cycle of life, while workers felt socially vulnerable. What distinguished the three non-manual groups was whether they were in the public or private sectors. Public employees referred to concepts of social order; those in the private sector had individualistic models. Pierret concluded: ‘In France, persons’ relations to the state, and in particular whether they work in the public or private sectors, seem to be as important as social origin’ (p. 22).

The risk society

These new types of occupational classification begin to explore class as an explanatory factor in health, rather than simply a descriptive category. Concepts of risk, resources and social control become relevant. Risk, in the form of risk factors for disease, relative risks of mortality, or predictors of ill health, has always been one of the basic concepts of social epidemiology. After a long period when the focus appeared to be on individual risk factors, the concept of the ‘risk society’ (Beck 1992) is now being found particularly fruitful. In part, this is a consequence of the limitations for epidemiology of the conventional individualised approach: even in one of the best cases, for instance, when all known risk factors for coronary heart disease are considered together, they account for only about 40 per cent of the incidence of the disease (Marmot and Winkelstein 1975).

In the conventional model, risk factors tended to be defined largely in terms of behavioural characteristics, and at one time factors such as smoking were offered as the most important part of social class differentials. Without denying that of course lifestyles and behaviours are socially distributed, it is now seen as less simple: in the longitudinal study of British civil servants of different grades known as the Whitehall study, for instance, the social gradient in coronary heart disease mortality was clearly not explained away by smoking, since gradients were similar among smokers and non-smokers (Marmot 1986). There is also a strong suggestion that such behavioural factors have different significance for different social groups: in the large-scale Health and Lifestyle Surveys in England, Wales and Scotland, for instance, ‘healthy’ behaviour was found to be more protective against ill health in better environments and more favourably placed social groups (Blaxter 1990). Measured lung function, among those who gave up smoking, was found to improve more over seven years in non-manual men than in manual, and among those who continued to smoke was found to deteriorate to a greater extent in manual men than in non-manual (Cox et al. 1993).

If this conventional individualised risk factor approach is found to be limited, attention has to turn to the characteristics of societies which foste...

Table of contents

• Cover Page
• Title Page
• Copyright Page
• List of Contributors
• Introduction—Health, medicine and society: Key theories, future agendas
• Part I: Rethinking social structure and health
• Part II: The body
• Part III: Risk and consumption
• Part IV: Emotions
• Notes