Black Report 6 Explanation of Health Inequalities

6.1 Standardised death rates in present day Europe have probably reached their lowest point in the entire history of human society. The twentieth century has witnessed a dramatic decline in the rate of infectious disease, as well as the introduction of powerful therapies for its treatment. Common causes of death which have thus greatly diminished, such as TB and diphtheria were often linked to poverty and material deprivation. They have been replaced by new diseases, some of which seem instead to be linked with affluence and material abundance. On that account inequalities in health might have been expected to diminish. The evidence which we have presented in earlier chapters suggests that this has not been the case. We must now address ourselves to the question of why social class should continue to exert so significant an influence on health in Britain,

6.2 There are a number of approaches to explanation of the relationship between social class and health, or at least some aspect of it, in the present day. In our view none of these approaches provides a wholly satisfactory explanation of the relationship, or one that can account for life cycle differences in the influence of social class on the risk of premature death. Indeed the variable of social class is in itself multi-faceted and its influence probably varies according to age or stage in the life cycle and according to the development sequence of a disease episode itself.


6.3 At the most fundamental level, theoretical explanations of the relationship between health and inequality might be roughly divided into 4 categories:-

i. artefact explanations;
ii. theories of natural or social selection
iii. materialist explanations; and
iv. cultural/behavioural explanations.

In some partial respect each one of these explanatory forms sheds some light on the observed relationships between class and health in present day Britain. We will first describe and discuss in general terms the four approaches and then go on, by reference to the problems of different age-groups, to show that a satisfactory explanation must build essentially on the ideas of the cumulative dispositions and experience of the whole life-time, and of multiple causation.

1.The Artefact explanation

6.4 In this approach to the analysis of the relationship between health and social class, emphasis is placed on the “artificial” nature of the correlated variables. Both health and class are artefacts of the measurement process and it is implied that their observed relationship may itself be an artefact of little casual significance. Hence the suggestion that continuities in the pattern of health inequality in the twentieth century may be more a reflection of changing trends in the occupational structure of British society than of a causal link between material welfare and health (see OPCS, 1978). Shifts in the occupational structure have led to the decline of certain “traditional” partly skilled and unskilled manual occupations (Registrar General’s Social Classes IV and V). Those men and women who continue to follow such trades tend therefore to be older on average than the rest of the active labour force because fresh entrants have been attracted to new job opportunities or they may be in ‘residual” occupations which have always had a poor health record. (This evolution in the social structure has of course been accompanied by some change in ethnic composition of the population.)

6.5 Thus the failure of health inequalities to diminish is believed to be explained to a greater or lesser extent by the reduction of the proportional the population in the poorest occupational classes (the implication being that the upwardly mobile either had better health than those who remained and that therefore the health of some of the latter could even have since improved; or that the health of the upwardly mobile subsequently improved).

6.6 One of the problems here is that because of changes in occupational classification the change in recent decades in the distribution of population between the classes, and especially the “decline” in the proportion in class V, has been believed to be larger than in fact it has been (see Table 3.16 above). Table 6.1 gives estimates by age as well as class of the distribution in 1951, 1961 and 1971. For men aged 25 and over the percentage in class V declined from 11 to 8 and in class IV from 27 to 18 in the 20 years 1951 to 1971 (Most of the corresponding increase taking place in class II, but also class I). However, the rate of change slowed down in 1961-1971 and it can be seen that between a fifth and a quarter of those in the youngest age-groups continued to be found in classes IV and V. Overall, substantially more than a quarter of economically active men remained in classes IV and V in 1971.

Table 6.1 Percentage of men aged 25 and over in five occupational classes in 1951, 1961 and 1971 (England and Wales) according to 1960 (and 1970) classification of occupations.

Age Year Professional I Intermediate II Skilled III Partly skilled IV Inskilled V
25-34 1951 2.1 7.9 54.5 26.8 8.6
1961 5.3 12.9 55.9 18.4 7.5
1971 7.5 18.1 53.2 15.2 6.1
35-44 1951 2.1 12.6 49.0 26.7 9.8
1961 4.3 16.6 53.4 18.8 6.9
1971 6.2 20.7 50.5 16.1 6.6
45-54 1951 1.9 13.1 43.5 28.5 12.9
1961 3.3 18.8 48.5 21.1 8.2
1971 4.8 21.3 48.8 17.9 7.3
55-59 1951 1.8 14.0 39.9 29.5 14.7
1961 2.9 17.8 45.0 23.5 10.8
1971 4.0 20.5 46.3 20.2 9.1
60-64 1951 1.7 13.6 38.3 30.0 16.3
1961 2.6 17.5 42.7 24.5 12.7
1971 3.7 18.7 45.3 21.4 10.9
65-69 1951 1.9 13.3 42.6 29.1 15.0
1961 2.7 17.9 41.8 24.5 13.1
1971 3.4 17.2 43.2 22.5 13.7
70+ 1951 1.9 14.7 46.1 25.4 11.8
1961 3.2 19.2 41.7 24.0 11.9
1971 3.3 19.4 44.0 21.6 11.8
Total 1951 1.9 11.9 47.1 27.5 11.5
1961 3.9 16.8 49.3 21.0 9.0
1971 5.2 1907 48.8 18.1 8.3

2. Natural and social selection

6.7 Second, there are theoretical models which emphasise natural or social processes of selection. In this approach, social class is relegated to the status of dependent variable while health itself acquires the greater degree of causal significance. The occupational class structure is seen as a filter or sorter of human beings and one of the major bases of selection is health, ie. physical strength, vigour or agility. It is inferred that the Registrar General’s class I has the lowest rate of premature mortality because it is made up of the strongest and most robust men and women in the population. Class V by contrast contains the weakest and most frail people. Put another way, this explanation suggests that physical weakness or poor health carries low social worth as well as low economic reward, but that these factors play no causal role in the event of high mortality. Their relationship is strictly reflective. Those men and women who by virtue of innate physical characteristics are destined to live the shortest lives also reap the most meagre rewards. This type of explanation has been invoked to explain the preponderance of individuals with severe mental disorders in social class V (a thesis which was reviewed critically in, for example, Goldberg and Morrison 1963). It is assumed that affected people inevitably drift to the bottom rung of the Registrar General’s occupational scale. Similar selective processes are thought to occur with other forms of disease even though the extent of drift may not be so great.

6 .8 Another problem to which this mode of reasoning has been applied is the distinct regional pattern exhibited by mortality rates. In this ease, the national economy is conceptualised as a mechanism of selection. Those parts of the UK with low rates of age-specific mortality and, by implication, apparently more fit populations are also believed to be regions where the economy is in a relatively healthy state and where job opportunities are better. This regional imbalance in economic vitality is associated with a trend in migration which systematically drains the depressed industrialised regions of their fittest men and women. The net result is believed to be reflected in the differential regional pattern of age specific mortality. Processes of selection may also influence the distribution of mortality between occupations. In certain types of manual work, the level of physical strength and stamina required may in itself act as a means of selecting or rejecting individuals. As a consequence although a given occupation might carry obvious hazards to the physical welfare of a worker, its tendency to attract and keep only the physically robust may disguise these inherent dangers (cf Fox and Collier, 1976). Thus it may be misleading to infer, on the basis of death rates, the level of health risk attached to different occupations; to do so is to assume that the pattern of recruitment is random rather than systematically exclusive.

3. Materialist or structural explanations

6.9 The third general type of explanation of the relationship between class and health emphasises the role of economic and associated socio-structural factors in distribution of health and well-being. There are many separate strands of reasoning within its explanatory rubric which can be more or less ordered according to the extent to which the primary causal significance is assigned directly or indirectly to the role of economic deprivation. Amongst explanations which focus on the direct influence of poverty or economic deprivation in the production of variation in rates of mortality is Marx’s radical critique. This theory of political economy provides a theory of history linked to an explanatory account of the contemporary form and inevitability of economic exploitation.

6.10 Those associated with such a radical approach see health or the physical welfare of workers as a key dependent variable determined by the system of production which also gives a particular character to the culture and the ideology of society. Not all follow Marx and those who do risk interpreting modern problems in a way which he would not necessarily have approved. Crudely expressed in its original form, the argument was as follows. Capitalism is in essence a system of economic organisation which depends on the exploitation of human labour. The accumulation of profit, the guiding principle not only of the economic system, but of the whole form of capitalist social organisation, is the storing up in tangible form of the human effort and resources expended by individual workers over and above what they either require or have been allowed, to maintain their bodies in a fit and healthy condition. Marx did not use the modern concept of health in his analysis and was primarily concerned with the material welfare of human beings. This is reflected in his theory of “immiserisation” which assumes a minimum subsistence level which capitalist social organisation systematically violates through its greed for profit. In the process of immiseration the worker experiences economic deprivation on an ever increasing scale until finally he is left with insufficient resources to maintain bodily health.

6.11 Placed in its historical context Marx’s analysis can be seen, at least in part, as a counter critique to Malthusian theory which saw the relationship between death, disease and poverty as a natural phenomenon: the demographic safety valve of the fixed relationship between population size and the natural level of material wealth in the world. Marx’s antidote to this “naturalistic” theory of social inequality stressed the potential elasticity of material production under the capitalist mode of production, while at the same time drawing attention to its dependency on an unequal distribution of resources as well as its inherent tendency to promote material inequality in health as well as property between human beings.

6.12 Most modern proponents of Marxist theory do not interpret the process of material exploitation in terms of human bodily resources, and tend to measure surplus value in terms of wealth or property, the factors into which human labour is transposed. With the benefit of a century’s hind-sight the validity of much of this nineteenth century theory of the relationship between health and material inequality has been accepted today especially for the earlier phase of competitive industrial capitalism (G F Stedman-Jones 1971; Thompson, 1976). Exploitation, poverty and disease have virtually become synonymous for describing conditions of life in the urban slums of Victorian and Edwardian cities, as they are today for the shanty towns of the underdeveloped world.

6.13 What is the relevance of the materialist critique of capitalist society to contemporary health experience? Can the premature mortality of the working class documented in Chapter 2 still be directly attributed to subsistence poverty and exploitation? It is true a relationship between material deprivation and certain causes of disease and death is now well-established, but then so is the capacity of the capitalist mode of production to expand the level of human productivity and to raise the living standards of working people. Economic growth of the kind most readily associated with the European style of industrialisation has in itself been credited with the decline in mortality from infectious disease during the nineteenth and twentieth centuries (cf McKeown: 1976; Powles: 1975). Today death rates for all age groups in, Britain are a fraction of what they were a century ago and many of the virulent infectious diseases have largely disappeared, (cf Morris 1975; OPCS 1978), and the “killer” diseases of modern society – accidents, cancer and heart disease – seem less obviously linked to poverty. Against this background, the language of economic exploitation no longer seems to provide the appropriate epithet for describing “Life and Labour” in the last quarter of the twentieth century. Through trade union organisation and wages council machinery it is now argued that labour is paid its price and, since health tends to be conceptualised in optimum terms as a fixed condition of material welfare which, if anything, is put at risk by affluent living standards, it is assumed by many that economic class on its own is no longer the powerful determinant of health that it once was.

6.14 The flaw in this line of reasoning is the assumption that material subsistence needs can be uniquely and unambiguously defined in terms independent of the level of economic development in a society. People may have too little for their basic physiological needs. But poverty is also a relative concept, and those who are unable to share the amenities or facilities provided within a rich society or who are unable to fulfil the social and occupational obligations placed upon them by virtue of their limited resources can properly be regarded as poor. They may also be relatively disadvantaged in relation to the risks of illness or accident or the factors positively promoting health.

6.15 The structure of living standards has been slow to change in Britain. Personal wealth is still concentrated in the hands of a small minority of the population, as the reports of the Royal Commission on the Distribution of Income and Wealth have most recently shown (see Appendix 6, Table A6.1).

6.16 The question of whether the richest men and women in Britain have maintained their economic position at the expense of less well endowed citizens eludes a categorical answer. The Royal Commission on the Distribution of Income and Wealth has referred to the “remarkable” stability of the unequal distribution of income over the past 2 decades. (See also Appendix 6, Table A6.2). However there is no doubt that the proportion as well as number of the population dependent on a subsistence or near-subsistence income from the State has grown. For some groups, and especially manual groups, relative life-time resources will have been reduced. Earlier retirement, unemployment and redundancies, single parent status and disablement – as well as the proportionate increase in the elderly population, all play some part in this development. For recent years Table 6.2 shows the tendency for those at the lowest relative income standards to increase in number and proportion.

Table 6.2 Numbers of persons in poverty and on the margins of poverty (Family Expenditure Survey)

Income relative to supplementary benefit
Britain (000s)
1960 1975 1976 1977
Under supplementary benefit standard 1260 1840 2280 2020
Receiving supplementary benefit 2670 3710 4090 4160
At or not more than 40 % above standard 3510 6990 8500 7840
Total 7740 12540 14870 14020
% of population 14.2 23.7 28.1 26.6

6.17 A paradox has thus to be noted. While we would not wish to assert that the evidence is consistent and complete, the proportion of the population with relatively low life-time incomes (in the widest sense of “income”) seems to have increased in recent decades, just as the proportion assigned to classes IV and V seems to have decreased. People with low incomes are less able to gain access to the facilities and the knowledge commanded by those with high incomes. Thus whilst economic growth has improved the access of both groups to income and other resources, other groups have gained in proportion and since neither facilities nor knowledge is a finite commodity, those with relatively low income (in increasing numbers) have remained relatively disadvantaged.

6.18 So it has been with health. Social class IV or V may in time catch up with the contemporary levels achieved by I and II but by that time, the latter groups will have forged even further ahead. What is clear from all this is that there is nothing fixed about levels of physical well-being. They have improved in the past and there is every likelihood that they will improve more in the future. Meanwhile class inequalities persist in the distribution of health as in the distribution of income or wealth and they persist as a form of relative deprivation.

Poverty and Ill-Health

6.19 The relationship between average means and descending occupational rank is well- established in the UK – and also in other countries like the US. Reports of national surveys in both the UK and the United States for example, testify to this relationship (GHS, 1972, pl88; FES reports; US, Vital and Health Statistics Series 10, no 70, 1968-69). However there is controversy about trends in the relationship between skilled manual and the lower non-manual workers (as shown in other income and earnings surveys). In a national survey carried out in 1968 the latter worked fewer hours, had larger employer welfare benefits in kind and longer holidays, had larger capital assets, were more likely to be owner occupiers and had fewer dependents. The former’s overall living standards tended to be lower. The highest earnings among skilled manual workers also tended to be temporary (Townsend, 1979, Chapters 10 and 18).

6.20 It is on the basis of knowledge of the relationship between occupational class and family income that inferences are made about poverty and health. Unfortunately the opportunities of examining directly the association between income and health are restricted. Relatively few data about personal health can be related even to occupation and fewer still to income and to wealth.

6.21 More frequent adoption in studies of a more reliable measure of income is desirable (as for example in the GHS) but such a measure will not always be sufficient as an analytic variable in explaining health. Extension of the concept of income must be developed in 2 respects. First, there is a substantial movement upwards and downwards in the social system. Goldthorpe and Llewellyn have shown just how widespread mobility is, not only between generations, but for an individual male moving through his work career (Goldthorpe and Llewellyn 1977). For example, they found that those born to fathers in their classes VI and VII (roughly equivalent to the Registrar General’s III, IV and V) 43 per cent had by the time of the survey (when they were adult) moved out of these classes. Looked at in another way, of those in their classes VI and VII 31 per cent had been born into a different class. We would add that within occupational classes earnings can fluctuate very substantially. While many non- manual occupations are “secure” in the sense that there is career expectation of incremental annual increases of salary, independent of any promotion, many manual occupations are insecure by virtue of dependence on overtime, the vicissitudes of trade and the economy generally and the tendency to move or be moved to lighter work in middle age. (see the evidence of substantial average loss of earnings during a three-year period on the part of the uppermost quartile of skilled manual earners, New Earnings Survey reports.) Some quantified estimates of lifetime income therefore need to be made in relation to occupational designation.

6.22 Second, during the decades since the war a growing fraction of family living standards has depended on benefits in kind. These include employer welfare benefits (sometimes in terms of accommodation and subsidised accommodation), social service benefits and owner-occupation. Methods of comprehensively and reliably valuing such resources and relating them to income have been devised but remain to be improved and accepted. The evidence relating financial poverty (causally) to ill-health is convincing, though only indirect.

6.23 The role played by material factors in the production and distribution of health and ill-health in contemporary times is a complex one. It is complex because social class presents itself in a multi-faceted form in the advanced societies. Apart from the variables most readily associated with socioeconomic position, income, savings, property and housing, there are many other dimensions of social class which can be expected to exert an active causal influence on health status. These other dimensions are encountered in the world of work, in the conditions under which men and women earn their livelihood, in the level of danger and risk, in the degree of security and stability, in the scope for self fulfilment and job satisfaction and in the physical or mental character of the task itself. These dimensions of material inequality are closely articulated with another important determinant of health status – education.

Education and Health

6.24 Educational attainment, or more fundamentally, the process of individual intellectual development, is closely linked to social class. To a large extent, it is on the basis of success or the lack of it at school that children are selected for manual and non-manual work and as we have seen, this occupational distinction plays an important part in measured health status differentials. But we can go farther. Bernstein has argued that distinctive patterns of child rearing and socialisation such as those which tended to differentiate between working and middle class families, produce quite different linguistic capacities which are in turn correlated with quite different intellectual approaches to the social world (Bernstein, 1971). The working class child is rendered at a particular disadvantage on account of these differences because of the fit which exists between middle class norms of socialisation and the dominant structure of the educational system. The outcome of this is that children from middle class homes enter the school system already equipped with the appropriate mode of communication and as a result, they have more successful educational careers and leave school with a greater facility to manipulate their social and economic environment (which of course includes health services) to personal advantage. These ideas carry the variable of education far beyond the simple idea of the transmission of knowledge and skills. They imply that education and linguistic skill are amongst the most important sources of advantage and disadvantage in both material and non-material spheres of human existence: determining the extent to which individuals are personally able to realise the level of human potential which advanced industrial civilisation has made possible. This dimension of class inequality in modern society brings us to the meeting point of the economic and the ideological in society. Should we ascribe the deprivation of linguistic, cognitive and communicative skills to economic or cultural factors? Socialisation experience is seen to play a significant role in the transmission of the crucial modern form of deprivation (and may also play a role, for example, in explaining the association between smoking and educational levels [GHS 1976] but is the practice of socialisation an entirely voluntaristic feature of social life or is it constrained by economic and occupational life? This “chicken and egg” debate cannot be easily settled (cf Rutter and Madge 1976).

Health and the National Economy

6.25 Two other lines of theoretical explanation require brief reference at this point. Each is concerned with the effects on the health of the population of macroeconomic variables: levels of production and of unemployment, sectoral distribution of economic production, and so on. Studies of this kind were actually carried out by Morris and Titmuss in the 1940s, in the attempt to examine the effects of the violent economic fluctuations of the 1920s and 1930s upon a variety of health and mortality indicators (see eg Morris and Titmuss 1944a, 1944b). First, Brenner, making use of time series data trends in the US economy and fluctuations in rates of mortality purports to show that recessions and wide-scale economic distress exert and impact on a number of health status indicators including foetal, infant and maternal mortality, the national mortality rate especially deaths ascribed to: cardiovascular disease, cirrhosis of the liver, suicide and homicide rates, and also rates of first admission to mental hospitals (Brenner, 1973,1976, 1977). In fitting the data on economic trends (essentially unemployment) to that of health status indicators Brenner lags the latter between 2-5 years choosing the lag to obtain the best fit. By doing this, he purports to be able to estimate both the initial impact of recession on the dependent variable as well as the cumulative impact over the space of several years. Brenner posits time lags of varying numbers of years between economic change and changes in his various health indicators (on a purely empirical basis) thus establishing the direction of causality in a temporal sequence as well as suggesting the length of time involved in the development process of disease. The problem with such research is that of casual mechanism. How does unemployment increase mortality? Brenner makes use of the somewhat ubiquitous concept of ‘stress’ to link the two.

6.26 The second approach, in opposition to Brenner’s. is concerned to disprove the common assumption (which Brenner’s work supports), that economic growth leads to an increase in general levels of health. The basic theme which runs through the ‘materialist epidemiology’ approach is as follows. In the advanced capitalist societies, surplus value, ie the excessive extraction or exhaustion of human bodily resources and effort, in the productive process, is realised not through the depression of real wages as Marx argued during the last century but through hazardous punishing and physically stressful work processes; human immiserisation is no longer manifested in terms of grinding poverty and deprivation, but finds expression in the spiritual and intellectual impoverishment of industrial workers. This situation is said to be the outcome of historical changes in capitalism itself. In the advanced societies, the exploitation of human labour for profit has become rationalised. Wasteful competition between small firms has been displaced by the emergence of large scale monopolies whose profits are realised not through a penny pinching approach to workers wages but through a two-fold process: massive expansion in the production of commodities linked to the artificial stimulation of demand for them through advertising and the like. (cf- Stark, 1977). Here again it is argued that many of the commodities upon which individual families as well as Western economies depend for their livelihood are either directly poisonous or potentially dangerous for human health. But in this theoretical approach, emphasis is placed more on the process of production than on what is produced. It is in the actual process of commodity production therefore that disease is produced, through physical stress engendered by tense competitive work relations and routines and through social stress manifested in neglected or disrupted networks and relationships in the realm of domestic and community life. As demonstration of this theory, epidemiologists working in this tradition have sought to show the way in which changes in the business cycle correlate with changes in mortality. But here, the correlation of time series data on employment and death are presented in such a way that high rates of mortality appear to follow on from low rates of unemployment and high levels of prosperity. (cf. Eyer, 1975, 1977a, 1977b). In periods of high unemployment death rates are said to fall. This is said to result from the fact that the institutionalised pressure to consume is lessened, workers are relieved from stressful work routines, social solidarity increases, supportive relationships and networks are stimulated and human existence acquires a more varied and more elevated meaning. While not at all associated with this general theoretical approach Draper and his colleagues in Britain have been aware of the importance of certain elements within it, and have argued the negative consequences for health of indiscriminate economic growth, which can mean higher pollution, encouragement of consumption of commodities hazardous to health,, and so on.

The Ameliorative Approach

6.27 Although we have so far interpreted ‘materialist’ or ‘structural’ explanations as being of a holistic, or macrostructural kind this is not wholly the case. Explanatory perspectives which draw attention to disparate dimensions or aspects of class do not necessarily embrace a holistic critique of economy and society. Such approaches, which include that of social medicine, often stress (or empirically demonstrate) the significance of single causal variables such as insufficient income, poor or crowded housing, large families, insecure employment, manual occupations, inadequate diet or low levels of educational achievement. Such factors either singly or in combination with each other can be shown to exert an important influence on health status indicators like mortality so that they offer scope for the amelioration of inequality through special policy initiatives (Cf for example, Morris 1975; Townsend 1974; Zola and Koza 1967; Brennan and Lancashire 1978; Rainwater 1968; Cartwright and Dunnell 1972). In our report we have made, and will continue to make, much of explanations of this kind. But is must be recognised that factors such as these can themselves be explained in macrostractura1 terms. At what point one chooses to stop the analysis – ie what one accepts as an explanatory variable – is a matter of intellectual preference and of the task at hand.

4. Cultural/behavioural explanations

6.28 Cultural or behavioural explanations of the distribution of health in modern industrial society are recognisable by the independent and autonomous causal role which they assign to ideas and behaviour in the onset of disease and the event of death. Such explanations, when applied to modern industrial societies, often focus on the individual as a unit of analysis emphasising unthinking, reckless or irresponsible behaviour or incautious lifestyle as the moving determinant of health status (cf. Fuchs, 1974). These personal traits appear to be relevant causal agents in an historical context where the majority of premature deaths are caused by degenerative disease or motor accidents. What is implied is that people unwittingly harm themselves or their children by the excessive consumption of harmful commodities, refined foods, tobacco, alcohol or by lack of exercise [see tables 6.3, 6.4 and 6.5] or by their underutilisation of preventive health care, vaccination, antenatal surveillance or contraception. Some would argue that such systematic behaviour within certain social groups is a consequence only of lack of education, or of shiftlessness, foolishness or other individual traits. More theoretically developed as the basis for cultural/ behavioural explanations is the ‘culture of poverty’ thesis which has much in common with the idea of ‘transmitted deprivation’.

TABLE, 6.3 Food Consumption by Income Group (oz/person/week) GB1977

Income Group
white bread brown, including wholemeal bread sugar potatoes fruit
A>£110 pw 18 4.8 9.3 29 33
B 25 3.4 11 39 24
C 28 2.8 13 49 20
D<£40 pw 31 3.3 15 52 17

Table 6.4 Cigarette smoking by socio-economic Group (males and females aged 16+) 1975-6

SEG Current smokers %
professional 25
managerial 39
intermediate non-manual 40
intermediate manual 51
semi-skilled manual 53
unskilled manual 59
all 47
professional 28
managerial 36
intermediate non-manual 36
intermediate manual 43
semi-skilled manual 41
unskilled manual 39
all 39

Table 6.5 Participation in active leisure pursuits: ratio of rates for male non-manual to manual workers by age (males aged 16 or over engaging in each activity in the 4 weeks before interview) GB 1977

Age Group
16-29 30-59 60+
Squash/fives 4.4 6.9
athletics 3.3 3.3
rugby 2.9
golf 2.8 3.2 4.9
badminton 2.8 2.8
cricket 2.4 1.7
tennis 2.4 4.1
table-tennis 2.4 4.1
swimming outdoors 1.6 2.1
walking (more than 2 miles) 1.6 1.8 1.7
bowls (indoor) 1.4 1.3 1.1
bowls (outdoor) 1.4 1.4 1.6
playing football 1.1 1.6
swimming (indoor) 1.1 2.2
dancing 0.9 1.1 1.2
gymnastics/yoga/keep fit 0.9 2.1

6.29 As originally proposed by Oscar Lewis, an anthropologist who studied poor communities in Central American and, later, migrant groups in New York, the “culture of poverty” was intended to apply only to market-organised social structure with poorly developed public systems of health, welfare and income maintenance (cf Lewis, 1967). Starting from a distinct cultural anthropological perspective, Lewis argued that human existence in any given environment involves a process of biological and social adaptation which gives rise to the elaboration of a structure of norms, ideas and behaviours. This culture over time acquires an integrity and a stability because of the supportive role it plays in helping individuals to understand and cope with their environment but, through its influence on socialisation practices and the like, it also comes to have an important autonomous influence in the social consciousness of individuals. The integrity of the culture ensures its autonomous survival even when the material base from which it emerged has changed or been modified. It is for this reason that people cling on to outmoded ideas or old-fashioned practices which do not seem to accord with the changed material realities of modern existence. In fact, the ‘culture of poverty’ thesis has been widely criticised by British social scientists, eg Holman (1978, Chapter 3), Rutter and Madge (1976) and Townsend (1979, pp 65-71).

6.30 Consider, for example, the diffusion of acceptance of the idea of family planning. There is general agreement that this practice was first adopted by the professional classes, from whence it diffused to classes beneath (Banks and Banks, 1964). Why should it have happened like that? To pose this question, is, in essence, to question the applicability of the ‘culture of poverty’ thesis. One answer is surely that men and women who felt secure and in control of their material lives were most reluctant to leave reproduction to the will of God – when a means of personal control was available. The question which must be posed a century later is why some young men and women, especially in the lower occupational classes, continue to leave the procreative dimension of their lives in divine hands? Is it lack of knowledge, outmoded ideas, or lack of access to the means of contraception – or is it due to an underdeveloped sense of personal control or self-mastery in the material world? It can certainly be argued that what is often taken for cultural variation in cognition and behaviour is merely a superficial overlay for differing group capacities of self-control or mastery, which are themselves a reflection of material security and advantage.


6.31 Seeking to apply, and to choose between, such complex approaches to explanation (some of which, like that we have labelled ‘materialist’ or ‘structural’ themselves include competing approaches) when applied to evidence as complex at that which we have assembled, is a daunting task. Intellectual honesty demands that we make clear our belief that it is in some form or forms of the ‘materialist’ approach that the best answer lies. But there can be little doubt that amongst all the evidence there is much that is more convincingly explained in other terms: cultural, social selection and so on. Moreover it may well be that different kinds of factors, or forms of explanation, are appropriate to different stages of the life cycle. This possibility has guided our presentation of data for explanation.

1. Birth and Infancy

6.32 The beginning of the human lifetime, birth itself and the first year of life, has witnessed amongst the most spectacular declines in mortality of all age groups during the last century (cf OPCS 1978). Between 1846 and 1976, death rates for infants of less than one year fell by more than 90 per cent and yet at the beginning of the 1970s the ratio of deaths for infants in social class V compared to social class I was between 4 and 5 to 1. How can this inequality amidst so much improvement be explained?

6.33 In the present day the cases of childbirth involving the greatest risk are those where the mother is older and where she has already produced several children. Such eases were, of course, more prevalent in Victorian society where knowledge about birth control was scant and where rather different ideologies about family size existed. In recent years the percentage of mothers in classes IV and V having a fourth or fifth child has decreased, but remains higher than in classes II and III. Class I produces a relatively large number of 4 and 5 child families. It may be that higher rates of stillbirth, or perinatal death, are a consequence of differences in maternal age and parity between the classes.

Table 6.6 Still births and post-neonatal mortality rates by occupational class, age and parity, 1975-6

Age and parity Stillbirths per 1000 live births Post neonatal deaths per 1000 live births
I&II IV & V Ratio I&II IV & V Ratio
No previous births
25 years 7.8 12.2 1.6 2.6 5.2 2.0
25-29 years 8.4 13.2 1.6 2.3 3.1 1.3
30 years 11.5 22.1 1.9 2.7 3.3 1.2
1-2 previous births
25 years 6.6 7.9 1.2 5.0 9.6 1.9
25-29 years 6.2 9.2 1.5 3.2 4.5 1.4
30 years 7.4 14.1 1.9 2.8 4.1 1.5
3+ previous births
25 years 7.1 12.3 1.7 14.3 15.1 1.1
25-29 years 11.5 15.6 1.4 5.3 8.4 1.6
30 years 11.4 22.1 1.9 3.0 6.2 2.1

6.34 However, as table 6.6 indicates, class inequalities in rates of death at birth and throughout the first year of life are found even when parity and maternal age are held constant. While age and parity exert an important influence on the risk of all stillbirths, the significance of these variables is much more evident among the wives of semi-skilled and unskilled manual workers. The risk of stillbirth for the wives of professional and managerial households is lower no matter what their age or their previous record of pregnancies. (Similarly for perinatal death rates: at all ages and parities the class differentials remain.) Rates of post neonatal mortality differ somewhat. Once again the variables of age and parity are more significant sources of differentiation among underprivileged women, but the pattern is different. Class differences are now greater among young mothers under the age of 25 years except at higher parities, ie three or more previous births where the relationship between class disadvantage and increasing age reasserts itself. The conclusion which must be drawn from Table 6.6 is that occupational class differences are real sources of difference in the risk of infant mortality and not merely proxies for the variables of maternal age and parity.

6.35 Also to be explained are similar differences in the class incidence of low birth weight for, as we shall see, this is a handicap which (except under the most advantageous conditions) has long-term implications for the health and development of the neonate.

6.36 Examining the historic decline in perinatal mortality rates Hellier (1977) has established that over half can be attributed to such factors as mother’s health and quality of obstetric care. Similarly, Doll, Hill and Sakula (1960) suggest that variations in the incidence of congenital malformations can largely be attributed to the nutritional status of the mother and the presence/absence of viral infections in pregnancy (eg influenza). In the light of the evidence which we have presented there can be no doubt that mothers’ health, nutrition, and obstetric care received (given differences in date of presentation) are all class related.

6.37 Other explanations of early death, low birth weight, and congenital abnormalities which the literature yields are more complex. Lawrence, Carter, and David (1968), for example, suggest that genetic factors play a ‘Predisposing’ role in giving rise to congenital defects – adverse environmental factors acting as a ‘trigger’. Janerich (1972) comes to a similar conclusion. Illsley’s work shows a mechanism for such genetic predisposition to become class-related.

6.38 Class differences in infant mortality have been ascribed to selective processes in mating and marriage. Illsley (1955) carried out a study of infant deaths in which he investigated patterns of hypergamy and hypogamy in relation to the physical characteristics of the mothers. He found distinctive patterns in relation to variables like height. Tall women showed a marked propensity to marry hypergamously (ie to be upwardly mobile at marriage, short women showed the opposite tendency – to marry below their father’s social class. These inherent physical characteristics of women were then translated into differential rates of infant mortality; the mothers of greater stature having lower rates of infant death. Illsley concluded that social class differences in infant mortality had a physical component: higher class men appeared to be recruiting as wives the most efficient child-bearing women.

6.39 Baird’s thesis of ‘transmitted nutritional deprivation’ offers another variant on the explanatory theme of selection (Baird, 1974). He suggests the existence of a vicious cycle of nutritional deprivation which leads to low birth weight and congenital malformation. This cycle is difficult to break because it originates in the nutritional deprivation of the mother, not at the time of giving birth, but at the time of her own birth. By this account perinatal death and low birth weight are seen as caused by the effects of nutritional deprivation upon the reproductive capacity of female infants. These explanations are based upon data accumulated on the childbearing population of Aberdeen over many years and they offer important insights into the mechanisms and processes whereby social class differences in mortality are produced and perpetuated.

6.40 If the relative importance of factors such as all of these in determining rates of perinatal death (and of handicap among survivors) is controversial, the situation is somewhat clearer in the post-neonatal period. When we look at the causes of death in infancy which exhibit the steepest class gradients, there seems to be much evidence to suggest that the important causal variables are contained within the contemporary socioeconomic environment and therefore subject to human intervention in the short term. As Figure 2 in Chapter 2 clearly indicates, the causes of infant death which are most likely to be associated with the stature or the nutritional status of the mother have the shallowest of gradients, whereas respiratory disease and accidents show steep class gradients. These observations lead us directly to consider the role of material deprivation on the life chances of the newborn.

6.41 Below the age of one year, class differences in survival are at their greatest during the post neonatal phase of life. This clearly represents a point of development when the individual is in need of almost continuous care and total attention from its parents and we may hypothesise that any factors which increase the parental capacity to provide adequate care for an infant will, when present, increase the chance of survival, while their absence will increase the risk of premature deaths The most obvious such factors fall within the sphere of material resources: sufficient household incomes a safe uncrowded and unpolluted home, warmth and hygiene, a means of rapid communication with the outside world, eg telephone or car, and an adequate level of man – or woman power (ie two parents would normally provide more continuous care and protection than one). In addition to these basic material needs must be added other cognitive and motivational factors which are not independent of the distribution of material advantage. Those factors would include knowledge, certain skills and resources in verbal communication and a high level of motivation to provide continuous and loving care. When all these factors are present the infant’s chance of survival is very good indeed. When some or even many of these are absent, then the outlook is less propitious. Moreover it should not be forgotten that these very same factors play a part in determining the development of the infant’s own cognitive/linguistic and other skills. Competences acquired at this stage of life can profoundly influence later intellectual (and hence educational) achievement.

Health Inequality in Childhood (1-14 years)

6.42 The evidence for health inequality between children in the different social classes presents something of a paradox. Rates of mortality throughout childhood form quite steep class gradients when parental occupation is used as a basis of presentation and yet, as the evidence in Chapter 2 shows, self-reported rates of morbidity for this age-group display a rather less consistent picture of disadvantage. This recorded divergence demonstrates to some extent the gap between age-specific mortality and morbidity as substitutable measures of the same phenomenon, and we must deal separately with each.

6.43 The most important causes of death amongst all children aged 1-14 were (in descending order): accidents poisoning and violence; respiratory disease; neoplasms; congenital abnormalities; and infections. Among 1-4 years olds, and this we wish to stress, almost all the differences in mortality rate between social classes I and V is due to: accidents, poisoning and violence; respiratory disease; and congenital abnormalities (which of course play a more important role at younger ages). Among older children, deaths from accidents, poisoning and violence remain highly class related (Table 6.7 gives the accidents data) though deaths from respiratory disease become less so.

Table 6.7 Fatal accidents in childhood by sex and social class (1-14 years)

Social Class (rates per 100,000 population)
All Ratio V/I
Boys 25.8 39 44.5 56.3 66.2 122 586 4.7
Girls 18.8 19 21.4 24.9 35.1 63.1 294 3.4
Sex ratio 1.36 2.05 2.08 2.26 1.88 1.93 1.99

6.44 Since a great deal of the class differential in mortality among 1-4 year olds is due to just two causes: respiratory disease, and accidents and violence one approach to explaining this inequality is by unraveling those aspects of social situation, of way of living, responsible for respiratory disease and accidents/violence – and which may then prove to be closely associated with social class. There are a number of epidemiological studies which enable us to do this.

6.45 Reviewing clinical examinations of the 1946 birth cohort up to the age of 15, Douglas and Waller (1966) found that the principal correlate of respiratory symptoms was the extent of air pollution in the children’s area of residence. There was no tendency for working class children to be more concentrated in high pollution areas, and social class had a small independent effect. A study of 2000 children living in Harrow (which did not permit consideration of a range of air pollution levels) concluded (Leeder, Corkhill, Irwig, Holland and Colley, 1976): “illnesses occurred much more commonly in infants born to families which had several other children already, and in those families where the parents had respiratory disability or were smokers”. When all relevant variables were taken together in a logistic model, the most important proved to be:

1. bronchitis/pneumonia in siblings
2. parental smoking
3. number of siblings
4. parental history of asthma/wheeze.

Social class had no effect independently of these factors. It may thus be that much of the class gradient in morbidity (if not mortality) from respiratory disease can be explained in terms of parental smoking (ie secondary inhalation of smoke) and size of family (and the likelihood of infection by siblings). Both smoking and family size are clearly related to class.

6.46 Colley and Reid’s (1970) study of over 11,000 children aged 6-10 in a number of urban environments was concerned with the interplay of social class and physical environmental variables. Here too, a past family history of respiratory disease was associated with chronic cough in the 6-10 year olds (who were examined by school medical officers). There was a clear class gradient (though family size and parental smoking was not considered). There was a rather complex relationship with air pollution levels. Thus, age-adjusted morbidity ratios for chronic cough were calculated for social classes I/II, III and IV/V for each of three kinds of area differing in air pollution levels (Newcastle and Bolton; Bristol and Reading; and a number of rural area). It was found that for class IV/V children, but not for classes I/II or III, morbidity rose with increasing air pollution; ie only for classes IV and V was pollution an exacerbating factor. (For all classes, morbidity was distinctly higher in Wales, even in rural Wales where pollution was lower, This was felt to require explanation in genetic terms or in terms of selective emigration). Colley and Reid argued that geographic variations within classes IV and V could not be explained by domestic circumstances such as differing levels of crowding (which were not great).

6.47 Finally we might refer to analysis of data from the US National Health Survey (which, like the GHS, is based on self-reporting). Speizer, Rosner and Taget (1976) found “a strikingly higher prevalence of reporting asthma and bronchitis in children over 4 years residing in the same household as a parent with CNSLD (chronic non-specific lung disease)….. [this] is not due to confounding by age, sex or smoking. Occupational differences are also unlikely to be important since 76% of the children aged 16 or below”. Their conclusion is that the principal factor is a “familial (household) aggregation of disease”. This they take to be due to a combination of genetic and environmental factors. Evidence for the relevance of the former was the six-fold excess of disease in children with both parents having CNSLD compared to those with only one. Higgins and Keller (1975) provide further evidence for some genetic factor in transmission. [In general, evidence on the genetic contribution to health inequalities tends to be indirect: little is really known of the importance of genetic factors].

6.48 The general implication of these studies seems to be that the class gradient in bronchitis among children is largely a consequence of parental smoking, family size (and the increasing likelihood of infection by siblings), and a parental history of lung disease (which may to some degree genetically place the child at risk). Parental history of lung disease is also, as shown independently, a function of type and severity of occupation. Environmental pollution is also implicated, and may be a particular danger for those children (from classes IV and V) rendered prone through other factors.

TABLE 6.8 Causes of death from accidents of children aged 1-14 years 1968-84

Type of accident or violence Total age 1-14 years
Age in years
1-4 5-9 10-14
Number % Number % Number % Number %
Total accidents and violence 10887 100 4371 100 3712 100 2794 100
All accidents 10204 93.8 4017 91.9 3566 96.1 2633 94.2
Motor vehicle collision with pedestrian 3656 33.6 1146 26.2 1739 46.8 771 27.6
Other motor vehicle collisions 1199 11 211 4.8 378 10.1 614 22
Other transport accidents 619 5.7 153 3.5 198 5.3 268 9.6
Accidental poisoning 276 2.5 181 4.1 35 0.9 60 2.1
Falls 625 5.7 285 6.5 155 4.2 165 6.6
Fires 904 8.3 629 14.4 210 5.7 65 2.3
Natural and environmental factors 47 0.4 20 0.5 6 0.2 21 0.8
Drowning 1401 12.9 608 13.9 524 14.1 269 9.6
Inhalation of food or other object 343 3.2 252 5.8 51 1.4 40 1.4
Accidental suffocation 375 3.4 211 4.8 47 1.3 117 4.2
Blows, cuts, explosions etc 460 4.2 195 4.5 135 3.6 130 4.7
Accidents caused by electrical current 150 1.4 60 1.4 42 1.1 48 1.7
Surgical and medical complications and misadventures 24 0.2 17 0.4 4 0.1 3 0.1
Other accidents 125 1.1 47 1.1 36 1 42 1.5
Total violence 673 6.2 356 8.1 156 4.2 161 5.8
Suicide 33 0.3 0 0 0 0 33 1.2
Homicide 417 3.8 248 5.7 95 2.6 74 2.6
Injury undetermined whether accidentally or purposely inflicted 223 2.1 108 2.5 61 1.6 54 1.9

6.49 The second condition responsible for much of the gradient in child mortality at ages 1-4 was accidents and violence of which, as Table 6.8 shows, accidents were the principal cause of death. The increased risk of death faced by lower class children in ordinary everyday activities such as playing at home or in the neighbourhood or (among older children) travelling to school by foot or bicycle, has to be viewed in terms of differences in the environment to which children from different classes axe typically exposed. Among child pedestrians for example, the risk of death from the impact of a motor vehicle is multiplied by 5-7 times in passing from class I to class V; for accidental death caused by fires, falls and drowning, the gap between the classes is even greater. These differences demonstrate the non-random nature of accidents as a collective class of events. While the death of an individual child appears as a random misfortune, the overall distribution clearly indicates the social nature of the phenomena. How is it to be explained?

6.50 From among the different modes of explanatory reasoning outlines above, we may quickly dispense with those which attribute causality to the influence of selection or which reduce the problem to an artefact of the measurement process. Accidents have two primary causes: either environmental hazard, or dangerous behaviour reflecting carelessness, adventure or irresponsibility. These primary causes involve both material and cultural factors and indeed full explanation of inequalities in the risk of death in childhood implicates each of them.

6.51 As Table 6.7 shows, while boys in all social classes bar I experience double the risk of a fatal accident of girls, the gap between the social classes is much more striking most especially for boys. When the two extremes of the occupational spectrum are compared it can be seen that children in the lowest social class are almost five times more likely to die before reaching school leaving age. The sons of other manual workers carry over twice the risk of accidental death. The loss of these lives so early in the lifetime, surely reflects the interaction of both material and cultural factors.

6.52 The consistency of the sex differential in each social class is a measure of differences in the way in which boys and girls are socialised. Patterns of child-rearing for all social classes with the exception of professional households, appear on the basis of Table 6.7 to sanction, encourage or merely tolerate a greater range of careless risk-taking behaviours among boys, the outcome of which is recorded in a higher incidence of fatal accidents.

6.53 Health inequalities in childhood expressed in the form of mortality rates can therefore be seen to be at least in part the reflection of cultural practices in socialisation. In the modern world there is nothing obligatory about the different preferences which parents show towards the behaviour of their sons and daughters: a fact which is perhaps reflected in the reduced sex ratio of social class I and which offers some scope for a reduction in this specific form of inequality. Class differences between children in the risk of accidental death, too, may appear in part as a manifestation of distinctive patterns of child-rearing on the part of parents in the different occupational classes. But such “patterns” have to be seen in the light of the great differences in the material resources of parents, which may place significant constraints on the routine level of care and protection that they are able to provide for their children. Children of parents in social classes IV and V are amongst the poorest members of their age group in the population. Their opportunities to play safely within eye – or earshot of their parents are far fewer than those of their better endowed peers higher up the social scale. Furnishings, including forms of heating in the home, are likely to be less safe, as are the other domestic appliances which they encounter.

6.54 These manifestations of the differences in the material resources of households mean also that the children of semi and unskilled workers are more likely to be thrown onto their own devices during holidays and out of schools hours and this alone would be sufficient to increase the probability of their being involved in an accident. In reflecting upon these forms of health inequality in childhood and adolescence it is impossible to escape the conclusion that the physical welfare of children is closely linked with material resources, and that the distribution of the former is a reflection of the distribution of the latter. In the context of childhood therefore, the most straightforward of materialist explanations is capable of providing a simple chain of causation by which the pattern of health inequality is illuminated. Households in social classes IV and V simply lack the means to provide their children with as high a level of protection as that which is found in the average middle class home. These resources consist of income, property and territorial space but they may also take an associated non-material form. As Brown and Harris concluded on the basis of their research in Camberwell, one of the reasons for greater prevalence of accidents in working class homes is the higher incidence of stressful life events experienced by mothers. Such women, who lack the means to resolve the recurrent setbacks which dominate their domestic lives, are less well equipped to provide continuous and vigilant protection and care to their children:

“The mother’s psychiatric state and the presence of a serious long-term difficulty or a threatening life event were related to increased accident risk to children under 16. These factors were more common among working class children, and in so far as they are causal, they go a long way to explain the much greater risk of accidents to working class children.” (Brown and Harris, 1978).

Because Brown seeks to link aetiological factors back to social structural differences which both generate them and determine the severity of their impact, his work provides a unique link with the second approach to the explanation of health inequalities.

6.55. The second, methodologically distinct approach, which also enables us to make sense of differences in morbidity among children (including self-reported data on both morbidity and GP consultations), focuses upon characteristics of life in the deprived household. A valuable recent such study has focused upon 6-7 and 10-11 year old boys from severely deprived large families in Birmingham, known to the social services department. Children were compared with a control group of similarly aged children, living in the same area but not under social service department supervision. The study fell into, and was published in, two parts.

6.56 In the first Brennan focused upon medical characteristics (Brennan, 1973), but interpreted broadly. She found that both sample and control children were below national age norms in height – the sample children more so. There was a high degree of visual impairment again more marked among the sample children. (Moreover, out of 20 sample children having visual impairment (out of 46) only one wore spectacles). There was a higher degree of hearing loss among children. Finally 78.2% of sample children (and 58.9% of control children) were diagnosed as having some illness on clinical examination (far higher than indicated for the city as a whole from school health records): the most important were respiratory disorders, orthopaedic defects, speech defects, skin disorders, and chest complaints.

6.57 The second study (Wilson and Herbert, 1978) considered also the 3-4 year old siblings of those older brothers, and made extensive Use of interviews, observation over a long period, and psychological test data. This study vividly illustrates the nature and the effects of considerable poverty on family life and on child development. It suggests that ill-health, inhibited cognitive development, and behaviour problems are associated in a general ‘poverty syndrome’. For example, of the 3-4 year olds in a specially established experimental play group: “The pervading impression of the group is one of bewildered, ‘lost’ children, who did not know what to expect, who continued to be worried, and who seemed unable to relax; an impression which was reinforced by their very poor clothing, unkempt appearance and smell” (p64).

6.58 Accidents to the children were common: 34 out of 56 families had experienced severe accidents (1 child had lost an eye, 16 suffered burns or scalding needing skin grafts … ).

Particularly striking (once again) is the extent to which ill-health was found to cluster in families. Of the 56 families studied:- “In forty families all, or most, members of the family were reported (ie by themselves) as having had much illness, or as suffering from defects or conditions which affected their activities. Respiratory diseases were most frequently mentioned, followed by gastric conditions and skin conditions.” Moreover, “only 4 among the 16 fitter families can be truly said to be healthy”; and obstetric problems were frequently mentioned.

6.59 The following quotations illustrate how these authors make sense of their findings:

“Thus the children, in the process of growing up, have many shared experiences. They live in overcrowded conditions, being members of large families; their homes are inadequate by current standards; the neighbourhoods are rough and disliked by most who have to live in them. They experience poverty, by which we mean that they go short of things considered essential or normal by others around them. Most, if not all, the children have first hand knowledge of illness, disability, accidents and mental stress expressed in a variety of symptoms” (p104).

“The objective is survival, the operative unit is the family. The needs of individuals must take second place. Decisions were made at family level and related to the main wage earner or recipient of benefit rather than to the needs of individual children.” (pl86)

6.60 The GHS shows that acute disorders are, on the whole, more prevalent amongst the sons of non-manual parents, a pattern which is reflected in the distribution of GP consultations in childhood. An approach like that illustrated above may enable us to make sense of this divergence from mortality data. Self-reported sickness is not a standard health indicator and it might be expected to vary between socioeconomic groups as much on account of variation in cultural values, in the meaning of everyday experience, and on the degree of behavioural flexibility available to men and women in different socioeconomic groups, as on the more obvious manifestations of sickness or disorder. Consultation is obviously easier for non-manual households. They are more likely to have a telephone readily to hand to make enquiries or arrange appointments and access to private motor cars means that a visit to the GP’s surgery causes less trouble and inconvenience. These sorts of differences represent inequalities in the distribution of opportunities to consult and, given that the act of consultation itself is likely to make an episode of sickness seem more “concrete”, certainly more memorable, that might even make an important difference to the incidence of “remembered” ie self-reported-sickness. Since self reports of sickness amongst children are generally provided by parents, the much higher rate among boys in non-manual households may reflect greater accuracy of recall on the part of their mothers or fathers. Perhaps professional parents might be more likely to remember or to acknowledge episodes of acute sickness among children because their own daily activities may be more flexible, making it easier for them to care for a sick child or attend for medical consultation. For the opposite reasons a working class parent would be more prone to scepticism of children’s complaints. If keeping a child at home presents greater disruption to the income generating activities of a household then parents might be expected to have a higher tolerance for minor disorders amongst their children.

6.61 At various points in this report we have indicated the importance which we attach to the promotion of health in its ‘positive’ sense. In childhood this may reasonably be interpreted in developmental terms, and it is relevant to consider finally research evidence relating to physical, behavioural and cognitive development in children.

6.62 In a series of studies over 2 decades, Pasamanick and his associates in the USA found associations between prematurity and birth complications and a subsequent history of cerebral palsy, epilepsy, mental deficiency, behaviour disorders, reading disabilities, strabismus, hearing disorders, and autism (see eg Pasamanick et al 1956: Pasamanick and Knobloch 1960). Douglas in an analysis of IQ, tests administered to children from the 1946 birth cohort aged 8 and 11 matched ‘premature’ babies (who had had birth weights of 5 ½ lbs or less) with controls similar in terms of sex, ordinal position in the family, mother’s age, social group, and degree of crowding in the home (Douglas, 1960). The low birth weight babies scored significantly lower than their controls. They also did less well in the ll+ examination. Douglas attributed this to significant differences, which he also found, in social and educational backgrounds of parents and grandparents, parental interest in school progress, etc. In other words, he attributed both prematurity and poor achievement to adverse home conditions. Drillien, however, found that children of birthweight less than 4 ½ lbs contained a high proportion of dull, retarded, and grossly defective children. Below 3 lbs birthweight only 13% were of average or above average intelligence (Drillien, 1959). Illsley (1966) found a clear association between birthweight and IQ within each social class at age 7: the effects of low birthweight being greater for children born into the lowest classes. Illsley does not feel it is possible to distinguish the possible routes by which this may come about: the parallel effects of environmental influences on reproductive health and cultural background; organic impairment in utero; and genetic inheritance.

6.63 Birch and Gussow (1970), reviewing numerous studies from many countries, conclude that whilst a good postnatal environment can compensate for the perinatal stress associated with prematurity, an environment which is not satisfactory can represent a cumulative hazard. They argue that for children born into poverty the circumstances surrounding their development in utero and their birth may be significantly responsible for “minor disorders of perception and cognition, an increase in impulsivity or distractability, a delay in the mastery of certain bodily functions … 11 (p67).

6.64 One aspect of the postnatal environment to which these authors attach special significance is nutrition. There is circumstantial evidence that nutritional deprivation can affect not only external physical growth, but also the brain and central nervous system. Eichenwald and Fry (1969), amongst others, have related nutrition to learning. Birch and Gussow suggest that the effect of malnutrition upon the brain during the period in which this is growing fastest (first 6 months after birth), unlike the effect on other growth, is not fully compensateable. Whilst in the case of humans there is no direct evidence that effects observed are due to malnourishment rather than other aspects of deprivation, laboratory studies of animals provide some confirmation. Here an effect of malnutrition during the period of rapid brain growth upon brain size and number of brain cells has been demonstrated, and an effect which cannot fully be made good later. Malnutrition appears also to produce abnormal EEG patterns, in both animals and human infants. Birch and Gussow conclude (p2l2): “Thus the evidence would seem to indicate that severe undernutrition during, the early months of a child’s life may reduce the number and size of cells in his brain as well as the extent of myelination, and that nutrition imbalance of a severe degree may produce structural changes in the brain and spinal cord, abnormal electrical activity, and pathological lesions of the central nervous system … However … the relationship between these findings and children’s mental development is an open question, because we know too little at present about the functional implications of most of what we can learn by looking at the brain, weighing it, or its constituents, or reading the pattern of electrical activity on its surface.”

6.65 It is, of course, less controversial to point out that “a hungry child is unlikely to be alert during lessons,” (DHSS Eating for Health 1978; 43). There is, it must be reiterated, ample evidence for the association of adequate nutrition with achievement of full potential for physical growth. The importance of milk in the nutrition (and hence growth) of children has been shown in a number of studies reviewed by the Sub-Committee on Nutrition Surveillance (Report, 1973). Moreover, the lessons of the food policy introduced during the Second World War, to ensure fair shares for all, and which resulted in one of the most rapid falls in perinatal mortality rate (between 1940 and 1948) should not be forgotten.

6.66 Social scientists, of course, have stressed the effects of different aspects of the impoverished early environment on intellectual development, including patterns of parenting (themselves relatable to the effects of economic stress, housing, and family size). Wilson and Herbert (1978), for example, refer to the effects of “delegation of mothering to older siblings; early severance of mother-child contact and play in the unsupervised stress play-groups, scarcity or absence of toys” for “the training of the young child in the achievement of age-appropriate behavioural standards … language development … the development of creative activities, power of concentration, manipulative skills” etc (p184-5).

6.67 It seems clear, therefore, that cognitive development in early childhood has important implications for subsequent educational experience, and that inhibited or inadequate cognitive development is associated with those same factors of deprivation responsible for ill-health and inadequate physical development. Moreover, it thus becomes possible to trace a link (all too often reinforced by social processes) to occupational ascription/attainment. At the same time, too, child ill-health has important implications for adult health (in the negative sense), as research shows.

6.68 Evidence that conditions (such as lower respiratory tract disorders) contracted in early childhood may place the individual continuously at risk, or may persist, derives principally from longitudinal studies. The study of 1,000 Newcastle families illustrates dramatically how repeated respiratory infections in the first 5 years, if inadequately treated, can lead to some degree of disability at age 15 (Miller, Court, Knox and Brandon, 1974: pp 94-128). This is sometimes due to permanent damage caused to the respiratory tract. In the Newcastle study 208 children aged 14 were identified as being ‘at risk’ of respiratory infection, on the basis of severe illness in the first 5 years and other indications, and compared with a group of 97 controls of similar age. Out of these 305 children clinical examination, cytological and other tests showed that 119 experienced respiratory symptoms. Of these 107 came from the ‘high risk’ group. In 45 cases disability was substantial.

6.69 Studies such as these (and there are many more) suggest, as the Court Committee put it, that inadequately treated bouts of childhood illness “cast long shadows forward”.

Health Inequalities in Adult Life

6.70 The rate of mortality in Britain today, once it has been standardized to take account of trends in the age composition of the population, continued to exhibit the steady pattern of decline which has been in evidence for the last 70 years, (though, as we showed in Chapter 3, this is not the case for men in certain social classes or age groups). Respiratory disease, which, among the major causes of death, has the steepest and most linear of class gradients has declined substantially for all age groups over the last two decades (Opcs, 1978: 11). Standards of health have continued to improve in post-war Britain, but differences between occupational classes have remained or even widened. The phenomenon is principally one of relative deprivation – the maintenance of a gap in life chances, against the dynamic background of improved prospects.

6.71 This feature of contemporary trends suggests that there is nothing natural about class inequalities in health. Men and women in social class V do not die before their fellow human beings in social class I as a matter of biological fate. Indeed, they are quite capable of collectively achieving the low rates of mortality achieved by social class I, the only difference being that it takes longer and by the time it is achieved, social class I has moved on to record an even lower rate, thereby ensuring the maintenance of the health gap.

6.72 Explanations of this failure to distribute the health benefits of social and economic progress evenly and contemporaneously among the occupational classes have taken a number of forms.

6.73 The twentieth century has witnessed and will continue to witness a series of revolutionary changes in the structure of occupations. To date these changes have resulted in a contraction in the size of the semi and unskilled manual labour force and an expansion in non-manual occupations (see Table 6.1). These changes have given rise to alterations in the age composition of each occupational class, older workers tending to be found in the contracting areas though not as emphatically as some people suppose – younger and more recent recruits to the work force tending to be found in the expanding area. Besides making comparisons between the occupational classes over time somewhat difficult, these shifts in the occupational structure in themselves offer an analytical solution to the continuing pattern of health inequality, for it is argued that the higher death rates of social class IV and V are at least, in part, a reflection of the older age structure of these occupational groups. The same argument can be used conversely to explain the relatively low death rates of social class I. This explanation which reduced class differences to an artefact of measurement has a certain plausibility. It is true that while social class V is today made up of workers who are older on average, social class I has a larger than average share of younger men. Nevertheless the grounds for asserting that age is the primary causal determinant of the higher mortality of social class V are weakened when class differences are examined for each age group. Figure 6.1 shows the relationship between age and occupation for some of the major causes of death. For all causes except malignant neoplasms class gradients are steepest in early adulthood and most shallow in the decade before retirement. The artefact explanation does not throw light on these observed patterns of fatal disease incidence amongst younger men.


Mortality Ratios by Occupational Class, Age and cause of Death (men 15-64; England and Wales)
Mortality Ratios by Occupational Class, Age and cause of Death (men 15-64; England and Wales)

6.74 These same distinctive trends also highlight the limitations of the thesis that the health gap is caused by age-related processes of social mobility. Occupational drift throughout the span of working life may help contribute to class differences among the over-fifties but it cannot be said to be the cause of class inequalities between the ages of 15 to 45 years. Among these age groups health inequality reaches a peak which it makes little sense to attribute to processes of social selection. In addition lack of continuity between children and adults in the class related incidence of fatal disease gives very limited plausibility to the idea of natural selection as an explanatory form.

6.75 The limitations of these paradigms especially among the male workforce of 45 years and under leads to a consideration of the direct role of material life on the production of health differentials. The most obvious starting point in the search for relevant causal relationships within a materialist perspective is the division between manual and non-manual occupations. Men engaged in manual occupations routinely confront a much higher degree of risk to health and physical well-being in their work task than their non-manual counterparts. These risks are manifold. They may result in direct loss of life either suddenly in the form of accidents or in an attenuated manner through long-term exposure to dust, dirt or poisonous substances in the workplace. The same eventualities may also entail antecedently, physical injury, disability and chronic illness. The direct risk of physical disability in manual work presents itself in a variety of forms. Amongst these, accidental injuries are the most obvious but other forms of physical impairment affecting vision, hearing and breathing are also common accompaniments of productive processes involving manual employment which it must be said, are in no way compensated for by way of financial reward or wages.

6.76 But significant as is occupational hazard in the production of broad differences in health between manual and non-manual workers and between men and women, it is not a sufficient explanation.

6.77 The influence of material deprivation in the aetiology of modern degenerative disease processes is poorly understood in the present day especially amongst workers below the age of 45. In the past poverty was an obvious antecedent in mortality associated with starvation, infection, and respiratory disorders but its possible influence in deaths traced to, for example, cancer or circulatory disease is less clear-cut. These causes of death, because they dominate the disease profile of the wealthy advanced industrial nations have been dubbed the diseases of affluence and, although their aetiology is in general poorly understood, it is thought that they have their origins in overindulgence rather than poverty and in behaviour which abuses and misuses the human body. The modern diet, with its emphasis on highly refined foods, and modern sedentary patterns of work and leisure are prime targets in the search for causes of premature death in the twentieth century and, at a rather more specific level, the mass consumption of tobacco products is blamed for the early onset of some diseases (eg lung cancer) (see Tables 6.3, 6.4 and 6.5).

6.78 This association between disease and death on the one hand, and the consumption of tobacco products on the other, provides an example of the way in which threats to health and well-being continue to be tied up with the productive activities of human beings on which they depend for material subsistence and on which, in the long run, social progress tends to be judged. The mass production and consumption of commodities (including cigarettes) is at the heart of economic life in the advanced societies, and cigarette production is a source of government revenue as well as industrial profits. Governments have been unwilling to bring in legislation to ban the production and the sale of tobacco despite the almost irresistible case that has been built up against it on health grounds in recent years.

6.79 The habit of cigarette smoking therefore provides a convenient focus around which to explore the causal relationships in patterns of health inequalities among adult men and women. This is so because of social life and specific patterns of behaviour which are often attributed to culture.

6.80 Among heavy smokers in Britain men outnumber women and, following trends over the last decade or so, manual workers have increasingly come to outnumber their non- manual counterparts.

6.81 This demographic picture of cigarette consumption reinforces the clinical view that smoking is damaging to health, since the statistical characteristics of heavy smokers are the same as those of the people who are most likely to lose their lives before retirement.

6.82 Can this demographic coincidence help us to understand the underlying cause of the health inequalities that have been observed among adult men and women? Given that cigarette manufacturers are now obliged to warn their customers of the dangers of smoking it would seem on the face of it that smoking is a wilful, self-destructive habit and that those people who fail to take heed of the risks have only themselves to blame if they forfeit their health as a result. Such a conclusion assumes that smoking is an entirely voluntaristic behaviour, the indulgence of the irresponsible. This idea does not square with the fact that the consumption of cigarettes depends on a multi-million pound industry, is sanctioned by Parliament, treated as an important source of taxation income, and freely permitted in public places, even on premises owned by the National Health Service. As health educationalists are fond of pointing out, tens of millions of pounds are spent every year in Britain on the promotion of smoking through advertising and sports sponsorship while only a fraction of this amount goes on publicity about the attendant health risks.

6.83 These facts about tobacco and its role in the economy are an indication that smoking remains part of material and cultural life in Britain. But changes are taking place and not surprisingly the avant garde of culture change are drawn from people in the higher social classes. If cigarette smoking is a major contributory cause of deaths due to cancer or heart disease, then the uneven response in the population to the news that it is dangerous is likely, in the future years, to make class differentials in health even wider than they are at present.

6.84 Moreover, in recent years sex differences in the pattern of smoking have begun to even out, perhaps a reflection of the increased earning capacity of women and their claims for equality with men. But the overriding characteristic of people who have continued to be heavy users of tobacco products despite the adverse publicity is manual work. Why is this so?

6.85 In general it seems likely that people who have the scope within work and leisure relationships and activities to find compensatory means of fulfilling the needs which smoking satisfies, will be more likely to take the warnings about cigarettes seriously and to amend their behaviour in favourable directions. This inference serves as both reminder and reflection of the fact that the structure of work opportunities and the associated levels of financial rewards (what sociologists term the social division of labour) remains fundamentally unequal in present day Britain. This structure of inequality which discriminated, between men and women, between races, regions and social classes, has many dimensions. Income, wealth, job security, pension rights, credit-worthiness, are among the most obvious, but equally significant, especially in their implications for health are education as a continuing lifelong process, and protection from threats to physical well-being. These characteristics of the work that people do and the context in which they do it, are reflected in the quality of the lives they lead. From this perspective smoking behaviour cannot be taken as a fundamental cause of ill-health, it is rather an epiphenomenon, a secondary symptom of deeper underlying features of economic society.

6.86 The drift of this section has been to argue that it is from the realm of the materialist perspective that the most plausible explanation of health inequality between adults in Britain is to be found. It has also been argued that it is not possible to distinguish clearly between cultural and material influences in human behaviour. It is no good treating cigarette smoking as an aberrant or irresponsible behavioural response while society as a whole permits, even depends on, the widescale production and promotion of tobacco goods. Human health is a part of the organisation of material existence. It is both produced and endangered by the work which men and women do in order to earn their livelihood. The manufacture and consumption of tobacco products and its effects on health provides a very clear example of the limitations of conventional health policy as a means to reduce health inequality. The prematurely lost lives of working class men and women will not be saved in the acute hospital or in the GPs surgery.

Health and Inequality in Old Age

6.87 The classification of elderly men and women into occupational classes is made more difficult by the fact of retirement. Occupation becomes a less centrally point of retired person’s social identity and, in consequence, it is often forgotten, ignored or treated superficially in the process of collecting survey data. These tendencies feed into the processes of generating statistical, rates, making it rather more difficult, for example, to compute class based rates of mortality. There is little doubt, however, that the class-based mortality gradients which follow the population from birth, through youth to middle age not suddenly disappear after retirement.

6.88 The new longitudinal survey being produced in the OPCS will, in time, provide a sound picture of class inequalities in rates of mortality among the retired. The early returns for this survey suggest that observed gradients will be more shallow for this age group. This trend is not unexpected. In adult life class inequalities in mortality become reduced in late middle age as the risk of death increases; the expected trends for old age merely complete this already established pattern.

6.89 If we were to argue along the lines of a “Selection of the Fittest” theory we might conclude that the flattening out of gradients amongst the elderly was the result of a filtering process which preserved only the most robust of social classes IV and V for survival into retirement. It is difficult either to prove or discount this inference but the data of morbidity differentials in old age do not support it. Chronic or longstanding disorder is more prevalent among retired manual workers and their wives suggesting, if anything, that the link between relative material deprivation and poor health is sustained throughout the whole lifetime. At the same time it is likely that some men or women engage in semi or unskilled manual work in the years immediately before retirement may well have arrived at their present jobs because poor or failing health caused them to forfeit a more demanding but better paid skilled manual occupation. In such cases, the descent into the lowest occupational classes on account of ill-health may well be a portent of an earlier than average age of death.

6.90 After retirement the appropriateness of mortality rates to health status measurement is increased by the fact that health may literally become a matter of life and death for the over sixty-fives. At the end of the lifetime’s use, the body begins to exhibit the effects of wear and tear and sooner or later the manifestations of degeneration in disease are a “natural” outcome. In general, it is reasonable to conclude that these processes occur sooner rather than later in men rather than women, and in manual workers rather than non-manual workers. An equally reasonable conclusion is that the timing of these outcomes are the end product of inequalities in the use made of, and the demands upon the human body earlier in the lifetime. In practical terms what is implied is that men in our society lead lives which are more punishing to the human frame, and that among men, manual work involves greater use of sheer physical resources and hence more ‘wear and tear’ on the body. Appropriate forms of redistribution between the sexes and the occupational classes might be expected to introduce a greater degree of evenness in these lifetime processes of physical growth, maturation and decay.

6.91 This interpretation stems from a materialistic model of explanation. It suggests that inequalities in health are the direct reflection in a dynamic sense of inequalities in the social division of Labour. In the collective effort of social production, some workers literally give more of themselves than others and hence their bodies wear out first. But inequalities in health at the end of the lifetime also emanate from the distribution of rewards associated with the social division of labour. Old age is a time of poverty, albeit poverty expressed in the form of relative deprivation. But relative deprivation among Britain’s aged can mean material scarcity in very real terms (as deaths from hypothermia among the old reveal in severe winters). A recent DHSS Report estimated malnutrition at 7% among a sample of the elderly who were studied (DHSS 1980 p3). In old age the relationship between income and the capacity to protect personal health is stronger perhaps than at any other time in the life cycle, and in general it is fairly certain that individuals who are well endowed through generous or index linked pension schemes will lead the healthiest, the most comfortable, and the longest lives after retirement. These material fortunes or misfortunes of old age are closely linked with occupational class during the working life. To have secure employment and an above average income when one is at work is to be better able to provide for one’s retirement. It is in this way that continuity in the distribution of material welfare is sustained, and inequalities in health perpetuated, from the cradle to the grave.


This consideration, seriatim, of different stages of the life cycle leads to several conclusions. In the first place, although different kinds of explanation have to be combined, and although cultural explanations and a genetic contribution of importance to early childhood conditions have their place (although knowledge, of the latter is inadequate), nevertheless we feel that more of the relevant evidence is explained by what we call ‘materialist’ or ‘structural’ explanations than by any other form of explanation. However this rubric, it will be recalled, covered many subsidiary approaches which need to be remembered and carefully set out.

6.93 The second important conclusion we wish to draw is this. Some of the evidence on class inequalities in health is adequately understood in terms of specific features of the socioeconomic environment: features (such as accidents at work, overcrowding, smoking) which are strongly class-related in Britain and also have clear aetiological significance. Since such features are recognised objectives of various areas of social policy we feel it sensible to offer them as contributory factors to be dealt with in their own right and not to go on to explain their incidence further in social structural terms. The same is true of other aspects of the evidence which we feel show the importance of measures related to the health services. Antenatal care, for example, is of clear importance in preventing perinatal death, and the international evidence presented in Chapter 5 shows that much can be done through improvement of antenatal care and of its uptake. The international evidence also suggests the importance of preventive health within health policy, despite studies (to which we have alluded earlier) which suggest that little of the differences in mortality either between nations, or between British regions can be explained in terms of health care provision. But beyond this there is undoubtedly much which cannot be understood in terms of the impact of so specific factors. Much we feel, can only be understood in terms of the more diffuse consequences of the class structure: poverty, work conditions (and what we termed the social division of labour), and deprivation in its various forms.

6.94 It is this acknowledgement of the complex nature of the explanation of health inequalities – involving access to and use of the health services; specific issues in other areas of social policy; and more general features of class, material inequality, and deprivation – which informs and structures the recommendations we make in Chapters 8 and9.

6.95 These recommendations draw in another way upon our interpretation of the evidence It is our view that early childhood is the period of life at which intervention could most hopefully break the continuing association between health and class. That is, not only may subsequent health (or propensity to ill-health) be to some degree determined in early life, but there may be some co-determination of subsequent educational (and hence to some degree occupational) achievement and future health status at that time. In our recommendations, therefore, we focus particularly (although not exclusively) upon measures directed towards reduction of health inequalities in childhood.


Abel-Smith, B and P Townsend  The Poor and the Poorest (London: Bell, 1965)

Banks, J and 0 Banks Feminism and Family Planning (London: Routledge, 1964)

Baird, D. “Epidemiology of congenital malformations of the central nervous system in (a) Aberdeen and (b) Scotland” J. Biosooial Sci 6 (1974) 113

Baird, D. “The epidemiology of low birth weight: changes in incidence in Aberdeen, 1948-1972” J. Biosooial Sci 6 (1974) 323

Bernstein, B. Class. Codes and Control (London: Routledge, 1971)

Birch, H 0 and J D Gussow Disadvantaged Children; Health, Nutrition and School Failure (New York; Harcort. Brace and World, 1970)

Brennan, M. E. “Medical characteristics of children supervised by the local authority social services department” Policy and Polities 1(1973)3. 255

Brennan, M.E. and R Lancashire “Association of childhood mortality with housing status and unemployment” J Epidem and Comm. Health 32 (1978) 28

Brenner, M H. “Fetal, infant, and maternal mortality during periods of economic instability” Int. J Health Services 3 (1973) 2 145

Brenner, M.H. “Estimating the social costs of national economic policy: implications for mental and physical health and criminal aggression” (Joint Economic Committee, U.S. Congress. U.S. Government Printing Office, Washington DC, 1976)

Brenner, M. H. “Health costs and benefits of economic policy” Int. J. Health Services 7 (1977) 4

Brown, G and T Harris Social Origins of Depression (London: Tavistock, 1978)

Cartwright, A and K Dunnell Medicine Takers. Prescribers. and Hoarders (London: Routledge, 1972)

Colley, J. R. T. and D. D. Reid “Urban and social class origins of childhood bronchitis in England and Wales” BMJ 2 (1970) 213

Department of Health and Social Security First Report by the Sub-Committee on Nutritional Surveillance (Reports on Health and Social Subjects, No. 6, London; HMSO, 1973

Department of Health and Social Security and Department of Education and Science Prevention and Health (Cmnd 7047, London HMSO 1977)

Doll, R., Hill, A. B. and J Sakula – Brit. J. Prey. Soc. Med. (1960)

Douglas, J. W. B. ‘”Premature” children at primary schools” BMJ 1(1960) 1008

Douglas, J W B and R E Waller “Air pollution and respiratory infection in children”Brit. J Prev. Soc. Med. 20 (1966) 1

Drillien, C. M. in Journal Obstetrics and Gynoecology of the British Commonwealth 66 (1939) 721

Eichenwald, H P and P C Pry “Nutrition and learning” Science 163 0969) 644

Eyer J “Hypertension as a disease of modern society” Int. J. Health Services 3 (1975) 4 539

Eyer J “Prosperity as a cause of death” Int. J. Health Services 7(1977)1 125

Eyer J “Does employment cause the death rate peak in each business cycle?” Int. J. Health Services 70977) 4 625

Fox A.J. and P. P. Collier “Low mortality rates in industrial cohort studies due to selection for work and survival in the industry” Brit. J. Prev. Soc. Med. 30 (1976) 4 225

Fuchs V. Who Shall Live ? Health. Economics^ and Social Choice (N. Y. Basic Books 1974)

Golderg, E.M. and S. L. Morrison  “Schitzophrenia and social class” Brit.J. Psychiatry 109 (1963) 785

Goldthorpe, J. and C. Llewllyn  “Class mobility” Brit. J. Sociology 3 (1977) 269

Hellier, J.”Perinatal mortality 1950 and 1973″ Pop Trends 10 (Winter 1973)
Higgins and Keller         “Familial occurrence of chronic respiratory disease and familial resemblance in ventilatory capacity” J Chronic Dis 28 (1975) .’39

Holman RT Poverty; Explanations of Social Deprivation. ( Lond on , Robe rt s on , 1978)

Illsley R “Social class selection and class differences in relation to stillbirths” BMJ (1955) ii 1520

Illsley R “Early prediction of perinatal risk” Proc. Roy. Soc Med. 59 (1966) 181

Janerich D. T. “Maternal age and spina bifida: longitudinal versus cross sectional analysis” Am J Epidem 96 (1972)

Kosa J. and I. K. Zola Poverty and Health; A Sociological Analysis (Cambridge Mass, Harvard Univ P., 1976

Lawrence, K. M., Carter, C. 0. and P. A. David “Major CNS malformations in South Wales II” Brit. J. Prev. Soc. Med. 22 (1968) 212

Leeder, S. R., Corkhill R., Irwig L. M., Holland, W. W.  and J. R. T. Colley “Influence of family factors on the incidence of lower respiratory tract illness during the first year of life” Brit. J. Prev. Soc. Med.                30 (1976) 203

Lewis 0. The Children of Sanchea (N.Y., Random House, 1967)

McKeown T The Modern Use of Population (London; Arnold, 1976)

Miller P. J.W., Court, S.D.M., Knox, E.G. and S. Brandon The School Year in Newcastle-upon-Tyne. 1952-1962 (Oxford Univ.P., 1974)
Morris, J. N.                 Uses of Epidemiology (London, Churchill, 3rd ed 1975)

Morris, J. N. “Social inequalities undiminished” The Lancet (13 January 1979) 87

Morris J. N. and R. M. Titmuss  ‘Health and social change: I The recent history of rheumatic heart disease” Medical Officer (26 August 2nd, 9 September 1944)

Morris J. N. and R. M. Titmuss “Epidemiology of peptic ulcer” The Lancet 30th December 1944) 841

O.P.C.S. The General Household Survey 1974, 1975, 1976, 1977 ( London, HMSO, annually )

O.P.C.S. Occupational Mortality 1970-72 (London, HMSO, 1978)

O.P.C.S. Social and Biological Factors in Infant Mortality (Occasional Paper No 12, 1978)

O.P.C.S. The Demographic Review (London, HMSO, 1978)

O.P.C.S Trends in Mortality (London, HMSO, 1978)

Paeamanick, B. et al “Pregnancy experience and the development of childhood speech disorders: an epidemiological study of the association with maternal and fetal factors” Am. J. Diseases of the Child 91 (1956) 113

Pasamanick, B. and H Knobloch “Brain damage and reproductive causality” Am. J Orthopsychiatry 30 (i960) 298

Powles, J. “Health and industrialization in Britain” (Proc. First World Congress on Env. Medicine and Biology, Paris, 1974)

Rainwater, L. “The lower class: health, illness and medical inEtitUtiOflB” in II PeVrischer and E. J. Thompson (eds) Among the People; Encounter with the Poor ^N. Y. Basic Books, 1968)

Rutter, M and N. Madge Cycles of Disadvantage (London; Heinemann, 1976)

Speiser, F.E., Rosner Band Tager I “ Familial aggregation of chronic respiratory disease: use of national health Interview Survey data for specific hypothesis testing” Int. J Epidem 5(1976) 2 167

Stark,  E. “The epidemic as a social event Int J Health Services 7 (1977)4 681

Stedman-Jones,   G. Outcast London (Oxford Clarendon Press,   1971)

Thompson,  P. The Edwardians    (London,  Paladin,   1976)

Townsend P “Inequality and the health service” The Lancet (1974)

Townsend P Poverty and Britain (London Penguin Books 1979)

Wilson H and Herbert GW Parents and Children in the Inner City (London Routledge 1978)





[The Socialist Health Association (SHA) regularly posts articles on its independent blog about the social determinants of health. Except where explicitly stated, these articles should not be taken as official statements of policy on such matters. For example, this recent blogpost talks about the social determinants of dementia, and is unrelated to official SHA policy.]