Goldthorpe, Halsey and Social Class

Edmund Leach on the survival of privilege

  • Social Mobility and Class Structure in Modern Britain by John Goldthorpe
    Oxford, 310 pp, £12.00, January 1980, ISBN 0 19 827239 1
  • Origins and Destinations: Family, Class and Education in Modern Britain by A.H. Halsey
    Oxford, 240 pp, £14.00, January 1980, ISBN 0 19 827224 3

I refer to the first of these items as ‘Goldthorpe’ and to the second as ‘Halsey’. Both are productions of the Oxford (Social) Mobility Project, a large collaborative exercise which has operated from a base in Nuffield College since 1969. For a long while, politicians and other interested parties are likely to cite them as authoritative sources, but in order to evaluate what is being said, the reader must penetrate a thick layer of mind-boggling numerical tabulations and pseudo-vector diagrams to the egalitarian value schema which lies beneath.

Most of the data derive from a nationwide questionnaire survey of 10,309 individuals representing a sample of all adult males on the Electoral Register in 1972 aged between 20 and 64. Goldthorpe also takes account of data from a much smaller follow-up survey conducted in 1974. The sample and questionnaires were designed by the research team but, in the 1972 case, the fieldwork was contracted out to a commercial agency (Mary Agar Field Services Ltd). Goldthorpe’s Appendix tells us a good deal about the sampling procedure. The general nature of the questions which produced the Goldthorpe data is evident from the text. Part of the set of questions which related to educational issues is given in Halsey, Appendix I.

Both books made extensive use of the H-G Scale,[1] though they do not use it in quite the same way. This scale graded the thousands of occupations recorded by the Office of Population Censuses and Surveys by first picking out 860 ‘representative’ occupations and then collapsing these into 36 categories. In both Goldthorpe and Halsey a set of loosely ranked ‘social classes’ is created out of a series of congeries of these latter occupational categories. Goldthorpe has seven such classes, Halsey eight. Halsey’s eight classes are then further aggregated into three major social classes with a managerial/administrative ‘Service Class’ at the top (Classes I and II), an ‘Intermediate Class’ (Classes III, IV and V), which is not at all like the ‘middle class’ of colloquial usage, and a ‘Working Class’ (Classes VI, VII and VIII). Goldthorpe has no Class VIII, but otherwise uses the same terminology. The reader needs to keep his wits about him. ‘Working class’ is a very loaded term, with strong Marxist connotations. The Goldthorpe/Halsey ‘working class’ and the Marxist ‘working class’ are far from identical, but it is not easy for the reader to keep the significant distinctions fully in mind.

Both books emphasise that the sorting of occupations into classes at two levels is not intended to imply that either the classes or the occupations can be ranked in a one-dimensional hierarchical continuum of prestige and influence from top to bottom. Social mobility need not be either ‘upwards’ or ‘downwards’: it could be sideways. Nevertheless, in practice, the authors concentrate most of their attention on constraints inhibiting upward movement out of the ‘Working Class’ into the ‘Intermediate’ or ‘Service Class’. Moreover, some features of the research design were influenced by a desire to replicate ‘Blau and Duncan’,[2] in which a one-dimensional continuum of prestige attaching to different occupations was assumed.

Goldthorpe/Halsey assume that social-class identity – ‘class-consciousness’ in the Marxist sense – is predetermined by occupation. The research into social-class mobility was thus concentrated on changes of occupation of the respondent informant during his own lifetime and a comparison between the respondent’s present occupation and that of his father at the time when the respondent was 14. These latter questions actually referred to the occupational status of the ‘head of the respondent’s family’ and no less than 8 per cent of the sample took this to refer to the respondent’s mother. Even so, says Halsey, ‘the data have been organised so that ... these variables are all labelled as characteristics of their non-existent father’!

If subjective phenomena such as social-class identity are to be made objective by conversion into numerical statistics, condensations of this general sort are unavoidable, but at the end of the day sceptics are bound to wonder what exactly it is that all this formidable apparatus has been set up to measure.

The male bias of the 1972 sample is discussed in the Appendix to Goldthorpe and in the text of Halsey. It is justified on several grounds: 1. the sample was concerned with individuals born between 1913 and 1952, and at those dates male occupations were, in fact, of much greater significance for the class status of households than female occupations; 2. ‘Blau and Duncan’ was an all-male study, so comparability required that the 1972 survey should also be all-male; 3. the team reckoned that in order to make it representative of the nation as a whole, they needed a sample of at least 10,000 units, but they had not the resources to go beyond this. Women could only have been brought into consideration at the cost of jeopardising the statistical virtues of the sample!

The authors’ eagerness to produce statements which are true for the nation as a whole may be attractive to politicians but could result in sociological nonsense. Statistics applied to large samples of very large populations effectively eliminate from consideration all exceptional cases. But if generalisations approximating to laws of nature are to be built up from empirical observations, the variations implicit in ‘exceptional cases’ may be of the utmost significance. After all, why should the changing class structure of Middlesbrough resemble in any way the changing class structure of Bristol or rural Dorset?

But these authors are not naive. They use their statistics with great sophistication. They aim to test out, against empirical evidence (sometimes confirming, sometimes refuting), a variety of politically-motivated theoretical assertions about what has been happening to the class structure and to educational opportunities in Britain over the past half-century. It is all quite impartial, with no concessions to readability. It is not that the authors have forgotten that their subject-matter is the behaviour of human beings: it is just that they genuinely believe that in the interests of science all human quirks must be reduced to a series of ciphers.

The best way for the reader to tackle Goldthorpe is to start with the first and last chapters, which really amount to a continuous single essay. They are by Goldthorpe solo and free of tabulations. The opening chapter sets out to explain the authors’ political orientation and to relate the Oxford study to other relatively recent surveys of the same general kind both in Britain and the United States, in particular to work by D.V. Glass at the London School of Economics and to American work by Lipset and by Blau and Duncan. These earlier writers, in their different ways, were concerned with individual status achievement within a system that was potentially completely fluid rather than with membership of social classes in the Marxist sense of social formation. By contrast, the Oxford authors intend that their use of the concept of class shall tie in with Marxist notions of class-consciousness and class struggle.

The underlying utopian ideal of an ‘open’ society in which there will be true equality of opportunity brought about by the political acumen of a ‘mature’ working class would certainly have appealed to Marx, but neither Goldthorpe himself nor any of his collaborators can properly be labelled Marxist. On some issues they could even be rated as anti-Marxist. ‘Classical’ Marxism was contemptuous of any interest in social mobility, which was considered a symptom of liberal nostalgia, and the Goldthorpe authors are clearly out to refute such recent forms of Marxist dogma as that which declares that the individual membership of particular classes is irrelevant to the ‘class structure [conceived] as a structure of positions constituted by the prevailing relations of production’, or that which holds that mobility across the dividing line between the bourgeoisie and the proletariat is too small to be significant.

Goldthorpe makes the point that Marx himself was by no means wholeheartedly committed to the ‘classical’ Marxist Verelendungstheorie, according to which capitalism must always produce a binary class struggle between a very small group of exploiters and a large mass of exploited, in the course of which members of the ‘intermediate classes’ of peasants, small entrepreneurs, artisans and the like will be increasingly forced downwards into the ranks of the protetariat. Indeed the definition, in the present study, of the Upper (‘Service’) Class as being as much professional and managerial as proprietorial and entrepreneurial was foreshadowed in some of Marx’s own writings.

Even so, the earlier part of this century saw the development of a clear distinction between the ‘negative’ socialist attitude which treated social mobility as a phenomenon blocking the achievement of socialism via the labour movement, and the ‘positive’ views of liberal writers such as Sorokin who eventually came to approve unequivocally of mobility as a value to be pursued. The socialists and the liberals shared a vision of an ultimately classless society: but classlessness was to be achieved in quite different ways. The socialists saw the rigid ossification of the existing class structure as the trigger which must lead to the revolutionary triumph of the working class whose classless virtues would then permeate the whole. The liberals, with an idealised version of American capitalism in mind, imagined a hierarchical system of such fluidity that social class and class-consciousness would never have time to develop at all, though some of them took the view that equality of opportunity in this form would impose intolerable psychological strains on the individual isolated from the social support of workmates, kin and neighbours.

In terms of this dichotomy Goldthorpe and his colleagues have a socialist view of Utopia, but they are empiricists. They recognise that at no time during the present century has the existing class structure of Britain been rigidly ossified to the degree that doctrinaire Marxist theory supposes. On the other hand, they insist that the fluidity of the class structure has not been such as to inhibit the development of class-consciousness and, furthermore (which will surprise many readers), that there has been no significant increase in this fluidity over the past fifty years. One way and another, despite much socialist legislation, the holders of privilege have been able to preserve their position.

As a result of technological and bureaucratic changes in the system as a whole, there has been a great expansion of the ‘Service Class’, both in absolute numbers and in proportional terms, and a corresponding proportional shrinkage in the size of the ‘Working Class’, as here defined. So it is not at all surprising that many of the present ‘Service Class’ have non-‘Service Class’ origins. But, if these shifts in scale are allowed for, the statistical probability of an individual born into a ‘Working-Class’ household later entering a Class I ‘Service-Class’ occupation is apparently no greater than it was at the beginning of the century. Goldthorpe and his collaborators find this shocking.

Goldthorpe’s own solution to the implicit political issue is spelled out in his final chapter, where he still puts his faith in a radicalised TUC and continuing legislation of the sort which characterised the ‘Social Contract’ era of 1974-76. This middle-of-the-road, left-of-centre, Labour Party line is much the same as that adopted in 1969 at the end of the final volume of the Affluent Worker series[3] where Goldthorpe was collaborating with a different team of research workers. There the ‘liberal’ view that affluence leads to an embourgeoisement of the skilled, wage-earning factory worker was totally rejected, but so was the alternative, neo-Marxist theory that affluence could only lead to an intensification of the frustration which the alienated worker must feel as result of his class-awareness of social deprivation.

All this assumes that the reader is persuaded that statistics based on mobility within male occupations alone provide an index of changes and rigidities in the class structure in the way that the authors suppose they do. Goldthorpe and his colleagues know the literature backwards, yet the ambiguities of Class Theory, both Marxist and other, are greater than they are here prepared to admit.

Raymond Williams has recently complained that in the whole range of contemporary discussion of this concept and the controversy that surrounds it three basic ‘meanings of class can be seen in operation, usually without clear distinction ... (i) group (objective): social or economic category, at varying levels; (ii) rank: relative social position, by birth or mobility; (iii) formation: perceived economic relationship; social, political and cultural organisation’.[4] Goldthorpe and Halsey are both concerned with (i) and (ii); Goldthorpe does not, it seems to me, make good his claim that he is also conducting a ‘debate with Marx’ relating to (iii).

In general, the Oxford Survey team has tried to evade the difficulties noted by Williams by playing down the distinction between self-identification (class-consciousness) – ‘I am a member of the working class’ – and reference: ‘He is a member of the working class.’ The tabulations throughout are concerned with reference categories: the hypothetical social classes, based on occupation alone, are constructs of the authors rather than of their informants.

Marxist class is oppositional – it presupposes feelings about ‘we’ and ‘they’; but the classes in the Goldthorpe/Halsey schema do not appear to have this character: they are fuzzy at the edges very much in the fashion of ‘the “new” working class’ described by Goldthorpe in 1969. But whereas the affluent workers of 1969 were said to maintain their Working-Class identity for many purposes, especially in relation to shop-floor politics, the ‘ “blue-collar” élite’ of the present studies are in an Intermediate Class category and thus not part of the Working Class at all. There is a slipperiness about the argument here that needs very careful watching. Insofar as the Goldthorpe/Halsey books do in places seem to enter into a dialogue with the Marxists, it is important for the reader to take note of exactly how ‘class’ concepts are being used in the passages concerned.

Goldthorpe’s second chapter is used to demonstrate that the extent of social mobility has usually been underestimated and that, on the face of it, three particular theses on this topic are refuted by the evidence. The figures show that: 1. Social Class I, at the top of the hierarchy, is not ‘closed’; 2. it is not the case that there is a ‘buffer zone’ between major social-class divisions which inhibits complete transfer from one class to another; 3. while mobility chances are becoming increasingly influenced by educational attainments, this is not correlated with a decline in the importance of experience and ‘training on the job’. The next three chapters then use sophisticated mathematical juggling to qualify these initial conclusions. Thus Chapter 4 shows that Class I, as represented by higher professional, administrative and managerial occupations, is indeed ‘open’, in the sense that it contains large numbers of individuals who ‘originated’ in other classes, but also that there is a section of Class I which consists of proprietors and entrepreneurs, which is relatively speaking ‘closed’. Similarly, Chapter 5 manages, up to a point, to restore credibility to the ‘buffer zone’ theory seemingly refuted in Chapter 2.

Chapters 6 and 7 endeavour to come to grips with the issue of how far a change of social class, as measured by the Goldthorpe/H-G criteria, corresponds to a change in ‘actual social relationships in which individuals are engaged and in their own definition of the social situation’ (i.e. ‘class-consciousness’ in a Marxist sense). The authors investigated these subjective issues by asking their respondents questions about how they would interact with ‘kin’, ‘neighbours’, ‘friends’ and ‘spare-time associates’ in various hypothetical situations. Social anthropologists, who are likely to have spent months of participant observation failing to disentangle the actualities of such behaviour, even in a community of quite minimal size, are not likely to be impressed, though they will note with (perhaps sceptical) interest the general conclusions that the data provide no evidence that ‘mobility is an inherently dissociative phenomenon which characteristically leads to a disruption of primary social relations of the mobile individual and in turn to his social isolation.’

Chapter 8, ‘The Experience of Social Mobility’, is non-statistical. It consists mainly of a large number of thumbnail autobiographies provided by a selection of respondents to the questionnaire. The Goldthorpe authors claim that these stories (which are, up to a point, personal expressions of ‘class-consciousness’) justify the emphasis which the survey placed on the market and work situation of individual respondents and their fathers as a measure of changes in ‘class’ as opposed to ‘status’ position. Here and elsewhere, the writers’ political commitment, their ideal of an open society in which, at birth, everyone starts from scratch and in which all factors which lead to any kind of nepotistic or hereditary conservation of positions of influence are excluded, has coloured the presentation. Such a system, if worked through to its logical conclusion, would entail free-for-all competition for all kinds of work positions in each generation. And this would seem to imply a pure meritocracy in which formal education would be entirely concerned with the allocation of certificates of competence that would be the sole means of obtaining any kind of job anywhere.

School as an instrument for furthering the development of a meritocracy provides the central theme of the Halsey study. At the outset the authors ask the deceptively simple question: can education (by which they mean schooling) change society? They then proceed to show that both the question and the possible answers are extremely complicated. They emphasise that both on the Right and on the Left there are influential writers on education, American, English, French, who regard educational meritocracy as specious. Boudon’s view that ‘society rather than school is responsible for inequality of educational opportunity’ is cited as being representative of a variety of writers. Yet the Halsey authors themselves, who emphasise that meritocracy has been the professed goal of the British educational system for many years, appear at times to defend that goal against its critics. As in Goldthorpe, the researchers present themselves as detached observers whose concern is simply to provide objective evidence for what has actually been happening. They are certainly not wholehearted admirers of what they have observed yet they seem to take it for granted that, in a good society, everyone would aspire to progress right to the top of the educational ladder and achieve a university degree. For those who have had such aspirations over the past half-century it has been a handicap race with all kinds of unadvertised obstacles cropping up on the way. The chances of success for the unusually bright boy of working-class origins have improved over the years, both in relative and in absolute terms, but even so ‘the boy from a privileged school or from a privileged school background had a much higher chance than his unprivileged contemporary.’ It is not surprising that the authors should find this regrettable but they seem unclear in their own minds as to why it should be regrettable.

At the very end of the book they distinguish two ways of looking at the problem. On the one hand, school education can be regarded as a consumer good to be valued as an end in itself. In that case, the more schooling the better. The state should expand the educational system to the limits that the economy can sustain and there should be equal access to the resulting places of secondary and higher education, regardless of class or parental influence. Here the authors seem to evade an issue which has received much attention earlier on: namely, that since the culture of the higher echelons of any schooling system will always be much more like that of the home background of the privileged classes than it is like that of the underprivileged working class, the children of élite groups will continue to be advantaged anyway; and they ignore altogether the inverse thesis that cultural diversity is to be valued for its own sake and since schooling seeks to impose cultural uniformity, and a culture which would inculcate disciplined subservience to the values of the ruling class, the less there is of it the better!

The Halsey alternative is that schooling may be seen as a stepping-stone to a job, a cog in the meritocratic machine. In that case, it is wasteful to offer an élite form of schooling to the great majority who will not end up in élite occupations. This was the official doctrine during the IQ test/‘11 +’ phase of selective education. The theory was that there was only a very limited pool of individuals who could hope to benefit from superior forms of secondary education, and this ‘natural ability’ could be detected by IQ tests, and that it was economically efficient to give an élitist education only to the select band who would be likely to end up in the higher echelons of the Service Class. The Halsey writers denounce the ethos of the 11 + examination much less vigorously than might have been expected, but one has to remember that their survey was concerned with men who had all completed their formal education before the emergence of comprehensive schools, sixth-form colleges and other possibly transient abberations.

Grand-scale longitudinal research of this general kind lies at the heart of one style of sociological thinking. By eliminating the human beings and replacing them by blocks of abstrusely calculated figures, social research is made to look like economic research – and becomes a social science (though the scientific status of economics is equally open to question). But sociologists, unlike economists, have to collect most of their own statistics and that is a serious handicap. The major cost of the enterprise is the cost of data collection, so you have to fix your questions in advance and if people start coming up with unexpected answers you are stuck: you can’t go back and start all over again. But it is now more than ten years since the Oxford team began to think about what questions they ought to be asking and, on the face of it, a great deal has been happening both to the class structure of Britain and to the class-related structure of British education during that period.

I do not myself have any faith in questionnaires anyway, partly because, in coding the answers, most of the interesting variables get eliminated and partly because the presumption that, in the mass, the lies told by informants A, B and C will be balanced out by the lies told by informants D, E and F is in no way borne out by the experience of the more intimate forms of investigation conducted by social anthropologists. But my major doubts about the Halsey exercise stem from the fact that it seems to have been overtaken by events. Problems of changing class structure and educational opportunity are still with us, but so also are problems of ethnicity and problems concerning the status of women. The critical issues of 1980 do not seem to be the same as those which the Oxford researchers considered crucial in 1969.

Whatever objectivity has been achieved by resort to statistical technicalities has been at the cost of relevance. There are dangers, too, in the very thoroughness with which the job has been done: politicians, like reviewers, will find the totality indigestible and simply pick out for praise or blame the snippets which fit in with their preconceived notions. I suggested at the beginning that these books are destined to be much quoted: but I do not think they are going to have much influence on those who do the quoting.

[1] The Social Grading of Occupations by John Goldthorpe and Keith Hope. Clarendon, 1974.

[2] The American Occupational Structure by P. Blau and O.D. Duncan. Wiley, 1969.

[3] The Affluent Worker in the Class Structure by David Lockwood, Frank Bechhofer and Jennifer Platt. Cambridge, 1969.

[4] Keywords by Raymond Williams. Fontana/Croom Helm, 1976.