Citation reference: Harvey, L.,  2011, Critical Social Research, available at qualityresearchinternational.com/csr, last updated
31 January, 2019, originally published in London by Unwin Hyman, all rights revert to author.
A novel of twists and surpises
The Affluent Worker in the Class Structure (Goldthorpe, Lockwood, Bechhofer, and Platt, 1969) provides an example of an embryonic critical social research project that used and adapted standard quantitative research methods. Goldthorpe and Lockwood, the principal researchers, initiated a research project into the sociology of the affluent worker in 1962 (Goldthorpe et al., 1969). Fieldwork started later the same year. Financial support came from the Department of Applied Economics of the University of Cambridge and later the Human Sciences Committee of the Department of Scientific and Industrial Research (a precursor of SSRC subsequently ESRC). Bechhofer and Platt were members of the research staff of the department throughout the fieldwork and analysis stage.
Several reports of the study of the affluent worker were published in journals (Goldthorpe and Lockwood, 1962, 1963; Lockwood, 1966; Goldthorpe, 1966) as well as two longer reports The Affluent Worker: Industrial Attitudes and Behaviour (Goldthorpe et al. 1968a) and The Affluent Worker: Political Attitudes and Behaviour (Goldthorpe et al., 1968b). The Affluent Worker in the Class Structure was the final and central book-length report and effectively marked the conclusion of the research.
Goldthorpe and Lockwood saw the Affluent Worker study as a contribution to the century-long debate about the working class within Western industrial society. They review the debate in order to provide a 'wider perspective' within which to locate their study. They are 'well aware that the conclusions from any single study, restricted in time and space, are likely to be of only limited application' and in no way suppose their work to resolve the problems they address. Their aim 'might more accurately be defined as that of providing controversy with more and better material on which to feed' (Goldthorpe and Lockwood, 1969, p. 1).
Capitalism, since 1945 had been marked by the growth of 'affluence' and consumerism; the replacement of intolerable factories; the increasing concern by management with 'human relations'; the general increase in automation with its changing work conditions; and the development of suburbanism. The middle-income group had swollen as a result of significant standard-of-living advances gleaned by large numbers of manual workers particularly in advanced technological plants.
In such circumstances, the working class was seen by liberals as in the process of decline and decomposition; as an anachronism belonging to the infancy of capitalism. This view was not new and repeated the long-running concern about the revolutionary potential of the working class.
Goldthorpe and Lockwood (1969) examined the various Marxist and non-Marxist contributions to the idea of the changing nature of the working class. They critically examined the widely taken-for-granted embourgeoisement thesis and revealed a number of different dimensions to the proposition that covered the work situation, social lives and cultural values, and aspirations. Briefly, the embourgeoisement argument was that the working class was no longer economically inferior, mass deprivation (Bernard, 1956) and alienated labour (Blauner, 1964; Woodward, 1958, 1964, 1965) no longer characterised capitalism. The traditional distinction between manual and non-manual labour and antagonistic worker-management relations were disappearing. Homogenisation of living standards had also meant the more affluent working class adopting bourgeois norms, value and attitudes in respect of a wide variety of cultural activities from fashion and eating to entertainment and parenting. Migration and the development of new housing estates further undermined traditional working-class communities and workers became more exposed to mass media and the values of other social strata. In short, the embourgeoisement thesis argued that status distinctions based on consumer power had replaced class distinctions based on productive role. Thus the Marxist notion of embourgeoisement reappeared not as temporary irregularity, as Engels (1895) suggested, but as an integral part of the evolution of capitalism. Instead of salaried workers joining the proletariat, production workers were believed to be joining the middle class.
There is a persuasiveness to the embourgeoisement thesis that gained considerable popularity after Labour's third successive electoral defeat in 1959. The thesis was accepted in Conservative circles as largely accounting for the party's unprecedented run of electoral success. Indeed the two-party system was thought to be in doubt. Labour supporters also accepted the thesis and it led to a radical reshaping of Labour policies and of the party's 'cloth cap' image.
2.5.3 The approach—a critical case study
Goldthorpe and Lockwood reviewed the contributions to the debate in detail in order to clarify what exactly was being claimed about the nature of embourgeoised worker in respect of these different dimensions. They then designed a study to test these different elements.
Given limited resources and the requirement of an in-depth analysis, Goldthorpe and Lockwood decided to adopt a critical case study approach. They determined criteria, based on the characteristics of an affluent working class community that they derived from the proponents of an embourgeoisement thesis, of a case study that would be as favourable as possible for the confirmation of the thesis. They argued that, if in this case the thesis was confirmed then they would have detailed material on workers who were in the process of changing their class situation. If the thesis were not confirmed in these favourable circumstances then they argued that they would be in a position to 'claim that a fortiori it was unlikely to be occurring to any significant extent within British society at large' (Goldthorpe et al., 1969, p. 32).
The criteria for the critical case study were that the population of workers should be physically mobile, affluent, economically secure, and consumption conscious. They should be in an industrial setting with 'progressive' employment policies, advanced technology and harmonious industrial relations. The community within which this population lived should be socially heterogeneous, economically expanding and open, new, and lacking tightly-knit kinship networks. While this ideal-type was not entirely attainable they showed that in Luton, workers at Vauxhall Motors Ltd., Skefko Ball Bearing Co. Ltd., and Laporte Chemicals Ltd., matched the criteria as nearly as possible. As a bonus, Luton had been identified as the prototype of the 'new middle-class' Britain, and, although they had not chosen Luton for that reason, it was fortunate that they were able 'to meet supporters of the embourgeoisement thesis on their own ground'. Thus they 'cannot be accused of seeking for workers turning middle-class where no-one had ever claimed or supposed that such a pattern of change was likely' (Goldthorpe et al., 1969, p. 47).
Testing the thesis required that the elements that constituted the various dimensions of embourgeoisement thesis be operationalised so that they could be checked against the descriptions of actual work practices, social activities and aspirations. For example, within the world of work, the newly embourgeoised worker was, according to the thesis, supposed to relate positively to the work situation reflecting white-collar careerism; teamwork; commitment to the job; derivation of a certain intrinsic satisfaction from the work; relative autonomy of action; a degree of social spin-off; and so on.
A substantial sample of workers were then analysed to see if their lives both at work and outside the work situation, and their attitudes and opinions reflected embourgeoisement. A small 'control' sample of white-collar workers, backed up by available empirical evidence from other studies, was used for comparative purposes. The population of workers was drawn up so that, at each of the three plants, workers who were central to the main production system were sampled (assembly line workers at Vauxhall; machine operators, setters and maintenance workers at Skefko; process and maintenance workers at Laporte). The population was limited to male workers between 21 and 46, married and living with their wives and regularly earning at least £17 per week gross (October 1962) and resident in Luton or the immediately adjacent housing areas. The sample was not a simple random sample of the population but, for practical reasons, one that was drawn from the major departments in the plants. All workers in the selected departments were interviewed except in the case of the two largest assembly divisions at Vauxhall: this covered 60 to 70% of the population. This does raise questions about representativeness especially among the assemblers; however, Goldthorpe and Lockwood checked the sample in detail and claimed that 'no grounds could be found for supposing that those men in our population in the excluded departments differed in their basic social characteristics from those in the departments studied (Goldthorpe et al., 1969, p. 48, footnote 2).
In all, they had an initial sample of 326 that yielded a final sample of 229 (a 70% response rate) made up of 86 assemblers at Vauxhall; 41 machinists, 23 setters and 45 craftsmen at Skefko; and 23 process workers and 11 craftsmen at Laporte.
The comparative sample of white-collar workers in the same age-range was drawn from general clerks and commercial assistants at Laporte and cost, correspondence and general clerks at Skefko. The initial sample of 75 men ended as a final sample of 54, a response rate of 72%. Material from this 'restricted' sample is only used for comparisons where it is supported by data from more 'extensive studies of members of white-collar strata'.
The main research instrument for obtaining material on social attitudes, behaviour and relationships was the scheduled interview. Respondents were interviewed first at work for about an hour and then again at home with their wives for about three hours (sometimes split into two sessions). The firm compensated for 'lost time' in the first interview and the project team paid £1 to the couples for the second (although some couples refused the payment). The interviews of the comparison sample of white collar workers also took in respondent's wives and were based on a schedule 'consisting of parts of both the "home" and the "work" schedules used with our manual respondents. Thus, over quite a wide range of items directly matching data for the two samples were obtained' (Goldthorpe et al., 1969, p. 52–53).
Interviewing was supplemented by observational studies of the respondent's working lives but no such observation of 'out-plant' life was possible as the respondents led very private lives. There was no local community of a 'public kind' which was accessible to the researcher as a fieldwork location. The social lives of the sample of affluent workers and their families 'were built around such essentially private occasions as the family walk or car-ride, the visit to relatives, or the couple's 'evening out' at a cinema or restaurant' (Goldthorpe et al., 1969, p. 50). All information relating to aspects other than work were based on the respondent's own account rather than 'direct study'. This raised the possibility of bias and distortion as respondents may present themselves in a particular way. Clearly, Goldthorpe and Lockwood noted, 'as the social anthropologists have traditionally insisted, it is wise to distinguish between what people say they do and what they do in fact'. However, this is not, an intractable problem and they found that their interviewees were both disinclined to attempt a 'front' and prepared to present themselves in what, for them, was often an unfavourable light. Where cross-checking of responses was possible 'no serious degree of inconsistency was found in the answers individuals gave'.
2.5.4 Testing the thesis
Goldthorpe and Lockwood maintained that there research design provided the basis for a descriptive account of the social lives of the affluent workers and thus the basis for an 'appropriate and cogent test' of the embourgeoisement thesis. The primary data that described those aspects of workers' lives relevant to the operationalisation of the embourgeoisement thesis were compared to expectations. This was cross-checked by using comparative material of white-collar workers.
For example, they used the reason why their respondents remained in their jobs as one indicator of the kind of satisfaction they derived from the work.
The reason that our respondents by far most frequently gave for remaining in their present jobs—and most appeared to be quite firmly attached to them—was in fact the high level of pay they could earn.... This reason was given by half the process workers, by two-thirds of the more skilled men and by three-quarters of the assemblers and machinists; and with 1 in 4 of the latter, this was the only consideration mentioned. The reason next most frequently advanced was security of employment (referred to by 38% of our respondents overall) and taking all economic factors together—level of pay, security, extent of social welfare and other fringe benefits—one or more of these was referred to by 87% of the craftsmen and setters in the sample and by 82% of the semi-skilled men. In contrast, in no occupational group did as many as a third of our affluent workers make any mention of staying in their jobs because they liked the work they did; and among the assemblers and machinists the proportion was as low as 1 in 8. (Goldthorpe et al., 1969, p. 56)
A table of results is supplied for greater detail. The analysis is further developed by way of a comparison with white-collar workers. This showed that among the blue collar workers 'affluence has been achieved only at a price: that of accepting work that affords little in the way of intrinsic rewards and that is likely to be experienced essentially as labour' . This is in contrast to white-collar work where a 'clear majority (70%) did not refer to pay as a factor attaching them to their present employment'. The comparison was further developed by assessing the reasons why the skilled workers showed dissatisfaction with their work. In became clear that, for them, affluence was bought at the price of less freedom in their work environment.
While it was evident that work experience was dissatisfying, it also became clear that non-economic, social, satisfactions did not compensate for inherently unsatisfying work by, for example, building up rewarding relationships with workmates, superiors or other work associates. Their observational studies indicated that 'tightly knit work groups were something of a rarity in the shops and departments in which we were concerned'. This conclusion was reinforced by 'the finding from our interviews that 76% of the more skilled men in our sample and 66% of the semi-skilled felt that they would be 'not much bothered' or 'not bothered at all' if moved to another job away from the men they presently worked with (Goldthorpe et al., 1969, p. 65).
Goldthorpe and Lockwood argued that among workers in their sample, 'affluence had been gained by these men sacrificing, directly or indirectly, the possibility of a higher level of intrinsic job satisfaction'. This sacrifice involved greater stress and deprivation in the work situation, overtime and shiftwork. The work experience was a fundamentally dissatisfying one for the workers.
In this way, each aspect of the embourgeoisement thesis was addressed and primary data used to test the expected attitudes and life experiences of the supposedly embourgeoised workers. The test is thus dependent upon an operationalised concept of embourgeoisement against which the actual lives and attitudes of workers can be compared. The operationalisation process reflects the standard quantitative social research process of identifying various dimensions of the concept, based on a literature review, and the determination of indicators of the various dimensions. The three dimensions identified were work, sociality, and social aspirations and imagery. The key difference from a standard quantitative approach was rather than construct 'neutral' operationalisations to be used to measure a random or otherwise 'representative' sample, the empirical context was selected to favour the thesis and the measures designed to address supposed material manifestations of the thesis.
The design of the research was thus directed to materially grounding an abstract theoretical debate. It was not a study of the process of embourgeoisement as it basically provides a static picture. It, thus, does not permit analysis of the mobility of the workers in the study. However, this is not the aim. What the study does permit is an assessment of whether, amongst affluent workers, middle-class values, and so on, have already been adopted. Thus, given their critical case, if the embourgeoisement thesis is a sound one, there should be a sizeable proportion of the sample that is indistinguishable from, at least, 'lower middle-class non-manual workers'.
Similarly, the research study, unlike most standard quantitative social research, does not attempt any causal or pseudo-causal explanations, or broad generalisations. It does not look for the factors that might account for an observed phenomenon. The principle aim of the primary data analysis is to critically examine, through a case study most likely to confirm the embourgeoisement thesis, the adequacy of this widely taken-for-granted view.
Goldthorpe and Lockwood conclude, by way of 'generalising from their findings', that in respect of the world of work the class situation of the affluent worker has not changed and the thesis is inadequate as it breaks down 'fairly decisively at any one of several points'. Workers lives and attitudes do not indicate a shift to middle-class values, attitudes and activities. The embourgeoisement of the worker is not as an inexorable thesis as it has been portrayed and has little relevance to contemporary British society. Although changes may have occurred in consumption, within the sphere of production there is still 'a fairly distinctive working class' (Goldthorpe et al., 1969, p. 83) even in progressive, modern industrial establishments. If anything, changes in white-collar work are tending to bring some non-manual workers closer to manual workers rather than the other way round. In contrast to the version of the embourgeoisement thesis that asserts affluent workers have a dual social identity (working class at work, middle class out of work), there is a tendency for workers to dissent from middle class conceptions of the status hierarchy. The aspirations of the affluent workers seemed to reflect neither working-class consciousness nor middle-class status consciousness. Differences in prestige and power were less significant for respondents than differences in income and standards of consumption.
2.5.5 Implications: the working class and party politics
The primary intention of the empirical analysis, the examination and testing of the thesis, has thus been completed. The authors do not finish here, however, but relocate their work in its wider framework. They argue that their resounding rejection of the thesis rebounds on the whole evolutionary theorising about the nature of class in advanced capitalist society. Clearly, their analysis shows that although economic development has had a considerable impact on working class lives, class and status relationships are relatively autonomous of changes in the economic, technological and ecological infrastructure.
Rather than an embourgeoisement thesis the authors point to a 'normative convergence' between certain manual and non-manual groups amongst their respondents, which blurs the distinction between middle and working class. In the case of white-collar workers there is a shift away from traditional individualism and towards collective, if instrumental, means of pursuing economic objectives. Manual workers are shifting from community oriented social life to recognition of the centrality of the conjugal family.
What then are the implications of their research for the debate on the working class as an historic agent of radical social change? Two alternative arguments can be constructed from the apparent decline of workplace and communal solidarity of workers who see labour as a means of sustaining a mode of social life dominated by home and family. The first, the 'liberal view', suggests that workers are becoming more individualistic, with a consequent decline in support for unionism and the Labour Party, and, as such, their potential as a radical force has declined. The second, the 'neo-Marxist view', takes up the alienation of the worker. Commodity consciousness and the lack of intrinsic satisfaction in work are indicative of the denial of the real needs of the affluent worker and thus such a worker is fundamentally in opposition to the system. The empirical material, when analysed in detail, leads Goldthorpe and Lockwood to reject both views.
Essentially, both approaches rely on an evolutionist position that takes basic premises for granted: on the one hand that 'middle class incomes' lead to a decline in support for labour politics: on the other that consumerism is alienating. On the contrary, affluent workers continue to adhere to traditional forms of working-class collectivism, although this tends to be instrumental and irrespective of any sense of participation in a class movement seeking structural changes in society, and are characterised by a growing 'commodity consciousness'.
Goldthorpe and Lockwood offer an alternative scenario, which they link firmly to Labour Party policy. Seeing a general disaffection with the anti-working class economic policies of the second Wilson administration and the particular drift of the new affluent worker away from Labour policies they predict that Labour would loose the next election (which, of course, they did a year later in 1970) and that a long period of Conservative government is likely to follow. This defeat will be because Labour has adopted a centrist position, not because they have ignored the new affluence. It is precisely the apparent abandoning of the working person in attempting to play the Conservative game of 'managing the economy' that Labour will lose support.
They conclude their praxiological analysis of what the Labour Party might do in the face of the growing affluence of the workers in order to sustain a class based political movement as follows:
if the working class does in the long term become no more than one stratum within a system of 'classless inegalitarianism', offering no basis for or response to radical initiatives, then this situation will not be adequately explained either as an inevitable outcome of the evolution of industrialism or as reflecting the ability of neo-capitalism to contain the consequences of its changing infrastructure by means of mass social-psychological manipulation. It will to some degree be also attributable to the fact that the political leaders of the working class chose this future for it. (Goldthorpe et al., 1969, p. 195)
Thus, Goldthorpe and Lockwood do not just draw on an existing wider theory, use it to inform their concept construction, undertake a study and suggest implications for a particular theory, as in the standard quantitative approach. Instead, they relate their analysis back to a much larger realm of debate and praxis. It is the framework in which it is situated and the nature of the enquiry into structural taken-for-granteds that gives Goldthorpe and Lockwood's study its critical edge.