Chapter 1
The death and resurrection of class in sociology
Introduction
The class system is dead.
(Prince Edward)
Proclamations of the death of the class system lack shock value when articulated by a member of the royal family. They can simply be dismissed as ignorant and arrogant mutterings of those who occupy social positions that lack any connection whatsoever with the everyday ârealityâ of most âordinaryâ peopleâs lives. Unfortunately those who opine about the death of the class system are not simply confined to the fine and well-bred specimens that constitute the aristocracy. A belief in the irrelevance of class is now so pervasive within the social sciences that issues of deprivation are treated with an indifference, or even disdain, that issues from an arrogance born of security (Charlesworth 2000). Contemporary sociology has become embroiled in ânewâ concerns (âindividualisationâ, âidentityâ, âdifferenceâ, âriskâ, âmobilitiesâ etc.) that reflect the lifestyle preoccupations of its middle-class intelligentsia, whose social position is parasitic on a discriminatory higher education system that reproduces the conditions of their class privilege (Bourdieu and Passeron 1977; Reay 2001a; Reay et al. 2001) and that, to add insult to injury, they then use to proclaim the âdeath of classâ (Charlesworth 2000; Skeggs 2004). As the sociological agenda has shifted towards these âpost-materialâ concerns of âlifestyle politicsâ (Giddens 1991) studies of working-class existence have sunk to the bottom of the hierarchy of intellectual subjects (Charlesworth 2000). The decline in the symbolic profits to be gained from scholarship on working-class existence has thus resulted in a recent dearth of publishing in this area (Charlesworth 2000), certainly when compared with the volume of scholarship on issues such as âriskâ, âidentityâ and âdifferenceâ.
This chapter commences by examining the validity of the scholarly claim that âthe class system is deadâ. This claim is based on empirical evidence of increasing standards of living and increasing levels of social mobility, which, it is suggested, undermines the idea that class barriers are impenetrable. The âdeath of classâ thesis is also based on empirical evidence that people no longer invoke social class labels as identifiers. The second part of the chapter examines critiques of the âend of classâ thesis. Critics of the âend of classâ thesis have presented more nuanced sets of empirical evidence that demonstrate a continuing lack of social mobility. They also argue that working-class people no longer invoke class labels to identify themselves because the stigma historically attached to those labels cause too much pain. The absence of âclass talkâ, in other words, is indicative of the endurance of class divisions that have never been more painful or relevant to working-class people.
The âend of classâ?
The âlogic of industrialismâ and âend of Ideologyâ
The âend of classâ thesis has a rich history that stretches back to the sociology of Emile Durkheim, who argued that the modern state has tended towards a maximal (rather than minimum) role designed
to provide for [individual] self-realization. This was not something which could occurâŚwhen the operations of the state were kept to a minimum. The self-realization of the individual could only take place in and through his or her membership of a society in which the state guaranteed and advanced the rights embodied in moral individualism.
(Giddens 1987: 105)
Durkheimâs notion, that the state represents a positive force for change in modern societies, has had a key influence on welfare state theory, where it found its most visible expression in the âlogic of industrialismâ thesis that emerged in the 1960s (Donnison 1967; Donnison and Ungerson 1982) and that continues to influence contemporary welfare state theory (Emms 1990; Power 1993). One of the most famous exponents of the âlogic of industrialismâ thesis is David Donnison (1967), who identified societies at three different stages of industrial development in the 1960s. He associated the first stage with countries such as Spain, Greece, Turkey and Portugal that (like Britain in the nineteenth century) were in the throes of a transition to industrial society and experiencing high levels of urbanisation. A second group of countries, which included Britain, were said to be at an intermediate stage of industrial development and so a âresidualâ form of social housing had emerged that was designed to cater for particular needs, such as those of slum dwellers. The final group of countries, which included Sweden and West Germany, were at a more mature stage of industrial development that had enabled governments to assume responsibility for developing a âcomprehensiveâ form of public housing, albeit not necessarily provided directly by the state.
For Donnison, the âlogic of industrialismâ operates as follows. At âless developedâ stages of economic development, government priorities are directed towards industrial growth and so intervention in the housing market tends to be passive and âhaphazardâ. This exacerbates social inequalities and, it follows, encourages social class divisions to develop around working-class resentment of a government that supports industry whilst doing little to address social inequalities. However, as societies attain an âadvancedâ level of industrial development, governments become âresource richâ and are able to broaden their âinterventionistâ horizons and, furthermore, have an incentive to do so. Since âhousing makes a fundamental contribution to economic developmentâ (Donnison and Ungerson 1982: 78), for example by promoting a healthy and productive workforce, âresource richâ governments tend to invest their new-found wealth in âhousing programmesâ. The logic of economic development thus pushes all governments towards a âcomprehensiveâ form of social housing provision. Donnison articulated this particular point by making a distinction between âleaderâ and âlaggardâ societies (Donnison 1967; Wilensky 1975; Gould 1993), with the implication being that the laggards would eventually have to follow the leaders largely because the âlogicâ of economic development dictated that this was what they would have to do. This historical point engenders an âend of ideologyâ because class politics loses its relevance as class divisions and tensions recede (Bell 1960; Lipset 1963; Marshall 1967; see also Giddens 1994 for a more contemporary argument that politics has moved âbeyond left and rightâ). This end to ideological struggle between the left and right is, apparently, nowhere more evident than in the way âadvancedâ welfare societies become more concerned with social cohesion rather than representing the interests of particular (dominant) social groups.
Governments which pursue [comprehensive] aims far enough find they are applying, in the sphere of housing, a new conception of the state. They are no longer regulating, supplementing or restraining the operations of the market. They have assumed responsibility for shaping the kind of world their people are to live in, and hence for mobilizing the resources and creating the conditions required for that purpose.
(Donnison and Ungerson 1982: 81)
Social mobility and demise of class barriers
Although the welfare sociology of Donnison is based on the apparently outdated perspective of functionalism, its claim about the âend of classâ is current in contemporary sociology. A key sociological literature here is that on âsocial mobilityâ, which has sought to âtestâ whether or not the idea of âmeritocracyâ (i.e. that one can âprogressâ through hard work, ability and strength of personality), which tends to elicit widespread support in surveys of attitudes to social mobility (especially Glass 1954), is valid. Glassâs respondents emphasised how individuals could âhelp themselvesâ to âget onâ by displaying the right attitude. Only a minority of working-class respondents thought that âcontactsâ (16 per cent) and money (12 per cent) affected the prospect of social mobility. Glassâs respondents therefore believed in the justice and legitimacy of the âreward systemâ. More recent evidence from the British Social Attitudes (BSA) survey suggests that little has changed, with hard work and education continuing to feature as the most common sources of social mobility. Although respondents to the 1987 BSA survey did not deny the value of social advantage to social mobility, they also did not stress it as a factor (Savage 2000).
Social scientists such as John Goldthorpe have subjected these widely held ideas about social mobility to empirical testing. Goldthorpeâs analytical approach (which is known as the âNuffield paradigmâ or âclass structural approachâ) is to establish the extent of mobility within the class structure, that is, the proportion and flow of individuals from one social class to another. This focus on âclass aggregatesâ means he is interested not simply in the extent to which British society is âopenâ, but also in the process of social class formation, that is, the changing composition of social classes as a result of social mobility. His studies have uncovered high levels of social mobility within British society and therefore a dynamic rather than static picture of class formation (Erikson and Goldthorpe 1992). For example, his study with Erikson (1992) found that two-thirds of sons were in a different social class from their fathers. For Goldthorpe, then, social mobility constitutes the norm rather than an exception, whereas there is little evidence of the inheritance of social positions.
The levels of mobility identified by Goldthorpe have not simply been found within the occupational structure, with people from working-class backgrounds progressing into âmiddle-classâ occupations. Sociologists such as Wynne (1998) have identified similar levels of social mobility within the housing market, with middle-class suburbs now increasingly populated by people from the ânew middle classâ, that is, people in middle-class occupations but from working-class origins. Others have argued that the level of social mobility is evident not only in the extent of this ânew middle classâ flight to the suburbs but also in its consequences. These consequences are identified as the alleged âdeclineâ of traditional working-class neighbourhoods in inner-urban areas of cities such as Manchester and Salford (Nevin et al. 1999). What better indicator of the level of social mobility in Britain than the apparent decline of inner-urban working-class neighbourhoods and concomitant growth of suburbia?
Nevertheless, sociologists such as Goldthorpe have not simply been concerned with understanding the process of class formation and the changing âclass structureâ. Goldthorpe regarded this as the âstarting pointâ of âclass analysisâ rather than a basis for drawing solid conclusions about social class formation and fragmentation (Crompton 1998). For Goldthorpe, the other key task for class analysts is to establish the extent to which members of social classes coalesce around their (new and changing) position within the class structure. The key question here is whether the social classes, which Goldthorpe claims are in a constant process of formation and fragmentation, possess a âdemographic identityâ that is associated with continuity of membership and individual membersâ occupation of particular sets of positions over time (Goldthorpe 1983: 467). According to this logic, social immobility is likely to be associated with strong and cohesive class identities whereas social mobility is likely to disrupt such cohesion and therefore the ability of social groups to construct and maintain a coherent class identity. So what do we know about class identities in this changing context of increasing social mobility?
Recent work has shown that people now seldom identify themselves with a particular social class category (âworking classâ, âmiddle classâ etc.). Thus Savage (2000) suggests that the current tendency is to present oneself as part of the âmainstreamâ of âordinaryâ people, which means that class is much less relevant than social surveys might initially indicate. A classic example of this can be found in Devineâs (1992) study of affluent manual workers in Luton. Devine found that most respondents defined their social position with reference to terms (such as âordinary working peopleâ) that she claimed to be reflections of a mainstream consciousness and thus indicative of the abandonment of discursive distinctions between working- and middle-class. This corresponds with Savage et al.âs own work, which found that two-thirds of their sample identified themselves with a social class category but that this identification was usually ambivalent, defensive and hesitant, e.g. identification with a class category was often prefaced by terms such as âI supposeâŚâor âIâm probablyâŚâ (Savage et al. 2000). Savage et al. (2000) suggest that a key reason for this was that significant numbers of their respondents had never thought deeply about their personal identification with class before (hence the hesitancy contained in âI suppose IâmâŚâ). It seems, then, that increasing levels of social mobility (which have seen many households abandon working-class socio-economic positions for employment in the service sector and homes in suburbia) have existentially disrupted most individualsâ sense of their social class positioning. That is to say, class no longer appears to be an important aspect of self-identity or to evoke the strong sense of group or collective allegiance that scholars writing in the 1950s and 1960s claim to have found in working-class communities (Young and Willmott 1957). For Savage (2000: 40), then:
Admittedly class is a widely understood term, and people do use the term to make sense of some aspects of British society. However, Britain is not a deeply class conscious society, where class is seen as embodying membership of collective groups. Although people can identify themselves as members of classes, this identification seems contextual and of limited significance, rather than being a major source of their identity and group belonging. Furthermore peopleâs social attitudes and views are too ambivalent to be seen as part of a consistent class-related world view. Finally, peopleâs own class location shapes only some of their views and even then in highly mediated and complex ways.
Widespread ambivalence towards the issue of âclassâ in late modern societies has led a number of sociologists to claim that âclass as a concept is ceasing to do any useful work for sociologyâ (Pahl 1989: 710), that it is âan increasingly redundant issueâ (Holton and Turner 1989: 194) and that it is âdyingâ (Clark and Lipset 1991); and even to agree with Prince Edward that class is âdeadâ (Pakulski and Walters 1996) and therefore to issue a âfarewell to the working classâ (Gorz 1982). A key theme in contemporary sociology and social and political science more generally, then, is the claim that class has lost its relevance either per se or, at the very least, vis-Ă -vis other identifiers of difference (Savage 2000; Eley and Nield 2000). This has led to a dramatic shift in the contemporary sociological agenda, in which concerns about class have been replaced in the mainstream âsociological imaginationâ by ânewâ concerns with âindividualisationâ, âidentityâ, âdifferenceâ, âriskâ and so on (Skeggs 2004). A key figure here is Anthony Giddens (1991), whose claim that an individualised concern with âlifestyle politicsâ has superseded the âemancipatory politicsâ of class has had a fundamental influence on the content of the contemporary sociological agenda. For Giddens (1994), then, politics has moved âbeyond left and rightâ to embrace new identifiers of âdifferenceâ such as gender and ethnicity (Harrison with Davis 2002) and sexuality (Binnie 2004). Ulrich Beck has made similar claims about the declining relevance of class politics, albeit from a different perspective. For Beck (1992), the issues of key import in contemporary societies (e.g. globalisation, environmental risk etc.) affect everybody and cannot therefore be addressed within a class politics framework. This results in the emergence of a global politics of âweâ rather than a class politics of âus versus themâ. More recently, Beck (2000) has documented the (apparent) emergence of a âglobalization of biographyâ that, he claims, stems from increasing levels of geographical mobility and, furthermore, is cementing the importance of the new global politics of âweâ.
The position of class analysis within sociology has therefore shifted from its âcoreâ position within the sociological mainstream of the 1950s and 1960s to one in which it is now seen and practised as a sub-field of sociology (Savage 2000). That is to say, class is increasingly a matter for sociologists who are centrally concerned with class rather than a matter for all sociologists, who are free to ignore it with impunity. Those who have retained an interest in class analysis, and who have kept it alive as a sub-field of sociology, fall into two camps. First, the line that writers such as Thrift and Williams (1987) and Bradley (1996) take is that more complex forms of stratification are now occurring based around factors such as race, religion, ethnicity, gender and sexuality as well as class. Class does not determine experience, consciousness and action, then, but neither is it irrelevant. Second, other writers have completely rejected the ânewâ concerns of sociology in favour of the continuance of a research programme in âclass analysisâ that seeks to understand class relations within a changing social, economic and political context (Goldthorpe and Lockwood 1968, 1969; Eley and Nield 2000).
A welcome back to class
Empirical evidence of the class society
First, it is instructive to note that movement of working-class people into service sector occupations does not necessarily constitute âsocial mobilityâ. For Braverman (1974), this is because the capitalist mode of production tends towards the deskilling of labour, which affects all workers and means that the apparent upgrading of the labour force, from manual to white collar, is more apparent than real (see also Wright and Singlemann 1982; Esping-Andersen 1993). Braverman (1974) refers to this as âlabour process theoryâ. These theoretical arguments aside, the empirical evidence that points to increasing levels of social mobility is less than convincing anyway. Although Goldthorpe identified increasing levels of absolute social mobility for example, he argued that analyses of relative rates of social mobility (which are based on a comparison of the relative life chances that members of one social class have when contrasted with members of other social classes) indicate that the sons of members of privileged social classes have better life chances than those of the working class. Furthermore he argued that these advantages have remained fairly constant over time, indicating, for him, the enduring importance of class, which has consistently been at the centre of his analysis of social mobility.
Further, albeit qualified, support for these claims has been provided by Savage and colleagues. Drawing on vast amounts of empirical data, Savage et al. (1992: 134) found that some middle-class groups (the service class) have higher levels of self-recruitment than others and that working-class mobility into the middle class is more restricted in some places than others. Specifically, middle-class households with âcultural capitalâ (e.g. higher education qualifications) were found to be better able to transmit their privileges than those working in management or administration work, who did not possess the same levels of cultural capital. Put simply, cultural assets such as educational competence can be transmitted within the family more easily than can a position within an organisation. This is borne out by empirical data, which shows greater levels of inter-class mobility into management positions (which require service to the organisation rather than education) than into the professions (where entry is based on credentialism), which are more self-recruiting. Thus Savage et al. (1992: 138) are able to present empirical data that shows that over half of industrial managers have working-class fathers compared with only 43 per cent of senior administrators and 39 per cent of professional employees. Put differently, over one-third of professionalsâ children move directly into professional work whereas only 19 per cent of managersâ children become managers (Savage et al. 1992: 148). Such claims have been substantiated by Wright (1997), who also found that it has been easier to progress into a middle-class position based on authority (i.e. management) than into a middle-class position that requires skills (i.e. a profession).
The crucial point is that while the sons of professionals appear to be able to follow their fathersâ footsteps into professional emp...