mobility, social

The movement-usually of individuals but sometimes of whole groups-between different positions within the system of social stratification in any society. It is conventional to distinguish upward and downward mobility (that is, movement up or down a hierarchy of privilege), and intergenerational from intragenerational or career mobility (the former referring to mobility between a family of origin and one's own class or status position, the latter to the mobility experienced during an individual career, such as respondent's first job compared to his or her present job). Other distinctions-most notably that between structural and non-structural mobility-are more contentious.

Most sociological attention has focused on intergenerational mobility, in particular the role of educational achievement as compared to that of social background or of ascriptive characteristics such as race , in explaining patterns of occupational attainment. Although there have been many case-studies of élite recruitment (for example P. Stanworth and A. Giddens's Elites and Power in British Society, 1974), the most popular research instrument has been the large-scale sample survey , and the most common points of comparison have been occupations . Some sociologists have studied social mobility in pre-industrial contexts (see, for example,, Historical Research on Social Mobility, 1977), and others in contemporary developing countries such as India (see, Caste, Class and Power, 1965), but the great majority of studies have dealt with the modern industrialized West and, to a lesser extent, the former communist states of Eastern Europe.

The study of social mobility has a long sociological pedigree, extending back to the mid-nineteenth-century writings of Karl Marx and John Stuart Mill , with major contributions in the early twentieth century from Vilfredo Pareto (who proposed a theory of the ‘circulation of élites’) and Pitirim Sorokin . The now vast literature on the subject is inextricably entangled with wider discussions of (among other things) education, gender, culture, power, statistical techniques, and the role of theory in social research.

It is possible, indeed, to trace many of the classic debates in modern sociology back to the early arguments about mobility. For example, in Social Mobility (1927) Sorokin wrote that ‘channels of vertical circulation exist in any stratified society, and are as necessary as channels for blood circulation in the body’. In an argument that prefigures the later functionalist theory of stratification , he suggested that these ‘staircases’ or ‘elevators’ are necessary to the efficient allocation of talents to occupations, and that failure to achieve this promotes inefficiency and disorder. However, unlike Kingsley Davis and Wilbert E. Moore writing two decades later, Sorokin did not then conclude that high rewards were necessary to motivate individuals to undertake training for functionally important positions in society. More plausibly, he maintained that the incumbents of these positions were able to exploit their strategic occupational roles, in order to attract material and other privileges. Sorokin was particularly interested in the role of educational institutions in allocating people to the various occupational positions. Anticipating the radical critiques of the new sociology of education of the 1970s, he argued that schools function primarily as ‘a testing, selecting, and distributing agency’; in other words, they merely certify children for particular positions in the labour-market , rather than promote each individual's abilities or encourage the development of talent.

Confronted by this potentially vast field of interest, it is useful for heuristic purposes to view the modern literature on social mobility as a dispute between two divergent research programmes which have set the terms of discussion for this subject since 1945, and continue to dominate the field even today. On the one hand, there are those investigators who view mobility in the context of a social hierarchy, within which individuals can be ranked according to income, educational attainment, or socio-economic prestige. On the other, there are those who set it within the context of a class structure, embracing social locations defined by relationships prevailing within labour-markets and production units. During the 1950s and 1960s, the former hierarchical perspective was dominant, culminating in the so-called status-attainment tradition of mobility studies emanating principally from the United States. This was increasingly challenged, during the 1970s and 1980s, by researchers schooled within or influenced by the European tradition of class analysis.

The status-attainment programme sees the principal interest of mobility studies as being an attempt to specify those attributes which are characteristic of individuals who end up in the more desirable rather than the less desirable jobs. Characteristically, these studies investigate the extent to which the present occupational status of individuals is associated with the status of their family of origin, rather than individual achievements such as educational attainment. One virtue of this approach, as compared to earlier cross-tabulations of father's occupation by son's occupation, was that it disentangled at least some of the processes that linked the generations. For example, researchers explored the effects of father's education on son's occupational attainment, and showed that these were distinct from the effects of father's occupation. Most studies maintained that son's education was the crucial link between family background and occupational success, arguing that as much as half of the association between the two was mediated via education, with children from more privileged family backgrounds being higher educational achievers than their poorer peers. Later researchers extended the field of interest to include data on income, with most concluding that the impact of family background on earnings is substantial, but operates entirely indirectly through educational and occupational attainment.

Most of these studies employed the statistical techniques of regression analysis (and in particular path analysis ). Most were also underpinned by a tacit adherence to a liberal model of industrial societies as increasingly homogeneous, middle class, meritocratic , and open. Typically, therefore, they tended to conclude that structural shifts in advanced industrial economies (especially the expansion of managerial, professional, and administrative occupations) created more ‘room at the top’ and so increased the opportunities for upward social mobility of individuals from working-class origins. This increasing social fluidity was reinforced by a progressive shift from ascriptive to achievement criteria as the dominant factors determining status attainment, a movement towards meritocratic selection that, together with the prevailing high rates of social mobility, undermined the potential for class formation and class conflict in industrial societies. Peter M. Blau and Otis D. Duncan's The American Occupational Structure (1967) is generally held to be the paradigmatic example of a study of social mobility within the status-attainment tradition.

The Blau-Duncan model prompted an enormous number of related and derivative studies. Whatever their differences and similarities, however, they all rested upon the assumption that occupations can be ranked within a status hierarchy about which there is a wide degree of consensus within and between societies. In some studies this social hierarchy was conceptualized narrowly as being one of occupational prestige . In others it was generalized to include additional wider aspects of socio-economic status . Rather than dispute the details of the occupational hierarchy, however, European class analysis came increasingly to challenge the basic premiss of the status attainment research programme; namely, that social mobility was most appropriately viewed as a matter of hierarchical occupational attainment among competing individuals.

The class analysis tradition starts from the rather different assumption that individuals are born into distinct social classes , membership of which has clear consequences for life-chances, values, norms, life-styles, and patterns of association. Representatives of this tradition argue that the socio-economic status scales at the heart of the status-attainment perspective display many unresolved methodological weaknesses. Most importantly, because these scales are a composite measure of popular judgements about the relative prestige or social standing of the various occupations, they rank alongside each other, as having similar levels of socio-economic status, occupations which have quite different structural locations. For example, skilled manual workers may have the same prestige score as routine clerical workers and self-employed shopkeepers, or office supervisors may be ranked alongside farmers and schoolteachers. In other words, the synthetic categories of the scale typically contain occupational groupings that are subject to different structural forces: because of sectoral and other changes in the occupational structure, some occupations will be in expansion, others in contraction, and some will be static. Such heterogeneity merely muddies the water of mobility: it is impossible to distinguish adequately the various structural influences on mobility from those which originate in other factors, and impossible also therefore to isolate hierarchical effects (family background, educational attainment, or whatever) from other effects of a non-hierarchical kind (such as changes in the occupational division of labour, industrial or sectoral growth and decline, government policies of protection, and so forth).

The class analysis programme of social mobility research, initiated in the 1970s, abandoned the Blau-Duncan form of occupational prestige-scaling in favour of discrete class categories whose members shared similar positions within labour-markets and production units. In Europe probably the most widely used class categories are those devised by John Goldthorpe (see Goldthorpe class scheme ) for the Oxford Mobility Study during the 1970s, a class scheme which attempts to aggregate occupational groupings whose members share similar ‘market situations’ and ‘work situations’ (a theory of class which Goldthorpe derived from his earlier collaboration with David Lockwood during the Affluent Worker Study of the 1960s). In the United States, the ‘ new structuralism ’ of the 1970s alerted some analysts of social mobility to the importance of labour-market influences on mobility trajectories, and led to the emergence of ‘multiple regression Marxists’ (such as Erik Olin Wright) who adapted the methods of Blau and Duncan to a theoretical stance which pointed to the importance of ownership, authority, and autonomy in the workplace.

Along with this new theory went new methods and conclusions. Class analysts argued that mathematical techniques of loglinear modelling were better suited to the analysis of mobility data, both because they did not require ordinal-level data (and therefore unsubstantiated assumptions) about a status hierarchy, and because they allowed researchers analysing a standard mobility matrix (a contingency table cross-tabulating class origins against class destinations) to distinguish absolute or total mobility rates (including those changes in mobility occasioned by shifts in the occupational structure) from changes in social fluidity or openness within the structure as such (relative rates). Applied to the same sorts of large-scale data-sets as were used within the status-attainment programme, the class analysis perspective and the new techniques of loglinear modelling suggested that the liberal assumptions of the earlier studies were unwarrantably optimistic. In most industrialized societies absolute levels of mobility have indeed increased significantly over the past three-quarters of a century, in accordance with the growth in skilled non-manual occupations, but relative mobility chances have remained largely unaltered throughout this period. More room at the top has not ensured greater equality of opportunity to get there, since proportionately more of the new middle-class jobs have been captured by the children of those already in privileged class locations. As a result, the association between an individual's class of origin and eventual destination has remained remarkably stable across successive birth-cohorts, despite economic expansion, educational reform, and redistributive social policies.

In the mid-1980s, Goldthorpe (together with collaborators in Sweden and Germany) designed the CASMIN (Comparative Analysis of Social Mobility in Industrial Nations) Project, an intensive comparative study of this problem. The project produced data which show that the mobility profiles of advanced societies are more complex than is allowed within either the liberal theory of industrial society or Marxist accounts of capitalist society. The most important findings suggest that, measured in absolute rates, the amount and pattern of mobility displays considerable variation across societies; that relative rates (or fluidity patterns) display a ‘large commonality’ across societies; and that changes in social fluidity over time follow a pattern of ‘trendless fluctuation’ rather than showing evidence of a general increase. In short, therefore, there is no long-term ‘loosening’ of the class structure, no increase in ‘fluidity’, and (by implication) no move towards meritocracy.

Each of these research programmes disputes the principal substantive conclusions arrived at by the other. The relevant journals are littered with acrimonious exchanges about matters of theory and method. Sometimes these leave outsiders puzzled or bemused: one sceptical observer has described the field as a ‘set of statistical techniques in search of a problem’. Others argue that, for a variety of different reasons, debates about social mobility continue to raise the fundamental issues of the discipline as a whole. For example, feminists have pointed to the fact that most mobility studies are based on samples of males only, and this has prompted wide-ranging discussion of the relevant ‘unit of mobility analysis’ (individual or family), the nature of so-called cross-class families (where two adult wage-earners are in different class positions and have different mobility trajectories), and the implications of the occupational division of labour by sex for mobility studies. The best overview of these and other related debates is Anthony Heath's Social Mobility (1981).

However, much dispute in the area is of a highly specialized and technical, rather than theoretical, nature. The longest-running controversies hinge on the possibility of distinguishing structural and non-structural sources of mobility. In earlier studies, some sociologists attempted to distinguish between structural (or net) and circulation (or exchange) mobility, the former being that amount of mobility required by the structure of the table itself (the fact that, if the marginal totals showing the distribution of fathers and that of sons are regarded as being fixed, then their differences mean that some respondents must fall into the off-diagonal cells in the table). The percentage of respondents who were mobile because of the very structure of the table was said to represent the amount of structural mobility in a society. Circulation mobility was then simply the difference between the total number of respondents who were mobile and those defined as being structurally mobile. However, both of these concepts are statistical artefacts with no clear substantive interpretation, so the somewhat artificial distinction between structural and circulation mobility has given way to a dispute about the concepts of absolute and relative mobility rates. In any origin to-destination mobility table, the row and column marginal totals (say, the distribution of fathers as compared to sons) will be different, an asymmetry that is due in part to changes in the occupational structure itself (such as, for example, sectoral shifts of the kind noted above). The use of loglinear techniques (based on the technique of odds ratios ) permits the calculation of relative mobility chances which allow for (exclude) that portion of total mobility that is due to changes in the marginal distributions of the table. Many class analysts insist that this technique therefore distinguishes meaningfully between mobility which is the result of changes in the shape of the class structure and that which reflects of changes in its degree of openness. Critics maintain that the concept of relative mobility is no less artificial than that of structural and circulation mobility since, whether or not social mobility is caused by sectoral shifts alone, absolute or overall mobility is ‘real’-whereas respondents do not experience the ahistorical and acontextual phenomenon of ‘relative mobility chances’. In part this is also a dispute about the relationship between occupational mobility and class mobility; and, therefore, is inescapably a debate about the very definition of social class itself. See also Benini coefficient ; contest and sponsored mobility ; Featherman-Jones-Hauser hypothesis.