The Blackwell Companion to Sociology (71 page)

further distinguished upper non-manual occupations (professional, technical,

and managerial positions) from lower non-manual occupations (clerical and

sales positions), skilled from semi-skilled and unskilled manual jobs, and

farm owners from farm laborers. Still, differences in the placement of specific

occupations, particularly sales and service occupations, rendered the compar-

ability of such schemes quite suspect even when the categories were nominally

similar.

However, in recent years a set of class categories devised by Erikson, Gold-

thorpe, and Portocarero ± and hence known as the EGP scheme ± has emerged as

the most widely accepted international standard. The EGP scheme organizes

302

Donald J. Treiman

jobs on the basis of the employment relations they entail, using information on

the occupation performed together with employment status and supervisory

responsibility (Evans, 1992). An important feature of the scheme is that it is

conceptually multidimensional. This point is discussed further below.

The result of these developments is that there are now three internationally

standardized scales of occupational status available to the research community,

all of which can be created from the categories of the 1968 and 1988 versions of the International Standard Classification of Occupations (International Labour

Office, 1969, 1990), a classification of several hundred occupational categories that forms the basis of many national census classification schemes. This is a

major accomplishment of the past quarter century, which makes possible valid

comparisons of occupational status attainment and mobility across nations and

over time.

Social Mobility and Status Attainment

The publication of The American Occupational Structure (Blau and Duncan,

1967) revolutionized the study of social mobility and created the field of status attainment. Although research on the extent to which socioeconomic differences

are reproduced from generation to generation had long been of central concern

to students of social stratification, the typical approach was to construct a two-variable cross-tabulation showing, for a representative sample of the male

population, the relationship between the occupations of men and those of their

fathers. In the same way, cross-tabulations of educational attainment by father's occupation and of occupation by educational attainment were constructed, but

three-way tabulations were seldom made, mostly because samples typically were

too small to permit reliable estimates.

Blau and Duncan radically altered the way intergenerational social mobility

was studied, both conceptually and methodologically. They reconceptualized the

problem of the intergenerational transmission of status from generation to

generation (the classic mobility formulation) as the problem of the determinants of the status attainment of individuals, where social origins (father's occupational status, parental education, etc.) were only some among several factors that accounted for the status outcomes of individuals. For example, in figure 21.1,

which expresses the core of Blau and Duncan's approach, social origins, mea-

sured by father's occupational status (SEI) and father's years of schooling,

influence the years of schooling of the respondent; father's occupational status plus years of schooling influence the status of the respondent's first job; and all three factors influence the status of the respondent's current job (at the time of data collection). Note that father's education has no indirect influence on the

status of the first and current job, but operates entirely through education. Of course, the explicitly identified factors do not completely explain any status

outcome; the unexplained portion is represented in the diagram by thè`residual''

paths ± the arrows with no determinants. Methodologically, Blau and Duncan

applied structural equation modeling and the algebra of ``path analysis'' to

Occupations, Stratification, and Mobility

303

Figure 21.1 Path coefficients in basic model of the process of stratification.

quantify the relative importance of various paths to status attainment. For

example, in Blau and Duncan's data the correlation between the status of the

respondent's occupation and that of his father when he was age 16 (a standard

measure of the extent of status reproduction) is 0.4. This correlation can be

decomposed into a portion that reflects the fact that the sons of higher status

fathers go further in school and hence obtain higher status jobs and a portion

that reflects the transmission of occupational status independently of education.

Such decompositions can reveal how the stratification system works. In Blau and

Duncan's data (for the USA in 1962) the correlation between the occupational

status of fathers and sons depended more on the fact that those whose fathers

had high status occupations tended to go further in school and those who went

further in school acquired higher status jobs than on the direct inheritance of

occupational status; 56 percent of the correlation was due to the paths that

involved education.

The combination of conceptual and statistical advances embodied in Blau and

Duncan's analysis unleashed an enormous body of subsequent work that refined

and expanded the status attainment model, on the one hand, and challenged it

on the other. Central findings include the following.

First, there is only a loose connection between the socioeconomic status of

parents and their children. With a handful of exceptions (for example, South

Africa, with its history of racial oppression), the correlation between the socioeconomic status of successive generations is quite modest in most nations: about 0.3±0.4 between the occupational status of men and that of their fathers when

they were growing up; about 0.4±0.5 between the years of schooling of parents

and children; and, for the limited number of places for which there are adequate data, about 0.2±0.3 between the incomes of men and their parents. Thus, from

fewer than 10 percent to no more than 25 percent of the variance in socio-

economic status characteristics is shared by successive generations. Moreover,

even when multiple determinants of each status outcome are considered, it is

seldom possible to account for as much as half the variance. Similar evidence is to be found in social mobility studies (discussed in more detail below). In

industrial societies fewer than 10 percent of the male population work in the

304

Donald J. Treiman

same specific occupation (for example, truck driver or school teacher) as their

fathers did when they were growing up. Moreover, even when jobs are grouped

into a small number of occupational classes (for example, the six-category

version of the EGP classification), the majority of men are in different categories from their fathers.

The implication of these facts is profound: modern society is quite open. The

achievement of socioeconomic status ± going far in school, obtaining a high

status job, earning a lot of money ± is determined as much by a combination of

individual traits (cleverness, charm, drive, etc. ± or their lack) and chance events (being in the right place at the right time, having an especially supportive teacher, stumbling upon crucial information about a job or financial opportunity ± or

their opposite) as by systematic social factors, despite the best efforts of parents to ensure the success of their children. A little reflection suggests that this is a happy state of affairs, when one considers the alternative; imagine what it would be like to live in a society in which (as noted by Blau and Duncan, 1967, p. 174) some people are destined to poverty and others to affluence, some to work as

laborers and others to enjoy professional careers, and so on, simply because of

the circumstances of their birth.

Second, education has become the primary vehicle for occupational alloca-

tion. Given that industrialized societies are bureaucratically organized, that offspring generally do not take up the same jobs as their parents, and that schooling has become extensive, education has turned from à`consumption good'' that

inculcates the style and manner appropriate to maintaining one's station into a

means of preparing people for work. To a considerable degree, people learn

occupationally specific skills in school. But even when they don't, their level of schooling is the most important determinant of the kind of occupation they will

be able to obtain.

Third, educational attainment promotes social mobility. The relatively modest

association between the education of parents and their offspring has already

been noted. Even when other attributes of parental status, such as father's

occupational status, are considered, less than half of the variation in educational attainment can be attributed to social origins. On the other hand, as also noted, the connection between education and subsequent occupational status attainment is quite strong. The implication is that education serves to weaken the

connection between social origins and destinations ± that is, to increase the

likelihood of social mobility.

Fourth, the more industrialized a society, the greater the importance of educa-

tion for status attainment and the smaller the importance of social origins. Over the course of the twentieth century virtually every nation in the world industrialized as rapidly as possible. The consequence has been a very substantial

decline in the proportion of the labor force in agriculture and, within the non-

agricultural sector, an increase in the proportion of the labor force in non-

manual jobs relative to the proportion in manual jobs. In most countries there

has been a concomitant increase in the level of education of the population and

the proportion of school age children enrolled in school as educational creden-

tials have become more important requirements for jobs. There is suggestive, but Occupations, Stratification, and Mobility

305

not yet definitive, evidence that as a result of these shifts, education has a

stronger influence on occupational status attainment in industrialized than in

non-industrialized societies, while the direct transmission of occupational status is weaker in industrialized societies.

Apart from these generalizations, which derive from application of the

core status attainment model to data from many countries and at different

points in time, the basic model has been elaborated in two additional directions.

One elaboration has been the inclusion of additional variables, which has

documented: the importance of intelligence, motivations, and other individual

level factors; the influence of teachers and peers; and the increasing stability of careers over the life course; and also has been used to assess the extent and

consequences of measurement error ± a topic that deserves more attention since,

contrary to the way most analysts behave, virtually all data used in social

research are measured with some degree of error. A second elaboration has

been the comparison of status attainment processes across subgroups within a

society. This work has demonstrated convergence in the average education of

men and women over time, a strong similarity between men and women in the

link between social origins, education, and occupational status, but also a

substantial gender gap in incomes, which is an unhappy but real cross-cultural

universal. It has also been used to disaggregate racial and ethnic differences in status attainment in order to identify the locus of such differences. In South

Africa, for example, there are very large racial differences in education, small racial differences in occupational status returns to education, but strong racial differences in income returns to education and occupational status. Thus, it is

evident that, contrary to previous assumptions, the main effect of apartheid

was not to block the access of non-whites to jobs for which they were qualified

but to block access to training and also to discriminate in pay rates on the basis of race.

Despite these successes, the status attainment approach, which relies heavily

on unidimensional measurements of status attributes, has not gone unchal-

lenged. The most important challenge has been mounted by Erikson and Gold-

thorpe, in the CASMIN (Comparative Analysis of Social Mobility in Industrial

Nations) project, which culminated in a 1992 monograph, The Constant Flux.

Erikson and Goldthorpe analyzed intergenerational occupational mobility tables

derived from 12 large-scale national surveys conducted mainly in the 1970s

which they had carefully recoded into the categories of the EGP classification.

By taking seriously the discrete nature of social stratification and social mobility, and using log-linear modeling techniques, Erikson and Goldthorpe, and others

working in the same tradition, were able to demonstrate, first, that occupational inheritance ± the propensity for men to work in the same occupational class as

their fathers ± was very strong for those categories that involved the transmission of property (farming, small business, and professional practice); and, second,

that there is a sharp distinction between agricultural and non-agricultural occupations, with substantial mobility out of agriculture but virtually no mobility

into agriculture. Still, it is clear that the dominant force driving the rate of mobility between different occupational categories is their similarity with