The conceptualisation and measurement of occupational hierarchies: A review, a proposal and some illustrative analyses

(2011) - Erzsebet Bukodi, Shirley Dex, John H. Goldthorpe

Journal: Quality & Quantity
Link:: http://link.springer.com/10.1007/s11135-010-9369-x
DOI:: 10.1007/s11135-010-9369-x
Links::
Tags:: #paper #SocialClass
Cite Key:: [@bukodiConceptualisationMeasurementOccupational2011]

Abstract

Occupational data are central to much research in the field of social stratification. Yet there is little consensus on how such data are most appropriately classified and scaled. We evaluate occupational scales currently in use on the basis of a fourfold typology. This crossclassifies scales, on the one hand, according to whether they are intended to be ‘synthetic’ or ‘analytic’ and, on the other, according to whether they are based on ‘subjective’ or ‘objective’ data. Focusing chiefly on issues of validity, we argue that scales of the analytic-objective type are those which, for most purposes, can be used to best advantage in stratification research. We illustrate our argument by applying scales of occupational earnings and occupational status in analyses of the worklife occupational mobility of men in Britain, using the data-set of the National Child Development Study.

Notes

"We propose decomposing social origins into parental class, parental status, and parental education." (Bukodi and Goldthorpe 2013:1024)

"we analyse data from three British birth cohort studies" (Bukodi and Goldthorpe 2013:1024)

"three components of social origins have independent and distinctive effects on educational attainment, and ones that persist or change in differing ways across the cohorts" (Bukodi and Goldthorpe 2013:1024)

"careers or in terms of the highest level of education that they eventually achieve. For example, researchers differ over whether they define and model educational attainment in terms of various transitions that individuals make in the course of their educational" (Bukodi and Goldthorpe 2013:1024)

"little explicit discussion has taken place of the theoretical grounds for treating social origins in one way rather than another" (Bukodi and Goldthorpe 2013:1025)

"difficult to avoid the conclusion that some notion of the 'interchangeability of indicators' (Lazarsfeld, 1939) has prevailed: or, in other words, that it has been assumed, if only implicitly, that however social origins are measured, it will make rather little difference in determining the extent of, or changes in, associated inequalities in educational attainment" (Bukodi and Goldthorpe 2013:1025)

"Jaeger (2007) has argued that the inadequate conceptualization and measurement of social origins is indeed a further likely source of divergent results regarding trends in educational inequalities." (Bukodi and Goldthorpe 2013:1025)

"First of all, we would note that the EGP or similar class schemata are not intended to serve as proxy variables in the way Jaeger believes they do, and that they certainly do not do so 'by definition', as he claims (2007: 528)." (Bukodi and Goldthorpe 2013:1025)

"both their criterion and construct validity have been extensively, and in general successfully, tested (see Goldthorpe, 2007, vol. 2, ch. 5; McGovern et al., 2008; Rose and Harrison, eds., 2009)." (Bukodi and Goldthorpe 2013:1025)

"maintain that parental economic resources, and especially as they might be used to support children's education, are well captured through such a concept of class" (Bukodi and Goldthorpe 2013:1025)

"Correspondingly, while we would agree that if class serves as the only indicator of social origins, it is likely to 'pick up' the effects of different, but associated, factors also influencing individuals' educational attainment, we would not see as the solution to this problem the ad hoc 'decomposition' of class" (Bukodi and Goldthorpe 2013:1025)

"would rather complement the concept of class, understood in the way indicated above, with a further concept, at a similar level of generality, intended to capture the sociocultural, as distinct from the economic, aspects of stratification" (Bukodi and Goldthorpe 2013:1025)

"Finally, and following on from the idea that what needs to be 'decomposed' is the concept of social origins rather than that of class, we would argue that where the dependent variable is educational attainment, the practice of including parental education as a further component of social origins is obviously appropriate" (Bukodi and Goldthorpe 2013:1026)

"parental status is also included in the analysis along with parental class, we would then wish to interpret parental education in a more specific way than do Jaeger and others working under the influence of Bourdieu. We would take parental education as indexing what might be described as 'educational resources': that is, parents' capacity to participate directly in furthering their children's educational careers as, say, by creating a supportive home learning environment and, further, by using their own knowledge of the educational system to provide informed guidance concerning choice of schools, subjects to study, courses and examinations to take" (Bukodi and Goldthorpe 2013:1026)

"each cohort, we restrict our attention to cohort members on whom we have complete information on all of the variables discussed below." (Bukodi and Goldthorpe 2013:1026)

"dependent variable of our analyses, we take individuals' 'completed' educational attainment: i.e. their highest educational qualification at age 34." (Bukodi and Goldthorpe 2013:1026)

"The independent variables that we introduce, in addition to cohort, are our three social origin variables treated as follows" (Bukodi and Goldthorpe 2013:1026)

"We use the 7-class 'analytical' version of the National Statistics Socio-economic Classification (NS-SeC), which can be regarded as a new and improved instantiation of the Goldthorpe class schema for Britain" (Bukodi and Goldthorpe 2013:1026)

"We use the status scale proposed by Chan and Goldthorpe (2004), which is derived from the occupational structure of close friendship relations" (Bukodi and Goldthorpe 2013:1027)

"The data available on cohort members' parents' education are less detailed than those available on their own education, so we are unable to use the scale shown" (Bukodi and Goldthorpe 2013:1027)

"Instead, we use seven ordered categories that take account of the level of both parents' educational qualifications—when their children were aged 10 in the case of the 1946 and 1970 cohorts and aged 11 in the case of the 1958 cohort" (Bukodi and Goldthorpe 2013:1027)

"When education is an explanatory variable, and especially in analyses extending over a period of time in which the distribution of education has changed substantially, we believe it preferable to treat education in relative rather than absolute terms. We therefore score each level of parental education for each cohort according to the percentage of parents falling below that level in the cumulative percentage distribution for the cohort." (Bukodi and Goldthorpe 2013:1027)

"Full details of the parental class, status, and education variables are provided in an Appendix." (Bukodi and Goldthorpe 2013:1027)

"Our first research question is that of whether parental class, status, and education have independent effects on children's educational attainment" (Bukodi and Goldthorpe 2013:1027)

"apply a series of binary logit models" (Bukodi and Goldthorpe 2013:1027)

"In this connection, it is of some further interest to note that for the 1970 cohort, we can also include in our analyses a reliable family income variable (as at child's age 10)—although with a 40% reduction in the cohort N because of missing data" (Bukodi and Goldthorpe 2013:1030)

"We now turn to our second research question, that of how far the effects on children's educational attainment of their parents' class, status, and education are constant over time or may change in similar or in different directions." (Bukodi and Goldthorpe 2013:1030)

"apply the same series of logit models as before but now introduce interaction terms between cohort and parental class, status, and education, respectively" (Bukodi and Goldthorpe 2013:1030)

"contrast, as regards parental status effects, it can be seen from Table 4 that while in the case of sons no significant decline is found between the 1946 and 1958 cohorts, such a decline does show up quite consistently between the 1958 and 1970 cohorts—that is, at each qualification threshold. And Table 5 reveals a similarly consistent decline in the case of daughters." (Bukodi and Goldthorpe 2013:1033)

"Our third research question—regarding the combined effects of parents' class, status, and education and their relative importance—is a complex one that we can treat here in only a limited way" (Bukodi and Goldthorpe 2013:1033)

"simplify the possible combinations" (Bukodi and Goldthorpe 2013:1033)

"this division as forming the dependent variable in a binary logistic regression model, with our four-level measures of parental class, status, and education being the independent variables, and with interaction terms included between cohort and the three parental variables and also between parental class and parental status and between parental class and education." (Bukodi and Goldthorpe 2013:1033)

"Comparing the extremes, and taking the cohorts together, both the sons and daughters of Type 4 HPs have less than a 20% chance of becoming what might be regarded as 'well-qualified', while the sons of Type 1 HPs have 60-70% chance and the daughters a 70-80% chance. Looking across the cohorts, no reduction in these disparities is evident for women and barely so for men." (Bukodi and Goldthorpe 2013:1033)

"The results we have presented show that parental class, parental status, and parental education cannot be taken as essentially interchangeable indices of social origins" (Bukodi and Goldthorpe 2013:1035)

"find that each has a significant independent effect on children's educational attainment" (Bukodi and Goldthorpe 2013:1035)

"n future discussion of, and research into, inequalities in educational attainment associated with individuals' social origins, the way in which social origins are conceptualized and measured needs to be given full and explicit attention." (Bukodi and Goldthorpe 2013:1035)

"seems a quite common practice, this is likely to mean that class effects will be overestimated, in that they will pick up different, but associated, social origin effects, while social origin effects in total will be underestimated. parental class is taken as the sole indicator of social origins, as now" (Bukodi and Goldthorpe 2013:1035)

"Thus, our finding of more or less constant class effects is not only consistent with evidence that the economic inequalities that our concept of class captures did in fact change little in Britain in the decades up to the 1990s (Goldthorpe and McKnight, 2006; McGovern et al., 2008) but can also be related to a more specific finding. It has been shown (Jackson et al., 2007; Jackson, in press) that in accounting for class-related inequalities in educational attainment in Britain, 'secondary' as opposed to 'primary' effects are of persisting importance: that is, effects resulting not from class differences in children's actual performance at particular stages in their school careers but from class differences in their subsequent educational choices, controlling for performance" (Bukodi and Goldthorpe 2013:1036)

"Finally, as regards the strengthening between the 1958 and 1970 cohorts of the effects of parents' education on their children's performance, at least two processes can be identified that would lead to greater importance attaching to what we have referred to as 'educational resources'. On the one hand, schools have placed a growing emphasis on involving parents in their children's education and in particular in taking a more active role in overseeing and reviewing homework—the required amounts of which appear to have steadily risen at all levels including the primary (Hallam, 2004)." (Bukodi and Goldthorpe 2013:1037)

"On the other hand, at the secondary and tertiary levels, examination and continuation procedures have become far more complex, and options for courses and qualifications to take and for institutions to attend have grown in number—again giving advantages to children whose parents are equipped with the kinds of knowledge needed to offer guidance through the system." (Bukodi and Goldthorpe 2013:1037)

"analyses of inequalities in children's educational attainment that 'decompose' social origins in the way we have proposed would seem essential to achieving a full understanding of the nature and extent of these inequalities" (Bukodi and Goldthorpe 2013:1037)