Beating the Odds (1): Inter-generational Social Mobility from a Human Capital Perspective

The objective of this paper is to use a human capital measure as the positional indicator in an investigation of trends in equality of intergenerational life chances. The BHPS retrospective work and employment history files provide a continuous account of respondents’ occupational positions throughout their adult lives, and also records of their parents employment circumstances (if any) when the respondents were aged 14. In a first series of analyses, a conventional Goldthorpe 3*3 mobility model is compared to a categorical model grouping fathers and children into human capital quintiles. The Goldthorpe analysis shows some increase in equality, over mid 20 century birth cohorts, whereas the human capital model shows no clear trend. The difference is explained by: (1) the exclusion of nonemployed women from the Goldthorpe model, where the human capital estimation covers all the BHPS respondents irrespective of employment status; and (2) the use of constant-sized quintile groups in the human capital model where the sizes of the Goldthorpe classes change over time. A second series of analyses relies on regression models, explaining children’s human capital scores at various ages in terms of their sex, birth cohort, father’s human capital and various transformations of, and interactions among, these variables. Instantiations of these models show diverging trends of growth in the predicted mean human capital of children of lowand of high-human capital fathers over successive birth cohorts from the 1930s to the 1960s—implying a regular increase in the inequality of intergenerational life chances in Britain over this period. This paper is part of the “Social Position and Life Chances” (SPLC) project, which aims to formulate and estimate new measures of social class appropriate for understanding life chances in modern Britain. An initial outline of the project is set out in ISER Working Paper 2001—20; the human capital measure used here is discussed in ISER Working Paper 2002—2. The relationship of wealth to human capital is discussed in ISER Working Paper 2002—16, and ISER Working Paper 2002—18 develops a Gini-type index of intergenerational mobility. The SPLC project is part of the Research Programme of the ESRC Research Centre on Micro-social Change. NON-TECHNICAL SUMMARY This paper uses a scalar measure of human capital to estimate intergenerational social mobility in Britain. The gap between this approach to social mobility and the established British tradition, which considers mobility in terms of class categories based on employment situations, is not as great as might initially be imagined. It is contended here that, since individuals’ labour market characteristics are determinants of their employment situations— and since it is now possible to operationalise a direct measure of human capital for empirical research purposes—so it is the human capital, rather than the employment situation, that should be considered constitutive of class position. The argument follows recent discussions on the normative basis of mobility theory, which suggest that the traditional British focus on equality of opportunity (concerning probabilities of transitions between parents’ and children’s social positions) should be supplemented by considerations of the level of rewards attached to those positions (otherwise, for example, improvements in one might be compensated-for by reductions in the equality of the other). Since human capital scores are simultaneously measures of social position and of individuals’ ability to obtain rewards, using of human capital measures permits the combination of the twin concerns with opportunities and rewards, into a single concern with “life chances”. The paper’s main substantive focus is on the use of the British Household Panel Survey to address the long-standing proposition of the “liberal theory of industrialisation” that, in the course of economic development, opportunities for mobility become more equal. The first step is to consider just equality of opportunity, comparing the results from Goldthorpe class measures, with categorical measures constructed by reducing the parents’ and children’s human capital scores into quintile groups. Using the same samples, some inconsistencies emerge, reflecting (1) the wider coverage of the human capital-based analysis (which includes the whole of the second generation sample rather than just those with recent employment records), and (2) the constant sizes of the “marginal” distributions of parents’ and children’s positions that results from the use of the human capital quintile groups (rather than class categories whose sizes vary over time). Overall, however, the results are similar to the traditional ones: there is no consistent trend supportive of the liberal theory. The second step uses the scalar measure of human capital to look at trends in intergenerational mobility in life chances, by constructing a regression model of the relationship between fathers’ human capital, and their children’s throug their life course. This model is instantiated to compare trends in human capital levels for children of low and of high human capital fathers, through successive mid-20 century birth cohorts. The conclusion is that while both groups show increases, those of high human capital fathers grow faster, implying that intergenerational mobility patterns are becoming more unequal. The “life chances” approach shows British mid-late 20 century mobility trends to be moving directly counter to the liberal theory prediction.