In this chapter, we consider the early-life experiences of each individual that help form a lifelong ‘learner identity’ that itself appears to become a determinant of their future patterns of participation. The first of these is family background, and the second is initial schooling.

Family as a determinant

Family background, assessed in terms of income, parents’ education, or parents’ occupation, is a key predictor of lifelong participation in education or training, as it is of success at school (Gorard et al, 1999c). In some periods and regions, it may be the characteristics of the father that are dominant (Cervero and Kirkpatrick, 1990), and in others those of the mother (Bynner and Parsons, 1997). In this study, the father's occupation, the mother's education and place of birth predominate, and there is no agreement with the finding of Gambetta (1987) that the significance of parental education varies by occupational class. There is little evidence from this study that the UK is moving towards a ‘classless’ society in terms of learning, or that cycles of advantage and disadvantage are reducing. Social reproduction is still a key factor, and the link between household of origin and later occupational status may have actually strengthened since 1945 (Gershuny and Marsh, 1994).

Several previous studies have examined educational attainment and occupation (for example, Halsey et al, 1980), while there is a smaller body of literature that considers participation in education or training as its primary variable, and an even smaller body that relates the participation of generations within the same families. In compulsory education, traditions of similar educational routes (or initial educational trajectories) within families are quite common. Models of reproduction involving post-compulsory education and training, on the other hand, have generally found much weaker links between the involvement of parent and child (for example, Cervero and Kirkpatrick, 1990). The correlations between parental education and the child's later participation in the longitudinal model created by Yang (1998) are so low that, even when combined with many other predictor variables, such as school attainment and attitudinal data, only 23% to 30% of the variance in participation is explained (and this model had only considered a restricted range of adults aged from 16-32).

Theoretical models of post-compulsory participation are generally complex, involving structural, socio-economic and psychological factors (see, for example, Cross, 1981), but nearly all are variants of an accumulation form.