Introduction
The essential advantage of the Becker approach is its clarity and support in the
formulation of hypotheses. It does not provide alternative hypotheses about the
processes by which resources impact on attainment but assumes that attainment
follows in a fairly straightforward way from the investments of parental resources.
The developmental model provides more insight into these processes and wider
constraints on them. However, while it recognises the importance of financial and
other constraints at the distal level it has been less explicit in formulating their
implications. The two approaches can thus be beneficially brought together within the
developmental framework adopted here.
1.5.10 The definition and meaning of education
In this study we focus on the effects of prior parental education assessed in terms of
qualifications and years of schooling. This reflects the approach adopted in the vast
majority of the studies reviewed. Therefore, we abstract from consideration of the
effects of vocational training and current learning by parents, either separately or
jointly with children in family learning programmes.
Moreover, there is an important distinction between participation in a learning
opportunity on the one hand and actual learning on the other. For some aspects of the
relation between parental education and children’s attainment the distinction may not
matter. For example, if parents with higher levels of education earn more purely
because of credential signalling advantages in the labour market then their children
may have access to better learning opportunities even if no parental learning has
actually taken place (Altonji & Pierret, 2001; Spence, 1973; Weiss, 1995). This is a
case where there is an inter-generational effect of education not channelled by
learning. However, most examples of theorised links from parental education to
children’s attainments will assume that education does include some genuine element
of learning and cognitive development. Quite what is learnt is an important research
question. Few studies, however, are in a position to assess empirically the
implications for our research questions.
1.5.11 Representativeness and causality
There are also problems in assessing the extent to which one can generalise from the
inter-generational returns to education for those who actually did participate in
education to benefits for others were they to do so. These issues have not been
resolved in the literature reviewed in this paper because, on the whole, the focus of
authors has been elsewhere. Indeed few papers focus on the full assessment of the
pathways for the inter-generational transmission of learning. Most of the papers
reviewed here are concerned with what we take to be steps in that process or links in
that chain of processes. Many authors have written on the effects of income on
attainments (see section 5.5) without considering the effect of education on income or
of education elsewhere in the process of transmission of income effects. Thus, even
where authors have been concerned to identify precise causal effects in an
econometric sense, their results do not allow for generalisation because they have
only been considering one link in the chain, as we see it.
19