Introduction
education but it is also the case that if one stratifies children by parental education
rather than traditional occupational measures of social class the attainment gradient is
every bit as steep. This is shown in Figures 1 and 2, reproduced from Feinstein
(2003).
Figure 1: Average rank of test scores at 22, 42, 60 & 120 months, by SES of parents
months
Dotted lines represent intervals of two standard errors. The definition of categories with sample
observations are as follows: High SES - Father in professional/managerial occupation and mother
similar or registered housewife (307 obs.) Low SES - Father in semi-skilled or unskilled manual
occupation and mother similar or housewife (171 obs.) Medium SES - Those omitted from the high and
low SES categories (814 obs.) Thus, children whose mothers were housewives were categorised by the
SES of fathers.
Figure 2: Average rank of test scores at 22, 42, 60 & 120 months, by schooling of parents
months
Dotted lines represent intervals of two standard errors. High schooling denotes families where both
parents have A Level or higher (474 obs.) Low schooling denotes families where neither parent has
qualifications (226 obs.) Medium schooling denotes those omitted from the high and low categories
(592 obs.)
Source: Feinstein, L. (2003) Economica, 73-98.
This suggests that the same common features underlay the attainment gap however
one stratifies parental background. Important gradients also exist in terms of other
aspects of family background such as family structure, income, neighbourhood or age
of mother (Haveman & Wolfe, 1995; Hobcraft, 1998, Hobcraft, 2003). Yet because