terms (and also the effect of having left school early). These models reflect, without
completely explaining, the low employment rates of lone mothers even after government
policy sought to encourage it through measures such as the National Childcare Strategy,
Working Families Tax Credit and New Deal for Lone Parents. Note the higher
propensity of mothers in the three Celtic Countries to be employed. This contrasts with
the negative terms for partners’ employment in Scotland and Wales.
Table 10 considers one binary measure of low income: those on means tested benefits the
four benefits involved are as close as possible to those used to rank areas for the
stratification of the sample, but include benefits available somewhat higher up the income
scale than the 60% line. It covers to whole sample for which the data are available,
15,819 cases, including both lone mothers and those in partnerships. The chances of
receiving means-tested benefits fall with rising age at motherhood and reach a floor for
those aged 31 or more. Including the antecedent variables reduces the effects while
maintaining the order in model 2. Including the current variables also roughly halves
these original estimates but maintains their significance and order. The high propensity
of earlier mothers to be on benefits is partly explained by their childhood circumstances,
their lack of qualifications and employment, the poor employment and education of their
partners if they have any, and where they live. The ethnic differentials apparent in model
2 seem to be mostly accounted for by the contemporary variables, possibly the ethnic
area term.
19