Distal family factors
5.4.2 The effects of prior parental education on teenage motherhood
Education affects the timing for women to become mothers through two main
channels related to those for the effect on family size. First, education increases the
opportunity cost of having children. Women with higher levels of education spend
longer in schooling and delay marriage and childbearing. High educational attainment
could increase future earnings and subsequently increase the opportunity cost of
having children. Secondly, education increases women’s agency, i.e. women’s ability
or sense of power to take control of their lives, empowering them over the choice of
fertility, partly through effects on self-esteem and aspirations (Hammond, 2002), but
also through changes in life possibilities. This may lead many women to delay child
rearing into later adulthood.
Empirical studies show that women with low levels of educational qualifications tend
to have children younger than their better educated counterparts (Rowlingson &
McKay, 1998). Statistics from the UK Labour Force Survey show that less than a
third of women with degrees had children by the age of 30 compared to four fifths of
women with no qualifications. The correlation is clear but this may be driven by a
number of underlying causal processes, which make it problematic for empirical
analysis to unpack the causal relationship between education and fertility.
The main difficulty in estimating the causal effect of education is the reverse causality
of fertility on education (Hobcraft, 1998). The presence of a child could prevent
mothers attending school and, consequently, decrease the likelihood of high school
completion; therefore fertility causes low educational attainment. In order to deal with
the problem of reverse causality, Hobcraft estimates the effect of early educational
tests scores on the likelihood of becoming a teenage parent, using normalised tests of
educational attainment at seven, eleven and sixteen added together into a single
variable. For both males and females, the odds of becoming a young parent - either a
father before the age of 22 or a teenage mother - are more than three times higher for
children attaining the lowest reading and maths test scores than children with the
highest test scores. However, this is an effect of low cognitive attainment or ability
not of educational participation.
Ermisch and Pevalin (2003) investigate the family background and childhood factors
that are associated with teenage pregnancy using two types of longitudinal datasets:
the NCDS and the BHPS. The age of the mother at the time of birth as the mother’s
education have strong effects on the likelihood of becoming a teenage mother even
after controlling for a large range of child specific variables (hence reverse causality)
and family variables later in childhood. Their results show a consistent association
between low parental education and high likelihood of teenage pregnancy both in the
NCDS and in the BHPS.
Another problem that remains in estimating the causal relationship of education on
teenage parenthood is the role of unobservable factors that affect both education and
mother’s age, for example labour market ambition. Women with high levels of
ambition tend to both choose higher schooling and delay childbearing, leading to an
68