In earlier work on the cohort in Scotland Joshi and Wright (2004) showed a consistent
association of age at motherhood with large range of variables recording education,
partnership, employment, occupation, income, neighbourhood, housing, the plannedness
of the cohort pregnancy, infant feeding, mother’s mental health and subjective well-
being. This paper documents, in greater detail, how the disadvantages of the earlier
starters also apply in the rest of the UK, and provides an examination of degrees of delay
beyond age 30, finding little evidence of advantages accumulating when motherhood is
postponed beyond a certain point.
After presenting cross tabulations in which a number of variables are ranked by age at
motherhood, this paper undertakes two strands of multivariate analysis. The first strand
considers some observed variables, which may help to explain the age at which the
mothers had their first child. The second strand considers age of motherhood as a
predictor of the current circumstances of the mother and her child(ren) when the cohort
member was nine months old. Selected outcomes studied are: current partnership status,
partner’s earning status, own earning status, on means tested benefits, equivalised
household income, life satisfaction and depression. We control first for the identifiable
set of characteristics used in strand 1 which we treat as pre-dating the entry to
motherhood, and then include a number of correlates describing current or intervening
circumstances, whose causal connection with the age at motherhood and the outcome is
less easy to disentangle. These correlates include educational attainment and current
location, and as the models build up, some of previously analysed outcomes. Thus