Table 11 analyses another binary indicator of low income, being below 60% of median
income threshold used in the official ‘child poverty statistic’ and for MCS data by
Bradshaw and Mayhew (forthcoming). The sample of 14,723 of those with non-missing
data is a bit smaller than table 10. The age profile of being below the 60% threshold
reaches a minimum for those entering motherhood between 31 and 33. The small
contrast with table 10 at the highest ages is within estimated margins of error. While the
antecedent variables are also predictive of being below the poverty line they make only a
modest contribution to explaining the age-at-motherhood profile in ‘poverty’. The more
recent variables in models 3 reduce the age of motherhood terms by more than half. The
extra information about the partner in model 4 makes only modest further differences, so
that the otherwise unexplained association of age at motherhood with this poverty
indicator is still significant for all age groups except for those 19-21. The association
with childhood factors is more reduced than in table 10. However the individual ethnic
differentials are larger and more robust for having income below the 60% threshold than
for claiming means tested benefits. However the ethnic area is more significant for means
tested benefits than for being below the 60% threshold. Nevertheless the message from
these two attempts to classify objective poverty is that we can say that early mothers are
more likely to be ‘poor’ than those who deferred to their mid twenties or thirties partly
because they have less education, more older children, fewer employed partners, less
employment themselves, live in less advantaged areas, are somewhat more likely to have
(or to have partners with) a longstanding illness and are more likely to belong to ethnic
minorities.
20