Labor force participation rates of married women have increased substantially since 1950s; and
the educational gender gaps for cohorts born after 1950 have also witnessed a shrinkage and, in fact,
a reversal. In this paper, we advance a novel connection between the two trends: the decrease in the
maternal time input in the children’s education as a result of the increase in maternal employment
has a stronger negative impact on the skill production on boys than on girls. We present three sets of
evidences supporting our hypothesis. First, we find that in the US, at both the state and the individual
level, the educational gender gap is significantly and positively correlated with the LFPR of married
women in the birth state within five years after birth. Cross-country evidence also shows a strong
correlation between female LFPR of the mother generation and the educational gender gap among
their children’s generation. Second, we directly examine and find evidence of the asymmetric effects
of maternal employment on their children’s educational achievement which favors girls than boys in
the Norwegian registry data. Third, we propose a model showing that the gender asymmetric effect
of maternal employment implies that a rational and altruistic mother would work more if she has
higher ratio of girls, controlling the total number of children. We find supporting evidence for this
prediction in both the US and the Norwegian data.