Ugo Bolletta : ugo.bolletta2[at]unibo.it,
Mathieu Faure : mathieu.faure[at]univ-amu.fr
Timothée Demont : timothee.demont[at]univ-amu.fr
Alice Fabre : alice.fabre[at]univ-amu.fr
This paper documents the evolving impact of childbearing on the work activity of mothers between 1787 and 2014. It is based on a compiled data set of 429 censuses and surveys, representing 101 countries and 46.9 million mothers, using the International and U.S. IPUMS, the North Atlantic Population Project, and the Demographic and Health Surveys. Using twin births (Rosenzweig and Wolpin 1980) and same gendered children (Angrist and Evans 1998) as instrumental variables, we show three main findings: (1) the effect of fertility on labor supply is small and often indistinguishable from zero at low levels of income and economically large and negative at higher levels of income; (2) these effects are remarkably consistent both across time looking at the historical time series of currently developed countries and at a contemporary cross section of developing countries; and (3) the results are strikingly robust to other instrument variation, different demographic and educational groups, rescaling to account for changes in the base level of labor force participation, and a variety of specification, sampling, and data construction decisions. We show that the negative gradient in female labor supply is consistent with a standard laborleisure model augmented to include a taste for children. In particular, our results appear to be driven by a declining substitution effect to increasing wages that arises from changes in the sectoral and occupational structure of female jobs into formal non-agricultural wage employment as countries develop.