PRELIMINARY

This paper analyzes the degree to which flextime reduces fertility-related career interruptions. In particu-lar, I ask whether women with flextime return to work sooner and remain employed when they have young children. I quantify the resulting reduction in the earnings penalty from periods of non-employment due to child-care responsibilities. To answer this question, I develop a structural dynamic discrete choice model for the fertility and labor supply decisions of married and cohabiting women. The model allows flextime to directly affect preferences, the arrival rate of job offers and offered wages. I estimate the model using a sample drawn from the National Longitudinal Survey of Youth 1979. The estimates suggest a small positive willingness to pay for flextime in a full-time job for just over half of women with no children. Moreover, a woman values flextime more strongly as her number of children increases or if she has an infant. Her willingness to pay for flextime if she has one child rises to between 1.3 and 6 % of full-time earnings. Jobs with flextime are also relatively scarce: only 37 % of job offers are for full-time jobs with flextime. A counterfactual experiment shows that if flextime were available to all women with infant children, fertility would be largely unaffected while women would work re-allocate 2 quarters of full-time work into jobs with flextime. Although participation and total full-time work experience remain largely unchanged, realized earnings and the present discounted value of utility between marriage and age 40 each increase by 1.0%. This welfare improvement is equivalent to around 13 % of annual full-time earnings