Essays on the economics of gender

Abstract

This thesis studies gender differences in educational and occupational outcomes. Chapter 1 examines whether the long work hours required by high-paying professions are a barrier to entry for women, who may face a tradeoff between market time demands and family formation. I study the introduction of a policy that capped the average workweek for medical residents. I find that a reduction in a specialty's weekly residency hours induced women to enter, but yielded little change in men's entry. To shed light on why women and men responded differently to the reduction in hours, I analyze the effect of the reform on residents' fertility. I find that a reduction in a specialty's weekly hours increased the specialty's female fertility rate in California, and had no effect in Texas. I discuss these results using a model in which physicians choose between career and family investments during residency, trading off long term incomes, investments in children, and leisure. In Chapter 2, coauthored with David Autor, David Figlio, Krzysztof Karbownik, and Jeffrey Roth, we use birth certificates matched to schooling records for Florida children in order to assess whether family disadvantage disproportionately impedes the pre-market development of boys. We find that, relative to their sisters, boys born to disadvantaged families have higher rates of disciplinary problems, lower achievement scores, and fewer high-school completions. Evidence supports that this is a causal effect of the post-natal environment; family disadvantage is unrelated to the gender gap in neonatal health. We conclude that the gender gap among black children is larger than among white children in substantial part because black children are raised in more disadvantaged families. Chapter 3 explores why women remain underrepresented in elective offices, by investigating whether there are gender differences in the persistence of politicians in response to an electoral loss. Using California local election returns and a regression discontinuity design, I analyze the subsequent political involvement of men and women who ran in close elections. I find that losing an initial election induces substantially more attrition among female than male candidates.