Working Papers

(Received Best Third Year Paper Award at NYU and Best Paper Prize at 2023 HILDA Survey Research Conference)

  • This paper studies how fertility and children reshape couples’ relocation decisions and their gendered consequences for geographic job mobility, wages, and welfare. Using Australian panel data, I document that before children, both spouses’ preferences equally predict long-distance moves; after the first birth, husbands’ preferences dominate regardless of relative education and income, and the gender wage gap widens following relocation. To interpret these patterns, I develop a dynamic collective household model with endogenous fertility, joint job search, and bargaining over relocation under limited commitment, in which decision weights respond endogenously to spouses’ outside options. The model highlights two mechanisms: children raise the payoff to specialization and weaken the primary caregiver’s outside option, tilting decisions toward the male partner’s preferences and gains. I validate these mechanisms using a quasi-experimental design based on a 2015 reform to Australia’s Family Tax Benefit, which tightened eligibility for single-earner couples for a benefit paid directly to the non-working spouse. Estimates from the model show that fertility is a key margin shaping couples’ geographic job mobility and that bargaining frictions induce moves that disproportionately lower women’s wages and welfare. Eliminating this friction would not reduce aggregate mobility but would instead reallocate moves to those with more equal gains. Finally, I show that policies ignoring these internal dynamics—such as relocation subsidies—amplify within-household inequality, whereas family policies that strengthen women’s bargaining position—such as childcare subsidies—promote more equitable moves without sacrificing mobility.

(with J. Albrecht, P-A Edin, R. Fernández, P. Thoursie, S. Vroman), NBER Working Paper. (Submitted)

  • The distribution of parental leave uptake and childcare activities continues to conform to traditional gender roles. In 2002, with the goal of increasing gender equality, Sweden added a second “daddy month,” i.e., an additional month of pay-related parental leave reserved exclusively for each parent. This policy increased men’s parental leave uptake and decreased women’s, thereby increasing men’s share. To understand how various factors contributed to these outcomes, we develop and estimate a quantitative model of the household in which preferences towards parental leave respond to peer behavior. We distinguish households by the education of the parents and ask the model to match key features of the parental leave distribution before and after the reform by gender and household type (the parents’ education). We find that changed incentives and, especially, changed social norms played an important role in generating these outcomes whereas changed wage parameters, including the future wage penalty associated with different lengths of parental leave uptake, were minor contributors. We then use our model to evaluate three counterfactual policies designed to increase men’s share of parental leave and conclude that giving each parent a non-transferable endowment of parental leave or only paying for the length of time equally taken by each parent would both dramatically increase men’s share whereas decreasing childcare costs has almost no effect.

(with K. Babiarz, G. Miller, T. N. Peng, C. Valente), Center for Global Development Working Paper. Media Coverage: The Economist, Malay Mail

  • Although family planning programs can improve women’s welfare directly through changes in realized fertility, they may also have important incentive effects by increasing parents’ investments in girls not yet fertile. Exploiting the staggered implementation of family planning programs in Malaysia during the 1960s and 1970s among girls of varying ages, we study these potential incentive effects, finding that family planning may have raised raise girls’ educational attainment substantially. We also find that these early investments are linked to gains in women’s paid labor at prime working ages and to greater support for women’s elderly parents (a marker for women’s bargaining power within the household). Notably, these incentive effects may be larger than the direct effects of family planning alone.

(with K. Babiarz, G. Miller, T. N. Peng, C. Valente), Center for Global Development Working Paper

  • There is longstanding debate about the contribution of family planning programs to fertility decline. Studying the staggered introduction of family planning across Malaysia during the 1960s and 1970s, we find modest responses in fertility behavior. Higher (but not lower) parity birth hazards declined by one-quarter—but imply only a 5 percent decline in the overall annual probability of birth. Age at marriage rose by 0.48 years, but birth spacing conditional on this did not otherwise change. Overall, Malaysia’s total fertility rate declined by about one quarter birth under family planning, explaining only about 10 percent of the national fertility decline between 1960 and 1988. Our findings are consistent with growing evidence that global fertility decline is predominantly due to underlying changes in the demand for children

Work in Progress

Increasing Returns to Social Skills and Long Hours: Implications for the Gender Gap in Occupation and Pay

  • This paper studies how the rising importance of social-skill–intensive work and the expansion of long hours and overtime jointly shaped the evolution of gender gaps in occupations and pay in the United States since 1980. Using U.S. Census/ACS, ATUS, and O*NET data, I document a large increase in employment and wage premia in social-skill–intensive occupations, driven disproportionately by women’s sorting into these jobs and accompanied by a marked decline in occupational segregation. Over the same period, these occupations increasingly adopted longer and more irregular hours and non-linear (convex) pay schedules, with rising premiums for long hours and steeper penalties for working less than the occupation-specific average. This combination creates a trade-off for women between leveraging their comparative advantage in social skills and facing career penalties for not meeting intense hours requirements. To formalize this mechanism, I develop an extension of a Roy model of occupational sorting that incorporates gendered comparative advantage in social skills, endogenous time allocation through heterogeneous preferences over non-work time, and occupation-specific non-convexities in the mapping from hours worked to efficiency units of labor. The model quantifies how the interaction of social skills and non-linear pay explains observed sorting patterns and accounts for both the early convergence and the subsequent stall in the narrowing of the gender wage gap.

Baby Steps? Motherhood and Occupational Change (with Lara Bohnet)

  • We study how motherhood shapes occupational choice. Using administrative matched employer–employee data from Germany, we document that more than one-third of mothers switch occupations upon re-entry into the labor market after maternity leave, and that those who switch experience larger child penalties in earnings than those who remain in their pre-birth occupation. We find that switching mothers predominantly move into occupations with shorter or more flexible hours, suggesting that occupational changes are driven by the demand for work-hour flexibility. We provide suggestive evidence that the additional earnings penalty reflects both human capital losses from horizontal moves and a shift toward lower-paying occupations on average. To quantify the channels through which occupational switching amplifies the child penalty, we develop a dynamic discrete choice model of labor supply and occupational choice in which occupations are bundles of attributes including flexibility, and human capital is occupation-specific. The model provides insights into the value of work flexibility through the lens of protecting mothers’ post-birth wages by enabling them to retain occupational tenure and human capital.

Incentives within Household (with Chris Flinn)

  • Standard household models typically assume perfectly observed actions and enforce cooperative choices, ruling out the possibility that spouses cannot observe each other’s effort. In this paper, we relax this by allowing hidden actions: spouses observe each other’s wages and hours, but not effort or idiosyncratic shocks, so earnings serve as a noisy signal. We characterize optimal sharing rules that make private consumption contingent on observed earnings, yielding incentive-compatible allocations that differ from both standard noncooperative outcomes and full-information Pareto-efficient benchmarks. The framework delivers testable implications for the joint distribution of spouses’ earnings, effort, and consumption, and we estimate the model using panel data to quantify how informational frictions shape intra-household risk sharing and labor supply incentives.