Research

Working papers

• Job Market Paper Worker Side Discrimination: Beliefs and Preferences–Evidence from an Information Experiment on Job-seekers, with Mehreen Mookerjee and Sanket Roy
  Abstract

  We provide novel evidence on the distribution of workers' preferences on manager's gender and their beliefs on manager's mentoring ability. We design and conduct a novel within-worker information experiment and embed it in a hypothetical job choice survey for job-seekers. We find that on average there exists an underlying preference of workers to work for female managers, willing to give up 1.3-2.2% of average annual wages. However, the lack of information on manager's mentoring ability leads both preferences and beliefs of workers to influence job choice. On average workers believe that male managers are better mentors counteracting their underlying preferences. This lack of information does not allow job choices of workers to reflect the average underlying preference to work for female managers. The distribution reveals that 60% of workers prefer to work for female managers, and in the absence of information on manager mentoring ability 62% believed male managers to be better mentors, with policy relevant heterogeneity across maternal education and majors of workers. An ex-post survey where workers beliefs are directly elicited corroborate this finding. Our results suggest that this information gap could lead to sub-optimal promotion of females to managerial positions, and firm executives who do not prefer female managers could use these information rents to generate glass ceilings for females.

  Awards: Richard E. Stockwell Dissertation Fellowship, UW-Madison. Conferences: SOLE 2021, North American Econometric Society 2021, ESA 2022, CEA 2022; Upcoming: EEA 2022, SEA 2022

• Female Inheritance Rights and Household Sanitation, with Monica (Draft coming soon.)
  Abstract

  Health hazards due to open defecation are most prominent in India. Females benefit from toilets in households more than males. In this paper we estimate the impact of increased inheritance rights of females on the presence of a toilet in the household. Daughters being usually married away to the household of the groom, available household level nationally representative data do not have all original (natal) household characteristics – which determines treatment eligibility. Under generic assumptions, we show that when the treatment is partially observed to the researcher, we can derive a lower bound on the average treatment effect in a difference in differences framework. We estimate that the policy increased the probability of the presence of a toilet in the household a woman is married into, by at least 4.3% points. We uncover conditional treatment effects by the age of the daughter at the time of policy implementation and find it to be the largest for the group of females who were the youngest at the time of policy implementation.

Selected Works in Progress

• “Preference Estimation and Welfare Loss Associated with Centralized Assignments Based on Preferred Sets” with Monica (Administrative datasets obtained. Survey data collected)
• “Optimal Place-Based Redistribution” with Jesse Gregory and Morris Davis (IRB in process. Pilots to begin)
• “Racial Gaps in Wage Growth: Discrimination, Selection and Search Frictions” (Administrative datasets obtained.)
• “Identification in Models of Discrimination”

Publications

• “Increases in shared custody after divorce in the United States.” (w/ Daniel Meyer and Marcia Carlson) Demographic Research 46 (2022): 1137-1162.
  Abstract

  This paper provides new evidence on the time trend in shared physical custody after divorce in the U.S., using eight waves of data from the Current Population Survey - Child Support Supplement. We find that the likelihood of shared custody more than doubled between divorces that occurred before 1985 and those in 2010-2014, from 12% to 28%. We show that non-Hispanic Whites and those who are more socioeconomically advantaged are more likely to have shared custody. Using more formal methods we show that the increase cannot be explained by changes in the characteristics of those divorcing; instead, we infer that this is the result of changing norms and policies that favor shared custody. Finally, this paper complements previous analyses using court record data from Wisconsin and shows that while the rate of shared custody in Wisconsin is higher than the national rate, a large increase over time has occurred in the nation as well as in Wisconsin. These changing patterns have important implications for children’s living arrangements and for the parental investments that children receive after their parents’ divorce