Working papers

Gender Diversity Improves Academic Performance

Does gender diversity benefit students? I answer this question with a field experiment in which 2,580 university students are randomly assigned to 645 study groups. The results show that group gender diversity significantly improves course performance: Being assigned to a gender-balanced group instead of a homogeneous one raises final grades by 0.14 standard deviations. Exploring the mechanisms, I find that students in more-diverse groups report increased social interactions and improved well-being. Gender diversity also reinforces gendered roles in group learning: While women ask more questions, men explain more. My findings highlight the value of increasing gender diversity in higher education.

Information-Optional Policies and the Gender Concealment Gap 

(with Christine Exley, Raymond Fisman, Judd Kessler, Louis-Pierre Lepage, Corinne Low, Xiaomeng Li, Mattie Toma, and Basit Zafar) 

We analyze data from two universities that allowed students to replace a letter grade with “credit” on their transcript. At both schools, we observe a significant and substantial gender concealment gap: women are less likely than men to conceal grades, particularly grades that would harm their GPA. This gender concealment gap produces differential GPA gains from the policy with men benefiting nearly 50% more than women. Additional complementary data, including surveys and experiments with students and employers, suggest why women may conceal less: women may expect observers to have more negative inferences about their concealed grades.

Disconnecting Women: Gender Disparities in the Impact of Online Instruction 

(with Ulf Zölitz and Uschi Backes-Gellner)

This paper experimentally studies gender disparities in the impact of online instruction. We randomly assign 1,300 university students to varying proportions of online versus in-person lectures in different courses. The results show that increased exposure to online instruction significantly reduces women’s academic performance but has no impact on men. The negative impact on women is concentrated in math-intensive courses and leads to decreased educational attainment in the longer run. We also find that women hold a lower preference for online lectures and evaluate them as less informative, interactive, and accessible. Our findings highlight that introducing online education disproportionately disadvantages women.

Publications

Peers Affect Personality Development (with Ulf Zölitz) 

Accepted, Review of Economics and Statistics

Lowering the Playing Field: Discrimination through Sequential Spillover Effects (with Judd Kessler and Corinne Low)

Forthcoming, Review of Economics and Statistics

Access to Pensions, Old-Age Support, and Child Investment in China (with Albert Park)

Forthcoming, Journal of Human Resources 

Selected work in progress

Father Involvement and Family Well-Being - Fieldwork in progress

(with Anne Brenøe, Pietro Biroli, and Victoria Baranov)

Gender Differences in Salary Requests - Analysis in progress

(with Ryotaro Hayashi and Norihiko Matsuda)