The Persistence of Female Political Power in Africa, with Siwan Anderson (University of British Columbia), Sophia Du Plessis (Stellenbosch University), and James A. Robinson (University of Chicago) • May 2025

Abstract: Research on female political representation has tended to overlook the traditional role of women as leaders across many societies. Our study aims to address this gap by investigating the enduring influence of historical female political leadership on contemporary formal political representation in Africa. We test for this persistence by compiling two original datasets: one detailing female political leadership in precolonial societies and another on current female representation in local elections. Our findings indicate that ethnic groups historically allowing women in leadership roles in politics do tend to have a higher proportion of elected female representatives in today's formal local political institutions. We also observe that institutional, rather than economic, factors significantly shape the traditional political influence of women. Moreover, in accordance with historical accounts, we uncover evidence of a reversal of female political power due to institutional changes enforced by colonial powers.

Project website: cournot.sun.ac.za/fppssa

Writing Style as Cultural Change, with Camilo García-Jimeno (Federal Reserve Bank of Chicago) • May 2025 (First version November 2024)

Abstract:We study how network structure and peer influence shape cultural change by tracking writing styles in economic theory from 1970 to 2019. We estimate a discrete‐choice model where individual preferences, peer effects, and co‐author bargaining drive gendered pronoun choice, leveraging variation in each author’s feasible co‐author network as a source of exclusion restrictions. Estimation reveals a profession of conformists with strong peer influence: when an author's peers shift from 20 percent to 70 percent feminine-only pronoun use, the author's odds of adopting the feminine form more than double. Counterfactual simulations show that absent external societal trends, the early masculine norm would have persisted; however, once those pressures appeared, peer influence magnified their impact boosting long-run stylistic diversity. We also find that homophily in co-authorship sustained writing style diversity by allowing minority preferences to express freely. Demographic shifts—entry of women and new cohorts—did not initiate cultural change but accelerated it once under way by amplifying peer effects. Cultural change depended less on demographic turnover than on how newcomers rewired the network structure, turning peer influence from an initial drag into the engine of norm evolution.

Mentorship and the Gender Gap in Academia, with Anders Kjelsrud (Oslo Metropolitan University) • May 2025 (First version December 2024)

Abstract: This paper examines how the absence of female professors affects graduates from top-50 U.S. economics Ph.D. programs. We leverage quasi-random variation in sabbatical timing and detailed data on advisor relationships and career outcomes. When a female professor goes on leave, third-year female Ph.D. students become 19 percent less likely to secure academic placements and publish 31 percent fewer early-career papers. Remarkably, male students in the same cohort experience corresponding gains, fully offsetting the losses of their female peers. We parse publication effects from placement outcomes. Treated women continue to publish less even after accounting for placement, while men's publication gains appear entirely driven by improved placement. Gender homophily in mentorship—female students are 51 percent more likely than male students to have female advisors—helps explain these patterns, along with zero-sum dynamics in the junior job market. Adding one senior female professor to each top-50 department would close one-third of the assistant professor gender gap at top-25 schools.

Immigration and Innovation: Lessons from the Quota Acts, with Petra Moser (NYU Stern) and Shmuel San (Hebrew University of Jerusalem) • Revision requested at Econometrica.

Abstract: In the 1920s, the United States introduced immigration quotas to limit arrivals from Eastern and Southern Europe (ESE). Intended to discourage low-skilled immigration, the quotas exempted scientists and students. We use biographical data on more than 80,000 American scientists to investigate the quotas' effects on innovation. These data show that the quotas discouraged ESE-born scientists from studying and working in the United States. To investigate effects on innovation, we use the world-wide universe of publications to identify research fields in which ESE-based science was prominent before the quotas. Difference-in-difference analyses show that, after the quotas, US innovation experienced a large and persistent decline in ESE fields. A decomposition exercise reveals that the quotas reduced innovation by lowering the productivity of incumbent scientists and by replacing immigrants with less productive natives. As US scientists produced fewer innovations in ESE fields, Canada gained relative to the United States.

Media coverage: New York Times, Washington Post, WSJ, Behavioral Scientist, and Marginal Revolution.

From Winning an Election to Political Engagement • 2020

Abstract: How do people react when the candidate they supported wins or loses an election? I explore the consequences of having supported a winning (or losing) candidate on future political participation. I present two empirical patterns: 1) there is a positive relationship between county vote share to the winning party and future voter turnout in that county, 2) there is no significant change in party composition of votes. This suggests that increased political participation arises from both supporters and opponents. I provide evidence for both an "individual" and "community" effect. While winning voters are encouraged by the victory, losing voters appear galvanized by being surrounded by winners. These offsetting mechanisms highlight the importance of peer dynamics in shaping political engagement.