Research

• World War II, the Baby Boom and Employment: County Level Evidence, with Abel Brodeur

Journal of Labor Economics, Apr. 2022, Vol. 40, Issue 2: p. 437–471:

This paper examines the impact of male casualties due to World War II on fertility and female employment in the United States. We rely on the number of casualties at the county-level and use a differences-in-differences strategy. While most counties in the U.S. experienced a Baby Boom following the war, we find that the increase in fertility was lower in high casualty rate counties than in low casualty rate counties. Analyzing the channels through which male casualties could have decreased fertility, we provide evidence that county male casualties are positively related to 1950s female employment and household income.

• Collective Memory and National Identity Formation: The Role of Family and the State, with Joanne Haddad and Bjorn Brey

State-led repression of minority identities is a well-documented phenomenon, yet its long-term effects on national identity remain understudied. We examine how the Soviet state-induced famine (1932–33) shaped contemporary Ukrainian national identity through both vertical (familial) and horizontal (community/state) transmission. Using newly geocoded micro-level survey data, we find that individuals from high-famine-exposure areas are more likely to identify as Ukrainian. We document that under Soviet rule, family networks preserved identity, while church closures weakened community transmission. After Ukraine's independence, we find that state-led remembrance efforts, such as Holodomor memorials, revitalized horizontal transmission. Our findings highlight how repression and commemoration shape identity persistence, with implications for the pre-war trade relations and ongoing Ukrainian-Russian conflict.

• Job Market Stars: Statistical Significance and Academic Hiring in Economics, with Abel Brodeur and Marco Musumeci

This study investigates recruitment bias in the academic job market. We examine the link between statistical significance and placement outcomes for 150 empirical job market papers from 2018-2021. Marginally significant results are associated with higher academic placement likelihoods, suggesting a preference for statistical significance. This bias may incentivize questionable research practices. The COVID-19 pandemic intensified these dynamics, with marginal significance gaining greater weight in hiring decisions during competitive periods. Evidence of publication bias indicates recruiters may use statistical significance to predict future publications, shaping hiring decisions. These findings highlight systemic challenges to research integrity in economics.

• Dynamic Effects of Affirmative Actions: Experimental Evidence, with Gary Charness, Ramón Cobo-Reyes, Simone Meraglia, and Ángela Sánchez

This paper examines how Affirmative Action (AA) policies affect workforce sorting, productivity, and firm performance using a lab experiment. Participants chose between firms with tournament-based pay or piece-rate compensation, with some treatments introducing AA in one of the competitive firms. The study finds that AA significantly increases female participation in competitive environments without lowering male engagement or overall productivity, thereby closing the gender gap in tournament selection. However, AA firms experience mild challenges in retaining top talent. The policy also leads to gender- and productivity-based sorting: highly productive women tend to prefer AA firms, while highly productive men favor non-AA competitive firms. Overall, AA boosts gender diversity in competitive settings without harming efficiency but may lead to gender-based workplace clustering.

• Strategic Timing in the Appearance of Scandals 

[New version coming out soon] 

This paper relies on a novel dataset on politicians’ misconducts and examine the timing of revelation of these misconducts in the two-party political system of the United States. In order to minimize negative publicity, politicians may strategically manipulate the timing of uncovering their own unpopular actions to coincide with other important events that distract the mass media and the public. I examine the time when misconducts were uncovered and became a scandal, and the intensity of its coverage by local newspapers. I find that evidence of strategic timing is present only for predicted newsworthy events and for politicians who were still in office at the time of the scandal appearance. I also examine the heterogeneity of political parties in the timing of revealing scandals and find that republican politicians seem to time the appearance of their sex and political scandals to minimize negative publicity.