Research

Can shared social identities help overcome online political divides? We investigate this question through a field experiment with 4,620 unique Twitter users conducted over six months during the 2022 Brazilian elections. Although both political congruence (supporting the same candidate) and social non-political congruence (rooting for the same football team) increase follows and reduce blocks, the positive effect of shared social identity weakens substantially when political identity information becomes available. The effect of political congruence remains strong even after the election and is unaffected by the Brazilian national team's positive results during the 2022 FIFA World Cup, despite the team being a quintessential national symbol. Text analysis of live-streamed tweets of Brazilian nationals during the tournament suggests that this shared national experience failed to reduce political polarization in our setting because polarization had extended to the players themselves. Overall, our results indicate that political polarization can undermine the potential of other shared identities to reduce political divides and foster social cohesion.

AEA RCT Registry

Media Coverage: Esto no es Economía

We study the impact of Black mayors’ election in Brazil on Black students’ educational attainment. Using a regression discontinuity design on close elections, we find that Black students from municipalities where Black candidates won are more likely to enroll in the National High School Examination, attend universities, and graduate. We find suggestive evidence that students' aspirations play a role: secondary/tertiary education is not a mayor’s primary responsibility; Black mayors do not perform better in policies that affect our outcomes; and effects are strong for Black students from both public and private schools, while weaker for White students from public schools.

Media Coverage: Foco Economico

This paper documents discrimination in the formation of professional networks among academic economists. We created human-like bot accounts that claimed to be PhD students in economics, differing in gender (male or female), race (Black or White), and university affiliation (top- or lower-ranked). The bots randomly followed users who belong to the #EconTwitter community. Follow-back rates were 12% higher for White students compared to Black students; 21% higher for students from top-ranked universities compared to accounts of lower-ranked institutions, and 25% higher for female compared to male students. The racial gap persisted even among students from top-ranked institutions.

AEA RCT Registry

Media Coverage: Promarket; Folha de São Paulo