Working Papers

Breaking the Bubble - The Determinants and Effects of Contact (Job Market Paper)

with Vlasta Rasocha

Abstract: We study how endogenous selection into intergroup interactions shapes social segregation and the optimal design of policies to address intergroup hostility. We conduct a field experiment in partnership with several non-profit organizations in the context of political polarization in Brazil. Our experiment recruits over 5,000 individuals through apolitical social media ads, combining incentivized measures of aversion to outgroup interactions and beliefs about their costs and benefits with causal evidence on their impact. The most hostile partisans exhibit the strongest aversion to outgroup interactions and overly pessimistic beliefs about their outcomes. We randomize participants into cross-partisan conversations, offering twice the daily minimum wage for compliance. The intervention strongly reduces hostility, with effects twice as large among participants with higher degrees of baseline hostility and larger effects among those with higher outgroup aversion. Participants' reported conversation outcomes strongly exceed predictions, indicating that pessimistic prior beliefs were largely mistaken. The conversations reduce aversion to outgroup interactions, correct beliefs about their outcomes, and reduce political extremity and labor market discrimination. Our findings suggest that endogenous segregation exacerbates cross-partisan hostility, motivating future policy interventions that target individuals with high degrees of hostility and aversion. We show that a scalable, cost-effective alternative policy intervention has similar short-run effects.

Does Contact Reduce Affective Polarization? Field Evidence from Germany (2023)

with Martin Koenen

Abstract: We analyze whether and how exposure to political opponents can impact attitudes toward political opponents (affective polarization) and extremity of political opinions (ideological polarization). We present findings from a quasi-experiment in Germany that matched 15,000 participants for a virtual one-on-one conversation with a stranger. Leveraging staggered treatment assignment, we find significant reductions in affective polarization among treated participants in both incentivized economic interactions and survey outcomes. In contrast, we do not find corresponding effects on ideological polarization, suggesting that exposure increases tolerance but not support for opposing positions.

Work in Progress

Contact on the Job

with Vlasta Rasocha and Pedro C. Sant’Anna

Funded by the Weiss Fund, J-PAL JOI Brazil, Stanford King Center, Berkeley CEGA

Know Your Place - Employment Decisions among Couples 

with Shakil Ayan, Nina Buchmann, Pascaline Dupas, and Muriel Niederle

Funded by the IGC

Hidden Identities, Stereotypes, and Discrimination

with Vlasta Rasocha

Funded by J-PAL JOI Brazil, Stanford King Center, Stanford SIEPR

The Effects of Salary Range Reporting Requirements

with Mariana Guido and Carl Meyer

Funded by J-PAL North America Social Policy Research Initiative, Stanford SIEPR

Task Allocation and Labor Market Decisions among Couples 

with Mariana Guido, Sahana Subramanyam, and Jason Weitze

Funded by the Weiss Fund, Stanford King Center, Stanford SIEPR