A large-scale field experiment to disentangle sources of statistical discrimination in a social setting

This paper considers two types of statistical discrimination: individual (productivity) and collecetive (team fit). We conduct a large-scale correspondence study in 15 Latin American countries in the context of sports to test their influence on individual behavior. We send over 10,000 applications to male amateur soccer clubs and ask them to participate in a practice session. Each club receives one application, randomly varying the applicant’s origin and signals about individual and collective productivity. We find no evidence of discrimination against immigrants overall, but we observe heterogeneity that is consistent with individual statistical discrimination. Productivity signals have no significant influence.