Neighborhood labeling and youth schooling paths

In 2015, the Swedish police unexpectedly started publishing their internal list identifying "vulnerable areas." While the public release of the list gave rise to a massive increase in media reports about the listed neighborhoods, often portraying them negatively, it did not lead to affect police resources or public interventions. Exploiting geocoded data on the police list, neighborhoods, and schools, we explore how the release affected the performance and sorting of students completing compulsory school and applying to high schools. Our research strategy compares the change in the outcomes of individuals living in listed neighborhoods with those in non-listed areas with comparable sociodemographic characteristics. Our main finding is that listing a neighborhood did not affect student performance but substantially increased sorting. Students from listed neighborhoods were less likely to sort into high-quality high schools and were more likely to apply for high schools within their neighborhoods. Our results are consistent with models of neighborhood labeling wherein residents in neighborhoods with a negative public image adjust their behavior in response to anticipated or actual discrimination.