Background: Sweden is a major host nation for asylum-seeking immigrants, and residential placement of these immigrants is an important policy concern. This quasi-experimental study estimated of the impact of being placed into an “immigrant enclave” on risk of officially-recognized drug involvement (ORDI) among asylum-seeking immigrants over a 15-year period. Methods: All data come from Swedish registries. The sample consisted of (a) asylum-seeking immigrants aged 5–35 years old at arrival (N = 51,017) that were subject to a nationwide policy (enforced 1987–1991) that dispersed asylum-seeking immigrants across municipalities, and (b) native-born Swedes aged 15 and older during this same period (N = 1,040,311). Neighborhood immigrant composition was quantified using the Reardon Index; residents of “immigrant enclave” neighborhoods (n = 960) were compared to residents of all other neighborhoods (n = 2,471). Cox proportional hazards models assessed the relationship between living in an enclave and risk of ORDI, identified by national registries, through 2015. Results: Overall, 29.7% of immigrants were assigned to, and 25.5% of Swedes lived in, an enclave. Cumulative incidence of ORDI in enclaves was 6.34% as compared to 6.89% in other neighborhoods. Immigrants living in an enclave had lower risk of ORDI (Hazard ratio (HR): 0.86, 95% Confidence Interval (CI): 0.77 – 0.96). This protective association was marginally stronger in lower poverty areas. Native-born Swedes living in an enclave had higher risk of ORDI (HR: 1.05, 95% CI: 1.03–1.08), a relationship that was exacerbated by neighborhood poverty. Conclusions: Neighborhood immigrant composition is associated with risk of ORDI, with differential associations for immigrants and native-born populations. © 2019 Elsevier B.V.

Neighborhood Immigration segregation Substance abuse

social segregation immigrant quasi experimental study register human controlled study priority journal policy neighborhood asylum seeker Sweden Adolescent male female substance use Article adult quantitative analysis Child