Do immigrants affect crime? : evidence from panel data for Germany

The paper analyses the empirical relationship between immigrants and crime using panel data for 391 German administrative districts between 2003 and 2016. Employing different standard panel estimation methods, we show that there is no positive association between the immigrant rate and the crime rate. We assess the robustness of this result by considering the heterogeneity of immigrant groups with respect to gender, age, country of origin and - if applicable - refugee status, and study naturalized immigrants. We also take into account possible spillover effects of immigrants on criminal activities by Germans, omitted variables and spatial correlation. Furthermore, taking advantage of the panel-structure of the data set we employ an instrumental variable approach that deals with the possibly endogenous allocation of immigrants and allows for causal interpretation of the estimates. There is no evidence that immigrants increase crime.