Abstract:
Background: Many refugees arrived in Germany in 2015 and 2016. At the same time, anti-refugee attitudes among Germans increased. This indicates an association between immigration and attitudes. Now, ten years later, renewed public concern about immigration - while not many immigrate to Germany - highlights the need to identify factors that shape attitudes towards refugees. Aim: Thus, I tested whether individuals living in counties with high refugee shares and large positive changes in refugee shares tend to hold more negative attitudes towards refugees. Data: The SOEP and the INKAR provided representative individual-level and countylevel panel data for 2016, 2018, and 2020 from Germany. The target population consisted of adults living in Germany who had not moved counties and had not been refugees in 2015 or after. After listwise deletion, 30,266 individuals with 61,444 observations across 398 counties remained for analysis (MAge = 53.07 years, %Men = 49). Design: An average five-item score of expected cultural and economic refugeeimmigration risks or opportunities formed the outcome variable. Apart from regional-level variables, several individual-level predictors were included in longitudinal and cross-sectional linear multi-level regression models. Findings: The longitudinal analysis indicated no relationship between regional refugee share or change in refugee share and individual attitudes towards refugees. Although the coefficients pointed in the expected direction, they were very small and only marginally statistically significant, with Ø11 = 0.023, 95%CI [0.001,0.044] for the share variable and Ø12 = 0.011, 95%CI [-0.002,0.024] for the change variable. The cross-sectional analysis showed that solely in 2020, the share variable was statistically significant, albeit marginally and unexpectedly negative. Thus, the small longitudinal effect was not stable over time. Instead, education and whether someone had been socialized in East or West Germany were the strongest predictors of attitudes towards refugees. Conclusions: Negative attitudes towards refugees are independent of the real regional share and change in the share of refugees. This indicates that restricting immigration would not reduce public concern.