Immigration and Nationalism in the Long Run

During recent waves of immigration, support for nationalist parties has increased in many countries, but the political backlash against immigration differs strongly across regions. We identify an underlying cause for these differences by studying how local experience with immigration shapes nationalist sentiment and electoral reactions to current immigration in the long run. Our analysis draws on a natural experiment in post-war Germany, where a short-term demarcation of occupation zones led to a discontinuous and quasi-exogenous distribution of forced migrants. Across this border, the population share of migrants differed by 12 percentage points. Applying a spatial regression discontinuity design, we combine historical migration records with panel data at the municipality level for the 1925-2021 period. The results reveal a substantially weaker backlash against contemporary immigration in regions where more migrants settled in the late 1940s. This historical experience reduces the nationalist backlash by about 20 percent. High levels of immigration activate this effect over a period of at least 70 years. To study the mechanisms, we conduct a geocoded survey with a randomized experiment and open-ended questions in the study region. We find that both family history and local collective memory of successful immigrant integration contribute to these effects. The results of the randomized experiment are consistent with the natural experiment, revealing how experience with immigration can curb nationalism.