Biased beliefs about immigration and economic concerns : evidence from representative experiments

We investigate the link between biased beliefs about immigrants, economic concerns and policy preferences. Conducting representative survey experiments with more than 8000 respondents, we first document substantial biases in respondents' beliefs about the immigrant population in various domains. Exposure to different types of signals about immigrants reduces concerns about adverse effects of immigration on the welfare state. On the contrary, different types of signals offset their effects on concerns about increasing labor market competition. Employing a data-driven approach to uncover systematic effect heterogeneity, we find that prior beliefs about immigration explain conditional average treatment effects. While attitudinal change is thus more pronounced among individuals with pre-intervention biases about immigrants, education and attitudes towards cultural diversity are additional drivers of heterogeneity. Treatment effects on welfare state concerns persist in a five to eight week follow-up.