Abstract
Using ESS and Eurostat data, we investigate attitudes to immigrants in the perspective of a dynamic process of belief formation. We want to verify the existence of a learning effect on personal attitudes to immigration by observing the different effects of past and recent inflows of immigrants. Furthermore, we investigate whether these can be explained as effects of stereotyping and/or by contact theory. We find evidence of a learning effect, since past flows prove not to be significant while recent flows are significant and negative. Stereotyping and contact theory partly explain personal attitudes to immigration, but they do not seem to explain the negative effect correlated to the presence of immigrants and the subsequent learning effect. Finally, we look at the interaction between migration flows and demographic and socio-economic characteristics of the population. Income is the only factor that explains the learning effect, as wealthier social groups are more averse to the presence of immigrants in their neighbourhood but also display a tendency to learn faster.