Discrimination and Immigration: Field Experimental Evidence from Austria

We study ethnic discrimination by private firms and public institutions in Austria. We send out more than 25,000 e-mail inquiries to the universe of communities with less than 100,000 residents, randomizing the name of the sender as either local or foreign. Differences in response rates are used to create a nation-wide index of discrimination. We document several empirical facts. First, discrimination is widespread and exists in virtually all branches and public institutions. Second, discrimination is substantially lower in the public than in the private sector (10 versus 17 percentage point differences in response rates). Third, the signaling of social status via spelling mistakes and PhD titles does not affect response rates, suggesting that discrimination is tasted-based rather than statistical. Fourth, the magnitude of discrimination is unrelated to economic and socio-demographic factors and also to vote shares for the far-right. Fifth, the index does significantly correlate with average attitudes towards foreigners in a community as measured by survey instruments. Additionally, we exploit the random allocation of refugees during the 2015/16 crisis to estimate the effect of inter-group contact on discrimination. We find no evidence that contact to refugees affects discrimination. The large sample thereby allows us to rule out economically meaningful effect sizes. Instead, we find evidence for a sorting effect: Foreigners seem to sort into communities with more positive attitudes towards migrants.