Not All Benefits Are the Same. the Welfare Magnet Hypothesis Reconsidered

International migration and in particular welfare-induced migration have always been controversial topics and the enlargements of the European Union since 2004 as well as increased migration to and within Europe intensified the debate.  Politicians express their concerns like the British Prime Minister David Cameron who stated that “Ending the ‘something for nothing’ culture is something that need to apply in the immigration system as well as in the welfare system” or the Ministers of the Interior from Austria, Germany, the Netherlands and the UK who jointly wrote a letter to the European Commission in 2013 expressing their worries over increased welfare-induced migration.

Underlying these concerns is the claim that individuals base their migration decision on the welfare generosity of their home as well as potential destination country. The academic equivalent is the welfare magnet hypothesis. According to this hypothesis, low-skilled migrants, as they are more likely to be net receiver within a redistributive welfare state, self-select into countries with high welfare generosity. At the same time, high-skilled migrants, being most likely net-payer within a redistributive welfare state, self-select into countries with low welfare generosity.

What neither the political statements nor the welfare magnet hypothesis in its rationale address is the heterogeneity of welfare generosity. This paper theoretically and empirically analyzes the self-selection of migrants emphasizing on the different dimensions of welfare generosity with a focus on the effects of unemployment insurance generosity and pension generosity. To my knowledge this is the first paper that explicitly differentiates between various dimension of welfare generosity when studying the self-selection of migrants.

Including the different aspects of welfare generosity into an extended version of the Roy model, I analyze how they influence the self-selection of migrants who differ by educational level. While unemployment generosity influences the selection ratio (i.e. the number of high-skilled over low-skilled) negatively, the results are ambiguous when it comes to pension generosity.

These theoretical claims are tested empirically using data on the bilateral stocks of migrants by educational level as well as generosity scores for pension and unemployment systems in the countries of the European Union from 1980 to 2010. The focus on migration amongst EU countries prevents that the results are mainly driven by restrictive migration policies as Intra-EU migration is not or less restricted.

By examining the various dimensions of welfare generosity as determinants of the selection ratio of migrants, this paper suggests that the self-selection reasoning of the welfare magnet hypothesis is too simplistic.