Purpose
The purpose of this paper is to provide a summary of empirical research on welfare magnetism and to assess the size and scope of the welfare magnet effect on the non-EU migrants in selected immigration countries of the European Union.
Design/methodology/approach
A conditional logistic regression model with interactions is used to estimate the strength of the welfare magnet effect, while controlling for demographic characteristics of the migrants and country-specific economic indicators. Data, used for estimation, comes from the Immigrant Citizen Survey, which provides a large, representative sample of first-generation (i.e. non-EU born) migrants. Various measures of welfare generosity are tested to assure the robustness of the results.
Findings
The coefficients suggest that the welfare magnet effect is present and significant in some immigrant groups, although it can have a negative impact on location decisions in other cases. Similar results are obtained for wage and unemployment indicators.
Research limitations/implications
Results corroborate the welfare magnet hypothesis, which states that more generous welfare states should expect greater clustering of negatively-selected (i.e. lower educated) migrants. One potential limitation comes from the sample size, which does not allow for more general conclusions.
Practical implications
Heterogeneous effects of basic economic indicators in different demographic groups show that aggregate immigrant flows, used widely in the literature, can provide biased estimates of welfare magnet effect.
Originality/value
This paper adds to the available literature by using representative, recently collected data and employing a more complete list of controls in a quantitative analysis of migration decisions.