A longitudinal investigation of integration/multiculturalism policies and attitudes towards immigrants in European countries

Abstract

A number of recent studies find that integration and multiculturalism policies help soften anti-immigrant attitudes among the broader population. These findings, however, emerge from cross-sectional analyses and are potentially vulnerable to omitted variable bias. The analysis in this paper overcomes that limitation by adopting a longitudinal approach. This approach uses data from repeated cross-sections drawn from the European Social Survey and the European Values Survey. These data can be treated as panels in a longitudinal framework once it is recognised that the relevant variables (including the attitudes variables) can be handled effectively as country-level averages. Multi-level modelling (the default approach in existing research) is not necessary; in particular, there is no need to use individual-level control variables. In a fixed-effects analysis of country-level data, adoption of more open/accommodating integration and/or multiculturalism policies does not lead to a reduction in anti-immigration sentiment. The findings of the cross-sectional studies evidently suffer from significant omitted variable bias.