Introduction

The increasing level of ethnic heterogeneity in Europe has spurred a debate about the relationship between ethnic diversity and income redistribution by the state. The current debate is much inspired by Alesina and Glaeser's (2004) sceptical view of the possibility of reconciling ethnic diversity and generous welfare arrangements.

The debate about the relationship between ethnic diversity and the welfare state is of particular interest for the Nordic countries, which developed their comparatively generous welfare states in a context of ethnic homogeneity. Over the last three decades, however, non-Western immigration has made the Nordic countries less homogeneous. In Denmark and Norway, issues of integration of immigrants have been high on the public and political agendas for some time, and both countries have relatively large anti-immigration parties in their parliaments. In Sweden, the immigration issue was hotly contested in the 2010 election, and the Sweden Democrats, an anti-immigration party, won 20 seats in the Swedish parliament. Anti-immigration parties are, or have very recently been, represented in the Finnish and Icelandic parliaments as well. Therefore, the issues that this chapter discusses are highly important for all the Nordic countries.

This chapter focuses on two mechanisms through which ethnic heterogeneity can have an impact on the degree of redistribution. First, ethnic heterogeneity might decrease voters’ support for redistribution and thus shift the political centre of gravity on the redistributive dimension in a Rightist direction. Alesina and Glaeser (2004) argue that voter animosity towards minorities is an important reason why the US has a much smaller welfare state than most European countries: racist attitudes directly reduce the preferred level of redistribution because redistribution disproportionably benefits immigrants, or is perceived as doing so. Using the language of Roemer et al (2007), I label this potential mechanism the ‘anti-solidarity’ hypothesis, for which there exists weak support, at best, in studies of European cross-sectional survey data (Finseraas, 2008; Senik et al, 2009; but see Eger, 2010). This chapter expands on these studies by exploring the relationship between change in xenophobia and change in support for redistribution at the aggregate level between 2002 and 2008. Examining whether these variables move together over time is obviously a better test of the ‘anti-solidarity’ hypothesis than examining the association at one point in time, especially since there are some indications that changes in the size of the immigrant population is negatively correlated with changes in social spending (Soroka et al, 2006).