Do Parties “Playing the Race Card” Undermine Natives’ Support for Redistribution? Evidence From Europe

Abstract

In this article, we address the question of whether the policy statements of political parties with regard to migration affect the link between individual anti-immigrant sentiment and support for redistributive policies. While the effect of political parties “playing the race card” is well documented and repeatedly discussed in the American context, it has received little attention in comparative studies. We test our measurements of issue-salience with regard to migration and welfare-related matters by conducting multilevel models for a sample of 14 European countries. We also control for the potential effects of the countries’ welfare regimes—which is so far the most prominent contextual variable. Our results strongly indicate a moderating party-effect: The more parties accentuate crucial migration issues, the less general support there is for welfare programs by native anti-immigrant groups. In contrast, we find no effect of the repeatedly discussed welfare regime on this relationship, once controlled for party statements.

Keywords

welfare chauvinism redistribution migration party discourse welfare state multilevel analysis European Social Survey

Introduction

With regard to the United States, the relationship between racial attitudes and welfare support has been analyzed by a considerable number of studies (Alesina & Glaeser, 2004; Alesina, Glaeser, & Sacerdote, 2001; Gilens, 1995, 1999; Luttmer, 2001; Quadagno, 1994). The results suggest that racial prejudices of White Americans about their fellow African Americans’ work ethic can explain a lot about “Why Americans Hate Welfare” (Gilens, 1999). Large parts of the White population characterize African Americans as lazy and suspect them of abusing welfare programs. Combined with the belief that non-White minorities are grossly overrepresented among welfare recipients, White Americans tend to see the welfare state as something targeted toward others, while they are left in the role of mere payers (Gilens, 1995). The political and economic consequences of this race-coding of welfare are serious. Alesina and Glaeser (2004) stress the view that the comparatively low level of redistributive policies in the United States is mainly caused by the ethnic diversity of the society; a similar line of argument is put forward by Gilens (1999), Lipset and Marks (2000), and Quadagno (1994).

From a European point of view, this has raised the question of whether the relationship constitutes an “American exceptionalism” (Lipset, 1996) or whether it should be regarded as a harbinger of the continent’s own future. Will the comparatively generous levels of European welfare programs be cut back as growing number of immigrants exacerbate anti-immigrant sentiments and in this way undermine natives’ support for welfare? It is precisely this question we wish to address in this article.

So far, the comparative studies addressing the relationship between welfare support and anti-immigrant attitudes in Europe have produced two main findings. On one hand, in Europe, there appears to be a relationship between anti-immigrant attitudes and welfare support that reflects the race-oriented U.S. pattern. Immigrants—and especially non-Western immigrants—are perceived by many native Europeans as “not belonging to us,” less willing to work hard, and therefore less worthy to be supported by welfare benefits (Crepaz, 2006; Larsen, 2011; Senik, Stichnoth, & Straeten, 2008). On the other hand, the relationship between anti-immigrant prejudices and welfare criticism has been found to vary considerably among European countries and is potentially moderated by the political, economic, and institutional context. Following the literature on popular attitudes toward welfare programs among different welfare regimes (Brooks & Manza, 2007; Crepaz, 2008; Larsen, 2008; Van Oorschot, 2006), discussion has considered whether the effect of anti-immigrant attitudes on welfare support is strengthened in more U.S.-like, liberal regimes and is less conflict driven among the more universal, social-democratic regimes which might be able to “absorb immigrants with less groaning and moaning from their native populations” (Crepaz & Damron, 2009, p. 456; for a similar argument, see Eger, 2010; Finseraas, 2008; Larsen, 2008, 2011; Mau & Burkhardt, 2009).

In contrast to the welfare regime, the effect of another potentially relevant variable in this context, the party discourse on migration and the welfare state, has only recently been addressed in comparative studies even though it has long since attracted scholars’ attention. As Alesina and Glaeser (2004) have convincingly shown in the U.S. context, “playing the race card”—that is, the instrumental use and activation of natives’ racial prejudices to undermine redistributive policies—might be a valuable strategy for political actors to mobilize political support. For Europe, a similar pattern has been discussed extensively in the literature on Extreme Right Parties (ERP; Arzheimer, 2008; Arzheimer & Carter, 2006; Kitschelt, 1995; Mudde, 2007) but remains largely unaddressed in studies of immigration and welfare support (an exception is Larsen, 2011). We wish to contribute to this ongoing debate by analyzing the effect of party discourse on migration-related issues.

We test these complex interaction effects among individual attitudes, party discourse, and welfare regime, by applying a multilevel model to a sample of 14 Western European countries using data from the fourth round of the European Social Survey (ESS; 2008). Our results suggest that the relationship between natives’ anti-immigrant attitudes and their support for redistributive policies is strongly moderated by the discourse of political parties. Once controlled for this party-effect, we find no evidence that the welfare regime does influence the connection between anti-immigrant attitudes and support for redistribution, a result that challenges most of the recent literature.

Racial Diversity and Welfare Support in the United States

Scholars interested in the history and recent development of the U.S. welfare state have long since pointed to the ethnic heterogeneity of American society as one crucial explanatory variable for the residual level of welfare spending. This literature might be divided into studies addressing the overall amount of U.S. social spending (where ethnic heterogeneity is the explanans) and studies focusing on the individual welfare support of Americans (with racial attitudes as the explanatory variable). While both strands of literature are thematically closely related, the theoretical arguments behind the observed relationship(s) are somewhat different.

Studies claiming to explain the comparatively low level of U.S. social expenditure with the societies’ ethnic heterogeneity are theoretically linked to the wider literature on welfare spending and, therefore, to its dominant power-resource approach (Huber & Stephens, 1993; Korpi, 1983; Korpi & Palme, 2003). For this, redistributive policies and most prominent welfare spending are results of the historical strength of the political left, including socialist and social-democratic parties as well as trade unions. The theoretical link between ethnic diversity and welfare spending is seen as being mediated by the working classes’ potential for collective action. In an early comment, Karl Marx expressed his concerns that the mainly migrant U.S. working class would not be able to build up the class-consciousness necessary for socialist political action. More recent commentators on American history state that ethnic heterogeneity hampers the potential to form a class-based feeling of solidarity, preventing the rise of left-wing political parties based on this solidarity, and that this eventually results in less redistributive policies (Lipset & Marks, 2000; Stephens, 1979). At the same time, right-leaning parties might be able to undermine class-based solidarity by stressing such contradictory identities as nationality or race to gain electoral support. Quadagno (1994) and Alesina and Glaeser (2004) offer numerous examples of this strategy in U.S. history. Even if voters are mainly motivated by these noneconomic, cultural issues and not directly by welfare antipathy, their vote for parties on the right might result in less redistribution (Soroka, Banting, & Johnston, 2003).

Scholars interested in the microfoundations of redistributive policies analyze the impact of race-related attitudes and welfare support on the individual level (Alesina & Glaeser, 2004; Fox, 2004; Gilens, 1995, 1999; Luttmer, 2001). This research strategy seems justified, as welfare programs, once in place, are particularly resistant to policy reform and therefore their spending level does not usually change greatly over time (Pierson, 1996). Summarizing the early literature on the link between racial prejudices and welfare support, Luttmer (2001) not only concludes that a negative correlation between them has been clearly identified for the United States but he also remarks that “the evidence on the forces behind such relationships is limited” (p. 501). More implicitly, the common theoretical starting point of most studies seems to be a social-psychological concept of “racial based group loyalty” (Luttmer, 2001, p. 501), stating that individuals belonging to the same group will in general be more willing to support fellow group members than out-group members. Van Parijs (2004) argues that different traditional points of departure complicate the evolution of a common definition of social justice and social solidarity along distinct group boundaries. In addition, ethnic diversity is assumed to affect economic solidarity negatively by hampering intergroup communication and depressing social trust (see also Putnam, 2007; Soroka et al., 2003).

A much more political explanation for the theoretical link between ethnic diversity and welfare support is given by Gilens (1995). In his study of support for different U.S. welfare programs, he identifies racial prejudices as the strongest explanatory variable among White respondents, controlling for an impressive number of potential variables. According to Gilens (1995, 1999), most White Americans characterize minorities (and especially African Americans) as lazy and unwilling to work and suspect them of abusing welfare programs. Combined with the belief that African Americans are grossly overrepresented among welfare recipients, White Americans tend to see the welfare state as something targeted toward others, leaving them in the role of mere payers for these unjustified demands (Gilens, 1995). Continuing with Gilens’s argument, the effect of racial attitudes on welfare support is not a sociological necessity but relies on “the stereotype of the lazy black” (Gilens, 1995, p. 1011), which is deeply rooted in U.S. public discourse and continuously seen as salient by politicians and the media (Gilens, 1999). Alesina and Glaeser (2004), Neubeck and Cazenave (2001), and Quadagno (1994) also point to the strategies of political parties to explain the link between racial diversity and welfare support by offering several case studies from U.S. history. These convincingly show how conservative politicians activated and strengthened racial prejudices to undermine political support for redistributive policies.

Migration and Welfare Support in Europe

As Banting (2005) rightly states, the American case has become a kind of “master narrative” (p. 107) for analyzing the link between ethnic diversity and the welfare state. Consequently, the U.S. experience has motivated a rapidly increasing number of international-comparative studies, addressing the warning of Alesina and Glaeser (2004) that “Europe’s new immigrant-based heterogeneity may eventually push the continent toward more American levels of redistribution” (p. 175). So far, these studies have produced two major findings.

First, also in Europe, citizens with negative attitudes toward members of different ethnic groups are on average more critical of welfare. However, as European societies are far less divided by race, natives perceive non-Western immigrants as the relevant out-group. As Larsen (2011) reports after comparing the attitudes of American Whites toward African Americans with those of Europeans toward non-Western migrants, both in-groups rarely differ with regard to their perceptions of the out-groups’ work ethic, are equally hesitant to accept out-group members as neighbors, and oppose the idea of improving the conditions for out-group members by public spending.

While this general finding of a negative impact of attitudes critical to welfare support for immigrants is echoed in several other comparative studies (Banting & Kymlicka, 2006; Crepaz, 2006; Crepaz & Damron, 2009; Finseraas, 2008; Mau & Burkhardt, 2009; Senik et al., 2008; Van Oorschot, 2006), it has to be stressed that these authors differ greatly in their portrayal of anti-immigrant sentiment. This comes as little surprise, as antimigrant sentiment is a complex phenomenon, consisting of several related but distinct beliefs (Arzheimer, 2008; Finseraas, 2008; Mughan & Paxton, 2006). In our analysis, we will therefore distinguish between three repeatedly discussed facets of natives’ anti-immigrant sentiment: the perceived economic and cultural threats associated with immigration and suspicions about immigrants’ welfare abuse. We expect cultural threat and welfare abuse to have a negative impact on support for redistribution. On one hand, respondents feeling threatened by cultural heterogeneity might not be willing to develop a feeling of economic solidarity with nonnative citizens (see Boeri, 2006; Finseraas, 2008; Freeman, 1986). In addition, respondents thinking that migrants receive more welfare benefits than they deserve would be less willing to support further redistribution, as has been extensively discussed for the United States (Alesina & Glaeser, 2004; Gilens, 1995, 1999). On the other hand, we expect respondents fearing a negative economic impact of immigration might call for more redistribution to compensate for their expected losses (Finseraas, 2008; Soroka et al., 2003).¹

The second major finding of comparative studies is that there is significant variation in the relationship between anti-immigrant sentiment and welfare support among the countries of Europe and that these differences might be explained by the country-specific socioeconomic and institutional context. Relevant factors moderating the impact of anti-immigrant attitudes on welfare support include characteristics of the foreign-born population (Alesina & Glaeser, 2004; Eger, 2010; Mau & Burkhardt, 2009; Senik et al., 2008; Soroka, Johnston, & Banting, 2004), multicultural policies (Crepaz, 2006; Evans, 2006; Kymlicka & Banting, 2006), the economic context (Mau & Burkhardt, 2009; Soroka et al., 2003; Taylor-Gooby, 2005), and the welfare regime (Crepaz & Damron, 2009; Eger, 2010; Finseraas, 2008; Larsen, 2008, 2011; Van Oorschot, 2006). Finally, different studies have included measures of the political context (Larsen, 2011; Mau & Burkhardt, 2009; Soroka et al., 2003; Taylor-Gooby, 2005). In this article, we wish to analyze the impact of the political discourse on the link between anti-immigrant attitudes and welfare support as well as its interplay with the welfare regime. Without neglecting the importance of other contextual factors, our further discussion will therefore focus on the hypothesized effects of these two variables.

Starting with the welfare regime, the general idea is that the welfare state is not only a result of ethnic diversity (as discussed for the U.S. context) but—once in place—it does also affect the attitudes of citizens, including their perceptions of migrants’ eligibility for welfare. Summarizing the wider literature on public welfare support, Arts and Gelissen (2001) say, “welfare states vary in terms of which particular egalitarian justice principles they accentuate and which specific notions of social solidarity they pursue” (p. 283).²

Larsen (2008) offers a detailed theoretical framework to analyze how distinct welfare regimes frame the definition of recipients’ welfare deservedness and thereby welfare support by the majority. According to him, liberal welfare regimes with dominantly means-tested social assistance programs open this debate by discussing who is in need and whether these people are to blame for their situation. At the same time, recipients are stigmatized as a special group of pure recipients and distinguished from the majority who merely finance the formers’ benefits. A similar line of reasoning is put forward by Crepaz and Damron (2009) who argue that means-tested systems reinforce the effects of social categorization by helping to construct attitudes of stark separation between individuals deserving of public support and individuals perceived as undeserving. Thus, “the effect of targeted welfare delivery systems is one of separation, stratification, and difference” (Crepaz & Damron, 2009, p. 446) and in consequence, public support for means-tested welfare programs is rather limited (see also Esping-Andersen, 1990). According to this dominant view, we would expect to find a negative impact of individual anti-immigrant sentiment on welfare support in liberal welfare regimes. In contrast, universal welfare regimes are seen to close the discussion of deservedness by blurring the boundary between receivers and payers: “In such systems almost everybody contributes and almost everybody receives, making it more difficult to stigmatize receivers of government support as other” (Crepaz & Damron, 2009, p. 449). Consequently, several authors see universal—and most importantly tax-financed, social-democratic welfare regimes—as being more capable than liberal regimes of bridging ethnically motivated conflicts over the redistribution of scarce welfare resources and of downplaying natives’ chauvinistic attitudes to welfare (Crepaz & Damron, 2009; see also Eger, 2010; Larsen, 2011). This brings us to our first hypothesis.

Hypothesis 1: The negative effect of anti-immigrant attitudes on support for redistribution is less strong in social-democratic welfare regimes (compared with liberal regimes).

While our reading of the literature finds Hypothesis 1 to express the common theoretical expectation, it is not uncontested. On one hand, empirical support for a welfare regime effect in general is rather mixed. Comparing two social-democratic regimes (Denmark and Sweden) with two liberal ones (United States and United Kingdom), Larsen (2011) finds that while Danes and Swedes are less hostile to financing programs aimed directly at migrants, they are as ready to characterize out-group members as lazy in the same way as British and U.S. respondents. Analyzing the effect of different measures of migration on the county level, Eger (2010) concludes that “ethnic heterogeneity negatively effects support for social welfare expenditure—even in Sweden” (p. 203), the ideal example of a social-democratic regime. Finally, in his study of welfare attitudes toward four groups of needy people in 23 European countries, Van Oorschot (2006) identifies migrants as the group which other citizens feel least concerned about and remarks that this “solidarity rank order is basically the same for all European countries” (p. 37). On the other hand, there are theoretical considerations that lead us to assume that the regime effect stated in Hypothesis 1 might indeed go in the opposite direction. Finseraas (2008) argues that it is precisely the heavily tax-financed and rights-based welfare states of the social-democratic cluster that might promote native citizens’ welfare chauvinism. This is because, by strongly relying on and promoting a collective feeling of solidarity among native-born citizens, social-democratic regimes might find it harder to extend these feelings toward nonnative citizens. As Freeman (1986) puts it, “the concept of membership implies the existence of persons who are not members and who are, therefore, excluded from the process of sharing” and the stronger this concept of membership is, the more difficult might be an expansion of it (p. 52). In contrast, welfare regimes financed through earnings-related contributions are much less driven by a common feeling of solidarity but rather by members’ strong sense of entitlement, purchased by paid contributions. As long as migrants contribute their part to the financing of welfare programs, natives might be less concerned about the threat of migrants’ welfare abuse (Boeri, 2006; Finseraas, 2008). This leads us to our second hypothesis.

Hypothesis 2: The negative effect of anti-immigrant attitudes on support for redistribution is stronger in social-democratic welfare regimes (compared with conservative regimes).

In contrast to the repeatedly discussed effect of the welfare regime, the role of political elites’ discourse has received much less attention in comparative studies. This is somehow surprising, as U.S. authors have long pointed to the important role of the electoral strategies of parties (Alesina & Glaeser, 2004; Neubeck & Cazenave, 2001; Quadagno, 1994). Reporting several examples of how U.S. conservative politicians activated and strengthened racial prejudices to undermine political support for redistributive policies,³ these studies suggest that “playing the race card” might also in Europe have a relevant effect on the link between anti-immigrant attitudes and welfare support. While Alesina and Glaeser (2004) even state that “the extreme right in Europe is already playing the race card to oppose welfare policies” and predict that “as racial heterogeneity increases, even the more ‘respectable’ right will move in this direction” (pp. 218-219), only a few studies have analyzed such a party-effect for the European context.

These studies do usually account for the role of parties by including measures of the vote or seat share of different party families (e.g., left parties). Such measurements are included in Soroka et al. (2003) and Taylor-Gooby (2005), who analyze social spending, as well as in Mau and Burkhardt’s (2009) analysis on individual welfare support. However, the cited studies treat the party-related variables as mere controls and do not investigate possible interplay between them and other individual or contextual variables. In the European context, as far as we are aware, the most detailed analysis of a possible effect of “playing the race card” is given by Larsen (2011). He explicitly accounts for a party-effect in his case selection by analyzing Denmark and the United States—where, according to Larsen (2011), “bourgeois politicians have played the race card” (p. 4)—with Sweden and the United Kingdom where this was seemingly not the case. However, although we appreciate Larsen’s theoretical considerations, it remains largely unclear how he comes to his classification of countries. We therefore question his conclusion that political elites “actually have little control” of public attitudes regarding anti-immigrant sentiment and welfare support; we will offer a more fine-graded analysis of the potential effect of “playing the race card” for the European context.

In detail, we will analyze the welfare-related and migration-related statements that parties make in their manifestos. This is a finer and more appropriate measurement of what parties actually say than the vote- and seat-share measurements of different party families used so far. Borrowing from the theoretical insights of Budge and Klingemann (2001) and Klingemann, Volkens, Bara, Budge, and McDonald (2006) on the salience theory of party competition, we expect these party statements to have a direct effect on the attitudes of citizens. Salience theory states that parties do not compete with each other on all possible issues but put greater emphasis on issues they “own” while ignoring opposing party lines. Applied to our case, parties might either stress welfare-related or migration-related issues to mobilize political support. The crucial difficulty is that parties that stress welfare-related issues—in general the Socialists and Social-Democrats—do usually not “own” migration-related issues, which are mainly appropriated by parties on the right. Therefore, we expect a trade-off between the salience of welfare-related and migration-related issues. The overall relevance of these two issue-sets signals to citizens their relative importance and we hypothesize that attitudes toward welfare and migration will be directly affected by the party statements (Petrocik, 1996; Petrocik, Benoit, & Hansen, 2003). This leads us to our Hypotheses 3 and 4.

Hypothesis 3: The more salient positive welfare-related party statements are, the higher the individual support for redistribution.

Hypothesis 4: The more salient negative migration-related party statements are, the less the individual support for redistribution.

Beside this trade-off effect between welfare and migration-related issues that has also been referred to as a “crowding-out effect” (Banting & Kymlicka, 2006, p. 10; see also Evans, 2006), we expect a secondary effect of the salience of migration-related issues. Parties on the right, which usually occupy the antimigration stance but do not “own” the prowelfare issue, might try to use their ownership of the former, to undermine political support for the latter. In fact, this instrumental strategy is precisely what Alesina and Glaeser (2004) understand by “playing the race card.” For Europe, this strategy of linking two former unrelated issues is well documented for parties on the extreme right (Arzheimer & Carter, 2006; DiGiusto & Jolly, 2008; Kitschelt, 1995; Mudde, 2007; Spies & Franzmann, 2011). While welfare-related issues are subordinated to the culture-based, noneconomic ideology of nationalism of ERP (Mudde, 2007), these parties might be able to link the anti-immigrant sentiments of citizens to a more critical view on the welfare state by raising fears about migrants’ welfare abuse or by strengthening natives’ welfare chauvinism (Kitschelt, 1995). This strategy of undermining support for redistribution might further be supported and thereby legitimized by mainstream parties, taking up the xenophobic appeal of the extreme right to address these electoral challenges (Ignazi, 1992). This leads us to expect

Hypothesis 5: The more salient negative migration-related party statements are, the greater is the negative effect of individual anti-immigrant sentiment on support for redistribution.

We will also investigate whether this kind of instrumental strategy is related to the national welfare regime. On one hand, parties stressing the anti-immigration issue might find it easier to link anti-immigrant attitudes to welfare criticism in liberal welfare regimes that draw a sharp line between welfare recipients and those who finance welfare benefits, as Hypothesis 1 suggests. On the other hand and as suggested by Hypothesis 2, playing the race card might be a more successful strategy in social-democratic welfare regimes that rely heavily on a common sense of citizens’ social identity that might not extend to nonnative citizens.⁴ In these regimes, parties stressing national identity and homogeneity might find it easier to blame nonnative citizens for perceived welfare abuse.

Data and Method

We test for the hypotheses developed in the previous section by conducting a hierarchical mixed model with two levels. On the first level, we include individual data on attitudes toward welfare support (dependent variable) and migration (independent variable) as well as several other attitudinal and sociodemographic control variables. On the second level, we include dummies accounting for a country’s welfare regime as well as our measures of party statements on welfare and migration. All models are estimated with maximum likelihood which allows us to compare the model fit of nested models using likelihood ratio (LR) tests (Hox, 2002). Table 1 summarizes our sample and also reports aggregated values of our main variables. Table OA1 (online appendix) gives the bivariate correlations of all variables.⁵ As our data set contains only 14 countries,⁶ we have to be very cautious with regard to the estimates from multilevel random effects models. Therefore, we test our models for their robustness by conducting country fixed effects analysis, which allows us to relax the assumption of a random country sample. We also use a (delete-one) Jackknife procedure to test whether our results are due to a single country in our sample.

Table 1.

Sample Description.

	Observations (n)		Mean
Country	Total ESS	Sample	Welfare support	Economic threat	Cultural threat	Welfare abuse	Salience of prowelfare issues	Salience of antimigrant issues	Welfare regime
BE	1,568	1,312	3.78	5.23	4.19	6.29	2,227.1	−8.1	SD
CH	1,368	905	3.67	3.88	3.89	5.38	792.9	123.0	C
DE	2,501	1,822	3.71	4.85	3.91	6.23	1,110.5	−320.0	C
DK	1,505	1,192	3.08	4.61	3.96	5.58	1,312.1	204.6	SD
ES	2,341	1,097	3.99	4.86	4.40	5.67	1,125.5	−252.0	C
FI	2,132	1,819	3.98	4.40	2.64	5.73	2,238.4	−74.7	C
FR	1,907	1,594	4.06	5.21	4.59	5.64	1,456.0	92.6	C
GB	2,097	1,543	3.54	5.51	5.26	6.22	682.3	−0.5	L
GR	1,943	917	4.37	6.44	6.38	5.88	1,280.3	−17.2	C
IE	1,469	1,169	3.83	5.12	4.31	6.12	1,327.7	−307.3	L
NL	1,602	1,286	3.42	4.65	3.91	5.80	843.8	308.7	SD
NO	1,412	1,319	3.54	4.22	4.09	5.66	1,543.6	−81.9	SD
PT	2,225	481	4.15	5.08	4.49	5.17	1,390.0	−23.5	C
SE	1,611	1,336	3.71	4.52	2.96	5.43	2,529.9	−20.2	SD
Sum (n)	25,681	17,792
Total			3.75	4.88	4.12	5.82	1,450.2	−35.3

Source: European Social Survey 2008 (unweighted data).

SD = Social-democratic; C = Conservative; L = Liberal.

Individual-Level Variables

All individual-level data come from the fourth round of the ESS “Attitudes Towards Welfare Provision,” which was carried out in 2008. As we are interested in the welfare support of natives, we first excluded all respondents with migrant status from our analysis. We define migrants as either not having citizenship of the host country or as not being born in the host country.

The dependent variable in our analysis is the individual welfare support, measured by the single item “the government should take measures to reduce differences in income levels.” Respondents were asked to state their preferences on a 5-point Likert-type scale which we have recoded so that higher values indicate preferences for more redistribution.

Our main interest lies in the relationship between individuals’ anti-immigrant sentiments and our dependent variable. We measure three dimensions of immigrant-related attitudes, which we derived in the theoretical section: economic threat (“Would you say it is generally bad or good for this country’s economy that people from other countries come here to live?”), cultural threat (“Would you say that this country’s cultural life is generally undermined or enriched by people coming from other countries to live here?”), and welfare abuse (“Do you think people who come to live in this country receive more social services than they contribute or contribute more than they receive?”). All three items are measured on an 11-point scale ranging from 0 to 10 which we have recoded so that higher values indicate a more critical view of migrants. Factor analysis shows that the three items do not tap one underlying dimension and we also expect them to be differently related to the dependent variable.

We control for a series of variables that have been found to affect the welfare support of individuals: gender (dummy variable, 1 = male, 0 = female), age (and age squared), subjective health (dummy variable, 1 = fair, bad, or very bad; 0 = good or very good), family status (dummy variable, 1 = married/civilian partnership, 0 = all other states), formal education (5 categories, International Standard Classification of Education (ISCED) 0-1, ISCED 2, ISCED 3, ISCED 4 [reference] and ISCED 5-6), household income deciles and labor-market status (4 categories, paid work, in education, unemployed, not in labor force [reference]). Finally, we control for the respondent’s left–right orientation measured on a scale from 0 (most left) to 11 (most right). We have z-standardized all continuous variables.

Country-Level Variables

On the second level, we include dummies accounting for a country’s welfare regime (liberal, conservative, or social-democratic) based on Esping-Andersen (1990). In each case, liberal welfare regimes will act as the reference category.

Our measures of party statements with regard to welfare and migration were calculated by using the data offered by the Comparative Manifesto Project (CMP). CMP provides a transcription of party manifestos due to salience theory based on hand-coded content analysis and reports the relevance of predefined issue-categories in party manifestos, indicating their relative importance to single parties. In the first step, we calculated the difference between positive and negative party statements with regard to welfare and migration-related issues for each single party. For welfare-related issues, we sum up the categories 503 (social justice) and 504 (welfare state expansion) and subtracted the category 505 (welfare state limitation) from it. For migration-related issues, we sum up the categories 601 (national way of life: positive) and 608 (multiculturalism: negative) and subtracted the categories 602 (national way of life: negative) 607 (multiculturalism: positive), and 705 (underprivileged minority groups).⁷ Therefore, positive values of our measure of salience of welfare indicate positive net statements of parties with regard to welfare state expansion, whereas positive values of our measure of salience of migration indicate negative net statements of parties with regard to migration. In a second step, we multiply the percentages obtained by the vote share of this party as we assume that the statements of more successful parties will also receive more public attention. Finally, we sum up these values for all parties that have gained seats in the election closest to the year 2008. Therefore, our measures of the salience of welfare and migration are not restricted to the statements of a single party or party family but offer a measure of the respective issue-salience for the entire party system at a single election.

As reported in Table 1, the European party systems analyzed show great variation in the salience of anti-immigrant issues. On one hand, we can identify countries in which anti-immigrant issues are very prominent on the political agenda. In each of these party systems (Denmark, the Netherlands, and Switzerland), ERP have gained a substantive share of seats. On the other hand, party systems without a relevant ERP are far less migration critical, which is the case for Germany, Ireland, and most of Southern Europe. For the following analysis, we have z-standardized the continuous party variables.

Results

Table 2 reports a series of multilevel models in which we test the hypotheses developed in the previous section. We start with an empty model (M₀) to estimate the amount of variance that is due to between-country variation. About 9% of the total variance is accounted for by country differences.

Table 2.

Multilevel Regression of Support for Redistribution.

	M₀	M₁	M₂	M₃	M₄
Individual-level variables
Economic threat of migration		0.040***	0.040***	0.040***	0.037*
Cultural threat of migration		−0.027**	−0.027**	−0.026**	−0.026**
Welfare abuse of migrants		−0.005	−0.005	−0.005	−0.008
Left–Right Scale		−0.235***	−0.235***	−0.235***	−0.235***
Gender (man = 1)		−0.097***	−0.097***	−0.097***	−0.095***
Age		0.041***	0.041***	0.040***	0.041***
Age²		−0.021**	−0.021**	−0.021**	−0.021**
Household income		−0.130***	−0.130***	−0.130***	−0.128***
Education
ISCED 0-1		0.129**	0.129**	0.128**	0.129**
ISCED 2		0.136**	0.137**	0.136**	0.133**
ISCED 3		0.117**	0.118**	0.117**	0.110**
ISCED 4		Ref.	Ref.	Ref.	Ref.
ISCED 5-6		−0.061	−0.061	−0.062	−0.068
Bad health status		0.059***	0.058***	0.058***	0.059***
Married/Civilian partnership		0.041**	0.040**	0.040**	0.038**
Employment status
In paid work		−0.012	−0.012	−0.012	−0.011
In education		−0.111**	−0.111**	−0.112**	−0.108**
Unemployed		0.087*	0.087*	0.087*	0.082*
Not in labor force		Ref.	Ref.	Ref.	Ref.
Country-level variables
Salience prowelfare issues				0.177**	0.173**
Salience antimigrant issues				0.022	0.025
× Economic threat of migration					0.010
× Cultural threat of migration					−0.010
× Welfare abuse of migrants					−0.028**
Welfare regime
Liberal			Ref.	Ref.	Ref.
Sociodemocratic			−0.082	−0.326*	−0.330
Conservative			0.255	0.138	0.140
Constant	0.020	−0.008	−0.107	0.049	0.051
Variance components
Var(Constant)	0.0917***	0.0756***	0.0502***	0.0289***	0.0283***
Var(Economic threat of migration)					0.0025***
Var(Cultural threat of migration)
Var(Welfare abuse of migrants)
Var(Residuals)	0.9197***	0.8202***	0.8202***	0.8202***	0.8171***
Statistics
n	14	14	14	14	14
N	17,792	17,792	17,792	17,792	17,792
R² Level 2 (Snijders/Bosker)	0%	17.53%	44.90%	67.93%	67.35%
R² Level 1 (Snijders/Bosker)	0%	11.44%	13.94%	16.05%	16.09%
Log likelihood	−24,534.78	−23,515.12	−23,512.30	−23,508.52	−23,485.24
Wald chi-square (df)	—	2,160.84	2,169.27	2,187.43	2,162.46
AIC	49,075.57	47,070.24	47,068.59	47,065.05	47,026.49

Source: European Social Survey 2008 (unweighted data).

ISCED = International Standard Classification of Education; AIC = Akaike Information Criterion.

p < .05. **p < .01. ***p < .001 (one-sided tests). All continuous variables are z-standardized.

In Model M₁, we include our three measures of anti-immigrant sentiment (economic threat, cultural threat, and welfare abuse) and all individual-level control variables. We find economic threat is positively and cultural threat is negatively correlated with individual support for redistribution. In contrast, the effect of the perceived welfare abuse of migrants does not reach common levels of statistical significance. While these different findings justify our decision not to summarize these variables in one index, they call for further interpretation. In line with the findings of Finseraas (2008), we interpret the positive effect of economic threat as a call for compensation for the assumed negative economic effects of migration.

In contrast, the negative effect of cultural threat seems to indicate some sort of antisolidarity between the native population and migrants. This interpretation is in line with the argument made by Freeman (1986), who sees the welfare state as relying heavily on a common sense of national identity which cannot be extended easily to the foreign-born, and therefore culturally different, new members (Finseraas, 2008). The contradictory effect of economic threat and cultural threat on our dependent variable seems somehow surprising, as both independent variables are highly and positively correlated. From a theoretical perspective, this finding however supports those studies stressing the multidimensionality of the concept of anti-immigrant attitudes (Finseraas, 2008) arguing that the different subdimensions are—while highly correlated—influenced by distinct contextual factors and should therefore be analytically separated from each other (Hjerm & Nagayoshi, 2011).

Finally, the nonsignificance of the perceived welfare abuse of migrants does not reproduce the findings made in the U.S. context. European natives holding the attitude that “migrants receive more than they pay in” do not show lesser levels of support for general redistribution. Taken together, the individual-level variables in M₁ account for 11.4% of the individual- and 17.5% of the country-level variance.⁸

In Model M₂, we add the first set of country-level variables, that is, the welfare regime dummies. While this raises the explained variance on the country level to 44.9%, we find no statistically significant effects. Neither respondents in social-democratic nor those in conservative regimes show less support for welfare than respondents in liberal regimes. The difference between social-democratic and conservative regimes, however, is statistically significant and indicates that support for redistribution is higher in conservative welfare regimes.

In Model M₃, we include our measures of the salience of prowelfare and antimigration issues. As expected in Hypotheses 3, we find a positive effect of the salience of prowelfare, but we do not find a negative effect of the salience of antimigrant issues. The inclusion of the two salience variables results in a further and quite impressive model improvement raising the country-level R² to almost 68%. The corresponding LR test indicates a significant model improvement (p < .05). The effect of the salience of prowelfare party statements is theoretically directly related to individual attitudes toward welfare and can thus easily be interpreted: The more positive parties are with regard to welfare programs, the more supportive are also voters with regard to this issue. In contrast, the nonsignificance of salience of antimigrant issues does not confirm Hypothesis 4: Anti-immigrant party statements do not in general affect levels of support for redistribution.

So far we have discussed only the main effects of our issue-salience variables, but—as we expect in Hypothesis 5—their relevance might be further increased if they do moderate the effect of the individual anti-immigrant sentiment on support for welfare. As has been discussed extensively in the literature on party competition, the issues of welfare and migration are related to different attitudes to political competition (Enyedi & Deegan-Krause, 2010; Van der Brug & Van Spanje, 2009; Warwick, 2002). While the former addresses economic-related interests, the latter makes cultural-related identities more salient. By doing so, culture-related issues like migration might drive back the political as well as the individual importance of economy-related issues, which should result in less attention being paid to redistribution by the native population. We test for this possibility by introducing three cross-level interaction effects between the salience of anti-immigrant issues and our three measures of anti-immigrant sentiment in M_4. Consequently, the three slopes for economic threat, cultural threat, and welfare abuse were assumed to be a function of a random and a fixed component.⁹ Model M₄ leads to two interesting results.

First, and while the main effect of perceived migrants’ welfare abuse stays insignificant, its interaction with the salience of anti-immigrant issues is highly significant. Figure 1 presents the marginal effect of perceived welfare abuse conditional on the observed values of the contextual level variable salience of anti-immigrant issues. The effect of welfare abuse becomes significantly negative if the salience value lies about a quarter of a standard deviation above the sample average. In other words, natives’ belief that migrants take more from welfare programs than they contribute only limits natives’ general support for welfare if these two attitudes are linked by political parties. In our view, this is exactly what Alesina and Glaeser understand by “playing the race card” and therefore supports Hypothesis 5. We see this link to be mainly influenced by the programmatic appeals of right-wing parties, and especially by parties of the extreme right, as antimigration issues are most prominent in their party manifestos. Not only do these parties indirectly depress support for welfare by downplaying the welfare issue—which they do not “own” in terms of salience theory—their accentuation of antimigration issues is also directly related to a more critical view of welfare programs among the native population. We argue that this might be due to raising fears among natives about migrants’ welfare abuse.

Figure 1.

Marginal effect of perceived welfare abuse.

Second, M₄ shows no significant interaction between the salience of anti-immigrant issues and the perceived cultural threat of migration. We think that this might be due to the difficulty of modeling three cross-level interactions with the same contextual variable in a model with only 14 second-level units. We have therefore reestimated M₄ by introducing the cross-level interactions separately. These models identify a significant effect of cultural threat and are presented in the appendix (Table A2). Figure 2 present the marginal effect of cultural threat for the observed range of the contextual level variable salience of anti-immigrant issues. The more salient antimigrant issues are on the political agenda, the stronger the negative relation between culture-related antimigrant attitudes and support for redistribution. We also argue that this interaction effect is caused by the political agenda of ERP. Mudde (2007) identifies nationalism as the core ideological feature of this group of parties, understood as a political doctrine that “strives for the congruence of the cultural and political unit, i.e., the nation and the state” (p. 16). This view is supported by Carter (2005), who argues that ERP see politics as constituted by a boundary line dividing natives and immigrants and stressing the primacy of cultural homogeneity within national boundaries. In this way, the ideology of the European extreme right strengthens already existing threats about migration-induced cultural diversity, thereby undermining feelings of economic solidarity between native and nonnative citizens, that are seen as essential for the support for redistributive policies (see Boeri, 2006; Finseraas, 2008; Freeman, 1986).¹⁰

Figure 2.

Marginal effect of cultural threat.

So far, our analysis of party-effects has controlled only for possible direct effects of a country’s welfare regime on individual support for welfare. However, several authors have argued that a moderating effect can also be observed for a country’s welfare regime, as stated in Hypotheses 1 and 2. To test the interaction effects of issue-salience and welfare regimes against each other, we would therefore have to compare the explanatory power of models with different cross-level interaction effects which is not possible with LR tests because the models are not nested. Hence, in Table 3 we first estimated a reference model (M₅) which we use to indirectly compare the model which includes interactions with welfare dummies (M₆) with the issue-salience model (M₇).

Table 3.

Multilevel Regression of Support for Redistribution—Model Comparisons.

	M₅	M₆	M₇
Individual-level variables
Economic threat of migration	0.037**	0.016	0.038**
Cultural threat of migration	−0.026*	−0.004	−0.026*
Welfare abuse of migrants	−0.007	0.003	−0.008
Left–Right Scale	−0.235***	−0.235***	−0.235***
Gender (man=1)	−0.095***	−0.094***	−0.095***
Age	0.041***	0.040***	0.041***
Age²	−0.021**	−0.021**	−0.021**
Household income	−0.128***	−0.128***	−0.128***
Education
ISCED 0-1	0.128**	0.127**	0.129**
ISCED 2	0.133**	0.133**	0.133**
ISCED 3	0.111**	0.111**	0.110**
ISCED 4	Ref.	Ref.	Ref.
ISCED 5-6	−0.068	−0.069	−0.068
Bad health status	0.058***	0.059***	0.059***
Married/Civilian partnership	0.037**	0.037**	0.037**
Employment status
In paid work	−0.012	−0.012	−0.012
In education	−0.110**	−0.111**	−0.109**
Unemployed	0.080*	0.080*	0.081*
Not in labor force	Ref.	Ref.	Ref.
Country-level variables
Salience prowelfare issues	0.170**	0.171**	0.170**
Salience antimigrant issues	0.023	0.023	0.023
× Economic threat of migration			0.009
× Cultural threat of migration			−0.008
× Welfare abuse of migrants			−0.028***
Welfare regime
Liberal	Ref.	Ref.	Ref.
Social-democratic	−0.325*	−0.325*	−0.324*
× Economic threat of migration		0.031
× Cultural threat of migration		−0.044
× Welfare abuse of migrants		−0.040
Conservative	0.141	0.144	0.140
× Economic threat of migration		0.021
× Cultural threat of migration		−0.013
× Welfare abuse of migrants		0.009
Constant	0.048	0.048	0.047
Variance components
Var(Constant)	0.0287***	0.0233***	0.0232***
Var(Economic threat of migration)	0.0016***	0.0017***	0.0019***
Var(Cultural threat of migration)	0.0008***	0.0004***	0.0007***
Var(Welfare abuse of migrants)	0.0009***	0.0004***	0.0001*
Var(Residuals)	0.8167***	0.8167***	0.8167***
Statistics
n	14	14	14
N	17,792	17,792	17,792
Log likelihood	−23,489.057	−23,485.18	−23,484.37
Wald chi-square (df)	2,130.53	2,144.87	2,154.97
AIC	47,032.12	47,036.35	47,028.73

Source: European Social Survey 2008 (unweighted data).

ISCED = International Standard Classification of Education; AIC = Akaike Information Criterion.

p < .05. **p < .01. ***p < .001 (one-sided tests). All continuous variables are z-standardized.

The reference model includes all variables as main effects and also includes random slopes for the three variables accounting for individual anti-immigrant sentiment. In comparison with this reference model, the welfare regime model (M₆) does not lead to a significant improvement of the model fit by introducing cross-level interaction effects between the welfare regime dummies and the three items accounting for anti-immigrant sentiment (LR test p = .26). The results of M₆ therefore do not support the findings of several other authors reporting a moderating effect of the welfare regime (Crepaz & Damron, 2009; Finseraas, 2008; Larsen, 2011). Neither are the anti-immigrant attitudes of natives more negatively related to the support for welfare in liberal regimes nor is the relationship between anti-immigrant sentiment and support for redistribution significantly weakened in social-democratic regimes. Both, Hypotheses 1 and 2 therefore have to be rejected.¹¹

M₇ introduces cross-level interaction effects between the salience of anti-immigrant issues and the three items accounting for anti-immigrant sentiment.¹² This leads to a significant improvement in the model fit (LR test p < .05). Thus, the issue-salience model outclasses the welfare regime model in terms of model fit. We also conclude that in the European context, “playing the race card” does have a significant effect on the relationship between individual anti-immigrant sentiment and welfare support, a result which challenges Larsen’s (2011) argument that political elites “actually have little control over the attitudes of the public” with regard to these issues (p. 351).¹³

Conclusion

In this article, we have analyzed the impact of party statements on the link between individual anti-immigrant sentiment and support for redistributive policies. Borrowing insights from the salience theory of party competition, we have hypothesized that a high salience of antimigration issues will depress individual support for welfare, while positive party statements with regard to welfare will increase support for redistribution. As voters do not usually expect all parties to have the same problem-solving competencies with regard to welfare and migration, parties have distinct incentives to accentuate welfare and migration-related issues. On the level of the party system, this results in a trade-off between migration and welfare issues; the one can only be made more salient at the cost of the other. We have further argued that political parties of the right might instrumentally use their migration appeals to undermine the legitimacy of existing welfare programs. We have tested these theoretical arguments against the most frequently discussed relevant variable, the welfare regime.

Our analysis supports most of our theoretical expectations. We find that there is indeed a salience trade-off between welfare and migration-related issues but we find only the former being directly related to individual support for welfare. Furthermore, our findings strongly point to a moderating effect of antimigration issue-salience on the link between anti-immigrant sentiment and welfare support: In countries where parties stress policies which are strongly related to migration, natives with anti-immigrant sentiment do show much less support for welfare. We conclude that this effect is comparable with the political strategy of “playing the race card,” which several authors have identified for the United States. We argue that in particular, parties of the extreme right are responsible for this moderating effect, by raising fears about migrants’ welfare abuse among natives as well as by stressing the primacy of cultural homogeneity within national boundaries.

With regard to the welfare regime, we were not able to replicate the results of previous studies. We do not find welfare support or the link between anti-immigrant sentiment and welfare support stronger or weaker in different regimes. Within the restrictions imposed by our statistical model, we have also tested our party-related hypotheses against possible welfare regime effects. This comparison shows that the party variables do explain variation in welfare support better than the welfare regimes. While further analysis on this issue should be encouraged, our results let us state that compared with the frequently discussed effects of the welfare regime, party statements with regard to welfare and migration are the more promising explanation for the migration–welfare nexus.

Footnotes

Appendix

Table A2.

Multilevel Regression of Support for Redistribution (Additional Models).

	M_A4	M_A5
Individual-level variables
Economic threat of migration	0.037*	0.040***
Cultural threat of migration	−0.027**	−0.025**
Welfare abuse of migrants	−0.005	−0.008
Left–Right Scale	−0.235***	−0.234***
Gender (man = 1)	−0.095***	−0.097***
Age	0.041***	0.041***
Age²	−0.021**	−0.021**
Household income	−0.129***	−0.128***
Education
ISCED 0-1	0.128**	0.128**
ISCED 2	0.134**	0.135**
ISCED 3	0.110**	0.115**
ISCED 4	Ref.	Ref.
ISCED 5-6	−0.068	−0.064
Bad health status	0.058***	0.058***
Married/civilian partnership	0.039**	0.038**
Employment status
In paid work	−0.012	−0.012
In education	−0.110**	−0.110**
Unemployed	0.081*	0.084*
Not in labor force	Ref.	Ref.
Country-level variables
Salience prowelfare issues	0.173**	0.177**
Salience antimigrant issues	0.024	0.023
× Economic threat of migration	0.003
× Cultural threat of migration	−0.016*
× Welfare abuse of migrants		−0.030**
Welfare regime
Liberal	Ref.	Ref.
Sociodemocratic	−0.331*	−0.331*
Conservative	0.142	0.136
Constant	0.052	0.052
Variance components
Var(Constant)	0.0279***	0.0297***
Var(Economic threat of migration)	0.0025***
Var(Cultural threat of migration)
Var(Welfare abuse of migrants)		0.0008***
Var(Residuals)	0.8177***	0.8185***
Statistics
n	14	14
N	17,792	17,792
R² Level 2 (Snijders/Bosker)	67.74%	67.32%
R² Level 1 (Snijders/Bosker)	16.06%	16.08%
Log likelihood	−23,492.29	−23,496.11
Wald chi-square (df)	2,147.00	2,186.31
AIC	47,038.59	47,044.21

Source: European Social Survey 2008 (unweighted data).

ISCED = International Standard Classification of Education; Ref. = reference; AIC = Akaike Information Criterion.

p < .05. **p < .01. ***p < .001 (one-sided tests). All continuous variables are z-standardized.

Acknowledgements

We like to thank Hans-Jürgen Andress, Henning Finseraas, Maureen Eger, Christian Albrekt Larsen, Mikael Hjerm, Jan Mewes,Wim van Oorschot and two anonymous reviewers for fruitful comments on our work. We also thank the participants of the 2012 Annual RC 19 Conference “Welfare States Facing Global Turbulence, Ageing and Migration” in Oslo for their valuable discussion of an earlier version of this paper. We would also like to express our thanks to Julia Däumling for collecting the data and to Martin Spitzenpfeil for proofreading. Of course, the responsibility for all remaining errors in text or data are exclusively in the authors’ responsibility.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biographies

Alexander W. Schmidt is a PhD student at the Research Training Group SOCLIFE and a research associate at the Insititue for Sociology and Social Psychology at the University of Cologne. He teaches survey methodology and applied statistics. In 2013, he published a book on applied panel data analysis. His research focuses on determinants of welfare attitudes.

Dennis C. Spies is a postdoc research fellow at the Cologne Center for Comparative Politics at the University of Cologne. His research focuses on political economy, migration, and party competition.

References

Alesina

Glaeser

E. L.

(2004). Fighting poverty in the US and Europe: A world of difference. Oxford, UK: Oxford University Press.

Alesina

Glaeser

Sacerdote

(2001). Why doesn’t the US have a European-style welfare system? (National Bureau of Economic Research Working Paper Series No. 8524). Retrieved from http://www.nber.org/papers/w8524

Alonso

Fonseca

S. C.

(2011, May). Immigration, Left and Right. Party Politics. Retrieved from http://ppq.sagepub.com/content/early/2011/05/13/1354068810393265

Andreß

Heien

(2001). Four worlds of welfare state attitudes? A comparison of Germany, Norway, and the United States. European Sociological Review, 17, 337-356.

Arts

Gelissen

(2001). Welfare states, solidarity and justice principles: Does the type really matter? Acta Sociologica, 44, 283-299.

Arzheimer

(2008). Protest, neo-liberalism or anti-immigrant sentiment: What motivates the voters of the extreme right in Western Europe? Zeitschrift für Vergleichende Politikwissenschaft, 2, 173-197.

Arzheimer

Carter

(2006). Political opportunity structures and right-wing extremist party success. European Journal of Political Research, 45, 419-443.

Banting

K. G.

(2005). The multicultural welfare state: International experience and North American narratives. Social Policy & Administration, 39, 98-115.

Banting

Kymlicka

(2006). Multiculturalism and the welfare state: Recognition and redistribution in contemporary democracies. Oxford, UK: Oxford University Press.

10.

Bean

Papadakis

(1998). A comparison of mass attitudes towards the welfare state in different institutional regimes, 1985–1990. International Journal of Public Opinion Research, 10, 211-236.

11.

Blekesaune

(2007). Economic conditions and public attitudes to welfare policies. European Sociological Review, 23, 393-403.

12.

Boeri

(2006, April). Migration policy and the welfare state. Paper presented at the Reinventing the Welfare State Conference, Tilburg, Netherlands.

13.

Bornschier

(2009, September). Old cleavages and new actors in the formation of a new cultural divide: Why a right-wing populist party emerged in France but not in Germany. Paper presented at the 5th ECPR General Conference, Potsdam, Germany.

14.

Brooks

Manza

(2007). Why welfare states persist: The importance of public opinion in democracies. Chicago, IL: The University of Chicago Press.

15.

Budge

Klingemann

H.-D.

(2001). Finally! Comparative over-time mapping of party policy movement. In Budge

Klingemann

H.-D.

Volkens

Bara

Tanenbaum

(Eds.), Mapping policy preferences: Estimates for parties, electors, and governments 1945-1998 (pp. 19-50). New York, NY: Oxford University Press.

16.

Carter

(2005). The extreme right in Western Europe: Success or failure? Manchester, UK: Manchester University Press.

17.

Crepaz

(2006). “If you are my brother, I may give you a dime!” Public opinion on multiculturalism, trust, and the welfare state. In Banting

Kymlicka

(Eds.), Multiculturalism and the welfare state: Recognition and redistribution in contemporary democracies (pp. 92-117). Oxford, UK: Oxford University Press.

18.

Crepaz

(2008). Trust beyond borders: Immigration, the welfare state, and identity in modern societies. Ann Arbor: The University of Michigan Press.

19.

Crepaz

M. M. L.

Damron

(2009). Constructing tolerance: How the welfare state shapes attitudes about immigrants. Comparative Political Studies, 42, 437-463.

20.

de Lange

S. L.

(2007). A new winning formula? The programmatic appeal of the radical right. Party Politics, 13, 411-435.

21.

DiGiusto

G. M.

Jolly

S. K.

(2008, August). French xenophobia and immigrant contact: Public attitudes toward immigration. Paper presented at the annual Conference of the American Political Science Association, Boston, MA.

22.

Eger

M. A.

(2010). Even in Sweden: The effect of immigration on support for welfare state spending. European Sociological Review, 26, 203-217.

23.

Enyedi

Deegan-Krause

(Eds.). (2010). The structure of political competition in Western Europe. London, England: Routledge.

24.

Esping-Andersen

(1990). The three worlds of welfare capitalism. Cambridge, UK: Polity.

25.

European Social Survey Round 4 Data. (2008). Data file edition 4.1. Bergen, Norway: Norwegian Social Science Data Services (Data Archive and distributor of ESS data).

26.

Evans

(2006). Is multiculturalism eroding support for welfare provision? The British case. In Banting

Kymlicka

(Eds.), Multiculturalism and the welfare state: Recognition and redistribution in contemporary democracies (pp. 152-176). Oxford, UK: Oxford University Press.

27.

Finegold

(1988). Agriculture and the politics of U.S. social provision: Social insurance and food stamps. In Weir

Orloff

A. S.

Skocpol

(Eds.), The politics of social policy in the United States (pp. 199-234). Princeton, NJ: Princeton University Press.

28.

Finseraas

(2008). Immigration and preferences for redistribution: An empirical analysis of European survey data. Comparative European Politics, 6, 407-431.

29.

Fox

(2004). The changing color of welfare? How Whites’ attitudes toward Latinos influence support for welfare. American Journal of Sociology, 110, 580-625.

30.

Freeman

G. P.

(1986). Migration and the political economy of the welfare state. Annals of the American Academy of Political and Social Science, 485, 51-63.

31.

Gilens

(1995). Racial attitudes and opposition to welfare. Journal of Politics, 57, 994-1014.

32.

Gilens

(1999). Why Americans hate welfare: Race, media, and the politics of antipoverty policy. Chicago, IL: The University of Chicago Press.

33.

Green-Pedersen

Odmalm

(2008). Going different ways? Right-wing parties and the immigrant issue in Denmark and Sweden. Journal of European Public Policy, 15, 367-381.

34.

Hinnfors

Spehar

Bucken-Knapp

(2011). The missing factor: Why social democracy can lead to restrictive immigration policy. Journal of European Public Policy, 19, 585-603.

35.

Hjerm

Nagayoshi

(2011). The composition of the minority population as a threat: Can real economic and cultural threats explain xenophobia? International Sociology, 26, 815-843.

36.

Hox

(2002). Multilevel analysis: Techniques and applications. London, England: Lawrence Erlbaum.

37.

Huber

Stephens

J. D.

(1993). Political parties and public pensions: A quantitative analysis. Acta Sociologica, 36, 309-325.

38.

Ignazi

(1992). The silent counter-revolution: Hypotheses on the emergence of extreme right-wing parties in Europe. European Journal of Political Research, 22, 3-34.

39.

Jæger

M. M.

(2012). The effect of macroeconomic and social conditions on the demand for redistribution—A pseudo panel approach. Journal of European Social Policy. Manuscript submitted for publication.

40.

Kitschelt

(1995). The radical right in Western Europe: A comparative analysis. Ann Arbor: The University of Michigan Press.

41.

Klingemann

H.-D.

Volkens

Bara

Budge

McDonald

M. D.

(2006). Mapping policy preferences II: Estimates for parties, electors, and governments in Eastern Europe, European Union, and OECD 1990-2003. New York, NY: Oxford University Press.

42.

Korpi

(1983). The democratic class struggle. London, England: Routledge.

43.

Korpi

Palme

(2003). New politics and class politics in the context of austerity and globalization: Welfare state regress in 18 countries, 1975-95. American Political Science Review, 97, 425-446.

44.

Kymlicka

Banting

(2006). Immigration, multiculturalism, and the welfare state. Ethics & International Affairs, 20, 281-304.

45.

Larsen

C. A.

(2008). The institutional logic of welfare attitudes. Comparative Political Studies, 41, 145-168.

46.

Larsen

C. A.

(2011). Ethnic heterogeneity and public support for welfare: Is the American experience replicated in Britain, Sweden and Denmark? Scandinavian Political Studies, 34, 332-353.

47.

Lipset

S. M.

(1996). American exceptionalism: A double-edged sword. New York, NY: Norton.

48.

Lipset

S. M.

Marks

(2000). It didn’t happen here: Why socialism failed in the United States. New York, NY: Norton.

49.

Luttmer

E. F. P.

(2001). Group loyalty and the taste for redistribution. Journal of Political Economy, 109, 500-528.

50.

Mau

Burkhardt

(2009). Migration and welfare state solidarity in Western Europe. Journal of European Social Policy, 19, 213-229.

51.

Möhring

Schmidt

A. W.

(2013). MLT: Stata module to provide multilevel tools (Statistical Software Components S457577). Boston, MA: Boston College Department of Economics. Retrieved from http://ideas.repec.org/c/boc/bocode/s457577.html

52.

Mudde

(2007). Populist radical right parties in Europe. Cambridge, UK: Cambridge University Press.

53.

Mughan

Paxton

(2006). Anti-immigrant sentiment, policy preferences and populist party voting in Australia. British Journal of Political Science, 36, 341-358.

54.

Neubeck

K. J.

Cazenave

N. A.

(2001). Welfare racism: Playing the race card against America’s poor. New York, NY: Routledge.

55.

Petrocik

J. R.

(1996). Issue ownership in presidential elections, with a 1980 case study. American Journal of Political Science, 40, 825-850.

56.

Petrocik

J. R.

Benoit

W. L.

Hansen

G. J.

(2003). Issue ownership and presidential campaigning, 1952-2000. Political Science Quarterly, 118, 599-626.

57.

Pfeffer

M. J.

(1983). Social origins of three systems of farm production in the United States. Rural Sociology, 48, 540-562.

58.

Pierson

(1996). The new politics of the welfare state. World Politics, 48, 143-179.

59.

Putnam

R. D.

(2007). E Pluribus Unum: Diversity and community in the twenty-first century the 2006 Johan Skytte prize lecture. Scandinavian Political Studies, 30, 137-174.

60.

Quadagno

(1994). The color of welfare: How racism undermined the war on poverty. New York, NY: Oxford University Press.

61.

Rehm

(2009). Risks and redistribution: An individual-level analysis. Comparative Political Studies, 42, 855-881.

62.

Senik

Stichnoth

Straeten

(2008). Immigration and natives’ attitudes towards the welfare state: Evidence from the European Social Survey. Social Indicators Research, 91, 345-370.

63.

Skocpol

Finegold

(1982). State capacity and economic intervention in the early new deal. Political Science Quarterly, 97, 255-278.

64.

Soroka

Banting

Johnston

(2003, May and June). Immigration and redistribution in a global era. Paper presented at the annual meeting of the Canadian Political Science association, Halifax, NS.

65.

Soroka

S. N.

Johnston

Banting

(2004). Ethnicity, trust, and the welfare state. In Van Parijs

(Ed.), Cultural diversity versus economic solidarity (pp. 33-57). Brussels, Belgium: De Boeck.

66.

Spies

Franzmann

S. T.

(2011). A two-dimensional approach to the political opportunity structure of extreme right parties in Western Europe. West European Politics, 34, 1044-1069.

67.

Stephens

J. D.

(1979). The transition from capitalism to socialism. London, England: Macmillan.

68.

Svallfors

(1997). Worlds of welfare and attitudes to redistribution: A comparison of eight Western nations. European Sociological Review, 13, 283-304.

69.

Taylor-Gooby

(2005). Is the future American? Or, can left politics preserve European welfare states from erosion through growing diversity? Journal of Social Policy, 34, 661-672.

70.

Van der Brug

Van Spanje

(2009). Immigration, Europe and the “new” cultural dimension. European Journal of Political Research, 48, 309-334.

71.

Van Oorschot

(2006). Making the difference in social Europe: Deservingness perceptions among citizens of European welfare states. Journal of European Social Policy, 16, 23-42.

72.

Van Parijs

(Ed.). (2004). Cultural diversity versus economic solidarity. Brussels, Belgium: De Boeck.

73.

Warwick

P. V.

(2002). Toward a common dimensionality in West European policy spaces. Party Politics, 8, 101-122.