Public Choice DOI 10.1007/s11127-013-0094-6 Immigration, redistribution, and universal suffrage Raul Magni-Berton Received: 22 October 2012 / Accepted: 18 June 2013 © Springer Science+Business Media New York 2013 Abstract The effect of immigration on redistribution has been widely debated. This paper contributes to this debate by testing two explanations, which are that (i) immigration tends to reduce redistribution due to people’s higher levels of xenophobia, and that (ii) immigration affects redistribution because immigrants do not have the right to vote. Since the demand for redistribution depends on the (expected) gap between median voter income and mean income, immigrants affect the demand for redistribution because, as non-citizens, they do not change the median voter’s income, but, as economic stakeholders, they do affect the mean income. Four empirical consequences of (i) and (ii) are tested at the individual level. Evidence from the European Values Survey in 45 countries confirms (ii), showing that immigrants’ expected competitiveness on the labor market affects preferences for redistribution and that it is amplified when the perceived number of immigrants is high. In contrast, (i) is globally rejected since the impact of the citizens’ declared level of solidarity with immigrants tends to be weak and depends on the type of measurement or specification used. Keywords Immigration · Redistribution · Universal suffrage · Xenophobia · Median voter · Public opinion The rise in immigrant populations in the last decade has introduced a large minority of people into developed countries who are entitled to economic rights (such as property rights, labor rights, and so on), but not to political rights. When foreign immigrants are residents, they cannot vote, hold political and administrative offices, serve on a jury, or participate more broadly in the country’s political decisions, at least not in the national elections.1 But, contrary to other populations deprived of political rights, such as children or prison inmates, immigrants are stakeholders of the productive system. This implies that as the number of 1 The only exception being New Zealand, where all permanent residents may vote. Several countries grant such a right at local elections. B R. Magni-Berton ( ) Science Po Grenoble, PACTE, Department of Political Science, Grenoble University, Grenoble, France e-mail: [email protected] Public Choice immigrants increases, the proportion of residents entitled to vote decreases. Thus today’s societies are moving away from universal suffrage, not because of restrictions on political rights, but because of a spectacular increase in cross-border mobility. In such a context, citizens can make decisions knowing that other stakeholders cannot. This situation is particularly relevant for redistribution, because immigrants are taxpayers and tax consumers, but not tax decision-makers.2 This has been demonstrated by Razin et al. (2002) using the standard theory of the size of government under majority rule, and it can be intuitively illustrated here: when the number of immigrants grows, and assuming immigrants to be poorer than natives, the proportion of the fiscal revenues ending up in the hands of immigrants increases, which implies that voters will opt for lower taxes. However, this explanation has never been tested at the individual level. This article provides such a test. Note that the current path for non-citizens to be franchised is through naturalization. Naturalization is legally possible in all European states, but it takes time.3 These time requirements, associated with the rise in immigration, lead to a larger disenfranchised population, even though the immigrants themselves are not permanently disenfranchised. In 2010, in the European Union, 47 million residents are foreign-born and, among them, 69 % are not citizens (6.5 % of residents) (Vasileva 2011). While recent studies have provided evidence of a negative correlation between income redistribution and the rise in immigration (Roemer et al. 2007; Finseraas 2008; Senik et al. 2009; Eger 2010), only a small strand of the literature on this issue provides an explanation based on immigrant disenfranchisement (see Razin et al. 2002; Hansen 2003; Mayr 2007). Most studies tend to view immigrants as belonging to ethnic, religious or linguistic groups that differ from those of the natives and argue that heterogeneity, especially ethnic fractionalization, reduces the size of the welfare state (Mueller and Murrell 1986; James 1993; Lind 2007; Alesina and La Ferrara 2005). Empirically, focusing on the U.S. case, scholars have found evidence that racial stereotypes or lack of solidarity with immigrants influence white people’s support for welfare spending (Gilens 1999; Alesina et al. 2001; Luttmer 2001), even though in other countries the relation is weak or insignificant (Soroka et al. 2006; Van Oorschot 2008). Note that until very recently in the United States—unlike in Europe—race has not distinguished between citizens and immigrants, so, initially, the results had nothing to do with immigration. However, this effect leads Alesina and Glaeser (2004, 11) to predict that “if Europe becomes more heterogeneous due to immigration, ethnic divisions will be used to challenge the generous welfare state.” All in all, two mechanisms link immigration to the demand for income redistribution. First, the “anti-solidarity explanation” (Finseraas 2008) assumes that redistribution is the result of citizens’ altruistic preferences (Hochman and Rodgers 1969), and that individu2 In recent history, there has been a clear trend towards bringing democracies closer to universal suffrage. Women or different ethnic minorities have obtained the right to vote in the last century and several studies have proved that these events have led to an increase in the level of redistribution (Lott and Kenny 1999; Aidt et al. 2006; Aidt and Dallal 2008). In fact, franchise extension to poorer populations on average tends to increase inequalities among voters and, hence, leads to a greater political demand for redistribution. Note, also, that when the franchise is limited, franchise extension can reduce income taxation (Plümper and Martin 2003; Aidt and Jensen 2009). 3 For example, beyond special requirements that can be more or less demanding, in France, Turkey and the United Kingdom, immigrants can qualify for enfranchisement after five years of residence; in Germany, after eight years; in Italy, Poland and Spain after ten years; 12 years are required in Switzerland. Public Choice als are more altruistic toward their own ethnic group than toward other (Tajfel et al. 1971; Alesina and La Ferrara 2005). Therefore, it concludes that ethnic diversity reduces the demand for the welfare state.4 Second the “selfish-citizen explanation” assumes that redistribution is the result of citizens’ selfish preferences under majority rule (Meltzer and Richard 1981) and that immigrants are disenfranchised. It concludes that immigration affects the political demand for the welfare state in different ways.5 Both explanations differ mainly with respect to the role played by solidarity with immigrants, i.e., by whether the immigrants’ welfare is an argument in citizens’ utility function. The empirical consequences of these explanations are tested with a static micro-level analysis based on survey data. The next section details the hypothesis of a relationship between immigrants as nonvoters and redistribution. In Sect. 2, the dataset and the micro-level tests are described, and Sect. 3 reports the findings. Finally, Sect. 4 provides a broad discussion of the empirical and normative implications of the results. 1 Redistribution and universal suffrage This section develops the selfish-citizen explanation and provides four testable hypotheses to distinguish it from the anti-solidarity explanation. According to the selfish-citizen explanation, support for the welfare state can increase among citizens who feel disadvantaged in the labor market compared to immigrants (H1), while among other citizens this support decreases with immigration (H2). According to the anti-solidarity explanation, though, xenophobia should systematically reduce support for the welfare state (H3). These effects are expected to amplify when the number of immigrants is large (H4). 1.1 The Meltzer and Richard model The most famous political economy model for explaining the level of preference for redistribution has been provided by Meltzer and Richard (1981), hereafter M&R, who stated that the demand for redistribution depends on the gap between the mean income and the median income. The model starts with the assumption that agents are rational, self-interested, and fully aware, and it assumes that agents regard redistribution as a way to maximize their own incomes. M&R argue that, first, under a linear tax rate, with lump-sum and proportional redistribution, the agents’ preferences can be placed in a one-dimensional space: the richer the agents, the less favorable they are to introducing redistribution. Second, under majority rule, the median voter is decisive. So, the policy outputs as well as the majority’s preference correspond to the median voter’s preference. Third, under universal suffrage, a person who earns the median income is the median voter. Therefore, redistribution maximizes the earnings of the agent with median income. 4 Another variant is based on a concept of redistribution as a public good. Assuming that redistribution is a kind of public good and that individuals cooperate more with people who belong to their own ethnic group than with other people (Putnam 2007; Habyarimana et al. 2007), ethnic diversity will reduce the demand for the welfare state. 5 When immigrants have voting rights, immigration affects the demand for the welfare state as well. However, as I will show below, when immigrants are entitled to voting rights, the impact of immigration on redistribution is opposite to when immigrants cannot vote. Public Choice Hence the model asserts that the collective preference for redistribution is greater when the gap between median and mean income is wider, and that the optimal tax rate brings the median income closer to the mean income, while maintaining a level of inequality that provides taxpayers with incentives to optimally finance redistribution. Thus, the demand for redistribution is demand for redistribution = yK − ym , (1) where y is the mean income, ym the median income and K a coefficient that gauges the negative effect of a tax rate increase on productivity. The M&R model has been tested several times (Bradley et al. 2003; Iversen and Soskice 2006; Kenworthy and McCall 2008; Moene and Wallerstein 2003), but the results are at best uneven. However, recent public opinion research supplies evidence of a positive relationship between inequality and support for redistribution (Finseraas 2009; Georgiadis and Manning 2012). These differences in the evidence provided could be due to the complex causal chain implied in the model: inequality causes perceived inequality, which causes support for redistribution, which leads to redistribution. The simplest way to test M&R’s mechanism is by observing causality in the chain’s middle link: the perception of inequality leads to stronger support for redistribution. 1.2 Introducing partial suffrage In this study, I define “immigrants” as agents able to undertake economic decisions, but not political decisions, and “partial suffrage” as a situation in which such agents exist. The aim of this subsection is to analyze the consequences that can be expected when we replace universal suffrage with partial suffrage in the M&R model, i.e., when the population proportion of foreign-born residents increases. The first consequence is that the median income does not necessarily match the median voter anymore. Here, the median voter is the agent who earns the median income among citizens ymc . Hence, under partial suffrage, the demand for redistribution of ymc remains the same as under universal suffrage only if the mean income of citizens y c equals the mean income of immigrants yi . In this case, substituting it into the M&R formula y c nc + y i ni (2) K − ymc n does not produce any changes when immigrants arrive. However, when y c > y i , the mean income in the whole population falls below that in the citizen population. Thus the demand for redistribution should also decrease. In other words, the median income among citizens is—at the time of new immigrants’ arrival—too close to the general mean income, exactly as if too much redistribution has occurred. As a consequence, voters should ask for less redistribution. This mechanism suggests that the impact of the arrival of immigrants, who are on average poorer than citizens, is comparable to that of an over-redistributive situation under universal suffrage. Indeed, both configurations lead to declines in the demand for redistribution due to a reduction in the distance between the median and the mean income. Nevertheless, one difference stems from the fact that, on the one hand, over-redistributing leads to an increase in the median income, while the mean income remains stable and, on the other hand, low-income immigration decreases the mean income (at least temporarily) without affecting the median voter’s income, so the distance is reduced through a different mechanism. Demand for redistribution = Public Choice Moreover, when y c < y i , the mean income increases, and the median voter can maximize his/her income by demanding more redistribution. Thus, the immigration of high-income earners increases the demand for redistribution. Note that (2) also implies that the magnitude of this effect depends on the proportion of citizens in the general population. In other words, the smaller is the population percentage of immigrants (i.e., the closer the political system is to universal suffrage), the smaller the impact of y c = y i on demand for redistribution. 1.3 Dynamic models Until now, citizens have been considered only as people who assess their actual earnings before deciding their degree of support for redistribution. Yet, political economic explanations for redistributive policies argue that the level of exposure to labor market risks, especially unemployment, is the main determining factor of the individual preferences for redistribution (Cusack et al. 2006). So, in contrast to the M&R thesis, people are expected to anticipate their future income. In such conditions, two scenarios are possible. First, since redistribution is regarded as a simple transfer, individuals support redistribution when they expect to lose income and to have to deal with their own downward mobility (Benabou and Ok 2001). Alternatively, if redistribution is viewed as insurance, support for it increases when people perceive the risk of future income loss. In this case, people do not necessarily expect that their incomes will fall, but rather that the standard deviation of their future income will increase. Thus, assuming citizens’ aversion to risk, they will demand more redistribution. Both scenarios are compatible with a kind of perceived threat from immigrants. When the latter are regarded as competitive in the labor market, some citizens fear losing their jobs or part of their incomes, or, simply, suffering from downward mobility. In this case, immigrants introduce both risky competition with citizens and, on average, a downward mobility, sometimes associated with an income loss for citizens. Contrary to the static M&R model, immigrants can produce fear of labor market instability even when they are, on average, unskilled and relatively poor, because they compete for jobs with those citizens who benefit from redistribution. In contrast, when immigrants are regarded as non-competitive in the labor market, citizens who do not fear income losses, benefit from upward mobility due to the arrival of new low-income stakeholders who reduce the mean income. In particular, the median voter, who is a native-born member, would now lose under the current income tax rate because a larger share of the revenues they generate would benefit the migrants whose number has increased. So, citizens who earn the median voter’s income or above now support a lower tax rate (Razin et al. 2002). Note, however, that this argument can be formulated in two different ways. First, immigrants affect the desire of the median voter for redistribution according to whether immigrants are expected to earn above or below the median income. Second, immigrants may also affect the support for redistribution depending on whether they are expected to decrease citizens’ incomes through labor market competition or through tax-financed wealth transfers. These different analyses may have different psychological effects on, for example, xenophobia. People prefer richer immigrants who increase the size of the cake to competitive immigrants who make their own situations unstable. However, as far as redistribution is concerned, both analyses are conjectured to have the same impact: in both cases individuals become relatively poorer compared to the mean income, even though in the second case Public Choice they also risk becoming absolutely poorer, while this is not the case in the first interpretation. 1.4 Testable hypotheses Three testable hypotheses are provided to test the selfish-citizen explanation: (H1) when citizens believe that immigrants take jobs away from natives, their preference for redistribution increases and (H2) when citizens believe that immigrants strain the welfare state, their preference for redistribution declines. Notice that only H2 is compatible with the anti-solidarity explanation. In order to clarify this point empirically, it should also be expected that (H3) Demand for redistribution also depends on the lack of solidarity with immigrants. Moreover, the expected effects described in H1 and H2 should be amplified when the respondents believe that the proportion of immigrants in the population is large. (H4) The perceived population share of immigrants increases the validity of H1 and H2. Finally, the impact of specific beliefs about immigrants—such as those listed above—should exclusively influence citizens, at least in the expected direction. 2 Data and methodology I use data from the last round of the European Values Survey (EVS) conducted in 2008. These data are interesting for several reasons. First, they cover 45 countries and almost 70,000 respondents (countries and respondents per country are described in the appendix). Most countries are quite poor and receive “white collar” immigration. They are rarely studied and, for all these reasons, they are worth taking into account. Also, this database asks questions that are well adapted to the aim of this article. Only citizens are taken into account in the estimations, meaning people within the country who are thus entitled to vote. 2.1 Demand for equality and for social protection The demand for redistribution is the dependent variable. It is difficult, in the real world, to identify how to measure redistribution. It can be regarded both as a form of insurance or as a way to equalize incomes. For this reason, I have used two indicators. The first one measures support for redistribution as a way of protecting people from risks. The respondents had to place themselves on a ten-point scale: “The government should take more responsibility to ensure that everyone is provided for” (9) versus “People should take more responsibility to provide for themselves” (0). This question indicates the demand for social protection, meaning: insurance-type redistribution. The second indicator captures the preference for redistribution based on equalizing incomes. The scale is: “Incomes should be made more equal” (9) versus “We need larger income differences as incentives” (0). This second dependent variable is called demand for equality. Public Choice Table 1 Statistical description of variables* Mean s.d. min max Dependent variable Demand for equality 4.68 0.011 0 9 Demand for social protection 4.01 0.011 0 9 Immigrants take jobs away 4.94 0.011 0 9 Immigrants strain the welfare state 5.42 0.011 0 9 −0.49 0.011 −9 9 Concerned by fellow-citizens 2.00 0.004 0 4 Concerned by immigrants 1.75 0.004 0 4 Too many immigrants 2.16 0.005 1 5 −1 1 1 10 About immigrants Relative competitiveness Individual control variables Preference for freedom 0.095 0.004 Right-wing 5.38 0.01 Female 0.002 1 2 46.37 0.07 14 108 Education 3.08 0.005 1 Income 1.212 0.005 0 Religious attendance 2.390 0.007 0 Age 1.554 6 14.73 6 * For all variables but income, the unit is 1. The variable income includes 611 values computed from specific scales for each country, adjusted for PPP Surprisingly, neither indicator is correlated with the other. Besides, both of them appear to be continuous and normally distributed variables, in which the value 4 is the mean, the median and the mode of the demand for social protection, whereas this value is 5 in the case of demand for equality (see Table 1). 2.2 Perceived immigrant competitiveness and solidarity with immigrants According to the selfish-citizen explanation, the perception of immigrants’ labor market competitiveness has an impact on citizens’ attitudes toward redistribution, regardless of the extent of solidarity with foreign-born arrivals. The main concern of this prediction is separating the competition effect from the solidarity effect. Indeed, when immigrants are perceived as competitive, they are also more accepted by the people who earn less than the median income, because immigrants contribute to paying taxes, and by rich people who are keen to share the fiscal burden. This fact has been highlighted by Hainmueller and Hiscox (2010), who demonstrate that, regardless of their incomes, people are less xenophobic toward highly skilled immigrants. Also, Senik et al. (2009) have found a correlation between perceptions of immigrants’ contributions to national prosperity and attitudes toward immigration. In order to avoid this problem, I need to identify four groups: xenophobic and non-xenophobic people, on the one hand, and citizens who believe that immigrants are competitive in the labor market, or not, on the other. This means that people can be xenophobic because they think that immigrants are competitive with their fellow citizens and, conversely, they can be unxenophobic when they perceive immigrants as non-competitive. Public Choice Hence, I use two questions that ask respondents to place themselves on a ten-point scale: (A) Immigrants take jobs away: Immigrants take jobs away from natives (9) versus immigrants do not take jobs away from natives (0) (B) Immigrants strain the welfare state: Immigrants are a strain on the welfare system (9) versus immigrants are not a strain on the welfare system (0) On the first scale, immigrants are regarded as dangerous because they actually compete with citizens in the labor market. So, immigrants are considered to be at a potential advantage compared to citizens as far as income is concerned. On the second scale, immigrants are viewed as poor and disadvantaged when competing with citizens. Note that both scales can measure some kind of xenophobic attitude: whatever the immigrants’ level of competitiveness, they are perceived negatively. However, in case A xenophobia should lead to a demand for redistribution, because immigrants threaten citizens with downward mobility, while in case B citizens should oppose redistribution because noncompetitive immigrants facilitate citizens’ upward mobility. It is also worth noting that the correlation between both variables is high (r = 0.52). This could indicate that many people answer both questions in similar ways. In fact, there is a set of people that are systematically opposed or systematically favorable to immigrants. Hence, whatever immigrants do, respondents will keep seeing them either as a threat to or as a benefit for society. Nevertheless, in this paper, I argue that, whatever the level of xenophobia, only the relative perception of the immigrants’ average economic situation is relevant. That is why I also use a variable that is the result of subtracting immigrants take jobs away from immigrants strain the welfare state (called relative competitiveness hereafter). Thus, the larger is this value, the more competitive immigrants are deemed to be. This indicator does not consider the level of xenophobia, because people answering 9 for both questions have the same score as the individuals answering 0 for both questions, but it captures the intensity of the preference for one idea of immigrants over the other. Therefore, it measures why people are more sensitive to one issue than another, regardless of their degree of xenophobia. Regarding negative beliefs about immigrants, existing studies measure attitudes toward immigrants mainly by offering options about their perceived numbers (Mayda 2006; Hainmueller and Hiscox 2010; O’Rourke and Sinnott 2004; Givens 2004). In the EVS survey, the most widely used question is whether respondents agree with the statement “Today in my country, there are too many immigrants”. Respondents had to choose among five responses: strongly agree, agree, neither agree nor disagree, disagree and disagree strongly (see Simon and Lynch 1999; Givens 2004). This question is correlated with both Immigrants take jobs away (0.41) and Immigrants strain the welfare state (0.44), but not correlated with Relative competitiveness. However, it provides an ambiguous piece of information: it measures both the perceived number of immigrants and negative attitudes towards them (“too many”). So, this question can be viewed as a measurement of generic xenophobia, as well as the perceived proportion of immigrants in the country. When this variable is interacted with Immigrants take jobs away and Immigrants strain the welfare state (which encapsulate negative attitudes towards immigrants), its effect plausibly captures only the perceived number of immigrants.6 6 For interaction terms, I computed the items to obtain a variable using two values: 1 when respondents strongly agree or agree with the statement and null otherwise. Public Choice Finally, in line with the explanations based on solidarity, the demand for redistribution arises when people feel solidarity with fellow-citizens or with immigrants. I use two variables which measure the level of solidarity with those populations. Citizens were asked to what extent they are concerned by the living conditions of fellow countrymen (Concerned by fellow-citizens) and of immigrants (Concerned by immigrants), on a five-point scale. 2.3 Individual and contextual control variables Above all, I have to test that beliefs about immigrants’ labor market competitiveness is a cause, and not an effect, of attitudes towards redistribution. To identify the specific impact of beliefs about immigration on the demand for redistribution, the general attitudinal predisposition towards redistribution is held constant. Basic values in favor of the free market (hereafter Preference for freedom) are measured by the respondent’s choice between freedom and equality.7 Another dimension potentially able to account for both perceptions of immigrants and support for redistribution is the right-wing or left-wing identification (Van Der Brug and Van Spanje 2009). Within this framework, people located on the left side of the political scale are expected to favor immigration and to oppose the market. I used the classic selfplacement on the ten-point scale in which 1 is “left” and 10 is “right”. The variable is, therefore, called right-wing. Beyond these attitudinal controls, I enter several structural independent variables. Income is a decisive variable and the most appropriate indicator of this is the monthly household income in Euros (×1000), adjusted for purchasing power parity (PPP).8 Finally, I have taken into account three major demographic characteristics: Gender (1 = female), Age and Education (in six points given by the ISCED code). Older people and females are more likely to support redistribution (Blekesaune and Quadagno 2003; Keely and Tan 2008); religious attendance traditionally is associated with less support for the welfare state. At the aggregate level, studies of the impact of inequality on the demand for redistribution use the Gini coefficient to measure the skewness of the income distribution. Its effect has sometimes been observed—in line with the M&R hypothesis—as positive (Aalberg 2003; Finseraas 2009), sometimes as insignificant (Lübker 2007; Kenworthy and McCall 2008) and some studies point to a negative relationship (Bowles and Gintis 2000). According to the M&R hypothesis, Gini impacts the demand for redistribution only among the people who earn the median income or less. Therefore, the Gini coefficient is interacted with belonging to the median income or less category (Gini—low income) and it is expected to have a positive impact, especially on the demand for equality. On the other hand, the demand for social protection, which implies insurance-type redistribution, can be more influenced by the country’s degree of openness to trade (Rodrik 1998). Immigrants are not the only reason for income-loss risks: the domestic labor market is more vulnerable to turbulence on the world markets when the economy is more open. In line with Finseraas (2009), I measure 7 Respondents who have chosen the statement “I would consider personal freedom more important, that is, everyone can live in freedom and develop without hindrance” is coded 1, while people who believe that equality is more important, “that is, that nobody is underprivileged and that social class differences are not so strong” equals −1. Note that about 6 % of respondents did not agree with either of these statements. They are coded with 0. 8 PPP is derived from the IMF World Economic Outlook Database (October 2008 Edition). Public Choice trade openness (in natural logarithms), using exports plus imports as a share of GDP, in constant prices. However, since this mechanism concerns the people whose income is more affected by market turbulence, trade openness is interacted with low levels of education, i.e., the fraction of the population with a secondary education or less (trade openness (ln)—low education). Low education is here a proxy for vulnerability to market instability. 3 Results 3.1 Individual and country level correlations Before proceeding to a multivariate analysis, simple correlations at the individual and country levels are computed. Table 2 correlates the demands for equality and for social protection with the variables related to solidarity with immigrants and their labor market competitiveness. Each country’s score is the mean of the individual answers given by the citizens of that country. Interestingly, the results are ambiguous with respect to the demand for equality, while the demand for social protection clearly is affected by the perceived degree of immigrants’ competitiveness. More accurately, at an individual level, the demand for equality is associated positively with negative opinions about immigrants, but it is also positively correlated with feeling concerned by immigrants (and, to a lesser extent, by citizens), in line with the anti-solidarity hypothesis. No significant correlations are registered at the country level. The results are more interesting with respect to the demand for social protection. The latter is positively correlated with believing that immigrants take jobs away from citizens and negatively with believing that immigrants strain the welfare state. Subtracting one from the other (relative competitiveness) leads to a difference that is strongly and positively correlated with the demand for social protection. These results are confirmed both at the country and individual level. Solidarity with citizens and immigrants has a positive impact, while thinking that there are too many immigrants surprisingly remains positively associated with redistribution at the individual level. While these results have to be confirmed with a multivariate analysis, they suggest that when redistribution is regarded as a form of insurance, people assess the way in which their Table 2 Correlations at country level and at individual level in 45 European countries Immigrants take jobs away Immigrants strain the welfare state Relative competitiveness Concerned by fellow-citizens Concerned by immigrants Too many immigrants Demand for equality Demand for social protection Country level Country level Individual level 0.30∗∗ 0.02∗∗ Individual level 0.18 0.10∗∗ 0.15 0.06∗∗ 0.06 0.05∗∗ −0.05 0.03∗∗ 0.10 0.003∗∗ 0.23 0.05∗∗ 0.04 0.003∗∗ 0.13 0.05∗∗ −0.04 −0.27∗ 0.54∗∗ ∗∗ Significant 5 % level ∗ significant 10 % level Each country’s score is the mean of the individual answers given by the citizens of that country −0.05∗∗ 0.07∗∗ 0.01∗∗ Public Choice Table 3 Estimations of citizens’ demand for equality and social protection with structural control variables Demand for equality Demand for social protection OLS Probit OLS Probit Immigrants take jobs away 0.077∗∗∗ (0.011) 0.031∗∗∗ (0.005) 0.013 (0.012) 0.005 (0.005) Immigrants strain the welfare state 0.010 (0.016) 0.002 (0.007) −0.037∗∗ (0.014) −0.013∗∗ (0.006) Female 0.154∗∗∗ (0.043) 0.067∗∗∗ (0.019) 0.228∗∗∗ (0.028) 0.076∗∗∗ (0.014) Age 0.001 (0.001) 0.001 (0.001) 0.000 (0.001) −0.000 (0.001) Education −0.063∗∗∗ (0.010) −0.059∗∗ (0.009) −0.106∗∗∗ (0.026) −0.048∗∗∗ (0.011) Income −0.070∗∗∗ (0.015) −0.075∗∗ (0.014) −0.106∗∗ (0.026) −0.057∗∗∗ (0.012) Religious attendance 0.025∗ (0.015) 0.007 (0.007) 0.007 (0.016) 0.003 (0.007) Gini × low income 0.001 (0.002) 0.001 (0.001) 0.008∗∗∗ (0.002) 0.003∗∗∗ (0.001) ln (Trade Openness) × low education 0.183∗∗ (0.083) 0.063 (0.043) 0.230∗ (0.126) 0.119∗∗∗ (0.051) Constant 5.023∗∗∗ (0.165) 0.335∗∗∗ (0.085) 3.619∗∗∗ (0.217) −0.302∗∗∗ (0.088) N 47309 47309 47448 47448 R2 0.15 0.09 0.07 0.04 Standard errors, corrected with the cluster method (by country), are in parentheses We include country fixed effects ∗∗ p < 0.10, ∗ p < 0.05, ∗∗∗ p < 0.01 own situation in the labor market could be affected by immigration, regardless of their level of solidarity with immigrants. In contrast, when redistribution is simply a way of equalizing incomes, the results do not have a clear interpretation. 3.2 Multivariate estimations with structural controls In order to estimate our ten-point scale dependent variable, several methods have been used, but, whatever the method, the results do not change. I choose to mainly show here the results from a simple ordinary least squares method, as if the relationships were thus, because the results are more easily interpretable. However, Table 3 compares OLS and Probit estimations with structural controls.9 Both sets of regressions also include country fixed effects and error variances are corrected using the cluster method. This largely ensures the quality of the estimations. 9 In the Probit model, the dependent variable is dummy (I have considered the answers from 1–5 values as 0, and the values from 6 up to 10, as 1). Public Choice Table 3 offers two estimations of the demand for equality and two for the demand for social protection. In each case, the first column uses OLS estimation (take jobs away and strain the welfare state), while the second column uses Probit estimation. Only structural control variables are taken into account, to avoid any concern about endogeneity. In the four columns, the classic individual control variables have the expected impact: income, education, and being male have a negative impact on the demand for redistribution. Religious attendance and age have no influence. The impact of contextual variables globally confirms the hypotheses tested, with some caveats. The Gini coefficient significantly and positively affects the demand for social protection, while it has no impact on the demand for equality. This result confirms the political economy model even though it was expected to better explain the demand for equality than the demand for social protection. On the other hand, trade openness has the expected impact on both the demand for social protection and for equality (in the latter case, only in the OLS estimation). Concerning immigrants, the variable Immigrants take jobs away has a positive impact on the demand for redistribution, but it is significant only in the case of the demand for equality, while the impact of Immigrants strain the welfare state is significant and negative as far as the demand for social protection is concerned. This result partly confirms hypotheses H1 and H2. 3.3 Multivariate estimations with structural and attitudinal controls Tables 4 and 5 present several OLS estimations with different specifications that take into account structural and attitudinal control variables. Including attitudinal controls does not change the global findings reported previously. However, they allow us to shed light on hypotheses H3 and H4 and on the relationship between the different motives associated with redistribution. The results globally confirm the selfish-citizen hypothesis (H1, H2, H4). The anti-solidarity hypothesis (H3) is not clearly invalidated, even though the results are somewhat fuzzy. The control variables are steady and in line with what Table 3 already established. The attitudinal variables not related to the immigration issue (Preference for freedom and Right-wing) negatively influence the demand for redistribution in all estimations, as expected. Table 4 estimates the demand for social protection. Column 1 shows, as in Table 3, the opposite impact of Immigrants take jobs away and of Immigrants strain the welfare state, with only the latter significant at the 5 % level. The same effect is confirmed in column 2. Feeling concerned by fellow-citizens does not significantly influence support for social protection, but caring about immigrants does, at the 10 % level (column 1). However, thinking that there are too many immigrants (column 2) does not have the expected negative impact, but a positive one. These results go a small way to confirming the idea that solidarity with immigrants (H3) can explain support for redistribution. Columns 3 and 4 replicate the previous estimations, but use the Relative competitiveness variable, and the results do not change. Finally, columns 5 and 6 test H4. The impact of Immigrants takes jobs away and of Immigrants strain the welfare state is expected to be amplified when people perceive that there are “too many” immigrants. This expectation is confirmed only for the first hypothesis, not for the second. In particular, while the Immigrants take jobs away coefficient is not significant in standard specifications (columns 1 and 2), it becomes significant when −0.292∗∗∗ (0.032) −0.071∗∗∗ (0.021) 0.203∗∗∗ (0.032) −0.002∗∗ (0.001) −0.090∗∗∗ (0.026) −0.089∗∗∗ (0.024) 0.010 (0.015) 0.007∗∗∗ (0.001) 0.175 (0.025) 4.259∗∗∗ (0.228) 35718 0.09 4.136∗∗∗ (0.238) 36290 0.09 Constant N R2 4.018∗∗∗ (0.217) 36290 0.09 −0.291∗∗∗ (0.031) −0.071∗∗∗ (0.021) 0.204∗∗∗ (0.031) −0.002∗∗ (0.001) −0.095∗∗∗ (0.026) −0.089∗∗∗ (0.024) 0.006 (0.015) 0.007∗∗∗ (0.001) 0.191 (0.021) 0.021∗∗ (0.010) 0.017 (0.034) 0.056∗ (0.029) (3) coef. (s.e.) 4.100∗∗∗ (0.214) 35718 0.09 −0.292∗∗∗ (0.032) −0.072∗∗∗ (0.021) 0.207∗∗∗ (0.032) −0.002∗∗ (0.001) −0.085∗∗∗ (0.026) −0.087∗∗∗ (0.024) 0.010 (0.015) 0.007∗∗∗ (0.002) 0.174 (0.125) 0.031 (0.030) 0.021∗∗ (0.010) (4) coef. (s.e.) 4.190∗∗∗ (0.241) 36290 0.09 −0.290∗∗∗ (0.031) −0.071∗∗∗ (0.020) 0.200∗∗∗ (0.031) −0.002∗∗ (0.001) −0.094∗∗∗ (0.026) −0.090∗∗∗ (0.024) 0.007 (0.015) 0.007∗∗∗ (0.001) 0.183 (0.120) 0.003 (0.014) 0.030∗∗ (0.013) 0.018 (0.032) 0.053∗ (0.028) −0.009 (0.012) −0.039∗∗∗ (0.014) (5) coef. (s.e.) Standard errors, corrected with the cluster method (by country), are in parentheses. We include country fixed effects. ∗∗ p < 0.10, ∗ p < 0.05, ∗∗∗ p < 0.01 0.072∗∗ (0.028) 0.001 (0.012) −0.44∗∗∗ (0.014) −0.290∗∗∗ (0.031) −0.070∗∗∗ (0.020) 0.202∗∗∗ (0.031) −0.002∗ (0.001) −0.100∗∗∗ (0.026) −0.090∗∗∗ (0.024) 0.006 (0.015) 0.007∗∗∗ (0.001) 0.195 (0.121) −0.020 (0.033) 0.047∗ (0.028) 0.012 (0.013) −0.033∗∗ (0.013) (2) coef. (s.e.) Immigrants take jobs away Immigrants strain the welfare state Relative competitiveness Concerned by fellow-citizens Concerned by immigrants Too many immigrants Immigrants take jobs away × too many immigrants Immigrants strain the welfare state × too many immigrants Relative competitiveness × too many immigrants Preference for freedom Right-wing Female Age Education Income Religious attendance Gini × low income ln (Trade Openness) × low education (1) coef. (s.e.) Table 4 Estimations of citizens’ demand for social protection 4.016∗∗∗ (0.218) 36290 0.09 −0.291∗∗∗ (0.031) −0.071∗∗∗ (0.020) 0.204∗∗∗ (0.031) −0.002∗∗ (0.001) −0.095∗∗∗ (0.026) −0.089∗∗∗ (0.024) 0.006 (0.015) 0.007∗∗∗ (0.001) 0.191 (0.121) 0.007 (0.013) 0.018∗ (0.010) 0.017 (0.034) 0.056∗ (0.029) (6) coef. (s.e.) Public Choice Public Choice it is interacted with the perceived number of immigrants. When both variables are combined into Relative competitiveness, however, the results are also not significant at the 5 % level. Let us now consider the demand for equality (Table 5). The positive effect of Immigrants take jobs away is confirmed, but the variable Immigrants strain the welfare state is now insignificant. Taken together (columns 3 and 4), the expected effect is observed. Again, the most notable effect is provided in columns 5 and 6. When these variables are interacted with too many immigrants, the effect is clearly amplified in both cases. In particular, Immigrants strain the welfare state regains its significant and negative influence on the demand for equality when respondents think that there are too many immigrants. On the other hand, the coefficient Immigrants takes jobs away is simply larger when immigrants are viewed as too numerous. Finally, variables measuring hostility to immigrants do not have their expected signs. Solidarity with immigrants has a barely significant impact (10 % level), and its sign is positive. Hypothesis 3 is, however, better confirmed in terms of solidarity with fellow citizens, which positively influences the demand for redistribution. All in all, the most consistent finding, reinforced by the perceived number of immigrants, is the positive effect of Immigrants take jobs away. So, H1 is confirmed in most estimations. Immigrants strain the welfare state tends to maintain the expected negative sign, even though, with respect to support for equality, this effect is visible only when the variable is interacted with too many immigrants. So, H2 is also broadly confirmed. The evidence for H3 is weak. Generally negative attitudes towards immigrants do not seem to have the expected influence. On the other hand, solidarity with others (with immigrants and with fellow citizens) affects the demand for redistribution. However, solidarity specifically regarding immigrants seems to have no significant effect. According to hypothesis 4, H1 and H2 should be truer when they are interacted with Too many immigrants. This is confirmed in three cases out of four. Thus, H4 is globally validated. 4 Conclusion This article has compared two concurrent explanations for the observed link between immigration and attitudes toward income redistribution; it operationalized two concepts of redistribution: the first is based on equality of income, and the second on social insurance. The anti-solidarity explanation argues that lack of concern for immigrants’ welfare is responsible for the (negative) link between immigration and the demand for redistribution. Our data partly confirm this hypothesis since, on average, people are less concerned about the living conditions of immigrants than about those of fellow citizens. Moreover, the data show that, at an individual level, there is a positive correlation between solidarity for immigrants and the demand for more redistribution. However, after a multivariate analysis with appropriate controls, the anti-solidarity explanation turns out to be globally weak: indeed, the influence of solidarity is unsteady but robust only with respect to fellow citizens. On the other hand, the selfish-citizen explanation states that citizens’ support for redistribution depends on their expectations about future incomes and their anticipated position on the income scale. Those expectations are influenced by the economic changes produced 5.792∗∗∗ (0.185) 35776 0.16 5.621∗∗∗ (0.186) 36346 0.17 Constant N R2 6.085∗∗∗ (0.197) 36346 0.16 −0.219∗∗∗ (0.028) −0.129∗∗∗ (0.024) 0.094∗∗ (0.041) 0.001 (0.001) −0.142∗∗∗ (0.024) −0.136∗∗∗ (0.030) 0.039∗∗ (0.015) 0.001 (0.002) 0.207∗∗ (0.101) 0.037∗∗∗ (0.011) 0.076∗∗ (0.037) 0.022 (0.031) (3) coef. (s.e.) 6.071∗∗∗ (0.184) 35776 0.16 −0.231∗∗∗ (0.029) −0.134∗∗∗ (0.025) 0.101∗∗ (0.041) 0.001 (0.001) −0.127∗∗∗ (0.022) −0.137∗∗∗ (0.029) 0.042∗∗∗ (0.015) 0.001 (0.002) 0.180∗ (0.093) 0.081∗∗ (0.033) 0.035∗∗∗ (0.011) (4) coef. (s.e.) 9.616∗∗ (0.316) 36346 0.17 −0.220∗∗∗ (0.029) −0.134∗∗∗ (0.025) 0.097∗∗ (0.041) 0.000 (0.001) −0.119∗∗∗ (0.022) −0.132∗∗∗ (0.029) 0.037∗∗ (0.014) 0.001 (0.002) 0.180∗ (0.097) −0.035∗∗ (0.017) 0.062∗∗ (0.028) 0.062∗ (0.035) 0.062∗ (0.032) 0.045∗∗ (0.019) 0.019 (0.015) (5) coef. (s.e.) Standard errors, corrected with the cluster method (by country), are in parentheses. We include country fixed effects. ∗∗ p < 0.10, ∗ p < 0.05, ∗∗∗ p < 0.01 0.010 (0.032) −0.231∗∗∗ (0.029) −0.136∗∗∗ (0.025) 0.108∗∗ (0.041) 0.001 (0.001) −0.118∗∗∗ (0.021) −0.133∗∗∗ (0.029) 0.042∗∗∗ (0.014) 0.001 (0.002) 0.177∗ (0.091) 0.063∗ (0.035) 0.058∗ (0.032) −0.220∗∗∗ (0.029) −0.134∗∗∗ (0.025) 0.100∗∗ (0.041) 0.000 (0.001) −0.123∗∗∗ (0.022) −0.132∗∗∗ (0.029) 0.037∗∗ (0.014) 0.001 (0.002) 0.187∗ (0.098) 0.072∗∗∗ (0.011) 0.005 (0.013) 0.076∗∗∗ (0.011) 0.008 (0.015) Immigrants take jobs away Immigrants strain the welfare state Relative competitiveness Concerned by fellow-citizens Concerned by immigrants Too many immigrants Immigrants take jobs away × too many immigrants Immigrants strain the welfare state × too many immigrants Relative competitiveness × too many immigrants Preference for freedom Right-wing Female Age Education Income Religious attendance Gini × low income Trade Openness (nl) × low education (2) coef. (s.e.) (1) coef. (s.e.) Table 5 Estimations of citizens’ demand for equality 6.076∗∗∗ (0.198) 36346 0.16 −0.219∗∗∗ (0.028) −0.128∗∗∗ (0.024) 0.092∗∗ (0.041) 0.001 (0.001) −0.144∗∗∗ (0.024) −0.137∗∗∗ (0.030) 0.039∗∗ (0.015) 0.001 (0.002) 0.211∗∗ (0.101) 0.038∗ (0.022) 0.022 (0.015) 0.076∗∗ (0.036) 0.020 (0.031) (6) coef. (s.e.) Public Choice Public Choice by the entry of new workers into the labor market. Within this framework, solidarity with immigrants does not play any causal role, and citizens assess their present or future position in order to evaluate if they will benefit personally from more or less redistribution. This mechanism is broadly confirmed by our data. Indeed, when immigration is viewed as leading to more instability in the labor market, citizens opt for more redistribution. On the other hand, when immigrants are seen as being less educated and less skilled people who will benefit from income transfers, citizens anticipate, on average, that they themselves will gain from redistribution, leading them to display less support for it. Moreover, this double effect proved to be stronger when citizens believe that there are ‘too many’ immigrants. This psychological mechanism is the necessary condition for the selfish-citizen explanation. Its consequence is that the demand for redistribution changes with immigration, except when immigrants’ profiles are similar to those of citizens. To conclude to the idea that immigration affects not only the demand for redistribution, but actual redistribution, two assumptions are required. First, redistribution policies are a response to citizens’ demands. Some authors remain skeptical about this causality (Kenworthy 2009), which is why this paper does not take any stance on this point. Second, people correctly perceive immigrants’ numbers and labor market competitiveness, which largely also is questionable: according to our data, individuals’ perceptions of the number of immigrants are significantly, but weakly correlated with their effective number (0.04). Besides, according to recent studies, and contrary to what around one of three respondents in our sample believe, foreign labor likely does not have a relevant negative effect on the natives’ job opportunities (D’Amuri et al. 2010; Longhi et al. 2006; Schmidt and Jensen 2012). However, if we assume that these assumptions are valid, the question whether or not to give immigrants the right to vote surely would have a major impact on debates on the welfare state. For example, we can determine whether the optimal level of redistribution is more easily reached when immigrants are enfranchised, or when they are not. Also, if maximizing political inclusiveness leads to rising prosperity (Acemoglu and Robinson 2012), then extending voting rights to all economic agents can favor economic development. Whatever the answer, according to this empirical analysis on individual preferences for redistribution, the issue of immigrants’ voting rights certainly appears to be an important element in the debate about redistribution and the economy. Acknowledgements The author would like to thank André Blais and the other members of the Canada Research Chair in Electoral Studies, Abel François, Anna Jeannesson, and Robert Kowalenko for their helpful comments. Finally, I would like to stress the excellent comments from two anonymous reviewers who considerably improved this article. Appendix Column I: Observations included in each country (peoples who have answered both questions about redistribution). Column II and III: Available data on foreigners (persons who are not citizens of the country of residence) per country in the sample and according to Eurostat (Vasileva 2011). The correlation is r = .94. Public Choice Albania Azerbaijan Austria Armenia Belgium Bosnia Herzegovina Bulgaria Belarus Croatia Cyprus Czech Republic Denmark Estonia Finland France Georgia Germany Greece Hungary Iceland Ireland Italy Latvia Lithuania Luxembourg Malta Moldova Montenegro Netherlands Norway Poland Portugal Romania Russian Federation Serbia Slovak Republic Slovenia Spain Sweden Switzerland Turkey Ukraine Macedonia Great Britain Kosovo Respondents Foreigners (sample) (%) 1455 1157 1455 1446 1501 1475 1413 1445 1450 929 1733 1478 1483 1058 3039 1392 2041 1482 1486 797 965 1431 1474 1386 1549 1389 1402 1427 1543 1085 1475 1462 1380 1431 1458 1430 1341 1383 1111 1230 2226 1470 1428 2004 1521 1 0 2 1 6 1 1 2 0 6 1 3 16 0 4 0 4 7 0 2 4 0 22 0 47 2 8 5 2 5 0 4 0 0 0 0 6 9 4 19 0 1 1 8 20 Foreigners (Eurostat) (%) 11 10 16 4 6 16 3 6 8 2 7 9 7 17 1 43 4 4 7 0 4 1 4 12 6 22 7 References Aalberg, T. (2003). Achieving justice: comparative public opinion on income distribution. Leiden: Brill. Acemoglu, D., & Robinson, J. A. (2012). Why nations fail. New York: Random House. Aidt, T. S., & Dallal, B. (2008). Female voting power: the contribution of women’s suffrage to the growth of social spending in western Europe (1869–1960). Public Choice, 134, 391–417. Aidt, T. S., & Jensen, P. S. (2009). The taxman tools up. Journal of Public Economics, 93, 160–175. Aidt, T. S., Dutta, J., & Loukoianovac, E. (2006). Democracy comes to Europe: franchise extension and fiscal outcomes. European Economic Review, 50, 249–283. Public Choice Alesina, A., & Glaeser, E. L. (2004). Fighting poverty in the U.S. and in Europe: a world of difference. New York: Oxford University Press. Alesina, A., & La Ferrara, E. (2005). Preferences for redistribution in the land of opportunities. Journal of Public Economics, 89(5–6), 897–931. Alesina, A., Glaeser, E., & Sacerdote, B. (2001). Why doesn’t the U.S. have a European-style welfare system? Brookings Papers on Economic Activity, 32(2), 187–278. Benabou, R., & Ok, E. (2001). Social mobility and the demand for redistribution: the POUM hypothesis. The Quarterly Journal of Economics, 116(2), 447–487. Blekesaune, M., & Quadagno, J. (2003). Public attitudes toward welfare state policies: a comparative analysis of 24 nations. European Sociological Review, 19(5), 415–427. Bowles, S., & Gintis, H. (2000). Reciprocity, self-interest and the welfare state. Nordic Journal of Political Economy, 30, 33–53. Bradley, D., Huber, E., Moller, S., Nielsen, F., & Stephens, J. D. (2003). Distribution and redistribution in postindustrial democracies. World Politics, 55, 193–228. Cusack, T., Iversen, T., & Rehm, P. (2006). Risk at work: the demand and supply sides of government redistribution. Oxford Review of Economic Policy, 22(3), 365–389. D’Amuri, F., Ottaviano, G. I. P., & Peri, G. (2010). The labor market impact of immigration in western Germany in the 1990s. European Economic Review, 54, 550–570. Eger, M. A. (2010). Even in Sweden: the effect of immigration on support for welfare state spending. European Sociological Review, 26(2), 203–217. Finseraas, H. (2008). Immigration and preferences for redistribution: an empirical analysis of European survey data. Comparative European Politics, 6(4), 407–431. Finseraas, H. (2009). Income inequality and demand for redistribution. Scandinavian Political Studies, 32(1), 94–119. Georgiadis, A., & Manning, A. (2012). Spend it like Beckham? Inequality and redistribution in the UK, 1983–2004. Public Choice, 151(3–4), 537–563. Gilens, M. (1999). Why Americans hate welfare: race, media, and the politics of anti-poverty policy. Chicago: University of Chicago Press. Givens, T. E. (2004). The radical right gender gap. Comparative Political Studies, 37(1), 30–54. Habyarimana, J., Humphreys, M., Posner, D. D., & Weinstein, J. M. (2007). Why does ethnic diversity undermine public goods provision? American Political Science Review, 101(4), 709–725. Hainmueller, J., & Hiscox, M. J. (2010). Attitudes toward highly skilled and low-skilled immigration: evidence from a survey experiment. American Political Science Review, 104(1), 61–84. Hansen, J. D. (2003). Immigration and income redistribution in welfare states. European Journal of Political Economy, 19, 735–746. Hochman, H. M., & Rodgers, J. D. (1969). Pareto optimal redistribution. The American Economic Review, 59, 542–557. Iversen, T., & Soskice, D. (2006). Electoral institutions and the politics of coalitions: why some democracies redistribute more than others. American Political Science Review, 100, 165–181. James, E. (1993). Why do different countries choose a different public-private mix of educational services? The Journal of Human Resources, 28(3), 571–592. Keely, L. C., & Tan, C. M. (2008). Understanding preferences for income redistribution. Journal of Public Economics, 92(5–6), 944–961. Kenworthy, L. (2009). The effect of public opinion on social policy generosity. Socio-Economic Review, 7(4), 727–740. Kenworthy, L., & McCall, L. (2008). Inequality, public opinion, and redistribution. Socio-Economic Review, 6, 35–68. Lind, J. T. (2007). Fractionalization and the size of government. Journal of Public Economics, 91, 51–76. Longhi, S., Nijkamp, P., & Poot, J. (2006). The fallacy of “job robbing”: a meta-analysis of estimates of the effect of immigration on employment. Journal of Migration & Refugee, 1, 131–151. Lott, J. R., & Kenny, L. W. (1999). Did women’s suffrage change the size and scope of government? Journal of Political Economy, 107, 1163–1198. Lübker, M. (2007). Inequality and the demand for redistribution: are the assumptions of the new growth theory valid? Socio-Economic Review, 5, 117–148. Luttmer, E. F. (2001). Group loyalty and the taste for redistribution. Journal of Political Economy, 109(3), 500–528. Mayda, A. M. (2006). Who is against immigration? A cross-country investigation of individual attitudes toward immigrants. Review of Economics and Statistics, 88, 510–530. Mayr, K. (2007). Immigration and income redistribution: a political economy analysis. Public Choice, 131(1– 2), 101–116. Public Choice Meltzer, A. H., & Richard, S. F. (1981). A rational theory of the size of government. Journal of Political Economy, 89, 914–927. Moene, K. O., & Wallerstein, M. (2003). Earnings inequality and welfare spending: a disaggregated analysis. World Politics, 55, 485–516. Mueller, D. C., & Murrell, M. (1986). Interest groups and the size of government. Public Choice, 48(2), 125–145. O’Rourke, K. H., & Sinnott, R. (2004). The determinants of individual attitudes towards immigration. European Journal of Political Economy, 22, 838–861. Plümper, T., & Martin, C. W. (2003). Democracy, spending, and economic growth: a political-economic explanation of the Barro-effect. Public Choice, 117, 27–50. Putnam, R. D. (2007). E pluribus unum: diversity and community in the twenty-first century. Scandinavian Political Studies, 30(2), 137–174. Razin, A., Sadka, E., & Swagel, P. (2002). Tax burden and migration: a political economic theory and evidence. Journal of Public Economics, 85, 167–190. Rodrik, D. (1998). Why do more open economies have bigger governments? Journal of Political Economy, 106, 997–1032. Roemer, J. E., Lee, W., & Van der Straeten, K. (2007). Racism, xenophobia and redistribution, a study of multi-issue politics in advanced democracies. Cambridge: Harvard University Press/Russell Sage Foundation Press. Schmidt, T. D., & Jensen, P. S. (2012). Foreign labor and regional labor markets: aggregate and disaggregate impact on growth and wages in Danish regions. (Forthcoming) Annals of Regional Science. Senik, C., Stichnoth, H., & Van der Straeten, K. (2009). Immigration and natives’ attitudes towards the welfare state: evidence from the European social survey. Social Indicators Research, 91(3), 345–370. Simon, R. J., & Lynch, J. P. (1999). A comparative assessment of public opinion toward immigrants and immigration policies. The International Migration Review, 33(2), 455–467. Soroka, S., Banting, K., & Johnston, R. (2006). Immigration and redistribution in a global era. In B. Sam, P. Bardhan, & M. Wallerstein (Eds.), Globalization and egalitarian redistribution (pp. 261–288). Princeton: Princeton University Press. Tajfel, H., Billig, M., Bundy, R. P., & Flament, C. (1971). Social categorization and intergroup behavior. European Journal of Social Psychology, 1, 149–178. Vasileva, K. (2011). 6.5 % of the EU population are foreigners and 9.4 % are born abroad. Population and social conditions. EUROSTAT, Stat. Focus 34 Van Der Brug, W., & Van Spanje, W. (2009). Immigration, Europe and the ‘new’ cultural dimension. European Journal of Political Research, 48, 309–334. Van Oorschot, W. (2008). Solidarity towards immigrants in European welfare states. International Journal of Social Welfare, 17, 3–14.
© Copyright 2026 Paperzz