Sage Journals: Discover world-class research

Abstract

How are concerns about immigration related to demand for social protection? The “anti-solidarity” hypothesis states that immigration reduces demand for welfare policies because citizens lack solidarity with non-citizens who they fear will access these schemes. According to the “compensation” hypothesis, however, demand for welfare will increase as a reaction to concerns about immigration. This article argues that contextual factors explain which of these two effects will prevail and introduces the political regime as one such factor: Authoritarian regimes are more likely to restrict immigrants’ access to welfare benefits than democracies. Hence, persons holding immigration-related concerns in autocracies may more readily demand social protection policies knowing that immigrants will be excluded. Therefore, I hypothesize that in authoritarian contexts, immigration-related concerns are more strongly associated with demands for social protection policies than in democracies. I find evidence for this hypothesis using World Values Survey data (2017–2022) on 42 countries in a multi-level model.

Keywords

Immigration social protection welfare demand welfare state

Introduction

While international migration is a global phenomenon, research around attitudes regarding migration is still overwhelmingly focused on a narrow set of “Western” liberal democracies. A central question has been whether negative attitudes toward immigration may reduce support for the welfare state.¹ Two opposing hypotheses have been proposed in this regard: On the one hand, concerns about immigration may undermine the solidarity that underlies the democratic welfare state as it introduces ethnic heterogeneity that diminishes people’s propensity to support each other—an “anti-solidarity effect” (Roemer and Van der Straeten, 2006). On the other hand, immigration may stoke economic concerns such as worries about increased labor market competition. Persons who feel economically threatened by immigration may demand more, not fewer, welfare state efforts, which leads to a “compensation effect” (Finseraas, 2008).

Evidence on which effect will prevail is mixed but indicates that the perceived costs and benefits of expanding the welfare state are crucial: Where the foreign population is large, the welfare state is generous, and migrants often rely on the welfare state, the “anti-solidarity effect” should be stronger (Burgoon and Rooduijn, 2021). Yet, it is unclear whether these explanations hold in contexts beyond the Organisation for Economic Co-operation and Development (OECD) countries in which they have been developed. In fact, many countries outside of the OECD, such as the Gulf states, boast very high rates of immigration (UN DESA, 2020a) and in some cases, this has gone along with significant redistribution (Thiollet, 2022).

This article therefore investigates the relationship between concerns about immigration and demand for the welfare state in a more diverse sample of countries, arguing that the political regime is an important factor for the relationship between concerns about immigration and demand for social protection. This argument builds on recent insights from the literature on authoritarian immigration policy which has found that authoritarian regimes indeed combine open immigration policies with increased redistributive efforts (Shin, 2017). I argue that this is because in autocracies, immigrants’ rights are more limited than in democracies (Koopmans and Michalowski, 2017; Ruhs, 2018). Consequently, for citizens in authoritarian regimes, the “compensation effect” is likely to be stronger than the “anti-solidarity effect”: As immigrants are unlikely to be included in social protection schemes, citizens in authoritarian regimes who feel concerned about immigration will be more likely to have stronger incentives to favor welfare expansion, from which they benefit exclusively.

To test the relationship between concerns about immigration and demand for social protection across political regimes, I utilize data on 42 countries taken from the seventh wave (2017–2022) of the World Values Survey (WVS, Haerpfer et al., 2022). Descriptively, the data show that concerns about immigration are present across different types of political regimes. The subsequent multivariate analyses support the hypothesis that in more authoritarian regimes, concerns about immigration and support for social protection are more positively associated. Due to the limited availability of comparative survey data covering all world regions, the findings are correlational and therefore warrant further analysis. Nevertheless, they suggest that immigration may not be an impediment to the development of authoritarian welfare states. Immigration may, indeed, reinforce the need for securing the “authoritarian bargain” (Desai et al., 2009) by which a lack of political rights is compensated by the safe-guarding of economic security. The results also advance the literature on the relationship between immigration and the welfare state in general which has overwhelmingly been centered on Europe and North America. As migration to countries beyond these regions has increased substantially in recent years (UN DESA, 2020a), it becomes crucial to investigate the impact of migration on other host societies. The findings presented here do so, but also resonate with the existing literature that points to the importance of the degree of immigrant inclusion in the welfare state as a moderating factor. In the following sections, I refine my theoretical argument before formulating hypotheses and presenting the empirical evidence.

Concerns about immigration and demand for social protection: “anti-solidarity” versus “compensation”

Whether concerns about immigration are associated with higher or lower demand for social protection depends on the relative strength of the “anti-solidarity” and “compensation” effects.² What, then, determines which one will be dominant? Two key points from the extant literature will be elaborated upon in this section: First, the “anti-solidarity effect” is furthered by the perception that immigrants benefit (disproportionately) from social protection schemes. Second, the “compensation effect” may be triggered by economic concerns in particular, but not exclusively.

The literature positing an “anti-solidarity effect” finds that negative attitudes toward immigrants may reduce support for welfare policies if the native population thinks of them as disproportionately benefiting non-citizens. In this view, individuals are less inclined to extend their solidarity to persons whom they perceive as “outgroups.” This work builds on studies by Gilens (1995) who finds that, in the United States, white Americans are less likely to support increased welfare spending if they hold negative attitudes toward African Americans’ work ethic. This logic extends to immigrants: In fact, immigrants are seen by the public as the group that is least deserving of welfare support in comparison to other populations in need across European countries (Van Oorschot, 2006). The “anti-solidarity effect” thus does not hinge on direct competition for benefits, which occurs when immigrants and natives both try to access a limited supply of welfare provisions (Hooijer, 2021). Rather, assumptions about the beneficiaries and their deservingness play a central role.

Studies have found ample evidence for the existence of an “anti-solidarity effect”: An association between negative attitudes toward immigrants and restrictive welfare attitudes has been found in the European context (Senik et al., 2009) as well as in the United States (Alesina et al., 2023; Garand et al., 2017) and in Hong Kong (Yang et al., 2020). In a priming experiment, Haselswerdt (2021) finds that information on immigrants’ fiscal contributions made participants assume that a hypothetical social assistance scheme was benefiting immigrants. For those who made the assumption, the link between negative attitudes toward immigrants and the rejection of social assistance was much stronger than for those who did not (Haselswerdt, 2021). Furthermore, individuals are more likely to be worried about the fiscal sustainability of the welfare state if prompted to think of immigration and its effects on the welfare state (Goerres et al., 2020) and receiving negative information about the impact of immigration causes a reduction in support to the welfare state (Avdagic and Savage, 2021). Especially support for universal benefits has been shown to decline when individuals were informed that they would benefit non-nationals (Bay and Pedersen, 2006).

The “compensation hypothesis” posits opposite consequences of concerns about immigration: In this view, immigration may cause economic insecurity from which respondents wish to be protected. Substantial portions of people believe that immigration has detrimental effects on wages, employment opportunities, or the economy as a whole (see, e.g. Semyonov et al., 2008 for some European countries; International Labour Organization (ILO), 2019 for some Asian countries).³ In the face of perceived negative economic impacts of immigration, citizens may demand to be “compensated.”⁴ Immigration may cause citizens to support redistribution to make up for wage losses or social insurance to hedge against increased unemployment risks: Finseraas (2008) finds concerns about wage and unemployment effects of immigration to be positively related with support for redistribution. The propensity to demand for compensation is particularly strong among people facing immigrant competition in the labor market (Burgoon et al., 2012; Naumann and Stoetzer, 2018).

What is less clear is whether other kinds of immigration-related concerns, for example cultural or security-related, also lead to demands for compensation. Concerns that are unrelated to the economic situation of an individual may not engender demands for material compensation in the form of welfare provisions or they could be part of an underlying negative attitude toward immigrants that affects welfare preferences. Finseraas (2008) finds a positive relationship between economic concerns about immigration and demand for redistribution, whereas general animosity toward immigrants (operationalized by whether or not respondents were willing to grant equal rights to immigrants) is negatively related to redistributive preferences in his analysis. Similarly, Schmidt and Spies (2014) disaggregate concerns about immigrants and find that economic and cultural concerns have opposite effects on preferences for redistribution: While economic concerns foster support for redistribution, cultural concerns do not (a third type of concern, about welfare abuse by immigrants, has no direct effect). On the other hand, Senik et al. (2009) differentiate between concerns about immigrant usage of the health and welfare systems and general animosity toward immigrants and find that both impact welfare attitudes negatively. Burgoon and Rooduijn (2021) distinguish economic, cultural, and general concerns and find that all are moderated similarly by contextual factors in their effects on preferences for redistribution.

“Anti-solidarity” and “compensation”: the role of contextual factors

Research has also investigated which contextual factors are associated with a stronger “anti-solidarity” or a stronger “compensation” effect: Finseraas (2008) provides evidence that the anti-solidarity effect is more prevalent in Scandinavian countries (with social-democratic welfare states), where tax-financing is stronger, than in conservative welfare states with their focus on contributory benefits. Schmidt and Spies (2014) fail to find the same result, although this may partly be due to a different classification of welfare regime types.⁵ In addition, Burgoon and Rooduijn (2021) conclude that the extent of immigration, existing welfare state generosity, and benefit dependency among immigrants are important contextual factors. They find concerns about immigration to be related to reduced support for the welfare state where the rate of immigration is high, the welfare state is already large, and immigrants make use of welfare benefit relatively more often (Burgoon and Rooduijn, 2021).

The review of the literature makes clear that the link between concerns about immigration and demand for social protection depends on how respondents assess the costs and benefits of welfare expansions: Individuals are less likely to support an expansion of the welfare state if they face higher costs from doing so (e.g. through taxes, Finseraas, 2008) or where “outgroups” benefit disproportionately (e.g. through higher benefit dependency rates, Burgoon and Rooduijn, 2021). Political regimes have a bearing on these factors, in particular the latter: A nascent literature attests to the fact that immigrants’ rights are more likely to be restricted in authoritarian countries than in democracies (Koopmans and Michalowski, 2017; Ruhs, 2018). It is often said that this is one of the reasons for the higher rates of immigration in authoritarian countries, since withholding rights from immigrants decreases costs for employers (Breunig et al., 2012; Mirilovic, 2010). Authoritarian regimes have also been found to combine redistribution among citizens and high rates of immigration if they have resource rents to do so (Shin, 2017). An example of this is Saudi Arabia, which has one of the largest immigrant populations in the world and also ranks among the countries with the largest share of non-citizens among its residents (UN DESA, 2020a). The country excludes migrants from accessing many benefits (ILO, 2023), while at the same time increasing redistribution among citizens (Thiollet, 2022).

Figure 1 systematizes the insight that authoritarian regimes grant fewer rights to migrants. It shows all countries covered by the Migrant Integration Policy Index (MIPEX, Solano and Huddleston, 2020) and plots their score in the overall MIPEX index over the level of democracy. The MIPEX overall index is not only a measure of migrants’ access to the welfare state but includes eight different fields of migration integration.⁶ It can be seen as a measure for the overall inclusion of migrants in their country of destination. At the time of writing, MIPEX is arguably also the measure of migrants’ rights with the largest country coverage. The scatterplot shows a strong correlation between the level of democracy and the integration of migrants (Pearson correlation coefficient: r(52) = 0.68; p < 0.01). The countries with the lowest levels of immigrant integration all feature authoritarian regimes, whereas the highest levels of integration can be found in democracies.

Figure 1.

Migrant integration policy and level of democracy in 2019.

A lack of immigrants’ rights, in particular the right to access welfare benefits, will alter the cost–benefit considerations that Finseraas (2008) and Burgoon and Rooduijn (2021) put forward in their work: Where immigrants are excluded from social protection benefits, these benefits can be expanded without “outgroups” benefiting from it. Since immigrants are often excluded from social protection in authoritarian contexts, expanding social protection is a more likely reaction to concerns about immigration in these contexts than in democracies.⁷ This leads me to the core hypothesis that is investigated in this article: In more authoritarian contexts, concerns about immigration and support for social protection are more likely to be positively related. Conversely, in more democratic contexts, immigrants have access to the welfare state to a greater extent and therefore, the “anti-solidarity” mechanism can be expected to be stronger. The consequence is a more negative association between concerns about immigration and support for the welfare state in more democratic countries.

Data and methods

In order to investigate the relationship between concerns about immigration and demand for social protection across different political regimes, I use data from the WVS’s latest wave which took place between 2017 and 2022 (Haerpfer et al., 2022). For several countries or territories in the WVS, either some survey items are unavailable or macro data are lacking, and these are therefore removed.⁸

Dependent variable

As a measurement for the dependent variable, social protection preferences, I use a survey item asking respondents to rate their views on a 10-point scale between The government should take more responsibility to ensure that everyone is provided for (1) and People should take more responsibility to provide for themselves (10). The measure captures the propensity of respondents to call for government intervention and efforts to establish a social safety net for people. It could in this sense be seen as a measure for demand for protective social protection—policies providing support to those most in need (Devereux and Sabates-Wheeler, 2004). The measure is recoded so that higher values indicate more support for government social protection interventions.

Independent variables

For the main independent variable of interest, concerns about immigration, the section on “Migration” in the WVS offers nine potential items that ask respondents about their perception of immigration or immigrants. However, four items stand out as specifically capturing concerns about perceived negative effects of immigration: First, the WVS asks respondents whether they agree with the statement that immigration increases unemployment (answer options: agree/hard to say/disagree). While this is a concern related to the economic impact of migration, a further three items cover security concerns: They ask whether the respondent thinks that immigration increases the crime rate, that immigration increases the risk of terrorism, and/or that immigration leads to social conflict (answer options for all three items: agree/hard to say/disagree). These four items are not only theoretically those that capture concerns about immigration, they are also the ones that are most strongly correlated (see Table 4 in the Online Appendix).

To determine whether these four items are representative of a latent concept, concerns about immigration, I run a confirmatory factor analysis (CFA). Since the observed variables are ordinal, I use a robust diagonally weighted least squares estimator. This is necessary since ordinal variables do not follow a continuous and multivariate distribution as is required by standard maximum likelihood estimators (Li, 2016). I first conduct the CFA for the entire WVS sample of 55 countries. As can be seen in Table 10 in Online Appendix C, fit measures (the root mean square error of approximation [RMSEA], in particular) are suboptimal in a number of countries. I therefore exclude the 13 countries with a RMSEA higher than 0.1 from the analysis in the main text. The analysis presented here is therefore based on a smaller sample of 42 countries (henceforth: small sample).

Table 1 reports the measures of fit of a CFA using this smaller sample (42 countries). Along with the RMSEA, it includes the comparative fit index and the Tucker–Lewis index in their standard and scaled versions which all indicate acceptable fit. The table also shows the results of a chi-square test which is significant. In order to accept the model, chi-square in this case should not be significant. However, the chi-square test has been shown to frequently reject models with large sample sizes such as this one (Bentler and Bonett, 1980) and therefore, the other measures of fit are trusted. As can be seen in Table 2, the R² and standardized loadings for each of the items in the model are quite high, indicating that the latent variable explains a large share of the variance in these items. In order to establish if the constructs are present in all countries, I additionally conduct a multi-group confirmatory factor analysis. The results of this analysis can be found in Online Appendix C and confirm the appropriateness of the model across countries. For the multivariate analyses shown in this article, the immigration concerns variable that I use consists of the factor scores predicted on the basis of the CFA.

Table 1.

Results of the confirmatory factor analysis (sample of 42 countries): fit measures for the basic model.

Chi-squared scaled	df scaled	p-value scaled	RMSEA	RMSEA scaled	CFI	CFI scaled	TLI	TLI scaled
909.641	2	0	0.057	0.089	0.999	0.996	0.996	0.987

It should be noted that there have been recent developments in literature on fit measures for confirmatory factor analyses with categorical data (Savalei, 2021). These new “robust” indices show worse fit. However, no clear guidelines or cut-off points have been devised, yet. Therefore, this article sticks with the procedure put forward by Svetina et al. (2020) and reports the scaled measures (RMSEA scaled, CFI scaled, and TLI scaled).

Table 2.

Results of the confirmatory factor analysis (sample of 42 countries): R² and loadings.

	Survey item	R ²	Standardized loading
Q124	Immigration increases the crime rate	0.72	0.85
Q126	Immigration increases the risk of terrorism	0.72	0.85
Q128	Immigration increases unemployment	0.61	0.78
Q129	Immigration increases the risk of social conflict	0.67	0.82

The second important independent variable of interest is the political regime (democracy). The data for this variable come from the V-Dem dataset (Coppedge et al., 2024; Pemstein et al., 2024). The V-Dem project provides a range of indicators and indices on the quality of democracy in most countries of the world, and I chose the most aggregated one of them: The V-Dem Liberal Democracy Index. This index consists of 69 indicators of free and fair elections, freedom of association, freedom of expression and the press, equality before the law and individual liberties, as well as judicial and legislative constraints on the executive (Alizada et al., 2021: 42) and as such, it offers a comprehensive view of the fulfillment of democratic ideals. The index can vary between 0 and 1, with higher values indicating a more democratic regime. Using a continuous index of democracy is preferable to a dichotomous one as it captures more nuanced gradations within political systems. These gradations are relevant to this analysis: Strongly authoritarian regimes may find it easier to restrict migrants’ rights whereas in more competitive regimes, there is more leeway to advocate for migrants’ rights (Henninger and Römer, 2021). Therefore, the gradations are also likely to have an effect on the relationship between concerns about immigration and demand for the welfare state. The choice of democracy measure can be consequential for the analysis (Vaccaro, 2021) and therefore, the analyses were re-run using other indices (see the “Robustness checks” section).

Ideally, the correlation between democracy and immigrants’ rights that is shown in Figure 1 would be tested for the country sample at hand. A positive correlation would bolster the claim that in more authoritarian settings, citizens have reason to believe that immigrants will be excluded from the social protection that they are demanding. Unfortunately, suitable policy data with a sufficiently large coverage do not exist. Of the 42 countries in the sample, only 17 are covered by MIPEX. Among these 17 countries, the MIPEX overall index and the V-Dem Liberal Democracy Index show a strong correlation just like in the full sample in Figure 1 (r(15) = 0.62, p = 0.01). The relationship is also visualized by Figure 5 in the Online Appendix.

It should be noted that concerns about immigration are present across regime types. The weighted country means of the immigration concerns factor scores in each country do not correlate with the V-Dem Liberal Democracy Index (r(40) = –0.05, p = 0.75). Neither do the shares of persons agreeing with each of the four statements regarding the impact of immigration, by country.⁹ Figure 2 shows these shares using the original items from the WVS data (not the latent variable), by country. It shows a wide variation in the prevalence of negative attitudes toward immigrants. In Iraq, less than one fifth of the population attributes increases in crime, terrorism, unemployment, or social conflict to immigrants while at the other end of the spectrum, more than three quarters of Jordanians do. The figure also reveals no clear relationship between the level of democracy (as indicated by the dots) and average concerns about immigration. Rather, relatively democratic countries and more authoritarian countries can be found across the entire range of the prevalence of concerns.

Figure 2.

Concerns about immigration and level of democracy by country.

Control variables

In addition, I include a number of individual- and country-level control variables that affect demand for social protection as well as attitudes toward immigration. At the individual level, I include the respondent’s gender (female = 1) as women have been found to be more favorable to government welfare efforts (e.g. Svallfors, 1997), but also to have more positive attitudes toward immigrants (Semyonov et al., 2008). Similarly, age is an important predictor of welfare attitudes (Busemeyer et al., 2009), but may also impact immigration attitudes (Quillian, 1995). The age of the respondent is therefore also controlled for. Moreover, the model includes a dummy for whether or not the respondent has children (one or more children = 1, see also Burgoon and Rooduijn, 2021). Further important socio-economic factors for demand for social protection are: a person’s employment status (Svallfors, 1997), their highest level of education, and their income (Häusermann et al., 2015). These factors have previously also been linked to attitudes toward immigration (Manevska and Achterberg, 2013). Hence, one variable captures whether or not a respondent is unemployed (unemployed = 1). The models also control for the respondent’s highest level of education on the International Standard Classification of Education (ISCED) scale. They differentiate between persons with at most lower secondary (ISCED 3), upper secondary (ISCED 4), and tertiary education (ISCED 5 or more), with persons without secondary education (ISCED 1 and 2) forming the baseline for comparison. Subjective income is measured by asking which income decile persons think they belong to in their country between 1 (the poorest) and 10 (the richest).

I introduce several country-level control variables that could influence the relationship between concerns about immigration and demand for social protection. This includes, first, the level of social protection expenditure as a percentage of gross domestic product (GDP; based on Gethin, 2024). Low existing levels of welfare state effort could spur citizens’ propensity to demand social protection in response to concerns about immigration, while where the welfare state is already extensive, demands might be dampened (Burgoon and Rooduijn, 2021). Therefore, it makes sense to include social expenditures and their interaction with concerns about immigration, particularly since they are also correlated with the level of democracy (r(40) = 0.63, p < 0.001). Social expenditure has weaknesses as a measure for welfare state effort (Öktem, 2020). However, no welfare state regime classifications with a global coverage exist as of yet and therefore, expenditures are used. Second, migrant stock is taken into account since higher rates of immigration may induce citizens to assume that migrants who are present in their country will make use of public resources. The number of immigrants as percentage of the population is taken from the United Nations Department of Economic and Social Affairs (UN DESA, 2020b). Data for the years 2015 and 2020 are given in the dataset and I employ linear interpolation to obtain values for the years in between, also extrapolating to 2021.

But it could also be the case that the relationship between concerns about immigration and demand for social protection is impacted by the general economic conditions in the country. Adverse economic conditions could cause those who are concerned about immigration to be particularly interested in state-provided social protection since they fear not being able to make a living by participating in the labor market. The models therefore capture the unemployment rate (as percentage of the total labor force), which I take from World Bank (2023) and real GDP per capita taken from the Penn World Tables (Feenstra et al., 2015). I use their measure for real GDP on the expenditure side which Feenstra et al. (2015) suggest for comparisons across countries and time. Both the unemployment rate and GDP per capita are also interacted with concerns about immigration.

The year in which the WVS was administered differs by country; all country-level variables are measured in the year prior to the fielding of the survey in each respective country. This lag by one year is introduced to account for delayed responses of attitudes to contextual factors. In accordance with guidelines from the literature on multi-level modeling, the main individual-level variable of interest, concerns about immigration, is country-mean-centered since it is involved in a cross-level interaction (Bliese et al., 2018: 12; Enders and Tofighi, 2007: 132–134). Individual-level control variables, including dichotomous variables, are also country-mean-centered (Yaremych et al., 2023). The macro-level variables social protection expenditures, migrant stock, unemployment rate, and GDP per capita are log-transformed. Country-level variables are centered to the grand mean as suggested by Enders and Tofighi (2007: 135). Descriptive statistics (before transformations) can be found in Tables 14 and 15 in the Online Appendix.

Model

The models estimated in this article are multi-level models. These account for the structure of the data by the nature of which individual respondents are clustered in countries. Since respondents within each country are likely to share certain characteristics, a central assumption of regular regressions, the independence of observations, is violated: Therefore, in these cases, standard errors will be underestimated (Hox, 2010: 4–8). This assumption was tested prior to the complete analysis by comparing a standard linear regression on the dependent variable of interest with a two-level linear model without covariates (the variance-components model) which includes a separate intercept for each country. A likelihood-ratio test can determine if the between-country variance is significantly different from 0, in which case a multi-level model is necessary (Rabe-Hesketh and Skrondal, 2008: 69). The results of the test are given in Table 5 in the Online Appendix and indicate a necessity to use a multi-level model. Specifically, I use a random-coefficients model in which the effect of concerns about migration is allowed to vary by country. The inclusion of random coefficients is necessary for correctly estimating a cross-level interaction between the context (democracy) and the individual-level predictor (Heisig and Schaeffer, 2019). The model can be expressed through the following equation:

\begin{array}{l} y_{i j} = β_{00} + (β_{10} + ζ_{1 j}) {ImmigrationConcerns}_{i j} + β_{01} {Democracy}_{j} + β_{11} {ImmigrationConcerns}_{i j} \\ \times {Democracy}_{j} + β_{p} X_{i j} + β_{q} Z_{j} + ζ_{j} + ϵ_{i j} \end{array}

Here, ImmigrationConcerns_ij is the latent variable obtained from the CFA which reflects individual-level concerns about immigration and Democracy_j is the V-Dem Liberal Democracy Index for a given country. X_ij and Z_j are individual-level and country-level control variables, respectively. The random intercept is represented by $ζ_{j}$ and the random coefficient by $ζ_{1 j}$ . If not indicated otherwise, all models include weights. The WVS provides poststratification weights that adjust for over- or undersampling of certain groups as well as equilibration weights (which equalize sample sizes across countries) and these are multiplied.

Results

Table 3 shows the results of the analysis with demand for social protection as the dependent variable, again using the sample containing 42 countries. Model (1) includes no independent variables and therefore only partitions the variance between the second and the first levels. Model (2) introduces individual-level explanatory variables. The coefficients of the control variables are generally in line with the previous literature: Being female and being unemployed are associated with higher demand for social protection. Higher age, higher educational attainment, and having a high subjective income are related with lower demand for social protection. These effects stay stable over the remaining models.

Table 3.

Multi-level linear regression: immigration concerns and demand for social protection (smaller sample).

	Demand for social protection
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
Immigration concerns		0.021 (0.016)	0.021 (0.016)	0.020 (0.014)	0.021 (0.016)	0.020 (0.016)	0.020 (0.016)	0.020 (0.016)
Female		0.104*** (0.025)	0.103*** (0.025)	0.103*** (0.025)	0.104*** (0.025)	0.104*** (0.025)	0.104*** (0.025)	0.103*** (0.025)
Age		−0.003** (0.001)	−0.003** (0.001)	−0.003** (0.001)	−0.003** (0.001)	−0.003** (0.001)	−0.003** (0.001)	−0.003** (0.001)
Children		−0.012 (0.032)	−0.012 (0.032)	−0.011 (0.032)	−0.012 (0.032)	−0.012 (0.032)	−0.012 (0.032)	−0.012 (0.032)
Unemployed		0.138** (0.046)	0.138** (0.046)	0.138** (0.046)	0.138** (0.046)	0.138** (0.046)	0.138** (0.046)	0.138** (0.046)
Level of education (ref: primary)
Lower secondary education		−0.067 (0.045)	−0.067 (0.045)	−0.066 (0.045)	−0.067 (0.045)	−0.067 (0.045)	−0.067 (0.045)	−0.067 (0.045)
Upper secondary education		−0.118** (0.042)	−0.118** (0.042)	−0.119** (0.042)	−0.118** (0.042)	−0.118** (0.042)	−0.118** (0.042)	−0.118** (0.042)
Postsecondary education		−0.269*** (0.055)	−0.269*** (0.055)	−0.269*** (0.055)	−0.269*** (0.055)	−0.268*** (0.055)	−0.269*** (0.055)	−0.269*** (0.055)
Tertiary education		−0.205*** (0.043)	−0.205*** (0.043)	−0.206*** (0.043)	−0.205*** (0.043)	−0.205*** (0.043)	−0.205*** (0.043)	−0.205*** (0.043)
Subjective income		−0.149*** (0.007)	−0.149*** (0.007)	−0.149*** (0.007)	−0.149*** (0.007)	−0.149*** (0.007)	−0.149*** (0.007)	−0.149*** (0.007)
Democracy			0.756 (0.469)	0.105 (0.513)	0.523 (0.568)	0.471 (0.476)	0.637 (0.431)	0.825 (0.590)
Immigration concerns × democracy				−0.173** (0.055)
Social expenditure					0.104 (0.270)
Immigration concerns × social expenditure					−0.014 (0.027)
Migrant stock						0.121 (0.083)
Immigration concerns × migrant stock						−0.001 (0.010)
Unemployment rate							0.540** (0.207)
Immigration concerns × unemployment rate							0.007 (0.026)
GDP per capita								−0.107 (0.187)
Immigration concerns × GDP per capita								−0.020 (0.018)
Intercept	6.005*** (0.132)	6.011*** (0.131)	6.011*** (0.134)	6.011*** (0.131)	6.011*** (0.132)	6.011*** (0.128)	6.011*** (0.123)	6.011*** (0.133)
Random effects
SD (random intercept)	0.848	0.846	0.862	0.846	0.850	0.828	0.794	0.861
SD (random slope)		0.098	0.098	0.087	0.098	0.098	0.098	0.097
Correlation (intercept, slope)		0.367	0.463	0.421	0.444	0.448	0.474	0.461
Number of observations	54,725	54,725	54,725	54,725	54,725	54,725	54,725	54,725
ICC	0.109	0.125	0.128	0.122	0.125	0.120	0.113	0.127
Marginal R²	0.000	0.018	0.023	0.020	0.022	0.027	0.035	0.022
Conditional R²	0.109	0.140	0.148	0.140	0.144	0.145	0.144	0.147
AIC	282360.4	281345.8	281345.4	281342.8	281355.1	281357.6	281349.6	281356.1

SD: standard deviation.

**p < 0.01, ***p < 0.001.

The coefficient of the main variable of interest, concerns about immigration, is small and insignificant, indicating that there is no relationship between these concerns and demand for social protection. However, this masks potential variation between the countries: Since a random-coefficient model is used, the effect of concerns about immigration on demand for social protection is allowed to vary by country and the coefficient provided in the table is the average of the country slopes. Therefore, it is instructive to look at these individual slopes and see how they vary.

Figure 3 plots the country-level slopes over the V-Dem Liberal Democracy Index (Coppedge et al., 2024; Pemstein et al., 2024). As can be seen in this figure, there is a negative relationship between the level of democracy and the slope of the variable “concerns about immigration.” This is a first tentative indication that in more authoritarian countries, the relationship between concerns about immigration and demand for social protection is more often positive, that is the compensation effect prevails.

Figure 3.

Random coefficients in model (2) in Table 3, over the V-Dem liberal democracy score.

To further investigate in how far the political regime moderates the association between concerns about immigration and demand for social protection, model (3) includes the V-Dem Liberal Democracy Index (democracy) and model (4) adds its interaction with immigration concerns. The coefficient of this interaction term is negative and significant, indicating that in more democratic countries, there tends to be a more negative association between concerns about immigration and demand for social protection. This relationship is depicted in Figure 4. As can be seen, the marginal effect of concerns about immigration on demand for social protection is positive at lower values of the democracy index. This indicates that in authoritarian regimes, concerns about immigration are positively related to demand for social protection. At the other end of the spectrum, the effect becomes more negative for highly democratic countries, indicating that here, concerns about immigration are associated with less demand for social protection. The results thus further corroborate the hypothesis put forward in this article.

Figure 4.

Marginal effect of concerns about immigration at different values of the V-Dem Liberal Democracy Index (Coppedge et al., 2024; Pemstein et al., 2024) on the basis of Table 3, model (4). The index has been mean-centered.

Models (5) to (8) include four macro-level control variables one after another. They are also interacted with concerns about immigration to see whether they could constitute alternative explanations for the variation in the relationship between concerns about immigration and demand for social protection. Model (5) finds no moderating effect of social expenditures on this relationship. Neither the direct effect of social expenditures nor its interaction with concerns about immigration is significant. The same goes for migrant stock (model 6). These findings are contrary to those by earlier works who had found a moderating effect of the share of the migrant population and the design of the welfare state. In model (7), the unemployment rate is included along with its interaction with concerns about immigration. There is a significant and substantial positive effect of unemployment rates on demand for social protection as seen by the coefficient of the direct effect. However, the interaction of unemployment rate with concerns about immigration is insignificant, indicating that unemployment does not play a moderating role. Finally, the coefficient for GDP per capita (model 8) also is not significant. Thus, none of the alternative explanations seem to hold.

Robustness checks

Several checks were carried out in order to ascertain the robustness of the results. First, the models were checked for multicollinearity of the independent variables. For all predictors in model (3) of Table 3, the variance inflation factors are presented in Table 16 in the Online Appendix. They are all low by conventional thresholds. I also checked for several alternative model specifications. First, I tested whether the result would change if multiple macro-level controls and their interaction with immigration concerns were included at the same time. The results are reported in Table 17 in the Online Appendix. Even when including one or several macro-level controls and their interaction with immigration concerns, the interaction between democracy and immigration concerns stays negative and significant—the coefficient even increases in size. Second, I ran the models presented in Table 3 without using weights. The results are presented in Table 18 the Online Appendix and are similar to the ones in Table 3. Third, I ran the models from Table 3 with the large sample of 55 countries, using the results of the CFA with that same sample (Table 19, Online Appendix). The results are broadly similar to those obtained with the smaller sample. The interaction term between democracy and immigration concerns is highly significant and negative in these models, as well, lending support to my main hypothesis.

A second batch of robustness checks uses alternative variables as main variables of interest. On the one hand, I tested whether the results hinged on the indicator used for democracy (Tables 20 and 21 in the Online Appendix). The results do not change substantively when either using the Polity5 measure of political regimes (Marshall and Gurr, 2020) nor when using Boix-Miller-Rosato dichotomous coding of democracy (Miller et al., 2022). On the other hand, I investigated whether the four different items included in the measure for concerns about immigration have the same effect on demand for social protection, and importantly, are moderated by democracy in the same way as the overall latent variable. The results are shown in Tables 22 to 25 in the Online Appendix and the results do not differ from those models using the overall measure.

Finally, I checked whether the results are robust to the exclusion of certain countries from the sample. Table 26 in the Online Appendix contains the results of 42 models, with one country dropped in each of these models. While the models include the same variables as model (4) in Table 3, only the coefficients, standard errors, and p values for the variables of interest (concerns about immigration, democracy, and their interaction) are shown. The results show that dropping any one country from the analysis does not alter the results, lending credibility to the robustness of the results.

Conclusion

For many years now, evidence on the relationship between attitudes toward immigration and demand for social protection has been divided: Some posit that there is an “anti-solidarity effect” by which citizens are less willing to expand the welfare state if its beneficiaries become more diverse (Alesina and Glaeser, 2004; Roemer and Van der Straeten, 2006). Others have connected concerns about immigration with demands for compensation in form of more generous welfare benefits through a “compensation effect” (Finseraas, 2008). Hence, this article endeavored to shed light on the contextual factors that might cause one or the other effect to be stronger. I have argued that the political regime is such a factor, as it affects how citizens view the costs and benefits associated with social protection policy: Authoritarian regimes have been shown to restrict immigrants’ rights, including access to welfare benefits. This means citizens can demand increases in government social protection efforts without needing to fear that they will bear the cost for providing for “outgroups.”

I therefore hypothesized that the likelihood that persons who have concerns about immigration also demand social protection policies will be higher in more authoritarian countries. Data from 42 countries provided by the WVS allowed me to test this hypothesis since it covers a variety of political regimes. This analysis thus also transcends the traditional geographic scope of the extant literature which focusses on “Western” liberal democracies mostly and adds to an emerging literature on immigration and immigration policy in authoritarian contexts (e.g. Natter, 2018). I find support for the hypothesis that concerns about immigration and demand for social protection are more likely to be positively related in authoritarian countries than in democracies. This is in line with earlier works stressing the importance of moderating contextual variables (Burgoon and Rooduijn, 2021; Finseraas, 2008; Schmidt and Spies, 2014). However, this article also qualifies the results of these earlier works: Contrary to previous works, I do not find that the share of migrants present in a country and the welfare state effort are important moderators. Rather, the political regime emerges as the most important moderating variable. Finally, it is noteworthy that the effect that I find is stable across different kinds of concerns about immigration, both economic and security-related. Therefore, it is not only economic competition that prompts demands for compensation, but also other perceived threats.

Despite the clear results obtained in the analysis, there are some limitations to this article. First, the moderating effect of the political regime should be further examined in future research. With potential future extensions of the coverage of migrants’ rights indicators, it will become possible to ascertain if the reason for this moderating effect is really the propensity of authoritarian regimes to limit migrants’ rights. Second, since the data are cross-sectional and I analyze the association between attitudes, I can make no causal argument on the relationship between attitudes toward immigration and demand for social protection. In particular, the moderating role of assumptions about who stands to gain from benefit extensions (citizens only, or also immigrants) should be explored further. Therefore, further analyses with longitudinal data or quasi-experimental studies are necessary to test the theoretical framework and the hypotheses for which I find initial support in this article.

Nevertheless, the results of this article suggest that the high rates of immigration that can be observed in authoritarian regimes could be associated with an increase in social protection provisions rather than with their decline—if immigration-related concerns and claims for compensation arise with them and authoritarian rulers respond to these public demands. To this point, the article shows that high levels of concerns about immigration can be found in both democratic and authoritarian contexts. If regimes rely on an “authoritarian bargain” (Desai et al., 2009) that trades political rights for economic security, high rates of immigration may require them to provide social protection. To ensure their political survival, authoritarian regimes have to either limit immigration to reduce their citizens’ concerns about it (see also Joo, 2022) or compensate them by providing social policies (see also Shin, 2017). This indicates that authoritarian regimes are constrained in the way they design their immigration policy—open immigration policies may elicit demands for compensation from the regime. This in turn opens up further areas of investigation on the politics of immigration policy in authoritarian regimes—which are potentially more contentious than often assumed.

Supplemental Material

sj-docx-1-cos-10.1177_00207152251319747 – Supplemental material for Concerns about immigration and demand for social protection: The implications of political regimes

Supplemental material, sj-docx-1-cos-10.1177_00207152251319747 for Concerns about immigration and demand for social protection: The implications of political regimes by Jakob Henninger in International Journal of Comparative Sociology

Footnotes

Acknowledgements

The author thanks Friederike Römer, Susanne K. Schmidt, Nate Breznau, Elias Naumann, Patrick Sachweh, Eloisa Harris, and the participants of the doctoral colloquium of the Bremen International School of Social Sciences (BIGSSS) for very helpful comments. All remaining mistakes are his own.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This article is a product of research conducted at the Collaborative Research Center 1342 “Global Dynamics of Social Policy”. The center is funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation)—project number 374666841—SFB 1342.

ORCID iD

Jakob Henninger

Supplemental material

Supplemental material for this article is available online.

Notes

References

Alesina

Glaeser

(2004) Fighting Poverty in the US and Europe: A World of Difference (Rodolfo Debenedetti lectures). Oxford: Oxford University Press.

Alesina

Miano

Stantcheva

(2023) Immigration and redistribution. The Review of Economic Studies 90(1): 1–39.

Alizada

Cole

Gastaldi

, et al. (2021) Autocratization Turns Viral: Democracy Report 2021 (March). Gothenburg: V-Dem Institute, University of Gothenburg.

Avdagic

Savage

(2021) Negativity bias: The impact of framing of immigration on welfare state support in Germany, Sweden and the UK. British Journal of Political Science 51(2): 624–645.

Bambra

(2004) The worlds of welfare: Illusory and gender blind? Social Policy and Society 3(3): 201–211.

Bay

A-H

Pedersen

(2006) The limits of social solidarity: Basic income, immigration and the legitimacy of the universal welfare state. Acta Sociologica 49(4): 419–436.

Bentler

Bonett

(1980) Significance tests and goodness of fit in the analysis of covariance structures. Psychological Bulletin 88(3): 588–606.

Bliese

Maltarich

Hendricks

(2018) Back to basics with mixed-effects models: Nine take-away points. Journal of Business and Psychology 33(1): 1–23.

Borjas

(2003) The labor demand curve is downward sloping: Reexamining the impact of immigration on the labor market. The Quarterly Journal of Economics 118(4): 1335–1374.

10.

Brady

Finnigan

(2014) Does immigration undermine public support for social policy? American Sociological Review 79(1): 17–42.

11.

Breunig

Cao

Luedtke

(2012) Global migration and political regime type: A democratic disadvantage. British Journal of Political Science 42(4): 825–854.

12.

Breznau

Rinke

Wuttke

, et al. (2022) Observing many researchers using the same data and hypothesis reveals a hidden universe of uncertainty. Proceedings of the National Academy of Sciences of the United States of America 119(44): e2203150119.

13.

Burgoon

Rooduijn

(2021) “Immigrationization” of welfare politics? Anti-immigration and welfare attitudes in context. West European Politics 44(2): 177–203.

14.

Burgoon

Koster

Van Egmond

(2012) Support for redistribution and the paradox of immigration. Journal of European Social Policy 22(3): 288–304.

15.

Busemeyer

Goerres

Weschle

(2009) Attitudes towards redistributive spending in an era of demographic ageing: The rival pressures from age and income in 14 OECD countries. Journal of European Social Policy 19(3): 195–212.

16.

Coppedge

Gerring

Knutsen

, et al. (2024) V-Dem Dataset v14. University of Gothenburg. Available at: https://www.v-dem.net/data/dataset-archive/ (accessed 13 March 2024).

17.

Cusack

Iversen

Rehm

(2006) Risks at work: The demand and supply sides of government redistribution. Oxford Review of Economic Policy 22(3): 365–389.

18.

Desai

Olofsgård

Yousef

(2009) The logic of authoritarian bargains. Economics & Politics 21(1): 93–125.

19.

Devereux

Sabates-Wheeler

(2004) Transformative social protection. IDS working paper 232. Brighton: Institute of Development Studies. Available at: https://www.ids.ac.uk/download.php?file=files/dmfile/Wp232.pdf (accessed 18 December 2023)

20.

Dustmann

Schönberg

Stuhler

(2016) The impact of immigration: Why do studies reach such different results? Journal of Economic Perspectives 30(4): 31–56.

21.

Eger

(2010) Even in Sweden: The effect of immigration on support for welfare state spending. European Sociological Review 26(2): 203–217.

22.

Eger

Breznau

(2017) Immigration and the welfare state: A cross-regional analysis of European welfare attitudes. International Journal of Comparative Sociology 58(5): 440–463.

23.

Enders

Tofighi

(2007) Centering predictor variables in cross-sectional multilevel models: A new look at an old issue. Psychological Methods 12(2): 121–138.

24.

Esping-Andersen

(1990) The Three Worlds of Welfare Capitalism. Princeton, NJ: Princeton University Press.

25.

Feenstra

Inklaar

Timmer

(2015) The next generation of the Penn World Table. American Economic Review 105(10): 3150–3182.

26.

Finseraas

(2008) Immigration and preferences for redistribution: An empirical analysis of European survey data. Comparative European Politics 6(4): 407–431.

27.

Finseraas

Pedersen

Bay

A-H

(2016) When the going gets tough: The differential impact of national unemployment on the perceived threats of immigration. Political Studies 64(1): 60–73.

28.

Garand

Davis

(2017) Immigration attitudes and support for the welfare state in the American mass public. American Journal of Political Science 61(1): 146–162.

29.

Gethin

(2024) General Government Revenue and Expenditure by Function in 177 Countries, 1980-2022. Available at: https://amory-gethin.fr/files/datasets/gethin-public-finance-statistics.dta (accessed 7 April 2024).

30.

Gilens

(1995) Racial attitudes and opposition to welfare. The Journal of Politics 57(4): 994–1014.

31.

Goerres

Karlsen

Kumlin

(2020) What makes people worry about the welfare state? A three-country experiment. British Journal of Political Science 50(4): 1519–1537.

32.

Haerpfer

Inglehart

Moreno

, et al. (2022) World Values Survey Wave 7 (2017-2022): Cross-National Data-Set. Version: 5.0.0. World Values Survey Association. Available at: http://www.worldvaluessurvey.org/WVSDocumentationWV7.jsp (accessed 16 February 2023).

33.

Haselswerdt

(2021) Social welfare attitudes and immigrants as a target population: Experimental evidence. Perspectives on Politics 19(2): 442–459.

34.

Häusermann

Kurer

Schwander

(2015) High-skilled outsiders? Labor market vulnerability, education and welfare state preferences. Socio-Economic Review 13(2): 235–258.

35.

Heisig

Schaeffer

(2019) Why you should always include a random slope for the lower-level variable involved in a cross-level interaction. European Sociological Review 35(2): 258–279.

36.

Henninger

Römer

(2021) Choose your battles: How civil society organisations choose context-specific goals and activities to fight for immigrant welfare rights in Malaysia and Argentina. Social Policy & Administration 55(6): 1112–1128.

37.

Hooijer

(2021) “They take our houses”: Benefit competition and the erosion of support for immigrants’ social rights. British Journal of Political Science 51(4): 1381–1401.

38.

Hox

(2010) Multilevel Analysis: Techniques and Applications (Quantitative methodology series), 2nd edn. New York: Routledge, Taylor & Francis.

39.

International Labour Organization (ILO) (2019) Public Attitudes towards Migrant Workers in Japan, Malaysia, Singapore, and Thailand. Geneva: ILO. Available at: https://www.ilo.org/wcmsp5/groups/public/—asia/—ro-bangkok/documents/publication/wcms_732443.pdf (accessed 12 September 2021).

40.

International Labour Organization (ILO) (2023) Social Protection for Migrant Workers in Countries of the Cooperation Council for the Arab States of the Gulf (GCC). Beirut, Lebanon: ILO. Available at: https://www.ilo.org/media/358976/download (accessed 6 July 2024).

41.

Isaksen

Jakobsen

Filindra

, et al. (2016) The return of prejudice in Europe’s regions: The moderated relationship between group threat and economic vulnerability. Nationalism and Ethnic Politics 22(3): 249–277.

42.

Jean

Jiménez

(2011) The unemployment impact of immigration in OECD countries. European Journal of Political Economy 27(2): 241–256.

43.

Joo

(2022) Elections and immigration policy in autocracy: Evidence from Russia and Kazakhstan. Government and Opposition 59: 482–503.

44.

Khalil

Naumann

(2022) Does contact with foreigners reduce worries about immigration? A longitudinal analysis in Germany. European Sociological Review 38(2): 189–201.

45.

Koopmans

Michalowski

(2017) Why do states extend rights to immigrants? Institutional settings and historical legacies across 44 countries worldwide. Comparative Political Studies 50(1): 41–74.

46.

Latif

(2015) The relationship between immigration and unemployment: Panel data evidence from Canada. Economic Modelling 50: 162–167.

47.

C-H

(2016) Confirmatory factor analysis with ordinal data: Comparing robust maximum likelihood and diagonally weighted least squares. Behavior Research Methods 48(3): 936–949.

48.

Lim

Burgoon

(2020) Globalization and support for unemployment spending in Asia: Do Asian citizens want to embed liberalism? Socio-Economic Review 18(2): 519–553.

49.

Manevska

Achterberg

(2013) Immigration and perceived ethnic threat: Cultural capital and economic explanations. European Sociological Review 29(3): 437–449.

50.

Marshall

Gurr

(2020) Polity5: Political Regime Characteristics and Transitions, 1800-2018. Center for Systemic Peace and Societal-Systems Research, Inc. Available at: https://www.systemicpeace.org/inscr/p5manualv2018.pdf (accessed 13 March 2024).

51.

Mau

Burkhardt

(2009) Migration and welfare state solidarity in Western Europe. Journal of European Social Policy 19(3): 213–229.

52.

Miller

Boix

Rosato

(2022) Boix-Miller-Rosato dichotomous coding of democracy, 1800-2020. Harvard Dataverse. Available at: https://dataverse.harvard.edu/citation?persistentId=doi:10.7910/DVN/FENWWR (accessed 28 June 2024).

53.

Mirilovic

(2010) The politics of immigration: Dictatorship, development, and defense. Comparative Politics 42(3): 273–292.

54.

Natter

(2018) Rethinking immigration policy theory beyond ‘Western liberal democracies’. Comparative Migration Studies 6(1): 1–21.

55.

Naumann

Stoetzer

(2018) Immigration and support for redistribution: Survey experiments in three European countries. West European Politics 41(1): 80–101.

56.

Öktem

(2020) The welfare state as universal social security: A global analysis. Social Inclusion 8(1): 103–113.

57.

Ottaviano

GIP

Peri

(2012) Rethinking the effect of immigration on wages. Journal of the European Economic Association 10(1): 152–197.

58.

Özden

Wagner

(2014) Immigrant versus natives? Displacement and job creation. Policy research working papers no. 6900. Washington, DC: The World Bank. Available at: http://elibrary.worldbank.org/doi/book/10.1596/1813-9450-6900 (accessed 21 May 2019).

59.

Pemstein

Marquardt

Tzelgov

, et al. (2024) The V-Dem measurement model: Latent variable analysis for cross-national and cross-temporal expert-coded data. V-Dem working paper no. 21, 9th edn. Gothenburg: Varieties of Democracy Institute. Available at: https://v-dem.net/media/publications/wp21_2024.pdf (accessed 13 March 2024).

60.

Putnam

(2007) E Pluribus Unum: Diversity and community in the twenty-first century the 2006 Johan Skytte Prize Lecture. Scandinavian Political Studies 30(2): 137–174.

61.

Quillian

(1995) Prejudice as a response to perceived group threat: Population composition and anti-immigrant and racial prejudice in Europe. American Sociological Review 60(4): 586–611.

62.

Rabe-Hesketh

Skrondal

(2008) Multilevel and Longitudinal Modeling Using Stata, 2nd edn. College Station, TX: Stata Press Publication.

63.

Rehm

(2009) Risks and redistribution: An individual-level analysis. Comparative Political Studies 42(7): 855–881.

64.

Rodrik

(1998) Why do more open economies have bigger governments? Journal of Political Economy 106(5): 997–1032.

65.

Roemer

Van der Straeten

(2006) The political economy of xenophobia and distribution: The case of Denmark. Scandinavian Journal of Economics 108(2): 251–277.

66.

Ruhs

(2018) Labor immigration policies in high-income countries: Variations across political regimes and varieties of capitalism. The Journal of Legal Studies 47(S1): S89–S127.

67.

Savalei

(2021) Improving fit indices in structural equation modeling with categorical data. Multivariate Behavioral Research 56(3): 390–407.

68.

Schlueter

Davidov

(2013) Contextual sources of perceived group threat: Negative immigration-related news reports, immigrant group size and their interaction, Spain 1996–2007. European Sociological Review 29(2): 179–191.

69.

Schmidt

Spies

(2014) Do parties “playing the race card” undermine natives’ support for redistribution? Evidence from Europe. Comparative Political Studies 47(4): 519–549.

70.

Schneider

(2007) Anti-immigrant attitudes in Europe: Outgroup size and perceived ethnic threat. European Sociological Review 24(1): 53–67.

71.

Semyonov

Raijman

Gorodzeisky

(2008) Foreigners’ impact on European societies: Public views and perceptions in a cross-national comparative perspective. International Journal of Comparative Sociology 49(1): 5–29.

72.

Senik

Stichnoth

Van der Straeten

(2009) Immigration and natives’ attitudes towards the welfare state: Evidence from the European Social Survey. Social Indicators Research 91(3): 345–370.

73.

Shin

(2017) Tyrants and migrants: Authoritarian immigration policy. Comparative Political Studies 50(1): 14–40.

74.

Solano

Huddleston

(2020) Migrant Integration Policy Index 2020. Barcelona; Brussels: CIDOB & MPG. Available at: https://www.mipex.eu (accessed 13 February 2023).

75.

Svallfors

(1997) Worlds of welfare and attitudes to redistribution: A comparison of eight western nations. European Sociological Review 13(3): 283–304.

76.

Svetina

Rutkowski

(2020) Multiple-group invariance with categorical outcomes using updated guidelines: An illustration using Mplus and the lavaan/semTools packages. Structural Equation Modeling: A Multidisciplinary Journal 27(1): 111–130.

77.

The World Bank (2023) Unemployment, total (% of total labor force) (modeled ILO estimate). Available at: https://data.worldbank.org/indicator/SL.UEM.TOTL.ZS (accessed 30 June 2023).

78.

Thiollet

(2022) Migrants and monarchs: Regime survival, state transformation and migration politics in Saudi Arabia. Third World Quarterly 43(7): 1645–1665.

79.

United Nations Department of Economic and Social Affairs (UN DESA) (2020a) International Migration 2020: Highlights. New York: UN DESA. Available at: https://www.un.org/development/desa/pd/sites/www.un.org.development.desa.pd/files/undesa_pd_2020_international_migration_highlights.pdf (accessed 21 November 2023).

80.

United Nations Department of Economic and Social Affairs (UN DESA) (2020b) International Migrant Stock 2020. Available at: https://www.un.org/development/desa/pd/sites/www.un.org.development.desa.pd/files/undesa_pd_2020_ims_stock_by_sex_and_destination.xlsx (accessed 22 May 2023).

81.

Vaccaro

(2021) Comparing measures of democracy: Statistical properties, convergence, and interchangeability. European Political Science 20(4): 666–684.

82.

Van Oorschot

WJH

(2006) Making the difference in social Europe: Deservingness perceptions among citizens of European welfare states. Journal of European Social Policy 16(1): 23–42.

83.

Yang

Miao

(2020) Immigration and public attitudes towards social assistance: Evidence from Hong Kong. Journal of Economic Policy Reform 25: 28–44.

84.

Yaremych

Preacher

Hedeker

(2023) Centering categorical predictors in multilevel models: Best practices and interpretation. Psychological Methods 28(3): 613–630.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.29 MB