European Welfare States and Migrant Poverty

Abstract

In almost all European welfare states, immigrants face a higher risk of poverty than natives, but the gap between the two groups varies. In examining this variation, our article contributes to the nascent literature on the impact of welfare states on immigrants. We hypothesize that whether immigrants benefit from welfare generosity depends on three intervening factors: immigration policy, labor market regulation, and welfare eligibility rules. We use fuzzy-set analysis to examine the interplay of these determinants in 16 West-European states. The findings show that in most countries a high migrant disadvantage results from the combination of a large share of humanitarian and family immigrants and generous social policies. The underlying mechanism is that “unwanted” immigrants are institutionally impeded from full access to generous welfare states.

Keywords

migration social welfare programs European politics political economy

Introduction

In almost all European countries, immigrants face a higher risk of poverty than natives. Although in the 27 European Union (EU) member states the poverty rate (below 60% of median income) of natives was on average 15.3% in 2007, for non-EU27 born immigrants this rate was 26.4%.¹ Migrant poverty, however, varies greatly across European countries.² This article examines the institutional reasons for this variation across 16 West-European states. Several studies have already demonstrated that the difference in the socioeconomic position of immigrants and natives cannot be fully explained by the social composition of the migrant population (Barrett & Maître, 2013; Büchel & Frick, 2005). Therefore, this article focuses on institutional factors, in particular the interplay of the welfare state, immigration policy, and labor market regulation.

The article builds on recent work by Sainsbury and Morissens (Morissens & Sainsbury, 2005; Sainsbury, 2012) that shifted the focus of research on immigration and the welfare state. A considerable amount of studies have interrogated the effects of immigration on support for the welfare state (e.g., Alesina & Glaeser, 2004; Banting & Kymlicka, 2006). By contrast, the new strand of research examines the impact of welfare states on immigrants. Are the social risks that immigrants face adequately alleviated? Do immigrants profit from social policies to the same extent as the native population? Comparative welfare state scholars have only started to address these questions (Castles & Schierup, 2010; Sainsbury, 2012, Ch. 1). At the same time, scholars in comparative political economy have recently pointed to the vulnerability of immigrants within countries with otherwise inclusive systems of social solidarity (Iversen & Soskice, 2015; Thelen, 2014). Our article contributes to this agenda in three ways. First, we demonstrate that the interplay of different institutions is crucial for explaining the socioeconomic position of immigrants. Second, we focus specifically on the differentiation of social rights between immigrants and the native population by examining what explains differences in poverty rates between the two groups. Third, this is, to our knowledge, the first study to systematically examine institutional determinants of migrant poverty across a wider range of countries: 16 West-European states (the EU15, excluding Luxembourg but including Norway and Switzerland). We use fuzzy-set Qualitative Comparative Analysis (fsQCA) because this method allows us to identify the combinations of institutions linked to strong migrant disadvantage. We supplement the fuzzy-set analysis with three short case studies to examine the causal mechanisms.

Our research question is: What are the institutional determinants of immigrants’ disadvantage in terms of poverty? Our main argument is that the fate of immigrants in advanced welfare states cannot be explained by welfare generosity alone as several institutional mechanisms determine to what extent immigrants benefit from welfare generosity. Our empirical findings show that the combination of a large share of humanitarian and family immigrants with generous social policies implies a marked migrant disadvantage. A large proportion of family reunion immigrants and refugees is often not intended but the result of past policies (e.g., guest worker programs), legal constraints (e.g., seeking asylum as a human right), and the international environment (e.g., humanitarian crises). As these immigrants are “unwanted,” their socioeconomic integration is not sufficiently facilitated and they lack access to the labor market and the welfare state at the same level as the native population. Therefore, generous welfare states that receive a lot of unwanted immigration differentiate social rights in a way that leads to a high poverty gap between immigrants and natives.

The following section reviews the literature and presents our argument in more detail. After a brief explanation of methods and operationalization, we then report and discuss the results of the fuzzy-set analysis and the case studies. The concluding section summarizes the main findings and reflects upon the political driving forces behind the identified institutional configurations.

Boundaries and the Welfare State

As mentioned, there has been little comparative research so far on how immigrants fare in developed welfare states. In particular, we are not aware of any study that has systematically compared institutional determinants of migrant poverty across a larger set of countries. Social rights of immigrants originally were considered in the context of research on migration and citizenship in the early 1990s (Brubaker, 1989; Soysal, 1994). Among welfare state scholars most of the attention was on whether immigration undermines popular support for the welfare state (e.g., Alesina & Glaeser, 2004; Banting & Kymlicka, 2006). Since the mid-1990s, the literature started to appreciate the complex relationship between migration and welfare states (e.g., Dörr & Faist, 1997; Geddes, 2003), and research on immigrant integration began to look beyond dedicated integration policies by taking a broader political-economy perspective (Freeman, 2004). An important empirical examination of immigrants’ social rights comes from Morissens and Sainsbury (especially, Morissens & Sainsbury, 2005; Sainsbury, 2012).

In her seminal book, Sainsbury (2012) finds “there are considerable cross-national variations in immigrants’ social rights and that the type of welfare regime matters” (p. 280). In her analysis, consisting of six historical case studies and descriptive statistics from Luxembourg Income Study data, immigrants fare best in social-democratic welfare regimes, as opposed to conservative corporatist and liberal regimes (cf. Esping-Andersen, 1990). She attributes this to the high degree of decommodification in social-democratic welfare states. “The distinctive policy objective of the social-democratic regime has been equalitarian redistribution, and this objective should benefit immigrants” (Sainsbury, 2012, p. 14).³ Sainsbury also considers the effects of integration regimes and immigration entry categories. She does not reach a general conclusion for integration regimes, but finds for entry categories that the formal social rights of asylum seekers, undocumented immigrants, and reunited family members are often more restricted than those of recognized refugees, co-ethnic citizens, and EU citizens (Sainsbury, 2012, p. 130).

In this article, we focus on institutions that differentiate the social rights of immigrants. According to Ferrera (2005), national boundaries are an important condition for collective solidarity in welfare states. In addition, there are membership spaces within national welfare states, defined by internal boundaries that adjust social policy to the socioeconomic heterogeneity of the population in the national territory. Several scholars of migration and welfare states have highlighted how the social rights of immigrants are in various ways subject to external or internal boundary-building by welfare states. According to Geddes (2003), migration has created organizational and ideological “pressures to demarcate more tightly a community of legitimate receivers of welfare state benefits” (p. 150). Dörr and Faist (1997) discuss how immigration policy and social policy eligibility rules intersect in facilitating or impeding migrant access to European welfare states. Emmenegger and Careja (2012) argue that immigrants were particularly affected by recent social policy retrenchments.

Although there is good evidence that welfare state generosity is an important determinant of poverty reduction (Brady, 2003; Moller, Huber, Stephens, Bradley, & Nielsen, 2003), we concentrate on three institutions that may condition to what extent immigrants benefit from the poverty-reducing capacities of welfare states: immigration policy, labor market regulation, and social policy eligibility rules. First, immigration policy shapes not only the overall size of immigration but also the composition of the immigrant population. To simplify matters, we distinguish between migrants for work or study on the one hand and humanitarian migrants (asylum seekers and recognized refugees) as well as joining family members on the other hand. For several reasons, humanitarian and family immigrants may face a higher risk of poverty: their motivation for entering the host country is not employment (though they may still be eager to find work); refugees in particular may be traumatized by experiences in their country of origin; their skills are less likely to be adapted to labor market needs; and they may face higher institutional hurdles to socioeconomic integration (e.g., work permissions and benefit regulations). Some research has indicated that the skill profile of humanitarian and family immigrants contributes to their worse labor market outcomes compared with other immigrant groups (Connor, 2010; Constant & Zimmerman, 2005). However, this is insufficient to explain their lagging socioeconomic integration. Instead, a comparative perspective shows that integration is conditioned by institutional rules of access to the labor market and social policies (Bevelander & Pendakur, 2014; Büchel & Frick, 2005).

Immigration policy is subject to many constraints, which can make it hard for governments to implement precise immigration objectives. For several reasons, the immigrants a country receives might not reflect its policy objectives (Joppke, 1998; Messina, 2007). Family immigration, for instance, is often a result of policy legacies, especially guest worker programs, even after a state decided to minimize immigration. Humanitarian immigration, such as political asylum, is protected by international and often constitutional law. Even if countries, such as Sweden, are politically committed to receiving refugees, the eventual intake may exceed anticipated numbers due to humanitarian crises (e.g., the 1990s war in Yugoslavia). Hence, immigration by refugees and joining family members can be higher than politically endorsed. As a consequence, domestic institutions are not geared to supporting the integration of these immigrants. In some cases, integration policies are underdeveloped due to the unexpected immigration numbers, and in other cases governments have actively restricted the social rights of foreigners in response to migration trends (Emmenegger & Careja, 2012). The resulting differentiation of access to benefits and services with respect to natives has particularly notable social consequences in generous welfare states.

Second, the employment opportunities of immigrants depend in part on labor market regulation. Good-quality employment not only raises disposable income but also facilitates access to social insurance benefits. Labor market scholars widely share the view that strong employment protection for workers in full-time, permanent jobs has negative consequences for the labor market chances of those seeking employment or in unstable, precarious jobs (Emmenegger, Häusermann, Palier, & Seeleib-Kaiser, 2012; King & Rueda, 2008; Lindbeck & Snower, 1988). Immigrants are, by definition, entrants into the national labor market. Therefore, they will find it more difficult to get stable jobs in a highly regulated labor market. However, the less secure, peripheral jobs may still pay sufficiently to minimize migrant disadvantage in terms of poverty. Besides, a weakly regulated labor market may be more accessible to immigrants, but may offer mostly low-paid jobs (Kogan, 2006). Although, therefore, it is clear that employment regulation shapes the job opportunities of immigrants, the direction of the effect is ambiguous. For our outcome of interest, general socioeconomic disadvantage, much depends on the alternatives to employment income, especially welfare state provisions. For instance, in a state with a deregulated labor market as well as low social benefits, immigrants are exposed to low-wage work without solid social protection, and low wages are further facilitated by the lack or low level of reservation wages.

Third, eligibility rules can be expected to differentiate the impact of welfare generosity on the socioeconomic status of immigrants. Benefit access for immigrants is eased if eligibility is universal or means-tested as opposed to insurance-based because the latter requires the accumulation of insurance contributions over sufficiently long spells of regular employment (Banting, 2000). Immigrants have on average less time to acquire those contribution records. In addition, this is more difficult for them if they face higher hurdles to standard employment. Hence, a generous but strongly insurance-based welfare state may imply disadvantages for immigrants in terms of poverty.

Overall, therefore we hypothesize that a generous welfare state (denoted below by G) produces a high disadvantage for immigrants in terms of poverty (D) if it combines with an immigration policy that leads to a high proportion of humanitarian and family immigrants (H) or strongly insurance-based benefit eligibility (I). As mentioned above, high employment regulation (E) could matter here as well, but its effects are ambiguous. In the notation of fuzzy-set QCA, the hypothesis can be expressed as follows, where “*” stands for a logical AND, “+” for a logical OR, and “→” for a sufficient condition.

Hypothesis 1 (H1): G*(H + I) → D

So far we have focused on institutional mechanisms that limit immigrants’ access to generous welfare states. For lean welfare states, we follow the standard assumption in the literature (e.g., Sainsbury, 2012) that these are insufficient to ensure equally low poverty risks among natives and immigrants.

Hypothesis 2 (H2): ~G → D

In sum, although we expect welfare generosity to matter for immigrants, it is crucial to take the limitations of immigrant access into account. Arguments about boundaries of welfare excluding immigrants have been articulated in the literature before. Yet, we are the first to systematically assess their effects across a large number of states.

Methods and Operationalization

We analyze the impact of welfare state generosity, welfare state eligibility, immigration policies, and labor market regulation on migrant disadvantage with fsQCA. There are three reasons for this choice of method. First, we are interested in the potential joint effects of these four institutional conditions. Indeed, our first hypothesis is based on the conjunction between welfare generosity and two other institutions. One or both of these institutions may constitute, together with welfare generosity, a sufficient condition for migrant disadvantage. Regression analysis uses interaction effects for this purpose. However, the analysis of interaction effects in multiple regressions is less flexible than fsQCA and more complex (three- or four-way) interaction effects are difficult to interpret (Vis, 2012). Second, fsQCA allows for equifinality. This is important as different combinations of institutions may imply a high migrant disadvantage in different countries. For example, in lean and in generous welfare states respectively different institutions may impede migrants’ socioeconomic integration, as suggested by our two hypotheses. Third, this method deals well with the intermediate number of cases (16) in our study, for which the possibilities of statistical analysis are limited (Ragin, 2008).

QCA uses Boolean algebra to identify sufficient and necessary conditions for an outcome, a sufficient or necessary condition potentially consisting of a configuration of several single causal conditions (as in H1). Fuzzy-set QCA is a variant of QCA in which cases can be awarded partial membership scores, between 0 (full non-membership) and 1 (full membership), in the causal conditions and the outcome. This maintains the graduations in the underlying data, but still makes a qualitative distinction between cases being a member of a set or not, by scoring them above or below the 0.5 threshold. Two coefficients are used to assess solution formulas. The consistency coefficient measures to what extent the data are consistent with the assertion that a certain condition is necessary or sufficient. The coverage coefficient of a sufficient condition indicates how much of the outcome across all cases is explained by the condition. QCA can produce different solution formulas depending on how researchers deal with logical remainders, that is, the logical combinations of conditions that are not covered by empirical cases. We focus on the so-called intermediate solutions, which are derived with the help of relatively uncontroversial theoretical assumptions (Schneider & Wagemann, 2012).

Operationalization

Immigrants in this article are defined as those born in a different country than the one they reside in. We focus on immigrants from outside the EU27 as this group faces, on average, more difficulties than EU migrants. Developed capitalist countries differ widely in their histories and patterns of immigration. To contain some of this variation, we focus on Western Europe. Within Western Europe we have excluded states with less than a million inhabitants (Luxembourg, Cyprus, Iceland, and Malta) as migration dynamics in states with small populations can differ for reasons outside of our interest. This leaves us with the following 16 states: Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, and the United Kingdom. The financial and economic crisis that erupted in the autumn of 2008 has sent national economies, labor markets, and income levels into turmoil and decline. This has direct and indirect consequences for the relative poverty measure we use (see below). We avoid the peculiarities of these developments and focus the analysis instead on the last pre-crisis years (2007-2008).⁴

As measure of poverty, we deploy the commonly used indicator of relative income poverty: receiving disposable household income below 60% of the national median, equivalized by household size. Although poverty has more dimensions than low income, we use this measure as a general indicator of economic hardship. We reduce our population sample to those aged between 16 and 64 because old-age poverty is strongly determined by just one policy: the structure and generosity of pensions.

The poverty figures in this article are based on data from Eurostat, European Union Statistics on Income and Living Conditions (EU-SILC), years 2007 and 2008.⁵ This article does not seek to account for migrant poverty as such, which is influenced by many economic and institutional conditions that affect also the native population. Rather, we are interested in cross-national variation of the extent to which immigrants are more often poor than those born in the respective country. We have estimated these differences with the EU-SILC microdata, controlling for the basic demographic composition of the immigrant population (age, gender, and education; for details see methodological online appendix). Figure 1 displays the distribution of migrant disadvantage in terms of poverty across the 16 cases of this study. We take the average of 2007 and 2008 to limit the noise in the data.

Figure 1.

Poverty differences in percentage points between non-EU immigrants and native population, 2007-2008.

In all countries of our sample, except Portugal, non-EU immigrants are more often in poverty than the native population. At the same time, the size of the effect of migrant status on poverty varies profoundly. In Belgium and Finland, non-EU immigrants are respectively 26 and 22 percentage points more likely to be in poverty than the native population. By contrast, the poverty risk is only between 7 and 11 points higher in countries like Spain, Italy, the United Kingdom, and Norway. The figures for Germany are underestimated because the data do not distinguish between EU and non-EU immigrants, and EU immigrants are generally less often poor than non-EU immigrants (we adjust for this in the fsQCA, see online appendix). Portugal stands out for its low migrant poverty risk. The impact of migrant status in Portugal is statistically not significantly different from zero. All other poverty differences are highly significant.

The distribution in Figure 1 does not immediately reflect a clear clustering into welfare regime types (Esping-Andersen, 1990). The disadvantage of migrants in terms of poverty is smaller in most South-European welfare states and the United Kingdom, but at the same time, it is very high in Ireland and relatively high in Greece. It is therefore important to carefully examine the causes behind the marked variation in migrant disadvantage.

In this section, we only present the operationalization of the outcome and causal conditions, while the calibration, that is, transforming the original data into set membership scores, is discussed in the online appendix. Having reported the outcome of interest above, we turn to the causal conditions. These are based on average values for the years 2003 to 2007—except the share of humanitarian and family immigrants due to data limitations.

To measure welfare state generosity, we use social expenditure (public and private mandatory) as a percentage of GDP (Organisation for Economic Co-Operation and Development [OECD], 2013). We exclude pension spending because this strongly depends on the age structure and because we focus our analysis on the working-age population. Since Esping-Andersen (1990) and the discussion on measuring welfare retrenchment (Allan & Scruggs, 2004), expenditure has fallen from grace among welfare state researchers as a measure of generosity: the amount of spending can be seen as less important than the actual entitlements of recipients, the indicator can be biased by changes in the denominator (GDP), and spending responds to fluctuations in needs as well as policy changes. However, the advantage of this measure for our purposes is that it is comprehensive, while excluding unsuitable aspects of other indicators. The commonly used decommodification index by Scruggs (2004) would have several drawbacks for our study: it includes pensions; it includes aspects of eligibility, which we prefer to measure separately; it is constructed around the average production worker, which is not necessarily relevant for immigrants; and it does not take into account social assistance and social services.

We operationalize welfare state eligibility by the extent to which social expenditures are financed from social contributions (own calculations with data from OECD, 2013). Although this is obviously a simplification, it is a good proxy. Funding from social contributions typically goes along with insurance-based eligibility rules (Bonoli, 1997). For individuals with low income and weak labor market attachment, like immigrants, entitlement based on insurance is more exclusive than needs- or residence-based entitlement.

Comparative indicators of immigration policies are not available (Bjerre, Helbling, Römer, & Zobel, 2014).⁶ As a proxy, we therefore take the composition of the non-EU immigrant population in terms of entry categories. We argue that this share reflects the long-term immigration regime of a country. We use the Eurostat Labor Force Survey ad-hoc module from 2008, which included a question on reasons for migration.⁷ We merge the reasons “international protection,” “accompanying family/family reunification,” and “family formation” to calculate the share of humanitarian and family immigrants.

Our measure of labor market regulation is based on the OECD indicator of employment protection, in particular protection against individual dismissals (OECD, 2013). This is a widely used indicator based on an assessment of legislation, collective agreements, and case law. We focus on dismissal protection of regular contracts as a crucial element of regulation that can lead to a dualization of the labor market.

In the following analysis, the causal conditions are defined as follows: generous welfare states (G), insurance-based welfare states (I), states with a high share of humanitarian and family immigrants (H), and states with strict employment regulation (E). The outcome, states with a large migrant poverty disadvantage, is denoted by D. The letters indicate the presence of a condition (fuzzy score above 0.5), a ~ sign before the letter indicates absence (fuzzy score below 0.5). The number of causal conditions that can be included in a QCA of 16 cases is limited. Due to the number of logically possible combinations of the conditions (see also below), including more than four conditions would not be sensible. Although potentially many factors can influence the socioeconomic status of immigrants, we focus on the composition of the immigrant population (demographics and entry categories) and institutions that can give rise to differences between immigrants and natives.

Findings: Determinants of Disadvantage

Although our main interest is in sufficient conditions for a high migrant disadvantage, checking for necessary conditions is both substantively informative and relevant to take into account when implementing the analysis of sufficient conditions. We have tested the presence and absence of all four single conditions that may be linked to a high migrant disadvantage. Schneider and Wagemann (2012) recommend that conditions should only be considered necessary if their consistency is higher than 0.9. By this standard, none of the tested conditions is necessary for the occurrence of marked migrant disadvantages (the highest consistency score is 0.78 for G).

As described above, the analysis of sufficient conditions for a high migrant disadvantage in terms of poverty (our outcome D) includes four institutional causal conditions (H, I, E, and G). These conditions can combine in 2⁴ = 16 different ways. The cases in our analysis provide only data for eight of these logically possible combinations, leaving us with eight so-called logical remainders (see online appendix, Table A6). Not all of the logical remainders are equally likely to contribute to a high migrant disadvantage. Therefore, we introduce assumptions of conditions likely to contribute to the outcome. These assumptions are the presence of insurance-based eligibility (I) and a high share of humanitarian immigrants (H). They do not change the results drastically.⁸ The results are presented in Table 1.

Table 1.

Sufficient Conditions for Migrant Disadvantage (D).

Solution	H^*G	~E^*~G
Countries covered	Sweden	Ireland
	Denmark	United Kingdom
	Belgium	Switzerland
	Finland
	Germany
	Norway
	Austria
	Netherlands
	France
Consistency	0.84	0.90
Raw coverage	0.72	0.43
Unique coverage	0.45	0.16

Solution consistency	0.83
Solution coverage	0.88

Consistency threshold is set at 0.89. Countries are listed in the order of their membership scores in the given pathway. Norway and the United Kingdom are contradictory cases as they do not have a high migrant disadvantage. D = states with a large migrant poverty disadvantage; H = states with a high share of humanitarian and family immigrants; G = generous welfare states; E = states with strict employment regulation.

The results show that two distinct pathways imply poverty rates among immigrants that are clearly higher than among natives. The first pathway, covering the majority of cases, indicates that a high share of humanitarian and family immigrants in a generous welfare state implies a high migrant disadvantage. This situation occurs in Nordic states (Sweden, Denmark, and Finland) and Continental states (Belgium, Germany, Austria, Netherlands, and France). A deviating case is Norway, which also has a high humanitarian and family share and a generous welfare state, but not a high difference in migrant poverty. The second pathway is a combination of low employment protection with lean social policy. It covers Ireland, Switzerland, and the United Kingdom. The United Kingdom is a deviating case here. It shares the causal conditions, but it does not have a high migrant disadvantage.

The findings show that conjunctural causation and equifinality matter for explaining migrant disadvantage. Although a generous welfare state or a high humanitarian and family immigration share by themselves do not imply high migrant disadvantage, the combination of these conditions does. Moreover, there are two different ways to high migrant disadvantage. In countries with generous welfare states, the composition of the immigrant population resulting from immigration policy is important for migrant disadvantage, confirming H1. The second pathway conforms only partly to H2. Not a lean welfare state by itself but in combination with low labor market regulation implies migrant disadvantage. In these countries, migrants are probably employed in low-paid, dead-end jobs, and meager welfare state provision does not counterbalance the low employment income.

The XY plot in Figure 2 shows the membership score of cases in the solution formula as well as in the outcome. As a rule of thumb, a condition is sufficient if most cases are located above or on the diagonal line (Ragin, 2008, p. 48). The narrow majority of cases are indeed located above or near the diagonal line. Importantly, cases should be differentiated beyond this rule of thumb by considering where cases lie with respect to the 0.5 crossover points on either axis (Schneider & Rohlfing, 2013). Typical cases for our solution formula are those above the diagonal and in the upper-right quadrant: France, Switzerland, Belgium, Finland, and Ireland. They are members of the solution set and outcome, while also fulfilling the consistency criterion. Sweden, Denmark, Germany, Austria, and the Netherlands are deviating only in degree. They are within solution and outcome set, but miss the consistency criterion by being more clearly members of the solution than of the outcome. All the cases in the lower-left quadrant are in line with our solution because they are neither part of the sufficient condition configuration nor of the outcome: Portugal, Italy, and Spain. The case in the upper-left quadrant of the plot, Greece, does not contradict the solution because it is not a member of the path that is posited as a sufficient condition. Yet, Greece also has a high migrant disadvantage and therefore should be explained by different causal conditions (see below). Overall, this leaves us with only Norway and the United Kingdom in contradiction of the solution formula.

Figure 2.

XY plot of sufficient condition for migrant disadvantage (D).

Norway has a high humanitarian and family immigrant share as well as a generous welfare state, but does not have a strong migrant disadvantage. Although research is lacking to explain the better outcome for immigrants in Norway, studies have found that the occupations of non-EU immigrants are more similar to natives (Dustmann & Frattini, 2013), and that the earnings gap between non-Western immigrants and natives in the lower half of the income spectrum is lower compared with Denmark and Sweden (Jakobsen, Korpi, & Lorentzen, 2014). It seems likely that these outcomes are brought about by full employment as result of the oil boom as well as better coverage of immigrants in the wage-setting system. The United Kingdom has a deregulated labor market and a lean welfare state, but immigrants do not suffer substantially more often from poverty than natives. This is probably linked to the fact that most non-EU immigrants in Britain come from former parts of the Empire and are more easily integrated due to lower language, cultural, and legal barriers. By contrast, Ireland and Switzerland do not have any post-colonial immigration. Although the “postcolonial bonus hypothesis” has not been tested extensively, Oostindie (2011) finds evidence for it in the Dutch case.

As advised by the methodological literature, we ran several robustness checks for the identified solution (Emmenegger, Kvist, & Skaaning, 2013; Schneider & Wagemann, 2012, pp. 284-287). Here we list only the most important ones. We lowered the crossover point of the outcome calibration so that Norway becomes a member of the set with high migrant disadvantage (as well as Spain). This only adds a third pathway to the baseline solution, namely I*~G, covering Spain, Greece, and Italy (where Italy is a contradictory case). This suggests that the strong insurance principle in Southern Europe diminishes immigrants’ access to, anyway, lean welfare states. It provides an explanation for the Greek case, which was not covered by our baseline model. Next, we lowered the consistency threshold of the truth table analysis so that the truth table row of Spain, Greece, and Italy is included. This leads to the same additional solution path for Southern Europe (I*~G; Spain, Greece, and Italy). Furthermore, we have excluded first Germany and then Denmark and Finland to test whether our imputed figures for missing data (see online appendix) influenced the findings. The first does not change the solution, whereas the latter becomes a subset of the baseline model, indicating relevance of insurance principle and employment regulation in addition to immigration policy (E*H*G + I*H*G + ~E*~G). Finally, adding the assumption (when deriving the intermediate solution) that labor market regulation contributes to migrant disadvantage does not alter our results. Although these robustness checks provide some interesting further insights regarding Greece and the relevance of insurance principle and employment regulation, they bolster our confidence in the baseline result above.

In QCA research, it is often recommended to analyze the absence of the outcome in addition to its presence. Due to the set-relational foundation of the method, the results can differ. However, the explanation of low migrant disadvantage would also require a different theory, which is beyond the scope and research interest of this article. For the sake of completeness, we report the results of the sufficient condition of the absence of a notable migrant disadvantage in the online appendix (Table A8). The solution yields one causal path (E*~H*~G) that covers Southern Europe (with Greece as contradictory case). It conforms to our results above as it includes the reverse of the main causal path above (H*G). That it additionally contains high employment regulation suggests that immigrants manage to generate enough income in the secondary segment of the labor market to minimize their disadvantage in terms of poverty. The solution has a low coverage score as it does not cover the United Kingdom and Norway, which also have low migrant disadvantage.

Case Studies: Checking the Causal Mechanisms

Due to the limited number of conditions that can be included in a QCA model with 16 cases, the analysis had to remain at a fairly abstract level. Accordingly, it is generally recommended (e.g., Emmenegger et al., 2013) to complement formal QCA with case studies to verify the causal mechanisms. Given the space limitations, this can only be very brief. We consider one typical case from both pathways of the solution formula: France as a case with a high humanitarian and family share of immigrants as well as a generous welfare state and Ireland as a case with a deregulated labor market and a small welfare state. In addition, we examine Sweden to learn more about the first pathway in the Nordic context.

Migrant Disadvantage in a Generous Continental Welfare State With Humanitarian and Family Immigration: France

France has received large numbers of post-colonial migrants before and after the Second World War. In addition, it attracted low-skilled immigrants through so-called guest worker programs in the 1950s and 1960s. Since the 1970s, immigration policy was tightened, but humanitarian and family immigrants remained able to enter, the latter as a consequence of previous guest worker programs and post-colonial immigration. France’s non-EU migrant population accounted for 8% of the total population (as of 2012; Eurostat, 2013a). More than half of these have entered as humanitarian or family reunion immigrants. Most are from sub-Saharan and North Africa. Despite its long experience with migration, France had no formally defined integration policy for a long time. “Because all French citizens are to be treated equally under the Republican model, there has been a great reluctance to acknowledge any ethnic divisions” (Algan, Dustmann, Glitz, & Manning, 2010, F7). Since the mid-2000s, restrictive integration policies have been introduced in the form of obligatory training sessions and self-sufficiency as a condition for naturalization.

The French welfare state is today the largest in terms of social expenditure per GDP even if Denmark and Sweden are larger when disregarding pension expenditure. Social contributions are crucial for gaining access to the generous parts of the French welfare state. As social contributions are paid from insurance-covered employment, the weak labor market integration of non-EU immigrants, especially of those who entered as humanitarian or family immigrants, is a drawback not only in terms of employment income but also access to social insurance. Indeed, large employment and earnings gaps have been found for non-EU immigrants (Algan et al., 2010, F24). Language proficiency is less of a problem for most immigrants in France, but lacking recognition of foreign qualifications hinders labor market integration, especially for highly educated migrants (OECD, 2008, p. 132). The public sector, which accounts for almost a quarter of all jobs, is practically closed to non-EU migrants (Schnapper, Krief, & Peignard, 2003, p. 39). Moreover, even immigrants’ access to the relatively generous minimum income protection is restricted: Only recognized refugees and immigrants that have resided in France for at least 5 years are eligible (OECD, 2008, p. 123).

This discussion shows how access to the generous French welfare state is limited in particular for humanitarian and family immigrants. They have substantially worse labor market positions, partly due to lacking recognition of their qualifications. In addition, insurance-based eligibility and legally restricted access to minimum income protection constrain this group’s access to the public benefits that support the native population.

Migrant Disadvantage in a Generous Nordic Welfare State With Humanitarian and Family Immigration: Sweden

Until the 1970s, Sweden attracted many labor migrants from other European countries, Yugoslavia, and Turkey. However, immigration policy changed after the oil crisis in the 1970s. Since then, refugees and family migrants from all over the world made up the largest share of immigration (OECD, 2007a, pp. 253-254). Sweden differs from many other European states by having relatively liberal refugee policies. Yet, the size of refugee inflows was larger than intended in the wake of humanitarian crises such as the war in Yugoslavia in the 1990s. In consequence, Sweden has one of the highest proportions of non-EU immigrants in Europe, namely 8.8% of the total population. Among these, a full 82% have migrated for humanitarian or family reasons, the largest share in our sample. Sweden was relatively early to develop integration policies and in 1979 enshrined multiculturalism in its constitution. It also introduced several programs to improve migrants’ labor market outcomes (OECD, 2007a, p. 255).

According to the OECD (2007b, p. 73), the generous benefits in Sweden create strong disincentives (unemployment, inactivity, or low-wage traps) for one-earner married couples with children (see also Blume, Gustafsson, Pedersen, & Verner, 2007, p. 391). This family type is not common in the native population, characterized by the dual-earner/dual-carer model, but it is frequent among low-skilled immigrants. Mood (2011) finds that immigrants receive social assistance more often and for longer periods. Therefore, there is some evidence that the composition of immigration together with the benefit structure lead to reliance on means-tested benefits and may disincentivize more active labor market integration.

At the same time, a main reason for reliance on social assistance among immigrants is that they do not have access to contribution-based benefits, even after years of residence in Sweden (Mood, 2011, p. 64). This is linked to a lacking foothold in the labor market as even in Sweden the employment record is important to gain access to generous social benefits. Labor market participation is lower among humanitarian and family immigrants than among labor immigrants. In addition, the Swedish labor market can be difficult to enter for immigrants with low educational attainments due to high minimum wages and the high-skill occupational structure (Blume et al., 2007, p. 375). However, even immigrants with high educational attainments experience that their qualifications are often not recognized. Although Sweden has had language training in place for a long time, it is voluntary and the participation rate of migrants is low. There are also indications that high employment protection reduces the chance that employers will take a risk and hire migrants (OECD, 2007a, 2007b).

Sweden is an example of a country where non-EU immigrants face a systematically higher poverty risk in spite of the best policy intentions in terms of a liberal refugee policy and supportive integration policies. We find that the combination of the generous Swedish welfare state and the composition of the immigrant population lead to lower household income vis-à-vis the native population. Apart from a generally lower labor market attachment by humanitarian and family migrants, this has to do with differences between the structure of immigrant and native households interacting with the design and incentives of the Swedish benefit system. There is also evidence of additional exclusionary mechanisms at play, such as the importance of employment record for access to benefits and the lacking recognition of qualifications.

Migrant Disadvantage in a Lean Welfare State With Deregulated Labor Market: Ireland

Ireland only transformed into a country of immigration since the mid-1990s. During its economic boom, the migrant population grew rapidly. The major share of immigration consisted of EU migrants, especially after Ireland opened its borders to migrants from the new EU member states in 2004. Yet, non-EU migrants still accounted for 3.9% of Ireland’s total population in 2012 (Eurostat, 2013a). Only a third of these were humanitarian and family immigrants. Hence, non-EU immigration was mostly motivated by employment or education. An official integration policy was lacking until 2008, as all immigration, except recognized refugees, was expected to be temporary (McGinnity, Quinn, Kingston, & O’Connell, 2013).

Immigrants on the Irish labor market, especially those from new EU member states and from outside the EU, suffer from a wage penalty with respect to the native population. Yet this penalty is in fact smaller in low-paying jobs as wages in this segment are often regulated by bipartite bodies, the Joint Labor Committees (Walsh, 2013). However, immigrants are also found systematically more often in lower occupational positions, and their occupational status does not improve significantly with the length of their stay (Barrett & Duffy, 2008). Most labor migrants work in the low-skilled and low-wage service sector where unstable employment and slim social contribution records prevent them from becoming eligible for contribution-based benefits (Timonen & Doyle, 2009, p. 168). Overall immigrants therefore fare worse on the Irish labor market than the native population.

Social expenditure in Ireland has always been below the EU15 average. Similar to the British welfare state, there is a strong emphasis on means-tested benefits, even if contribution-financed benefits have become more important in the 10 years before the crisis, aided by economic growth and low unemployment. Social policy development in this period tended to be regressive with respect to income distribution as benefit levels did not keep up with the rise in average incomes. This left the level of many benefits below the poverty line (McCashin & O’Shea, 2009). For immigrants who, as we have seen, have lower employment income, this means that their income situation is not effectively mended by the Irish welfare state.

Moreover, the Irish government has limited migrants’ rights on two fronts. Access to citizenship was markedly restricted after a referendum in 2004 (McGinnity et al., 2013). And Ireland has limited migrants’ access to its already lean welfare state through a two-year residence condition for receiving means-tested benefits (Timonen & Doyle, 2009, p. 162). This particularly affects asylum seekers who are not legal residents until granted refugee status. The reduced accessibility of the Irish welfare state for immigrants is reflected in welfare receipt being lower among migrants than among natives even controlling for demographic characteristics (Barrett & McCarthy, 2007).

In sum, we find evidence that the combination of a loosely regulated labor market and a lean welfare state constrain the socioeconomic integration of immigrants in Ireland, as indicated by the QCA analysis. Although immigrants are not excluded from the Irish labor market, they are more often found in lower quality and low-paying jobs. This disadvantage in work income is not adequately corrected by the Irish welfare state.

Discussion and Conclusion: Not Just Welfare Generosity

In all West-European countries, except Portugal, non-EU immigrants face a higher risk of poverty than natives also when statistically controlling for the composition of the immigrant population. However, the relative position of immigrants varies greatly across countries. Our article has made three main contributions to the study of the socioeconomic integration of immigrants. First, we focused explicitly on the disadvantage that immigrants encounter vis-à-vis the status of the native populations. Second, we highlighted how different institutions combine in affecting the socioeconomic outcomes of immigrants. And, third, to the best of our knowledge, we are the first to explore the institutional determinants of migrant disadvantage across a large group of West-European states. Employing fuzzy-set QCA as a novel methodological approach has proven useful because the analysis identified that different conditions combine in generating immigrants’ socioeconomic disadvantage and that more than one path brings about this outcome.

We have hypothesized that three institutions can hinder migrants’ access to generous welfare states: immigration policy conditioning the composition of the immigrant population, especially when unintendedly leading to a high share of humanitarian and family immigrants; strongly insurance-based welfare state eligibility; and labor market regulation shaping the types of jobs accessible to immigrants. Among these, we have found immigration policy to be particularly important, in line with H1. In many West-European welfare states, the higher poverty rates among immigrants are explained by the fact that these states have generous social provisions while the humanitarian and family immigrants that are principally allowed to enter the country suffer from diminished access to these provisions. Hence, a generous welfare state in itself is not enough to help immigrants. Rather, generous welfare states that receive a lot of unwanted immigration differentiate social rights in a way that leads to a high poverty gap between immigrants and natives.

In a smaller group of countries, high migrant disadvantage is linked to lean welfare states in conjunction with a deregulated labor market where mostly poor jobs are available to immigrants. In line with the literature, we had hypothesized that meager social provisions may by themselves imply a worse situation for immigrants (H2). Yet, our findings show that it is the combination of the low level of welfare generosity with low income from the labor market that is decisive.

Overall these results caution against previous findings that have highlighted the benefits of generous, social-democratic welfare states for immigrants (e.g., Sainsbury, 2012). By contrast, our study shows that institutional boundaries can keep immigrants from benefitting from welfare generosity. Moreover, we have discovered that the relation between employment regulation and immigrant labor market access is not as straightforward as a simple reading of the dualization literature would imply. We have found no general negative effect of tight employment regulation on the socioeconomic integration of immigrants. Instead, we have seen that in some countries lean welfare states and deregulated labor markets jointly result in migrant disadvantage.

Further research should look into the political factors that bring about the institutional configurations we have identified as affecting migrant disadvantage in terms of poverty. Here we offer some preliminary reflections. Historically the emergence of the combination of a generous welfare state and an immigration policy that leads to a large share of humanitarian and family immigrants is likely to have been influenced by strong labor movements. We know that strong social-democratic parties and trade unions have helped to build generous welfare states (Huber & Stephens, 2001). At the same time, labor has in principle little economic interest in an increase of labor supply through migration, whereas social democracy is committed to international solidarity when it comes to forced migration (cf. Boräng, 2012). These political dynamics can be expected to have contributed to low labor but high humanitarian immigration in Nordic states. At the same time, the more recent rise of right-wing populist parties reflects some of the distributional conflicts around the combination of a generous welfare state with a high share of humanitarian and family immigrants who often experience social exclusion (cf. Iversen & Soskice, 2015; Thelen, 2014).

Social-democratic parties were relatively influential also in Continental Europe. In addition, Christian democratic parties have actively contributed to relatively generous, even if more stratified, Continental welfare states (van Kersbergen, 1995). It seems likely that the socially conservative Catholic ideology of these parties has favored limiting immigration, but it would ideologically be hard for them to actively advocate closing borders to humanitarian and family immigrants. The configuration of meager welfare states with low employment regulation, on the other hand, is typical of liberal states where conservative or liberal parties are politically stronger (Esping-Andersen, 1990; Manow, 2009).

In our analysis, the migrant disadvantage in terms of poverty tended to be lower in South-European states with high employment regulation, low shares of humanitarian and family immigrants, and lean welfare states (when excluding pensions). In the post-war history of these countries, polarized domestic politics and “incomplete Fordism” have led to strong protection of labor market insiders and underdeveloped social services (Ferrera, 2010; Picot, 2012). Consequently, immigrant labor is used to make up for the lack of publicly provided social care (van Hooren, 2012). Although their care work is clearly in the secondary, insecure segment of the labor market, it still bolsters immigrants’ income. In addition, it leads to a higher share of labor immigrants (even if this may not always be the formal entry category). The less disadvantaged but still weak position of immigrants in this configuration illustrates a limitation of the outcome our analysis focused on, which is concerned with avoiding a higher risk of poverty but not with more ambitious goals of socioeconomic integration of immigrants.

Footnotes

Acknowledgements

We thank Anne Britt Djuve, Patrick Emmenegger, Stefania Marino, Paul Marx, Diane Sainsbury, Franca van Hooren, Barbara Vis, the editors of Comparative Political Studies, and the four anonymous reviewers for their helpful comments.

Authors’ Note

Previous versions of this article were presented at a workshop at University of Southern Denmark 2013, the ESPAnet annual conference in Poznan 2013, and the International Conference of Europeanists in Washington, DC, 2014.

Declaration of Conflicting Interests

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

Funding

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

Supplemental Material

The supplementary appendices are available at

Notes

Author Biographies

Gerda Hooijer is a DPhil (PhD) candidate at the Department of Politics and International Relations and Lady Margaret Hall, University of Oxford. Her current research focuses on immigration and preferences for redistribution in advanced democracies.

Georg Picot is an associate professor at the Department of Comparative Politics, University of Bergen. His main research interests are in welfare states, labour markets, and political parties.

References

Alesina

A. F.

Glaeser

E. L.

(2004). Fighting poverty in the U.S. and Europe. New York, NY: Oxford University Press.

Algan

Dustmann

Glitz

Manning

(2010). The economic situation of first and second-generation immigrants in France, Germany and the United Kingdom. The Economic Journal, 120(February), F4-F30.

Allan

J. P.

Scruggs

(2004). Political partisanship and welfare state reform in advanced industrial societies. American Journal of Political Science, 48, 493-512.

Banting

(2000). Looking in three directions: Migration and the European welfare state in comparative perspective. In Bommes

Geddes

(Eds.), Immigration and welfare: Challenging the borders of the welfare state (pp. 13-33). New York, NY: Routledge.

Banting

Kymlicka

(Eds.). (2006). Multiculturalism and the welfare state. New York, NY: Oxford University Press.

Barrett

Duffy

(2008). Are Ireland’s immigrants integrating into its labor market? International Migration Review, 42, 597-619.

Barrett

Maître

(2013). Immigrant welfare receipt across Europe. International Journal of Manpower, 34, 8-23.

Barrett

McCarthy

(2007). Immigrants in a booming economy: Analysing their earnings and welfare dependence. Labour, 21, 789-808.

Bevelander

Pendakur

(2014). The labour market integration of refugee and family reunion immigrants: A comparison of outcomes in Canada and Sweden. Journal of Ethnic and Migration Studies, 40, 689-709.

10.

Bjerre

Helbling

Römer

Zobel

(2014). Conceptualizing and measuring immigration policies: A comparative perspective. International Migration Review. Advance online publication. doi:10.1111/imre.12100

11.

Blume

Gustafsson

Pedersen

P. J.

Verner

(2007). At the lower end of the table: Determinants of poverty among immigrants to Denmark and Sweden. Journal of Ethnic and Migration Studies, 33, 373-396.

12.

Bonoli

(1997). Classifying welfare states: A two-dimension approach. Journal of Social Policy, 26, 351-372.

13.

Boräng

(2012). National institutions, international migration: Labour markets, welfare states, and immigration policy (Doctoral thesis). University of Gothenburg, Sweden.

14.

Brady

(2003). The politics of poverty: Left political institutions, the welfare state, and poverty. Social Forces, 82, 557-588.

15.

Brubaker

W. R.

(1989). Immigration and the politics of citizenship in Europe and North America. Lanham, MD: University Press of America.

16.

Büchel

Frick

J. R.

(2005). Immigrants’ economic performance across Europe—Does immigration policy matter? Population Research and Policy Review, 24, 175-212.

17.

Castles

Schierup

(2010). Migration and ethnic minorities. In Pierson

Obinger

Lewis

Leibfried

Castles

F. G.

(Eds.), The Oxford handbook of the welfare state (pp. 278-291). Oxford, UK: Oxford University Press.

18.

Connor

(2010). Explaining the refugee gap: Economic outcomes of refugees versus other immigrants. Journal of Refugee Studies, 23, 377-397.

19.

Constant

Zimmerman

K. F.

(2005). Immigrant performance and selective immigration policy: A European perspective. National Institute Economic Review, 194, 94-105.

20.

Dörr

Faist

(1997). Institutional conditions for the integration of immigrants in welfare states: A comparison of the literature on Germany, France, Great Britain, and the Netherlands. European Journal of Political Research, 31, 401-426.

21.

Dustmann

Frattini

(2013). Immigration: The European experience. In Card

Raphael

(Eds.), Immigration, poverty, and socioeconomic inequality (pp. 423-456). New York, NY: Russell Sage Foundation.

22.

Emmenegger

Careja

(2012). From dilemma to dualization: Social and migration policies in the “reluctant countries of immigration.” In Emmenegger

Häusermann

Palier

Seeleib-Kaiser

(Eds.), The age of dualization: The changing face of inequality in deindustrializing societies (pp. 124-150). Oxford, UK: Oxford University Press.

23.

Emmenegger

Häusermann

Palier

Seeleib-Kaiser

(Eds.). (2012). The age of dualization: The changing face of inequality in deindustrializing societies. Oxford, UK: Oxford University Press.

24.

Emmenegger

Kvist

Skaaning

S.-E.

(2013). Making the most of configurational comparative analysis: An assessment of QCA applications in comparative welfare state research. Political Research Quarterly, 66, 185-190.

25.

Esping-Andersen

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

26.

Eurostat. (2013a). Migration and migrant population statistics. http://epp.eurostat.ec.europa.eu/statistics_explained/index.php/Migration_and_migrant_population_statistics

27.

Eurostat. (2013b). Statistics. Retrieved from http://epp.eurostat.ec.europa.eu/portal/page/portal/statistics/search_database

28.

Ferrera

(2005). The boundaries of welfare: European integration and the new spatial politics of social protection. Oxford, UK: Oxford University Press.

29.

Ferrera

(2010). The South European countries. In Pierson

Obinger

Lewis

Leibfried

Castles

F. G.

(Eds.), The Oxford handbook of the welfare state (pp. 616-629). Oxford, UK: Oxford University Press.

30.

Freeman

G. P.

(2004). Immigrant incorporation in Western democracies. International Migration Review, 38, 945-969.

31.

Geddes

(2003). Migration and the welfare state in Europe. The Political Quarterly, 74(Suppl. s1), 150-162.

32.

Huber

Stephens

J. D.

(2001). Development and crisis of the welfare state: Parties and policies in global markets. Chicago, IL: University of Chicago Press.

33.

Iversen

Soskice

(2015). Democratic Limits to Redistribution: Inclusionary versus Exclusionary Coalitions in the Knowledge Economy. World Politics, 67(2), 185-225.

34.

Jakobsen

Korpi

Lorentzen

(December 2014). Immigration and integration policy and labour market attainment among Scandinavian immigrants (Working Paper 11). SFI Danish National Centre for Social Research.

35.

Joppke

(1998). Why liberal states accept unwanted immigration. World Politics, 50, 266-293.

36.

King

Rueda

(2008). Cheap labor: The new politics of “bread and roses” in industrial democracies. Perspectives on Politics, 6, 279-297.

37.

Kogan

(2006). Labor markets and economic incorporation among recent immigrants in Europe. Social Forces, 85, 697-721.

38.

Koopmans

(2010). Trade-offs between equality and difference: Immigrant integration, multiculturalism and the welfare state in cross-national perspective. Journal of Ethnic and Migration Studies, 36, 1-26.

39.

Lindbeck

Snower

D. J.

(1988). The insider-outsider theory of employment and unemployment. Cambridge, MA: MIT Press.

40.

Manow

(2009). Electoral rules, class coalitions and welfare state regimes, or how to explain Esping-Andersen with Stein Rokkan. Socio-Economic Review, 7, 101-121.

41.

McCashin

O’Shea

(2009). The Irish welfare system. In Schuber

Hegelich

Bazant

(Eds.), The handbook of European welfare systems (pp. 260-276). London, UK: Routledge.

42.

McGinnity

Quinn

Kingston

O’Connell

(2013). Annual monitoring report on integration in 2012. Dublin, Ireland: Economic and Social Research Institute and The Integration Centre.

43.

Messina

A. M.

(2007). The logics and politics of post-WWII migration to Western Europe. Cambridge, UK: Cambridge University Press.

44.

Moller

Huber

Stephens

J. D.

Bradley

Nielsen

(2003). Determinants of relative poverty in advanced capitalist democracies. American Sociological Review, 68, 22-51.

45.

Mood

(2011). Lagging behind in good times: Immigrants and the increased dependence on social assistance in Sweden. International Journal of Social Welfare, 20, 55-65.

46.

Morissens

Sainsbury

(2005). Migrants’ social rights, ethnicity and welfare regimes. Journal of Social Policy, 34, 637-660.

47.

Oostindie

(2011). Postcolonial Netherlands: Sixty-five years of forgetting, commemorating, silencing. Amsterdam, The Netherlands: Amsterdam University Press.

48.

Organisation for Economic Co-Operation and Development. (2007a). Jobs for immigrants (Vol. 1): Labour market integration in Australia, Denmark, Germany and Sweden. Paris, France: OECD.

49.

Organisation for Economic Co-Operation and Development. (2007b). OECD economic surveys: Sweden 2007. Paris, France: OECD.

50.

Organisation for Economic Co-operation and Development. (2008). Jobs for immigrants (Vol. 2): Labour market integration in Belgium, France, the Netherlands and Portugal. Paris, France: OECD.

51.

Organisation for Economic Co-Operation and Development. (2013). Statistics. Retrieved from http://www.oecd.org/statistics/

52.

Picot

(2012). Politics of segmentation: Party competition and social protection in Europe. London, UK: Routledge.

53.

Ragin

C. C.

(2008). Redesigning social inquiry: Fuzzy sets and beyond. Chicago, IL: University of Chicago Press.

54.

Sainsbury

(2012). Welfare states and immigrant rights: The politics of inclusion and exclusion. Oxford, UK: Oxford University Press.

55.

Schnapper

Krief

Peignard

(2003). French immigration and integration policy. A complex combination. In Heckmann

Schnapper

(Eds.), The integration of immigrants in European societies: National differences and trends of convergence (pp. 15-44). Stuttgart, Germany: Lucius.

56.

Schneider

C. Q.

Rohlfing

(2013). Combining QCA and process tracing in set-theoretic multi-method research. Sociological Methods & Research, 42, 559-597.

57.

Schneider

C. Q.

Wagemann

(2012). Set-theoretic methods for the social sciences: A guide to Qualitative Comparative Analysis. New York, NY: Cambridge University Press.

58.

Scruggs

(2004). Welfare state entitlements data set: A comparative institutional analysis of eighteen welfare states (Version 1.2). Retrieved from http://sp.uconn.edu/~scruggs/

59.

Soysal

(1994). Limits of citizenship. Chicago, IL: Chicago University Press.

60.

Thelen

(2014). Varieties of liberalization and the new politics of social solidarity. New York, NY: Cambridge University Press.

61.

Timonen

Doyle

(2009). In search of security: Migrant workers’ understandings, experiences and expectations regarding “social protection” in Ireland. Journal of Social Policy, 38, 157-175.

62.

van Hooren

F. J.

(2012). Varieties of migrant care work: Comparing patterns of migrant labour in social care. Journal of European Social Policy, 22, 133-147.

63.

van Kersbergen

(1995). Social capitalism: A study of Christian democracy and the welfare state. London, UK: Routledge

64.

Vis

(2012). The comparative advantages of fsQCA and regression analysis for moderately large-n analyses. Sociological Methods & Research, 41, 168-198.

65.

Walsh

(2013). Labour market regulation and migration in Ireland. The Economic and Social Review, 44, 85-102.

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

0.28 MB