Making regional citizens? The political drivers and effects of subnational immigrant integration policies in Europe and North America

ABSTRACT

This special issue provides the first internationally comparative analysis of regional immigrant integration policies. The introduction defines socioeconomic, cultural–religious and legal–political domains of integration, expecting regions to be most active policy-makers in the first. Regional politics drives policy orientations: leftist regions develop more inclusive policies than their right-wing counterparts, and Rokkan regions with strong regionalist parties adopt more assimilationist policies than ordinary regions. Through policy feedback, regional policies also influence immigrants’ political integration, shaping their prospects of becoming ‘regional citizens’. Six empirical contributions assess these arguments for five federations (Belgium, Germany, Switzerland, United States and Canada) and two quasi-federal systems (Italy and Spain).

KEYWORD:

• immigrant integration policy
• citizenship
• decentralization
• party politics
• Europe
• North America

JEL:

• H75
• J68
• K37
• R23

INTRODUCTION

In January 2019, the Italian regions of Piedmont and Tuscany announced that they would ask the Constitutional Court to review whether the recent ‘Salvini decree’ on the reception and integration of asylum seekers was compatible with the Italian constitution. Earlier, a network of mayors, led by the mayor of Palermo, had refused to implement the decree, arguing that the foreseen measures were ‘inhuman’ and that the decree interfered with local and regional authority. On the other side of the Atlantic, 16 US states sued US President Donald Trump in February 2019 for invoking emergency powers to construct a wall along the border with Mexico. These two recent episodes show that lower governmental levels increasingly compete with national and supra-national actors for ownership of the regulation of immigration, asylum and integration, setting their own distinctive policy priorities.

Largely unnoticed by scholars of social policy, migration and territorial politics, the rescaling of social and cultural policies to the subnational level has turned immigrant integration into a major competence of regional authorities. Across the globe, decentralization reforms have decisively increased regional authority over policy areas directly related to immigrant integration processes (Hooghe et al., 2016; Joppke & Seidle, 2012; Keating, 1998; Paquet, 2014; Spiro, 2002). This trend has not been restricted to federal systems. Even in unitary states, central governments have enabled subnational authorities to ‘put their spin on social policies and thereby on migrant integration strategies’ (Ireland, 2006, p. 378). Reversing Marshall’s (1950/1992) classical argument – that citizenship progresses from civic to political rights, culminating in social citizenship – regional governments hold the necessary competencies, and increasingly also use them, to equip immigrants first with social rights to then foster their political participation – even in the absence of formal citizenship (Joppke & Seidle, 2012).

We conceptualize immigrant integration policies as those steering the socioeconomic, the cultural–religious and the legal–political integration of immigrants, reflecting Entzinger’s (2000) argument that immigrants need to find their place in the market, the nation and the state, respectively. We are concerned with the integration policies of regions, defined as jurisdictions located between national and local governments (Hooghe et al., 2016, p. 24). Distinct regional policies targeting immigrants and their children constitute a steadily growing realm of legislation, in both classical and more recent immigration countries. For example, by the 2000s, all but two of Italy’s 20 regions had passed their own immigrant integration laws (Zuber, 2019, in this issue). In the United States, state-level legislation in the area of immigration and integration (regulating access to social benefits, cultural diversity or law enforcement) almost doubled in just five years, from 123 bills enacted in 2005–06 to 218 bills in 2009–10 (Filindra & Manatschal, 2019, in this issue). The fact that regional integration policy variance occasionally exceeds variance between countries – as observed in Swiss cantons (Manatschal & Stadelmann-Steffen, 2013) – shows that regional heterogeneity in this policy field is far from a marginal phenomenon.¹

In spite of this impressive policy-making activity at the regional level, we lack a systematic overview of the content of these policies, as well as sound explanations for their variance. We also do not know whether regional policy differences matter for immigrant integration (for an exception, see Manatschal & Stadelmann-Steffen, 2013). The literature on territorial rescaling of social policy has so far neglected immigrant integration as a policy field. And where migration scholars have turned to the regional level, they have either focused on aspects of multilevel governance (Adam, 2018; Adam & Hepburn, 2019; Caponio & Jones-Correa, 2018) or have studied only a specific subset of regions, focusing on national minority regions in multinational countries (Hepburn & Zapata-Barrero, 2014; Jeram et al., 2016).

With the goal of overcoming these shortcomings in previous scholarship, this special issue provides the first systematic, internationally comparative analysis of regional immigrant integration policies in Europe and North America. It asks under which conditions regions turn from pragmatic providers of decentralized social services into active creators of regional citizenship, and aim to become alternative loci of belonging for immigrants as well as for their autochthonous citizens (Henderson et al., 2013).

The multidimensional nature of integration policy, which comprises policies regulating immigrants’ socioeconomic, cultural–religious and legal–political integration, entails that immigrant integration cuts across policy areas normally dealt with in separate literatures. In order to understand how regions govern immigrant integration, we therefore need to connect the perspectives of migration studies, social policy research and territorial politics. In doing so, this introductory paper moves each of these literatures forward in a significant way. For the field of migration studies, it suggests a multidimensional perspective on immigrant integration that combines the focus on citizenship and cultural identity (from research on national models of integration) with the emphasis on social service delivery (from the literature on local integration and spatial rescaling of social policy). For the field of social policy, it tests the expectation of welfare regionalism (Keating, 2017) on the critical case of a particularly vulnerable target group neglected by the literature so far, showing the conditions under which regions open their social services to non-citizens. For territorial politics, this paper incorporates the field’s traditional subject matter of minority nationalist regions into a broader perspective, comparing the policy responses of regions with a distinct history of statehood and/or a distinct cultural and linguistic identity (hereafter termed Rokkan regions, following Hooghe et al., 2016) to those of ordinary regions that lack such a distinct identity. This leads to sound conclusions on what is, and what is not, special about how Rokkan regions respond to immigration.²

The theoretical concepts and scenarios developed in this introductory paper are confronted with findings from six empirical analyses of subnational integration policies in five federations (Belgium, Germany, Switzerland, United States and Canada) and two quasi-federal systems (Italy and Spain).³ The selected settings represent so-called classical settler states (United States and Canada), post-Second World War immigration countries (Germany, Switzerland and Belgium), as well as former emigration countries, which only recently became countries of immigration (Italy and Spain). Consequently, the regions covered in this special issue show great variation in terms of immigration realities. While some regions such as Extremadura in Spain or Apulia in Italy (3% share of the foreign-born population) are hardly affected by immigration, the foreign-born population amounts to almost one-third of the resident population in other regions (California). Variation within countries is also substantial, ranging from 2% (Newfoundland and Labrador) to 30% (Ontario) in Canada, or from 3% (Extremadura) to 14.5% (Catalonia) in Spain. This variance allows us to probe whether regional policy choices can be explained by structural (demographic and economic) conditions, or if, as we will argue, by the political agency of regional governments.

The paper is structured as follows. The next section introduces the three-dimensional conceptualization of immigrant integration and discusses theoretical arguments for regional policy choice in each of the three domains. The third section describes the research design of the special issue’s empirical contributions and discusses their main findings in light of the theoretical expectations. The fourth section concludes and points to avenues for further research.

REGIONAL INTEGRATION POLICIES: FROM SOCIOECONOMIC TO CULTURAL–RELIGIOUS AND POLITICAL INTEGRATION

Despite the manifold definitions and controversies surrounding the concept of integration, there is consensus that integration is a transversal and multidimensional policy field (Boswell, 2003; Entzinger, 2000; Joppke & Seidle, 2012; Manatschal, 2011). Several scholars agree that integration processes occur in three distinguishable domains. Joppke and Seidle (2012, p. 9) distinguish between economic, social and civic/political integration; Penninx and Garcés-Mascareñas (2016, pp. 5ff.) refer to legal/political, socioeconomic and cultural/religious integration; and Boswell (2003, p. 75) talks about immigrants’ incorporation into the economy, society and political life. Integration policies can then be defined as policies that ‘intend to guide and steer  …  integration processes of immigrants’ (Penninx & Garcés-Mascareñas, 2016, p. 10) in socioeconomic, cultural–religious and legal–political domains.

Regional policies in the socioeconomic domain might regulate non-citizens’ access to healthcare, social benefits or housing subsidies. Cultural–religious policies typically range from culturally monist (or assimilationist) to culturally pluralist (or multiculturalist) measures (Manatschal, 2011; Penninx, 2005, p. 139; Bloemraad et al., 2008). Pluralist measures could include promoting the preservation of migrants’ cultures of origin, for example, by subsidizing immigrants’ cultural associations, or allowing teachers to wear a veil. Typical examples of culturally monist policies are language policies prioritizing one official language (e.g., ‘English first’ policies) or policies making access to certain rights, such as residence, citizenship or social benefits, conditional on language proficiency.

The legal–political domain is where regions tend to hold least of the relevant competencies. Formal access to citizenship remains a prerogative of the nation-state.⁴ Nonetheless, subnational authorities can choose to grant voting rights to non-citizens for subnational elections (Arrighi de Casanova & Bauböck, 2017) and they can be more or less active in informing immigrants about opportunities to naturalize. They can further establish migration councils that represent non-citizens and advise regional governments on their specific concerns.

In all three domains, policies can have different orientations. In the socioeconomic domain, inclusive policies are those which ensure that immigrants are treated equally and can access social benefits and services with ease; exclusive policies restrict immigrants’ access to social benefits and services compared with the native population.⁵ In the legal–political domain, inclusive policies enable immigrants to access their legal rights and foster their political participation, guiding them on their path towards achieving citizenship, whereas exclusive policies restrict immigrants’ access to rights and do not encourage their political participation. In the cultural–religious domain, scholars typically distinguish orientations according to the criterion of whether integration policies accommodate cultural diversity (i.e., culturally pluralist policies) or not (i.e., monist or assimilationist policies) (e.g., Manatschal, 2011; Penninx & Garcés-Mascareñas, 2016).

The following sections develop three central arguments. First, regions are most active in the socioeconomic policy domain. Second, regional politics shapes regions’ policy orientations in the socioeconomic and cultural–religious domains. Third, while regional competencies to directly steer legal–political integration are limited, regional integration policies influence migrants’ political integration indirectly through policy feedback effects.

Immigrant integration policies in the socioeconomic domain

Earlier contributions in the field of migration studies overlooked the socioeconomic domain when identifying distinct national models of integration. Focusing on legal–political and cultural–religious aspects of integration, they saw these models as driven by historical legacies and the symbolic construction of national identity (Brubaker, 1992; Favell, 1998; Koopmans et al., 2005). Classical distinctions were made between exclusionary and inclusionary models of integration (Penninx, 2005), or between the ‘French Assimilationist’ and ‘Dutch Multiculturalist’ models (Bertossi et al., 2015). This notion of national models has recently been challenged in several ways. First, by the observation of trends towards cross-national policy convergence (e.g., Goodman, 2010). Second, through the discovery of local integration policies, which are much more concerned with the socioeconomic domain, and sometimes thoroughly at odds with an overarching national model (Scholten, 2016). This critique of ‘methodological nationalism’ (Charlie, 2008; Wimmer & Glick Schiller, 2006) in classical migration research highlights the neglect of considerable subnational policy heterogeneity, and the fact that the policy-making context at the national level may be very different from the context prevailing at the regional or local levels.

The discovery of distinct local policies triggered a vital series of contributions investigating whether they display similar characteristics to each other, even if they are nested within different national settings (e.g., Caponio & Borkert, 2010; Cinalli & Giugni, 2011). At the local level, political decisions directly meet with policy implementation and the delivery of public services, especially within the socioeconomic domain. This proximity between deciding and implementing, and the stronger involvement of client-oriented, immigrant-supportive non-governmental organizations (NGOs) in local policy-making and implementation, explains why local authorities tend to follow a less symbolic, more pragmatic approach than nation-states (Caponio & Borkert, 2010, pp. 188–190).

While not as close to clients as local administrators are, regional policy-makers are still closer to the day-to-day challenges of managing integration processes than their national counterparts. We might therefore expect regions to be more concerned with pragmatic questions of social inclusion and access to regional labour markets than they are with the grand narratives of identity and symbolic belonging, as reflected in national policies (e.g., Schmidtke & Zaslove, 2014). However, legislation and implementation at the regional level are not as closely intertwined as they are at the local level, and it is not obvious that regional legislators display a pragmatic client orientation in the legislative arena (e.g., Piccoli, 2019a). For example, Swiss cantons’ integration policies are shaped by path-dependent notions of belonging, which do reflect symbolic national discourses (Manatschal, 2012). Hence, although nested in the same national setting, not all cantons developed inclusive policies facilitating immigrants’ political participation.

The imperfect correspondence between local and regional policy-making calls for an engagement with the literature on regional models of social welfare; this will help theorize regional responses to immigrant integration in the socioeconomic domain. Several recent contributions highlight important processes of spatial rescaling of social policy to the subnational level – processes that have been ongoing since the mid-1970s (e.g., Kazepov, 2010; Keating, 2013; Kleider, 2018; Vampa, 2016). These contributions emphasize that whereas the nation-state is still an important redistributor of income, in many places regional governments are now the key actors responsible for the allocation of public services (such as healthcare or social assistance for the elderly). Thanks to these competencies, regions now decide on the socioeconomic issues most relevant for immigrants’ everyday lives.

This literature has shown that regions use social policies to strengthen distinctive regional identities and create models of regional social citizenship (Keating, 2013; Vampa, 2016). In explaining these decisions, authors focus on party politics as a key driver of regional welfare priorities (Vampa, 2016). While leftist parties generally tend to promote social solidarity at all levels – and hence also at the regional level – this tendency increases when they are excluded from national government, but in control of regional governments. And in case of a salient centre–periphery conflict, regionalist parties,⁶ when in government, use social welfare to strengthen the regional distinctiveness of Rokkan regions (Béland & Lecours, 2008; Mulvey, 2018).

The literature has not yet addressed the question of whether regions strive to also open these regional conceptions of social welfare to immigrants. Do regions use socioeconomic integration policies to engage in a race to the bottom – avoiding becoming ‘welfare magnets’ (Borjas, 1999) for immigrants as expected by economic theories of federalism? Or are we rather observing signs of ‘welfare regionalism’, as regions show social solidarity also to immigrants and build a cohesive regional society in a race to the top (Keating, 2017, pp. 11–13)? Party politics might explain different regional responses to immigrants, as systems of multilevel governance can create additional arenas for the politicization of migration policy, especially if national and regional governments are incongruent (Adam & Hepburn, 2019; Filomeno, 2016; Newton, 2018).

Based on the above discussion, we argue that regions should be predominantly concerned with the socioeconomic domain of integration, drawing an analogy from local to regional pragmatism. However, when it comes to the substantive policy choices that regions actually make in this domain, we follow the literature on regional models of social welfare and hence expect these choices to be influenced by party politics. We expect regions governed by the left to be more inclusive than centrist or centre-right regions, particularly when partisan conflict on immigration is played out across levels of government, and leftist regional governments confront their political opponents at the national level, and vice versa.

Immigrant integration policies in the cultural–religious domain

While the nexus between regional welfare models and immigrant integration is still under-researched, the relationship between immigrant integration and specific regional cultures, traditions and languages is a prominent topic in the literature on multinational states. Kymlicka’s (2001) seminal work on Quebec argued that having control over immigrant integration led the Canadian province to adopt an overall inclusive approach towards immigrants; it also allowed for assimilationist cultural policies that expected immigrants to adopt the French language and culture. Scholars have shown that the link between identity politics and immigration, and hence the cultural–religious domain of integration, is indeed particularly salient in Rokkan regions (e.g., Barker & Zapata-Barrero, 2014). Nevertheless, not all Rokkan regions behave like Quebec. While Catalonia opted for a similar mix of culturally assimilationist and socioeconomically inclusive policies (Zuber, 2014), a variety of alternative approaches has emerged in other Rokkan regions. They range from an emphasis on the multicultural Scottish nation (Hepburn, 2011), to constructing diversity as a new marker of Basqueness (Jeram, 2014), to combining multicultural policies with civic integration requirements in Flanders (Adam, 2013), and finally to excluding migration-related diversity from the construction of the South Tyrolean identity (Wisthaler, 2016).

Territorial politics scholars agree that the centre–periphery cleavage continues to be salient in multinational states, and that this also affects the issue of immigrant integration. Some Rokkan regions pursue their own nation-building projects that compete with the one advocated by the state. Concerned with peripheral nation-building and aiming to define their own (sub-)national identity, they have engaged in developing particularly pronounced and distinctive regional models of integration – often carrying as much symbolic weight as national models, and emphasizing the elements that differentiate the region from the state.

Studies in this field have predominantly focused on the immigration discourses of minority nationalist parties (Franco-Guillén, 2016; Hepburn, 2011; Jeram et al., 2016; Wisthaler, 2016), rather than their actual integration policies (Adam, 2013; Ruiz Vieytez, 2017; Zuber, 2014). Overall, however, this line of research has shown that Rokkan regions tend to behave like states and link immigration to questions of nation-building, using the cultural–religious domain to strengthen constructions of national identity (Adam, 2013; Barker, 2015).

We therefore argue, first, that Rokkan regions with strong regionalist parties will focus more on the cultural–religious domain of integration than ordinary regions and will, on average, also adopt more assimilationist policies than the latter. Second, party politics on the left–right dimension should again explain the orientation of ordinary regions, with leftist regions adopting more pluralist policies than centrist and centre-right regions.

Immigrant integration policies and the legal–political domain

In the legal–political domain, regional competences remain rather limited. With the exception of Switzerland, where naturalization is a three-tiered process involving the local, cantonal and federal levels (D’Amato, 2009), access to formal citizenship remains a prerogative of the nation-state. While municipalities have been quite active in granting voting rights to non-citizens (Arrighi de Casanova & Bauböck, 2017), very few regions allow non-citizens to vote in regional elections (Piccoli et al., 2017). However, many regions have developed alternative venues for the political incorporation of immigrants and have created regional ‘integration councils’ – consultative or advisory bodies representing the immigrant population (Manatschal, 2011; Schmidtke & Zaslove, 2014). The inclusion of civil society actors via these bodies can produce less politicized and more pragmatic policy approaches at the regional level (Schmidtke & Zaslove, 2014).

Notwithstanding the limited scope of regional integration policies in the legal–political domain, regional integration policies in general may be crucial in shaping immigrants’ political incorporation. According to the classical policy feedback literature, public policies can influence individuals’ attitudes about governments and toward societal groups. They can also enhance rates of political participation of groups that are targeted by a specific policy, for instance to protest against loss, or expand material gains, or they can undermine political participation leading to social alienation (Condon et al., 2016; Mettler, 2002; Pierson, 1993).

Migration research shows that integration policies have direct effects on immigrants’ civic engagement, and can explain political protest behaviour, or even provide (dis-)incentives to naturalize (Bloemraad, 2006; Cinalli & Giugni, 2011). In addition to these direct, material effects on specific target groups, integration policies also send strong symbolic signals of inclusion, exclusion or even threat, which can spill over into broader communities and peer networks (Condon et al., 2016). ‘Immigrant’ is a complex category, comprising, for instance, asylum seekers, undocumented individuals or economic and family migrants. It also intersects with questions of nationality, ethnicity and race. Closer attention to policy spillover effects, especially to family and descendants of immigrants (i.e., second generation) and co-ethnics, is thus essential. As US research shows, the threat from exclusionary policies towards undocumented migrants led to a strong mobilization of Hispanic immigrant voters and their US-born children (Pantoja & Segura, 2003; Zepeda-Millán, 2017).

Completing the policy cycle, immigrants’ increased political engagement or voter turnout constitute new policy inputs, which may eventually translate again into policy outputs. Information about voters’ migrant background is still scarce. Yet, the few existing studies clearly support the immigrant voter–policy output link; for instance, the share of immigrant voters has been shown to influence social policies in Swedish municipalities (Vernby, 2013). There is also international comparative evidence showing that the share of voters with an immigrant background influences national integration policy outputs (Koopmans et al., 2012).

Based on these theoretical reflections, our argument regarding the legal–political domain of regional integration policy is twofold. Overall, regional policies in this area will be scarce, and regional policy-making activity limited. At the same time, regional integration policies will have important feedback effects on individuals’ political engagement and incorporation. First, regional policies can increase or decrease the likelihood that immigrants will seek naturalization, and they can affect immigrants’ government support and their probability to vote. Second, regional policies affecting immigrants and non-citizens will also have spillover effects for their children and co-ethnics, thereby creating active regional citizens in the medium to long run.

REGIONAL PRAGMATISM AND THE DYNAMICS OF IN- AND EXCLUSION: FINDINGS FROM THE CONTRIBUTIONS

In line with the conceptualization of immigrant integration as a transversal and multidimensional policy field, the contributions to this special issue show how regions translate their formal authority into concrete policy outputs in several areas, such as education and culture, language policies and policies regulating religious practices, labour market inclusion, or access to social benefits and healthcare. To measure the quantity and orientation of policy output in these areas, authors provide an impressive array of original data based on the systematic coding of regional integration laws and other policy documents for all contexts studied.

In terms of research design, all contributions benefit from controlled comparisons across subnational units that are located within the same national context (Snyder, 2001). Subnational comparisons across ordinary regions and Rokkan regions with strong regionalist parties, and across regions governed by parties on the left and regions governed by parties on the right allow for sound inferences about whether party political factors make a difference for the orientation of immigrant integration policies. In addition, some of the contributions combine the strengths of subnational and cross-national comparisons. Such comparative multilevel analyses (Thomann & Manatschal, 2016) can indicate whether insights about regional policies hold only for specific national settings or can be generalized. To draw valid conclusions from such comparisons, contributors either include country fixed effects (as in Zuber’s, 2019, in this issue, quantitative study of German, Italian and Spanish regions) or choose most similar national systems (as in Piccoli’s, 2019b, in this issue, doubly paired comparison of two Italian and two Spanish regions). The contributions use regression analysis to detect systematic patterns in regional integration policies (Zuber) and to estimate the effect of integration policies on immigrants’ attitudes and behaviour (Bennour, 2020, in this issue; Filindra & Manatschal, 2019, in this issue). They turn to qualitative studies to trace the development of political discourse on immigrant integration over time (Xhardez, 2019, in this issue; Paquet & Xhardez, 2020, in this issue) and to reveal the causal mechanisms behind path dependent policy-making (Piccoli).

In more detail, the contribution by Christina Zuber studies how regions respond to immigration, and what drives their policy choices. Her analysis of integration laws of German, Italian and Spanish regions confirms the prevalence of socioeconomically inclusive measures, regardless of national context. Where exclusive and assimilationist provisions occur at all, they are associated with minority nationalist mobilization and the strength of anti-immigrant parties, while leftist regions facing right-wing national governments tend to adopt more inclusive policies.

Focusing on undocumented immigrants, Lorenzo Piccoli scrutinizes why access to healthcare beyond urgent treatment differs across regions within the same state. Comparing Italian regions and Spanish autonomous communities, Piccoli contends that traditions of regional citizenship and solidarity shape regions’ readiness to provide healthcare to undocumented immigrants. The relevance of path-dependent traditions of regional solidarity notwithstanding, the paper shows that regional party politics still matters, as it shapes how regional traditions are activated in policy formulation and implementation.

Analysing political elite discourses on regional integration policies in Belgium, Catherine Xhardez shows that policies can also converge across regions. Historically, francophone and Flemish political elites have taken opposing views on immigrant integration, confirming the distinctiveness of Rokkan regions and the influence of minority nationalist parties. However, regional policies have recently converged towards Flanders’ model, which involves a strong focus on compulsory civic integration. Xhardez attributes this convergence to two factors: politicization of immigrant integration by the francophone right and policy learning.

Focusing on Canadian provinces, Mireille Paquet and Catherine Xhardez confirm the focus on socioeconomic integration policy at the regional level, and do so even for the Rokkan region of Quebec. However, instead of party politics, economic competition between regions emerges as a driving force in creating inclusive regional labour market integration policies. This particular outcome must be understood within the Canadian constitutional context, which allows provinces not only to decide on immigrant integration policies but also to set their own criteria for immigrant selection (immigration policy). Under these conditions, economic federalism is resulting not in a race to the bottom but in interregional competition for the most talented, highly skilled immigrants.

In light of regions’ limited autonomy in the legal–political realm, the two final contributions scrutinize the feedback effects of regional integration policies on immigrants’ political behaviour and attitudes. Salomon Bennour’s contribution on immigrants’ naturalization intentions across Swiss cantons shows that the regional integration policy context affects individual intentions to naturalize. While inclusive cantonal integration policies directly increase naturalization intentions among European Union citizens, this positive effect only unfolds over time in the case of nationals from less developed countries. These findings suggest that, depending on immigrants’ cultural, linguistic or religious closeness to the Swiss context, policy effects may take more or less time to unfold their inclusive potential.

Alexandra Filindra and Anita Manatschal finally assess how increased policy-making activity across US states affects political attitudes and behaviour among immigrants and their larger social networks. Similar to the contributions on subnational regions in Europe, the contribution documents US states’ high activity in the socioeconomic realm. The study further reveals heterogeneous policy effects on voter turnout and governor approval among different ethnic and nativity groups. These results show that regional integration policies affect not only immigrants themselves but also spill over to their offspring and co-ethnics. In his concluding essay, Michael Tatham places the papers in the special issue in the broader perspective of debates about immigration, territory, community and solidarity in contemporary democracies. He concludes that subnational authorities ‘have transitioned from “spaces” to “actors”' who take an active part in the politics of immigration and who propose their own more in- or more exclusive definitions of community that impact on how solidarity plays out within their territory. Immigration thus serves as a core example of how territorial politics contributes to the transformation of democratic systems in their policy, politics and polity dimensions.

Taken together, the contributions lead to three main conclusions, reflecting the three domains of integration policy. First, across countries, both Rokkan and ordinary regions are most active in shaping the socioeconomic integration of immigrants. In this domain, regions governed by left-wing parties are particularly active providers of social citizenship, often exceeding minimum standards of social welfare set by national legislation. The expectations of economic theories of federalism were only confirmed in ordinary Canadian provinces; however, this did not manifest as a race to the bottom, but as interregional competition to attract and keep the most talented, highly skilled immigrants.

Second, comparing Rokkan regions with ordinary regions shows that the former place more emphasis on cultural–religious integration, and are more likely than ordinary regions to have assimilationist policies. Heightened concern with issues of regional identity, culture and language makes Rokkan regions more similar to the national level, whereas ordinary regions appear to be more similar to the local level of government in their focus on socioeconomic integration.

Finally, regions show the least activity in designing policies targeting the legal–political integration of immigrants. This is in line with the distribution of migration-related competencies in multilevel states. However, regional integration policies affect immigrants’ political integration indirectly through policy feedback, making immigrants and their offspring more or less prone to seek naturalization and to participate politically.

CONCLUSIONS AND OUTLOOK

The contributions to this special issue revealed that in Europe and North America alike, regional governments are shaping immigrants’ integration processes, making active use of their decision-making competencies. They further show that what immigrants receive (e.g., in terms of specific integration programmes and access to public services) and what they owe (e.g., in terms of civic duties and requirements for cultural adaptation) varies not only across countries but also, in many instances, across regions of one and the same country. This variance matters, as it shapes immigrants’ prospects of becoming ‘citizens of the region’ (Hepburn, 2011), altering immigrants’ attitudinal ties with the destination societies (naturalization intention, governor approval), and how they engage politically (voter turnout).

The findings further support the expectation that regional integration policies can be explained by regional politics, rather than by demographic or economic incentives. It is only when regions can also select immigrants, as in the Canadian context, that integration policies are depoliticized and universally geared towards the goal of fostering regional economic development. In all other cases, regional political actors can set their migration-related political priorities only through integration policies, and they do so in ways that underline the importance of party politics – whether in the form of strategic interaction between centre-left, centre-right and radical right parties, or between regionalist and state-wide parties in the case of Rokkan regions. Importantly, the regions studied here show a readiness to include immigrants in the provision of social services at the regional level, corroborating a logic of ‘welfare regionalism’, rather than the logic of a ‘race to the bottom’. Where we found differences in the orientations of policies, these were related to the ideology of regional governments and the strength of anti-immigrant parties. Regions governed by parties on the right, or regions where governments were challenged by the radical right, have been more prone to opt for restricting immigrants’ access to social benefits and services compared with the native population (socioeconomic domain). Furthermore, Rokkan regions with strong regionalist parties oriented their cultural–religious policies more towards the assimilationist model to protect their territorially based cultural distinctiveness. The contributions documenting policy feedback effects on the political attitudes (governor approval, naturalization intention) and behaviour (voter turnout) of immigrants and their broader communities further confirmed our argument that regions steer political integration processes rather indirectly.

The comparative design of this special issue combined systematic within- and cross-country comparisons of regional policies. This allowed us also to assess the importance of potential drivers at the contextual level. The cross-country comparison shows that fully developed federalism is not a necessary condition for active regional policy-making in the area of immigrant integration. Contributions to this special issue find early policy-making and diverse regional policies in the quasi-federal systems of Italy and Spain, contrasted by late initiative (Germany) or convergent policies (Canada and Belgium) in federations – though the United States and Switzerland behaved more in line with the expectation of vivid regional policy activity in federal states.

Beyond the corroboration and refinement of our theoretical expectations on regional integration policy, the contributions to this special issue show that subnational regions can assume diverse roles in the multilevel dynamic of integration policy-making. First, regions often assume the role of ‘policy laboratories’, inspiring policy-making at the central government level, as well as in other regions (Schmidtke & Zaslove, 2014). Region-to-centre learning occurred in Spain, for instance, where the autonomous community of Barcelona introduced a ‘Plan on Immigrant Integration’ in 1993; the Spanish government followed with a similar plan in 1994 (Zuber, 2019, in this issue). Region-to-region policy learning occurred in Belgium, where Wallonia followed the Flemish example and adopted more assimilative cultural integration policies (Xhardez, 2019, in this issue). Similarly, we observe regional policy convergence around integration programmes focused on socioeconomic integration in ordinary Canadian regions (Paquet & Xhardez, 2020, in this issue).

Second, regions may step in where no central policy exists, assuming the function of ‘gap fillers’. For example, the lack of an Italian national strategy for providing healthcare to undocumented migrants motivated regions such as Tuscany to provide specific services to this vulnerable group within the framework of their general welfare services (Piccoli, 2019b, in this issue). The Belgian case shows that although Wallonia and Flanders learned from each other, there is still no Belgian national model of immigrant integration (Xhardez, 2019, in this issue).

Third, regions can emerge as alternative loci of emotional belonging and identification (Bennour & Manatschal, 2019). They redefine thereby the contours of ‘affective citizenship’, which goes beyond cognitivist conceptions of citizens by considering the role of affect and emotions in political life (Fortier, 2016). By determining the degree of ease or difficulty of immigrants’ access to certain regional rights, entitlements and benefits, these policies send both material and symbolic signs of inclusion or exclusion, which in turn shape immigrants’ attitudes and behaviours, turning them into (re-)active regional citizens (Bennour, 2020, in this issue; Filindra & Manatschal, 2019, in this issue).

Summing up, the contributions to this special issue reveal the multifaceted ways in which increased integration policy-making at the level of subnational regions contributes to turning immigrants into regional citizens. These channels range from providing access to labour markets, social benefits or political rights and even enfranchisement, to recognizing cultural diversity or demanding the adaptation to a particular regional culture and language. As the contributions show, these regional integration policies matter, as they strengthen non-citizens’ identification with, and political engagement in, the respective regions.

To conclude this introduction, we highlight some avenues for future research. A first venue concerns the policy fields covered. The contributions to this special issue focused on access to social benefits (Zuber; and Filindra and Manatschal), healthcare (Piccoli) and regional labour markets (Paquet and Xhardez), as well as language policies (Zuber; and Filindra and Manatschal). Housing policies and non-citizen enfranchisement at the regional level warrant further study, in particular. In addition, future work could explore regional policies that target the host, rather than the immigrant population, most prominently in the form of anti-discrimination policies (but see Bennour, 2020, in this issue), as well as policies targeting subgroups of migrants more specifically, such as asylum seekers.

Second, this special issue focuses deliberately on one level of government: subnational regions, to explore systematically the extent, orientation, and political drivers and effects of regional integration policies. To be sure, subnational integration policy-making does not unfold in isolation, but forms part of a more complex and multilayered system of migration governance. National, supranational and even global regulations (Margheritis, 2013; Panizzon & van Riemsdijk, 2019), but also policies at the local level of government (Caponio & Borkert, 2010), clearly influence regional integration policy-making initiatives and priorities. Future research on regional integration policies should therefore also consider the impact of the larger local, national and international contexts in which they are embedded.

Related to this point, although international in its coverage, this special issue restricts its scope to post-industrialized countries in Europe and North America. Intra- and extra-continental immigration flows increasingly also shape other subnational regions around the globe. It would be interesting to see whether the theoretical arguments developed here can also help to explain subnational integration policy outputs and outcomes in South America or Asia. Eventually, future research may show whether our findings are generalizable at a large scale, or if they have the character of a middle-range theory (Castles, 2010) that holds only for post-industrialized regions.

ACKNOWLEDGEMENTS

Previous versions of this paper were presented at an authors’ workshop at the University of Konstanz, Germany, 14–15 September 2017, at the MIGPROSP – Prospects for International Migration Governance, ‘The Dynamics of Regional Migration Governance, European’ Conference, University Institute, Florence, Italy, 25–26 May 2017, and at the Dreiländertagung, ETH Zürich, Switzerland, 14–15 February 2019. We thank the participants on these occasions and, in particular, Nathalie Behnke, Nona Bledow, Marius Busemeyer, Susanne Garritzmann, Lorenzo Piccoli, Philip Rathgeb and Gerald Schneider, as well as three anonymous referees and the editors for very helpful comments. All authors contributed equally to this paper.

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the authors.

Additional information

Funding

This research was supported by the NCCR – On the Move, funded by the Schweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung (Swiss National Science Foundation) [grant number 51NF40-182897]. The authors thank the Department of Innovation, Research at the University of the Autonomous Province of Bozen/Bolzano for covering the Open Access publication costs.

Notes

1. Manatschal and Stadelmann-Steffen (2013, p. 681) compare interregional and international variance of the criterion ‘residence period required for naturalization’. They document that at the time, the length of residence periods varied more between 26 Swiss cantons (10 years, N = 26 cantons) than between 33 different countries (9 years, N = 33 countries).

2. Drawing on the seminal work of Stein Rokkan, Hooghe et al. (2016, p. 73) define Rokkan regions through distance (of the peripheral region from the core), difference (given if a majority in the region speaks a language other than the majority language of the country as a whole), and lack of dependence on the centre (given if a region had a prior history of independent statehood).

3. Both Italy and Spain have undergone processes of regionalization in recent decades. Central governments have granted a substantial degree of self-rule to regional parliaments and governments. Nevertheless, both countries are commonly labelled quasi-federal rather than federal due to the lack of shared rule (i.e., the involvement of regions in central decision-making, e.g. through a second chamber of regional representatives) (Loughlin et al., 2013). For more details on migration-related competencies and multilevel governance, see Franco-Guillén (2019) for Spain and Caponio et al. (2019) for Italy.

4. Switzerland with its three-tiered naturalization system being an exception here (D’Amato, 2009).

5. Depending on the generosity of the regional system of welfare, immigrants who are treated unequally relative to natives in region 1 might still end up receiving more in absolute terms than immigrants who are treated equally in region 2. However, we agree with Penninx and Garcés-Mascareñas (2016) that we should judge whether socioeconomic policies have an inclusive or exclusive orientation by looking at relative differences between immigrants and natives, not by measuring the absolute size of benefits immigrants receive.

6. Regionalist parties are those that contest elections and field candidates in only one or several regions, but not across the whole territory of the state, and whose core programmatic objective is to defend the identities and interests of ‘their’ region (Massetti, 2009).