Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Introduction

Migration and migrant integration policies have become increasingly dispersed over various levels of government. Besides the national level, the European Union (EU) level and the regional and local levels have become more involved. In the area of immigration, EU member states have handed over significant power to the EU, particularly in the context of the Common European Asylum System. The EU’s Family Reunification Directive, for instance, significantly limits member states’ policy discretion in family migration policies. With regard to migrant integration there has been some Europeanization as well, but this has been overshadowed by a sharp “local turn” in policymaking. Local governments, large cities in particular, are becoming increasingly entrepreneurial in developing their own integration philosophies and policies. This has led to cities having markedly different approaches to migrant integration, even within the same countries.

This chapter focuses on migration and integration as multilevel policy issues and explores the consequences in terms of multilevel governance. The fact that both migration and integration have become multilevel issues presents both opportunities and challenges. Immigration policymaking has been characterized by a constant struggle between national governments and the EU about the amount of discretion states have in interpreting EU directives. The involvement of local and regional governments in debates about intra-EU migration, particularly East–west migration from new member states, has further complicated the situation. With regard to migrant integration, even more complex relations have emerged between local, regional, national, and EU institutions. The superdiverse cities of Europe, such as Barcelona, London, Berlin, and Rotterdam, have taken policy directions very different from their national governments, resulting in a “decoupling” of national and local policies. While politicization of migrant integration continues to drive policies in many countries, the EU has developed various soft governance measures aimed at promoting policy learning between local governments.

This chapter offers an analysis of how migration and integration policies have evolved at various levels during the past decades, including at the EU, national, and local levels, as well as in some cases at the regional level. This enables us to understand the factors that drive policies at the different levels and the extent to which these lead to convergence or divergence between the levels. We analyse the relations—or absence of relations—between the levels of government. To make sense of these, we apply a framework that allows for different arrangements of the relations between levels of government. This is where the notion of “multilevel governance” comes in as one possible way of structuring relations between various government levels.

A Framework for the Study of Multilevelness

Regulation of international migration has traditionally been a competency of the nation state, with the voluntary transfer of competencies to the EU being only a recent exception to this rule. Migrant integration, similarly, has largely been a purview of the nation state, as ideas about how to integrate migrants are often strongly correlated with ideas about national identity or the “national imagined community”.

Various scholars have argued that such nation-based views (Favell 2005) have also affected migration research. Wimmer and Glick Schiller (2002: 301) describe this as “methodological nationalism”. Bommes and Thränhardt (2010) show that migration research has evolved in distinct national paradigms or national models of integration (Thränhardt & Bommes 2010). These models are national ‘not just because of their context dependency and insufficient clarifications on the conditions of generalizability, they are national because the modes of presenting and questions are politically constituted by the nation-states for which migration becomes a problem or a challenge’ (ibid.: 10). Favell (2005: 47) argues that national models have been sustained in policy and politics as ‘self-justificatory discourses’, and that this is to some extent also true of migration research because of the strong policy orientation during the development of this research field. Indeed, in some countries, nation state-centeredness has been reinforced by strong institutional relations between researchers and policymakers in the fields of migration and integration (Scholten 2011; Scholten and Verbeek 2014; Scholten et al. 2015). A national orientation, however, could hamper the comparative and theoretical development of migration research. National paradigms of migration and integration reduce complexity and also introduce a historical-institutionalist bias in explaining and inadvertently reifying national differences. Bommes (2010) in particular argues that this has restricted the urge of migration scholars to look for more generalizable theoretical accounts of differences as well as similarities between countries. Furthermore, national frameworks obscure views of developments at other levels. Only in the 2000s did, due to concerted efforts of EU institutions, attention to the European level increase. For instance, the European Integration Fund and European framework programmes have promoted cross-national comparative research in the European setting. Attention to the local level is of a more recent date, and many times seems to be supported by EU research funding in particular.

Thus, while our understanding of how policies develop at various levels has increased, there is still a layering of knowledge per level. Studies and literature, too, tend to focus on just one level, rather than seeking an understanding of the interactions between levels. A next step to widen the scope of studies of migration and integration policies at different levels would be to explore their consequences in terms of the relations between the different levels. What sorts of interactions or relations (or absence thereof) can be identified between various levels of government, and what are the consequences? The literature on governance in multilevel settings defines various ways of configuring relations between government levels. Scholten (2013) brings these different ways together in a typology that distinguishes between four ideal type configurations of relations between government levels: centralist (top-down), localist (bottom-up), multilevel, and decoupled.

First of all, the centralist ideal type exhibits a clear hierarchy and division of labour between government levels. In a multilevel setting, this involves a top-down relationship between the different levels of government, such as a clear central codification of the division of labour between levels and control mechanisms to ensure that policy implementation at the local level follows central rules and reflects the central policy frame. This implies a strong institutional structure for policy coordination, for instance, at the European or the national level. The centralist type is expected to produce policy convergence between the different levels of government. As such, this type of governance setting corresponds with the idea of national paradigms of migration or integration.

The second ideal type involves a more localist and bottom-up perspective on governance in multilevel settings. In this type, policy competencies follow the principle of subsidiarity; that is, what can be done locally should be done locally. Local governments do more than just implement policy; they formulate policies, respond to local policy agendas, and exchange knowledge and information horizontally with other local governments. The localist type may lead to greater policy divergence between the national and the local level. It speaks to what some scholars describe as “the local dimension of migrant integration policies” (Alexander 2007; Caponio and Borkert 2010; Penninx et al. 2004), which stresses that local governments are often confronted with integration problems in different ways than the national or European level. This leads them to frame migrant integration policies in a specific local way.

As distinct from these centralist and localist types, multilevel governance refers to interaction and joint coordination of relations between the various levels of government without clear dominance of one level. This means that “vertical venues” are needed where governments from different levels jointly engage in meaningful policy coordination. These might involve forums or networks in which organizations from different government levels meet. Multilevel governance is thought to be most effective when the idea of there being different government levels shifts to the background, or in other words, when in terms of power a degree of “levelling” takes place between the different government levels. In terms of policy frames, the multilevel governance type is likely to engender some convergence between policy frames at different levels, produced and sustained by their mutual interaction.

The fourth type is decoupled relations between government levels. Such a situation is characterized by the absence of any meaningful policy coordination between levels. Thus, in any single policy domain, policies at different levels are dissociated and may even be contradictory. This type can lead to policy conflicts between government levels. It can also send conflicting policy messages to the policy target groups, thereby diminishing policy effectiveness. It is associated with divergence between different levels of policy, reflected in studies finding that national and local integration policies have increasingly become “two worlds apart” (Jørgensen 2012; Poppelaars and Scholten 2008).

Immigration Policies

Classic immigration countries, like the USA and Canada, have defined themselves as nations of immigrants. In contrast, North-Western European countries have not seen themselves as immigration countries, although they received large numbers of newcomers from abroad between 1950 and 1974: refugees from the East, immigrants from onetime colonies, and guest workers. As we read in Chap. 3, after 1974, when the first oil crisis precipitated the restructuring of economies and labour markets and new hands were no longer needed, these countries responded by adopting restrictive immigration policies. These new policies were framed in the 1980s (regarding labour migrants and family migrants) and 1990s (on asylum migrants). Only very recently have countries like the UK and Germany adopted new active immigration policies—for the first time since the 1960s and early 1970s—to recruit labour for certain sectors suffering shortages of workers. The immigration policies developed in the 1980s and 1990s were mostly framed in a nation state-centred way. For instance, Nordic and North-Western European countries often framed immigration policies in relation to the welfare state. In the UK, they were framed in particular by the history of the British Empire. Germany’s immigration policies cannot be understood without reference to its long history as a “divided nation” and its consequential reluctance to become a country of immigration. In some countries, arguments of overpopulation (the Netherlands) or population decline due to ageing have played an important role.

Europeanization

A cornerstone of migration policy in the European setting is the principle of free movement for EU citizens. This principle, which applies to intra-EU migration only, has been at the heart of European integration since its inception. The European Coal and Steal Community (ECSC) established, already in 1951, a provision of free mobility for workers in this industry. Since then, the free movement principle has been extended and firmly anchored in EU treaties. It is a key supranational element of the Europeanization of immigration policies, and has had a clear binding effect on member states. Intra-EU mobility increased significantly after the accession of Central and Eastern European countries in 2004 and 2007. Europeanization of policies on immigration from outside the EU has occurred much more incrementally. First, immigration and border security were discussed intergovernmentally in the so-called Trevi Group in the late 1970s. A major step followed in the 1980s when a group of member states moved to abolish border controls and adopt joint immigration policy measures with the Schengen Agreement (1985). The Schengen group numbered 26 countries as of 2014. In 1999, the Schengen Agreement was incorporated into the Treaty of Amsterdam, though exceptions and opt-outs have continued to apply to several countries. Thus, a form of cooperation between nation states eventually arose and contributed to the anchoring of common regulations in the supranational treaties of the EU.

Asylum migration in the 1990s became an important impetus for the Europeanization of asylum and immigration policies. With the Dublin Convention of 1990, EU member states formalized arrangements to address the problem of “asylum shopping”. The Maastricht Treaty of 1993 established a broader framework for intergovernmental cooperation in the field of asylum and migration, under the so-called Third Pillar “non-binding” cooperation. Perhaps the most important step towards a common EU policy was taken in 1997, when asylum and migration were moved to the First Pillar, which involved a much stronger role for the European Commission and a legal basis for EU activity. This was further reinforced by the 2009 Lisbon Treaty, which “normalized” immigration policy as a core EU issue, introducing qualified majority voting in this domain and strengthening the role of the European Parliament and the European Court of Justice (ECJ).

It is fair to say that by the 2010s a strongly Europeanized policy field had emerged on asylum and migration. This encompassed numerous elements: joint border controls (Frontex), the Returns Directive regulating the return of illegal migrants, standardization of asylum procedures, the EU Blue Card Directive on selective labour migration, the Family Reunification Directive, which had strong impact on national family migration policies, and a series of cooperation agreements with migrant-sending countries to address the root causes of migration. Particularly important in terms of multilevel governance has been the step by step strengthening of the role of EU institutions like the ECJ and Parliament, marking a real transfer of competencies to supranational institutions. Nevertheless, via the EU Council, various intergovernmental working groups, and to some extent also the Commission, the member states continue to play a key role.

Patterns of Multilevel Relations in Immigration Policies

Geddes and Scholten (2014a) distinguish three patterns of Europeanization of immigration policies, closely corresponding to the different types of government relations discussed earlier. One of them is that Europeanization clearly involves loss of control for nation states, given the supremacy and direct effect of EU directives. In our typology, this is closest aligned to the centralist ideal type, with EU institutions exerting top-down control over immigration policies throughout Europe. Starting as a spin-off from major steps in European integration, like creation of the internal market and freedom of movement within the EU, immigration was Europeanized primarily for functional reasons; if there is freedom of movement, then there should also be a common immigration policy. This is in line with the broad literature on the gradual erosion of national control over borders and migration caused by globalization and economic and political interdependencies between nation states (see Sassen 1999). Here, it might also be mentioned that European institutions, in particular the ECJ and more recently the European Commission, have played important “activist” roles in the Europeanization of immigration policies.

A second pattern, described by Geddes and Scholten (2014a) as the “escape to Europe thesis”, counters the argument that states have lost control due to the Europeanization of immigration. Reflecting a literature on how European cooperation might rather strengthen the nation state (see Moravcsik 2013), countries may seek cooperation with their European neighbours to jointly fortify their grip on international migration. Thus, working together might increase their control rather than weaken it. Furthermore, seeking cooperation at the EU level might allow governments to find ways around the political and legal constraints they face within their own countries. The escape to Europe thesis provides a good account of the intergovernmentalist evolution of the EU’s immigration policies. Many EU migration and asylum measures were first introduced as forms of cooperation involving subsets of EU member states and discussed in intergovernmental working groups (such as the Trevi Group) rather than at the level of EU institutions. In our typology, this comes closest to what was termed the localist model, with the nation state being the “local” actor seeking cooperation in an EU setting for the benefit of the nation state while not ceding any substantial degree of control.

Finally, Geddes and Scholten (2014a) identify a third pattern of the evolution of EU immigration policies that stresses a transgovernmentalist form of Europeanization. This means that governments seek cooperation in a European setting, even ceding some power and control to EU institutions, in order to gain a firmer grip on immigration, to the benefit of the nations as well. In fact, this form of transgovernmentalism comes close to our ideal type of multilevel governance, with the national and European levels systematically connected rather than one or the other being in control. Such a transgovernmentalist account gives a very good explanation for the strong involvement of EU member states (rather than EU institutions) in development of several key EU directives in this area, such as the Family Reunification Directive. It also accounts for the delicate balancing of national and EU interests; for instance, the Dutch government, together with several other governments, recently tried to renegotiate the Family Reunification Directive in order to realign national and EU interests in this policy area.

Apart from the three patterns of Geddes and Scholten, we also observe our fourth type, “decoupling” in multilevel settings and absence of coordination. The struggles between nation states and EU institutions, and sometimes even between subnational governments and national and EU institutions, signal that policy interests are not always aligned. Conflicts do take place. An issue that has become particularly prominent in recent decades is that of intra-EU mobility, especially East–west migration within the EU after the accession of Central and Eastern European (CEE) countries such as Poland, Hungary, Bulgaria, and Romania. Migration from CEE countries is now by far the largest migration flow to some North-Western European countries, such as the UK and the Netherlands. Although transition arrangements were made which postponed free movement for a number of years, the borders have now opened to all new member states. Many CEE migrants appear to be permanently settling in other EU member states, raising concerns about how to incorporate these EU citizens into their new home countries. However, policy measures that would impose an obligation in terms of integration efforts (such as a language requirement) are considered at odds with the principle of freedom of movement of EU citizens within the EU. In France, this conflict was brought into sharp focus when the French government decided to deport large numbers of Roma migrants to Romania and demolish their camps, thereby engaging in direct confrontation not just with Romania but also with the European Commissioner on Immigration.

This brief review of types of relations demonstrates that rather than a single pattern, there are various patterns of interaction and relations taking place simultaneously between national and EU institutions. It is undeniable that some competencies have been transferred, but many of these transfers came about at the initiative or with the consent of national governments and in fact strengthened member states’ control over immigration flows (of third-country nationals). There is no clear dominance of the centralist or localist pattern. Rather, there appears to be a delicate balancing between nations and EU institutions, as evident in the recent efforts to renegotiate the Family Reunification Directive and the conflict around Roma deportations. Although this is to some extent a matter of interpretation, we propose that the evolution of patterns of interaction fits our description of multilevel governance. There is certainly a high degree of interaction between nation states and the EU in the formulation of immigration policies.

Besides national–EU relations, there are some indications of involvement of subnational governments in these already complex relations. Subnational governments rarely have immigration policy competencies, but they do have policy interests in this area. For instance, economic and demographic characteristics of regions may increase or decrease their demand for immigration. Scotland, for example, advocates a much more open and active immigration policy than the UK government. Cities, too, have been important actors, especially in relation to policy implementation, as they may be particularly affected by the consequences of immigration policies. For instance, the human consequences of deportation and irregular migration are often most evident at the local level. Hence, many local governments have offered forms of assistance to irregular migrants even though this may be distinctly at odds with national policies. Local governments have furthermore been important advocates of “pardons” or regularizations of undocumented migrants. Some cities have even developed their own “urban citizenship”, counterbalancing exclusionist effects of national definitions of citizenship (see Chauvin and Garcés-Mascareñas 2012).

Integration Policies

The multilevel dynamics of migrant integration policies have been very different from those of immigration policies. Rather than the turn towards Europe described above, a “local turn” seems primarily at play. This involves a shift away from historically rooted models of integration strongly related to nationally specific models of identity and belonging (see also Ireland 1994). Such models would imply, in our typology, strongly state-centric (centralist) modes of governance. Brubaker (1992), for instance, shows that French and German policies have their respective roots in deep historic notions of the French “Staatsvolk” (ius soli) and German “Volksstaat” (ius sanguinis). This idea of national models of integration has been strong not just in policy but also in academic discourse (for a critical discussion see Bertossi 2011; Bertossi et al. 2015; Joppke 2007). Yet, as argued earlier, this has led to an overemphasis on differences between national integration models, such as the British race-relations model, the German differentialist model, the French Republicanist model, and the Dutch multiculturalist model.

The politicization of migrant integration that took place in many European countries in the 1990s and 2000s revealed the resiliency of such national models. In this period, there was a revival of ideas of cultural integration, especially in national political and policy discourses. Throughout Europe this led to policy initiatives that strengthened the importance of national history, culture, values, and norms in relation to immigrant integration. For example, during this period the Netherlands, France, Germany, and the UK introduced civic integration programmes including tests of basic knowledge about society. Joppke and Morawska (2003) speak in this respect of an assimilationist turn in migrant integration policies.

The Local Turn in Migrant Integration Policies

Local governments, especially those in Europe’s larger cities, have become increasingly active in developing their own integration philosophies. From a sociological perspective, this development makes sense as it is at the local level that migrants meet others, find a job, have children, et cetera. It is also at this level that negative as well as positive aspects of diversity are experienced most concretely. Also, we know from research that migrants identify much more with the city they live in than with the nation. Hyperdiverse cities like Berlin, Amsterdam, and London embrace diversity as part of the city’s identity and as a positive anchoring point for local policies, sometimes in spite of their respective national models. Industrial cities like Manchester and Rotterdam have linked their traditional emphasis on work and housing to the new challenge of diversity. This supports sociologist Benjamin Barber’s suggestion that it is precisely the inability of national democracies to develop effective responses to migration and diversity that prompts cities to develop their own strategies with a much greater emphasis on pragmatism, trust, and participation.

Various scholars, including Alexander (2007) and Penninx et al. (2004), illustrate how cities in particular started developing their own integration philosophies, often in response to the specific local situation. For instance, various successive mayors of the Greater London Authority were particularly proactive on migrant integration. Similarly, the City of Berlin had an integration strategy in place long before Germany developed a national strategy. Penninx (2009) demonstrates that in many countries policies evolved in large and diverse cities before national integration policies were developed, as attested to by Birmingham and Bradford in the UK, Berlin and Frankfurt in Germany, Vienna in Austria, and the Swiss cities of Zurich, Bern, and Basel. In our typology, this fits best with the localist or decoupled models, depending on whether these local philosophies are in line with national policy contours (as in Germany) or contrast and possibly even conflict with national policies (as in various cases in the Netherlands). As we will read below, only in some cases has it led to what we describe as multilevel governance.

The local turn in migrant integration policies has several implications in terms of vertical relations between national and local governments. Under the centralist model, local governments would play a role but this would be confined primarily to policy implementation. Indeed, in many countries we find top-down structures for policy coordination. In France policy coordination is strongly state-centric, and countries including Sweden, Denmark, and the Netherlands have long had strong national policy coordination frameworks. Often, the way funds are distributed and allocated is indicative of the division of labour between the national and the local level. Even in the UK, a country with relatively active local actors, significant funds are allocated from the national level (including funding for courses in English for speakers of other languages). However, many studies suggest that the top-down or centralist model has become much less applicable to the practice of migrant integration policymaking in many European countries (see also Entzinger and Scholten 2014). Local integration policies tend to differ from national policies in various respects. Caponio and Borkert (2010: 9) even speak of a distinctly “local dimension of migrant integration polices”. Some scholars argue that local policies are more likely than national policies to be accommodative of ethnic diversity and work together with migrant organizations, due in part to the practical need to manage ethnic differences in a city (Borkert and Bosswick 2007; Vermeulen and Stotijn 2010). Thus, in contrast to the often symbolic tendencies of national policies, local policies are driven by pragmatic problem-solving (Poppelaars and Scholten 2008). For instance, cities might work more closely with migration representatives and organizations than a national government would (see also Bousetta 2000). Cities may also be more inclined to accommodate and support cultural and religious activities of minorities in response to migrants’ needs and demands.

Others contend that, rather than being characteristically more accommodative, local policies are driven by specifically local factors in very different directions. Significant variation in local policies can therefore be expected. Mahnig (2004) concludes that local integration policies in Paris, Berlin, and Zurich have very much responded to local political circumstances, often in ad hoc ways and leading to accommodation in some instances and exclusion in others. According to Alexander (2003, 2007), differences in local social situations have triggered different policy responses, with some cities adopting a more culturalist and others a more socioeconomic approach. A recent study of integration policies in Amsterdam and Rotterdam found that these two cities within the same country and with similar migrant populations produced very different policy outcomes in terms of migrant integration. Rotterdam stressed work and housing, whereas Amsterdam was much more oriented towards promoting intercultural relations (Scholten 2013). In other studies (e.g., Garbaye 2005; Bousetta 2000), a key factor identified as a trigger of specifically local responses is the political mobilization of migrants at the local level. Garbaye (ibid.), for example, found more significant political mobilization and ethnic elite formation in Birmingham than in Lille. This could not be explained only by differences between the groups involved (mainly South Asians in Britain and North Africans in Lille). Another factor was the difference between the liberal British citizenship regime and openness of the local labour party towards ethnic elite formation compared to the French citizenship regime, which had barred access to many Maghrébins, and the local socialist party, which had remained very restrictive in admitting migrants to local political elites.

The local turn of integration policy has a number of implications for governance. In some cities, it has led to what can be described as a decoupling of national and local policies. Thus, policies at these levels were not mutually coordinated and sometimes sent very different policy messages to the same policy target groups. Poppelaars and Scholten (2008) speak, in this respect, of national and local policies being “two worlds apart” in the Netherlands, because of their divergent logics of policy formulation (politicization at the national level and pragmatic problem-solving at the local level). Similarly, Jørgensen (2012) observes a growing disconnect between national and local integration policies. Collett and Gidley (2013) find that in several countries local governments feel they have to repair some of the centripetal forces unleashed by national political and policy discourses. As such, politicized debates at the national level can have a performative effect at the local level as well.

In other situations, more localist types of relations have emerged. Local governments have become increasingly active in what has been described as “vertical venue shopping” (Guiraudon 1997). This refers to efforts by local governments to lobby for policy measures at the national (and increasingly also European) level. Scholten (2013) cites the example of the City of Rotterdam, which managed to get a special law passed at the national level allowing it to adopt stricter policies aimed at spatial dispersal of migrants in the city. The city has also been active at the European level, lobbying for integration measures for intra-EU labour migrants. Establishment of networks among European cities has become a particularly powerful strategy for vertical venue shopping in the field of migrant integration. We will look at this in more detail later.

In contrast to the examples above, which fit the localist or decoupled types of relations, institutionalized relations between national and local governments have evolved in several countries over the past decade towards our definition of multilevel governance. Germany, in particular, has established multilevel venues for coordination of integration policies, with a key role for national integration conferences. These conferences bring together actors from various government levels as well as nongovernmental actors to align efforts to promote integration. The UK’s tradition of coordinated vertical relations includes its delegation of policy coordination at the national level to the Department of Communities and Local Government. Even France, a country known for its state-centric approach, has developed dedicated structures for organizing relations with local governments. Although often not framed explicitly in terms of coordinating migrant integration policies (still reflecting the French colour-blind Republicanist approach), integration clearly plays a role in France’s so-called Urban Social Cohesion Contracts and Educational Priority Zones. These allow the Parisian government to adopt tailored, localized approaches within the context of national policy. The Netherlands’ government has established a “common integration agenda” for national and local governments, though it appears to have been rendered hollow by a lack of central funding.

European Involvement and Nascent Multilevel Governance

Besides the local turn in migrant integration policies, the past decade has also witnessed a gradually increasing involvement of the European level. Nonetheless, compared to the strong trend towards Europeanization that we found in the field of migration and asylum, the Europeanization of migrant integration has come much later and been more modest and hesitant (Goeman 2013). There is as yet no common European policy aimed at migrant integration. This reflects the persistence of the connection between migrant integration and the nation state. The way that countries integrate “their” migrants appears strongly related to conceptions of national identity, history, culture, and values and norms—especially since the “assimilationist turn” described above. Several steps have been taken towards greater EU involvement in this area. Some of these involve EU directives, primarily as a spin-off of the communitarization of immigration policies. Because of the binding effect of EU directives, one could say that they to some extent signal our top-down centralist model of migrant integration, as significant policymaking power is transferred to the EU level. Particularly important in this respect are two 2003 directives: the Directive on the Status of Non-EU Nationals Who Are Long-Term Residents, which provides a framework for policies toward third-country nationals in the EU, and the earlier-mentioned Directive on the Right to Family Reunification, which provides a framework for admittance of family migrants to the EU. Both directives have been influential as a framework for development of civic integration policies for third-country nationals (many of whom are family migrants), as they stipulate what integration measures may be demanded of migrants.

An additional key area in which Europeanization has been significant is anti-discrimination policy. Two directives issued in 2000—the Racial Equality Directive and the Employment Equality Framework Directive—establish a binding structure within which member states can develop their anti-discrimination policies. These directives are yet another example of vertical venue shopping, as they were formulated in response to lobbying by the UK and Dutch governments in particular.

Besides such “hard” and “binding” measures, which may suggest an EU-centric approach (fitting our centralist type), specific frames and definitions have been developed and various non-binding measures put in place which can be described as softer or more open methods of coordination (see also Geddes and Scholten 2014b). In 2003, the European Commission formulated its first comprehensive view on integration policies in the Communication on Immigration, Integration and Employment (EC COM (2003) 336 final). It defines integration as ‘a two-way process based on mutual rights and corresponding obligations of legally resident third country nationals and the host society which provides for full participation of the immigrant’ (ibid.: 17). Integration is conceived as a ‘balance of rights and obligations’ (ibid.: 18). The holistic approach of policies encompasses all dimensions of integration, from economic, social, and political rights to cultural and religious diversity, citizenship, and participation.

In November 2004, the EU Conference of Specialised Ministers responsible for integration agreed on a set of 11 Common Basic Principles for Immigrant Integration Policy (CBP) as a first step towards a European framework for immigrant integration and a point of reference for implementation and evaluation of current and future integration policies. These principles define integration as a two-way process of accommodation and stress the importance of language, interaction, and participation. They furthermore call for the mainstreaming of migrant integration in other policy areas. Importantly, this step towards a more comprehensive framework was accompanied by continuation of the limited definition of the integration target group following directly from migration policies: integration policies are aimed at third-country nationals only and do not target immigrants who are citizens (or long-term residents) of another EU member state. They are supposedly already integrated, by definition, though this assumption has been criticized by local authorities in regions that have received numerous immigrants from the EU’s newest member states (e.g., Poland, Romania, and Bulgaria).

Although rather broad and not binding, the CBPs provide a foundation for more EU involvement in this policy area (primarily intergovernmentalist and thus, in the EU setting, fitting our “localist” type). Following the CBPs, the European Handbook on Integration was published in 2004. In 2005, the Common Agenda for Integration by the European Commission and The Hague Programme were formulated to promote implementation of the CBPs primarily via soft governance means like persuasion, networking, and exchange of best practices. In 2013 the Common Agenda for Integration was developed further into the European Agenda for the Integration of Third-Country Nationals, which stresses the importance of socioeconomic participation and the relevance of the local level in its promotion.

This evolving EU policy framework reflects the EU’s distinctive internal organizational setting for integration policies. First, there is DG Freedom, Security and Justice (also responsible for migration policies), which targets particularly the early reception and integration of recent newcomers, of refugees and accepted asylum seekers, and also of third-country nationals until they have become long-term residents. It is in this particular part of EU policies that West European countries have increasingly “uploaded” their cultural integration requirements for new third-country immigrants into EU integration policies (Goeman 2013). The second setting from which integration is promoted is DG Employment, Social Affairs and Equal Opportunities. Its programmes aim to promote social inclusion and cohesion. Its sizeable funding is—again—used quite extensively by local and regional authorities (and their policies) and by nongovernmental civil society partners at all levels. Equality and anti-discrimination are key concepts (for this reason the European Monitoring Centre on Racism and Xenophobia (EUMC) and its successor the Fundamental Rights Agency (FRA) were associated with this DG). Target groups include not only immigrants but also ethnic minorities and the disabled. Priority domains are equal access to and long-term integration in employment, education, housing, and health. The new Commission in place since autumn 2014 has complicated the picture even more: DG FSJ has been split into the DG Migration and Home Affairs (Immigration, Asylum, and Borders) and Justice and Consumers (Union Citizenship, Free Movement, Equality legislation, and Anti-discrimination).

In the absence of a clear division of formal policy competencies in the area of migrant integration, the very incremental Europeanization of this area of policy has been based on two main resources: expertise and cities (see also Penninx 2015). Regarding the first, migration scholars from the Netherlands and USA played a key role in formulation of the CBPs (ibid.). Furthermore, the EU has used various funding schemes to mobilize comparative research on policy topics that it considers relevant. From 2003 to 2006, this involved, in particular, the Integration of Third-Country Nationals (INTI) Fund and from 2007 to 2013 the European Integration Fund. As Geddes and Scholten (2014b) observe, the initial objective was mainly to promote the horizontal exchange of relevant information, knowledge, and policy best practices. Gradually, with the formulation of the CBPs and the Common Agenda for Integration, these funding schemes have increasingly mobilized expertise to help substantiate the nascent EU policy framework. A clear example in this respect is the EU-sponsored Migrant Integration Policy Index (MIPEX). Though first established to promote comparison and exchange of best practices, the MIPEX has evolved into a tool for monitoring member states’ compliance with EU integration principles, enabling “naming and shaming” of those that do not comply. In the context of our discussion of multilevel governance, this bears out the potentially strategic role that knowledge and expertise can play in multilevel governance, acting in “soft” but sometimes impactful ways (open method of coordination), especially in the absence of more structural “vertical” relations between levels.

Regarding the second EU resource deployed in this area, European programmes have sought to establish a strong relation between the EU and the city level. It is in these efforts that, according to our typology, we can distinguish the contours of an emerging multilevel governance framework. With various means, including some of the funding schemes mentioned above, the European Commission in particular has actively promoted various city networks on a European scale. These networks primarily involve cross-national horizontal forms of cooperation between cities, but with strong connections to the Commission. One example is the CLIP Network (Cities for Local Integration Policies), which since 2006 has brought together some 30 European cities in conferences to systematically exchange knowledge and experience regarding local integration policies.

Integrating Cities is another network (also established in 2006) organized under the Eurocities’ working group on migration and integration, a large network of some 140 major European cities. The Integrating Cities initiative includes a policy dialogue between Eurocities and the European Commission, a conference series, the Eurocities Charter on Integrating Cities, and other EU-funded projects.

Another example is Intercultural Cities, which is a joint activity of the Council of Europe and the European Commission. It emerged from the 2008 White Paper on Intercultural Dialogue contributed by the Council of Europe to the European Year of Intercultural Dialogue that same year. Intercultural Cities advocates pluralistic city identities that respect diversity. The Intercultural Cities Programme was developed and first applied in 11 European pilot cities and has since evolved. It has developed the Intercultural Cities Index for cities to evaluate and develop their policies, and it organizes international conferences for cities to exchange experiences.

Other more specifically horizontal cooperation initiatives have been undertaken as well, such as the European Coalition of Cities against Racism (ECCAR), established in 2004 at the initiative of the United Nations Educational, Scientific and Cultural Organization (UNESCO). The aim of this coalition of cities is to share experiences in order to improve policies against racism, discrimination, and xenophobia. Some 104 municipalities from 22 European countries have joined the network and adopted its 10-point plan of action.

Besides making a direct connection between the nascent European policy framework on migrant integration and the local level of government, thereby constructing the most distinct multilevel governance structures in this area today, the focus on the local level also feeds into the local turn in migrant integration policy described above. Horizontal exchanges of knowledge and best practices between cities, promoted by the EU, has increased cities’ entrepreneurship in developing their own integration philosophies. In a number of cases such integration philosophies encompass relations with cities from which migrants originated, as Chap. 10 will show. One might interpret this as the “three-way process” proposed by the European Commission in its 2011 European Agenda for the Integration of Third-Country Nationals, but one should be aware that the local policy actors involved might have quite different intentions and motives.

Conclusion

The analysis presented in this chapter shows that immigration and integration policies have not shifted unidirectionally upward. Rather, we observe a growing complexity of policies in both areas being formulated at various levels of government, including the EU and national levels as well as the local and in some cases also the regional level. We observe substantial fragmentation as well, imposing the risk of “layering”; that is, policies are being developed at different government layers without structural connections. We provided various examples of such “layering” leading to a decoupling of policies, resulting in potential policy contradictions and even conflicts between different levels. Regarding migration, we mentioned as one example the lack of acceptance of Roma as fully integrated EU citizens. As for integration, we mentioned the potential effect of national symbolic discourses on integration processes at the local level. Local governments may move to rectify such effects in order to prevent interethnic tensions within city boundaries. We also saw the tension that has arisen from the EU definition of integration being applicable only to third-country nationals, as local and national governments have expressed a desire to integrate EU migrants into their host societies in a similar way.

At the same time, various and increasingly effective efforts are being made to institutionalize vertical relations between different levels of government. Following our definition of multilevel governance—that it should involve real vertical structures for policy coordination—we believe that we can speak to some extent of a multilevel governance structure for migration that has come in existence in a rather long struggle between national and EU forces, though still in the absence of regional and local governmental agents. Even in the strongly Europeanized field of migration and asylum policies—where one would expect centralist policy relations—we observe that most policies have been developed in a strongly intergovernmental way. Rather than states losing control to Brussels, they are working together and institutionalizing their cooperation, particularly that aimed at better control over immigration flows. However, the coordinated multilevel governance structure described here pertains mainly to restrictiveness and control of migration. Efforts to establish a more comprehensive, proactive immigration policies, as envisaged and proposed by the European Commission, have failed.

With regard to integration policies, partners’ competencies at different levels are clearly different from those in the migration policy field, and there seems to be no dominant level. Local governmental agents have claimed and are acquiring a more prominent position in relation to their national governments, and the EU level seems to be playing a mediating role. Relations across levels have intensified over the past decade, and they are both horizontal and vertical, top-down and bottom up. Some countries are developing vertical structures between the national and local levels, such as localized policy measures and joint integration conferences. At the same time, some countries are transferring their strict integration policies to the European level. Cities are applying pressure on their national governments to support local integration policies, and they are “venue shopping” at the EU level. An intriguing direct relationship has developed between the European Commission and city networks on a cross-European scale. All of this is recent and difficult to evaluate, but in view of the absence of clearly centralist and localist dominance in this process, the result could be a multilevel governance structure that, more than in the field of migration, includes nongovernmental partners in the process.

A final observation on the state-of-the-art of the study of multilevel governance as surveyed in this chapter is that so far multilevel governance has been framed, by definition, as an EU-internal phenomenon—that is, it includes only levels and actors within the EU as relevant components. What has been called the “external dimension” of immigration and integration policymaking—that is, relations, negotiations, and agreements with countries of origin of migrants and with international organizations and institutions in the field of international migration and development—does not have a place in this frame (yet). Consequently, the EU’s (re-)definition of integration as a three-way process does not resonate in studies of the multilevel governance of migration and integration. The concluding chapter of this book comes back to this external dimension of EU policymaking and its relevance for both immigration and integration policies.Open Access This chapter is distributed under the terms of the Creative Commons Attribution Noncommercial License, which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.