Inter-administrative Relations in Migrant Integration: France, Germany, and Sweden Compared

Abstract

Migrant integration is a prime example of intergovernmental coordination and multilevel governance; first because no level of government can carry out this task alone, and second because its cross-cutting nature often leads to fragmented institutional structures that must be overcome. Within the research strand of intergovernmental relations (IGR), the focus has been on executive actors and governmental decision-makers, resulting in an underexposure of the role of public administration, known as inter-administrative relations (IAR). Against this backdrop, we aim to remedy some of the deficits in IGR research by (1) adopting an explicit IAR perspective which systematically addresses the role of local governments; (2) including a comparative dimension in IAR research that accounts for different administrative ‘starting conditions’ in European countries; and (3) using the policy area of migrant integration as a case in point to empirically investigate developments of institutional convergence and divergence in IAR patterns. It is argued that the coordination of migrant integration in the three countries examined has moved towards more intergovernmental coordination, on the one hand, and that the role of municipalities in this context has been enhanced—varying degrees of (de-)centralization notwithstanding. While certain convergent patterns of inter-governmental coordination have become apparent during the migration crisis, historical path dependencies and administrative cultures still appear to be factors that influence institutional development.

Introduction

In many countries, local authorities are at the forefront of the reception and integration of migrants, often even formulating their own integration policies (Franzke & La Ruano de Fuente, 2021). The multi-level and cross-cutting nature of migrant integration policies makes them a prime example of intergovernmental relations (IGR). On the one hand, local authorities are bound to coordinate their efforts with other administrative units across different levels of government; on the other hand, they must overcome fragmented, overlapping, and somewhat opaque responsibilities and institutional structures that may result from the cross-cutting nature of migrant integration (Bogumil & Kuhlmann, 2022).

Yet, there is a remarkable gap in empirical research regarding the role of public administration (PA) and inter-administrative coordination in the multi-level system (Scholten & Penninx, 2016). Within the research strand of IGR, the primary focus has been on political actors, thus neglecting the role of administrative actors.

This is particularly true for migrant integration, where no comparative research explicitly focuses on the inter-administrative dimension of coordination. Thus, this book chapter aims to fill a missing link in comparative research and addresses inter-administrative relations (IAR) as a neglected subdimension of IGR in multi-level systems. While IGR include both a political and an administrative dimension, the latter has not been at the centre of research interest. The expertise of technocrats at the bureaucratic and implementation levels (see Beer, 1978) has been less the focus of IGR studies than political decision-makers and actors involved in policy formulation and legislation. Therefore, the dimension of IAR merits more systematic and in-depth empirical investigation within IGR research, in particular from an international comparative perspective. With this chapter, we aim to fill this gap in IGR research. Like other authors (Penninx & Garcés-Mascareñas, 2018; Schiller, 2019), we note a lack of systematic, cross-national comparisons of integration processes that go beyond particular fields of activity (e.g., political participation, urban policies, and education). Consequently, our focus is on how IAR play out in this domain across three countries.

Drawing on migrant integration as a case in point, our study reveals how the interactions and conflicts across different levels of government affect IGR processes, and how IAR play out in a multi-layered system in implementing policy goals. We analyse IAR in Germany, Sweden, and France, given their significant number of newly arrived migrants, especially in 2015/16 from Syria and, more recently, from Ukraine. Drawing on the distinction of various types of IGR as outlined in the framework of the book, we will elaborate a more refined typology of IAR and scrutinise to what extent country-specific administrative traditions (and other factors) have shaped IAR over time. The chapter thus addresses the following guiding question:

How have intergovernmental systems in countries with different administrative traditions and local government systems responded to similar pressures, and to what extent has this changed pertinent patterns of IAR?

Selection of Cases

We have selected the policy area of migrant integration as a case in point for empirically investigating trends towards institutional convergence and divergence in IAR patterns from a temporal and cross-country comparative perspective. The public task of migrant integration refers to diverse services, including inter alia language training, labour market integration, housing, education, and cultural activities. In this chapter, we focus on the integration of refugees as a particularly topical and urgent issue for many municipalities in Europe.

France, Germany, and Sweden share commonalities regarding the extent to which they are affected by the migration crisis. All three countries were exposed to similar external pressures in receiving an exceptionally high number of asylum seekers over the last years in the European Union. According to immigration statistics, Germany received the most asylum applications between 2015 and 2019 in absolute terms, followed by France. In terms of relative population, however, Sweden received the most asylum applications, if the small island state of Cyprus is excluded (see Table 5.1).

Table 5.1 Reception of asylum seekers between 2015 and 2021

The municipalities in the three countries studied play an essential role in migrant integration. Following the sharp increase in migrants in 2015/16, the German federal states have massively promoted local integration measures through funding programmes. In Sweden, municipalities have been legally obliged, since 2016, to accommodate the number of recognised refugees assigned to them by the central government for at least two years. In France, the role of the local level in migrant integration has been reaffirmed by the national strategy introduced in 2018 for improved reception and integration of refugees.

Despite similarities between the countries in terms of how they are affected by immigration and the increased role of the local level, they represent different European administrative systems and cultures. Germany belongs to the so-called Continental European Federal PA profile, Sweden represents the Nordic and France the Continental European Napoleonic PA profile (for details, see Kuhlmann & Wollmann, 2019; Kuhlmann et al., 2022). From a comparative perspective, these varying ‘starting conditions’, administrative contexts, and local government traditions are assumed to significantly influence, shape, and limit subsequent institutional choices and corridors of action, particularly in times of crisis. Moreover, the high variance of PA traditions promises valuable insights into the relationship between PA cultures and the functioning of IAR in crises. Thus, we can gain insight into how different IGR systems responded to the sharp increase of newly arrived migrants.

The comparative study is based on a qualitative design. Between May 2021 and June 2022, a total of 120 semi-structured interviews were conducted in the three countries with representatives of local and deconcentrated administrations, politics, and NGOs, complemented by secondary sources.¹

Conceptual Framework: Inter-administrative Relations and New Institutionalism

Inter-administrative Relations as a Missing Link in Comparative IGR Research

In federalism research, intergovernmental relations (IGR) have been studied predominantly with regard to the interactions among political actors in decision-making, legislative, and policy formulation processes (Benz, 2021: 6). Although IGR are defined to include both political and administrative relationships, pertinent studies have largely neglected the implementation processes while mainly concentrating on policymaking. The role of public administration, technocrats representing the bureaucratic and expert views (see Beer, 1978), and inter-administrative relations (IAR) have not been at the centre of empirical investigation so far (see Bogumil & Kuhlmann, 2022; Heuberger, 2022). An international comparative perspective on IAR in specific policy processes is missing. In this contribution, we will, therefore, adopt an IAR perspective which systematically addresses the role of public administration and bureaucrats/technocrats in the implementation phase, focusing on a specific policy process, namely the integration of refugees.

Conceptually, we consider IAR as being a—though largely understudied—key component of IGR. They represent the dependent variable of our research. Rooted in the structure of multi-layered systems, IAR face a tension between the principles of decentralisation and subsidiarity, on one hand, and the principle of equalisation of living conditions, on the other. Furthermore, effective IAR have to balance standardisation and stability with maintaining necessary flexibility and some organisational autonomy. Still, these tensions are often not easy to resolve and sometimes lead to ‘blame shifting’ and control deficiencies if responsibilities are not clearly defined. Problems may also arise as a consequence of IAR. The excessive coordination and standardisation of administrative action may render the advantages of decentralised task fulfilment obsolete and, as crises demonstrate, tendencies towards ‘intergovernmental centralism’ become visible (Kuhlmann & Franzke, 2022). However, administrative interdependence cannot simply be abolished; it is an integral part of the distribution of tasks in multi-level systems. Systems differ in their design depending on the specifics of the policy fields and the institutional framework conditions of different country contexts.

Analytically, we can distinguish between horizontal forms of IAR (between administrative sectors) and vertical forms of IAR (between administrative levels), as well as IAR of specialised and cross-sectional functions, and lastly, institutionalised and informal IAR (Benz, 1997: 168 et seq.). In this chapter, we focus on vertical IAR—that is, between administrative levels. In doing so, we distinguish between two key dimensions: (1) the allocation of tasks in the multi-level system (decentralised vs. centralised) and (2) the degree of intergovernmental coordination/decoupling (coordinated vs. decoupled). The first dimension refers to the predominant level of functional responsibilities formally assigned to a particular unit of administration in the field of migrant integration (state/central government vs. local government). However, decisions on migrant integration policies and their subsequent implementation are not only centralised or decentralised; they are also shaped by the interactions between the various governments (Hegele & Schnabel, 2021). Therefore, the second dimension addresses the separation or coupling of levels with regard to the fulfilment of integration-related tasks, thus relating to coordination in the intergovernmental setting. Combining the two dimensions, the following ideal types of IAR can be derived (see Table 5.2), which serve as heuristics for classifying the institutional development logic in the countries studied. The traditional IAR types of the various countries, that is the situation before the European ‘refugee crisis’ since 2015, appear in brackets.

Table 5.2 Ideal types of inter-administrative relations in migrant integration

New Institutionalism(s) as an Explanatory Framework

Cultural, actor-specific, and external factors are regarded as independent variables in our research, necessary to explain how IAR in migrant integration, our dependent variable, play out in the various countries. Borrowing from new institutionalism, institutional changes or continuity, as well as cross-country convergence or divergence in institutional developments, can be explained on the basis of three major theoretical approaches, from which the following contrasting hypotheses about the evolution of IAR can be derived.

The ‘External Pressure Hypothesis’

It is hypothesised that institutional evolution in migrant integration follows progressively convergent institutional paths. This assumption is supported by the literature on crisis management (Bouckaert et al., 2020; Boin et al., 2016; Christensen et al., 2016), in which we can observe a general tendency towards centralisation and concentration in times of acute crisis (Hernes, 2021). International research on policy diffusion and policy transfer further supports this, showing that similar external pressures often result in convergent institutional developments regardless of different ‘starting conditions’ and institutional traditions. Exogenous forces such as the migration crisis are expected to prompt decision-makers to streamline their institutional arrangements, producing somewhat congruent IAR features (convergence hypothesis).

The ‘Historical Path Dependence Hypothesis’

From the viewpoint of historical institutionalism, the administrative contexts and ‘starting conditions’ merit close attention when it comes to explaining institutional trajectories (see Steinmo et al., 1992). Historical path dependencies and cultural imprints inherited from the past have been highlighted as influential institutional development factors (see Brubaker, 1992). We anticipate that these traditional institutional features of the three countries examined largely shape subsequent institutional developments since they can either support or inhibit specific reform steps. From this, one can derive the hypothesis that the given administrative cultures and traditional profiles of IAR will persist even in times of crisis insofar as they constitute institutional path dependencies and limit the corridor for change. This means that, once established, existing institutional patterns can only be changed at a high cost (so-called lock-in effect; see Hall, 1986). Therefore, we expect the distinctive institutional models in their political and historical contexts to result in divergent IAR in Germany, Sweden, and France (divergence hypothesis).

The ‘Actor Constellation Hypothesis’

The ‘convergence hypothesis’ may be further contested on the grounds that institutional choices are not only prompted by external pressures but depend heavily on ‘endogenous’ forces—in particular on country- or even city-specific actor constellations and power-seeking strategies. Here, we draw on the actor-oriented approach of new institutionalism (Mayntz & Scharpf, 1995; Scharpf, 1997, 2000). This approach argues that institution-building must be interpreted as the result of choices made by political actors within specific institutional settings. Even if it guides and constrains behaviour, the institutional framework provides considerable scope for strategic action, encouraging political actors to influence institution-building in the direction of their own political objectives and benefits. Irrespective of similar external pressures, in different political contexts, we should therefore expect clearly distinctive or even progressively divergent trajectories and effects of institution-building, depending on the specific constellations of actors, political interests, and political ‘will and skill’. In line with this reasoning, we assume that changes and continuity in developing IAR depend heavily on whether and how the relevant actors support these measures. Hence, the IAR trajectories in the countries under consideration can be assumed to be either convergent or divergent, depending on the endogenous forces and actor-specific interests (variety hypothesis).

Our analysis does not aim to test these neo-institutionalist theories, but rather to use them as a heuristic framework and explanatory background. The objective of the following sections is to provide empirical evidence for institutional convergence/divergence in IAR regarding migrant integration in the three countries and to reveal some of its major determinants. The analytical framework we draw on can be summarised as follows (Fig. 5.1).

Fig. 5.1

Explanatory model. Source: Own illustration

Shifting Inter-administrative Relations in Migrant Integration

Starting Conditions: Task Allocation and Coordination

France stands out with an integration management that is strongly determined by the (deconcentrated) state administration. At the centre of this is the French Immigration and Integration Office (OFII), subordinate to the Ministry of the Interior, which coordinates the main integration programme, the Republican Integration Contract (Contrat d’Intégration Républicain—CIR). Once a person has been granted international protection, they sign a one-year contract in one of OFII’s 28 territorial directorates in metropolitan France. This includes language and civic orientation classes given by OFII’s contract schools, which are selected by public tender. The state administration also plays a vital role when it comes to professional integration. Deconcentrated employment agencies (Pôle Emploi) operating under the Employment Ministry offer job-specific counselling and support in job placement to any registered job seeker. These services become particularly relevant for migrants after completing the CIR training. Thus, French municipalities are neither involved in the conception, implementation, and financing of the CIR training nor in the subsequent labour market-related counselling, which gives a clear indication of France’s state-centric approach. However, cities and bodies of inter-municipal cooperation (metropolises) play an increasingly important, albeit complementary, role in this regard. They partly fulfil a ‘catch-all’ function, especially for vulnerable groups who fall through the state net, for example persons without a residence permit. Given the predominance of central government agencies and the limited coordination with local governments, the French model of migrant integration initially (before the 2015/16 migration crisis) most closely matched the centralised-decoupled type of IAR of our four ideal-typical IAR configurations.

Sweden’s integration policy for newly arrived migrants is determined by a comprehensive two-year integration programme (Etableringsprogrammet). This state-funded programme was introduced as early as 1994, with the municipalities initially being solely responsible for its organisation. The municipalities were therefore free to decide in what form they would provide financial compensation to participants and whether they would impose sanctions for irregular participation. The only condition for municipalities to receive a fixed state allocation per person was to draw up an individual and needs-oriented activity plan with the participant (Hernes, 2021: 8). Against this background, and in line with the generally high degree of autonomy of Swedish municipalities (Kuhlmann et al., 2022), Sweden’s approach to integration can best be classified as belonging to the decentralised-decoupled ideal type. This type is characterised by a decentralised task allocation and relatively decoupled coordination. It should be noted, however, that even before the sharp rise in refugee immigration in 2015, labour market competencies and the coordination of the integration programme were transferred to the Swedish Public Employment Service (Arbetsförmedlingen). This was done with a view to levelling out differences between the municipalities in the programme organisation and outcomes to ensure a more uniform offer of employment opportunities and equal quality standards throughout the country. What this transfer of competencies in 2010 did not change, however, is the fact that the municipalities bear the financial responsibility for immigrants as soon as the integration programme ends. Consequently, the institutional setting before the 2015 migration crisis may still be regarded as predominantly decentralised.

Germany is closer to the Swedish than the French model with regard to the allocation of integration tasks. German municipalities assume a broad portfolio of integration-related tasks, such as providing welfare services, accommodation, and integration monitoring. At the same time, the municipal scope of action—in providing counselling, for example—is limited by state regulations, funding schemes, and various obligations to coordinate activities with higher levels of government. Despite this, the local governments may set different policy priorities and institutional arrangements. Moreover, due to the federal structure, there is a high degree of variation across the Länder, which mirrors different levels of crisis affectedness across the country, as well as different policy preferences among Länder governments in relation to migrant integration. Many municipal integration activities are financed by funding programmes of the Länder, whereas voluntary tasks of self-government are funded through the municipalities’ own budgets and depend on the will of the political actors. Nevertheless, certain integration-related tasks are allocated outside the municipal sphere of influence and carried out by authorities at the federal and state level. The federal level dominates labour market integration and language support for foreign adults. Labour market measures fall within the remit of the federal employment agency (Bundesagentur für Arbeit) and the job centres, while language acquisition, including job-related language support, as part of the integration programme, is the responsibility of the federal agency for migration and refugees (BAMF). Since state actors are obliged to coordinate their activities with local governments in all these fields, and in light of the strong emphasis on decentralised task fulfilment, the German integration model resembles the decentralised-coordinated type of migrant integration.

IAR Shifts and Continuities: Responding to Pressure and Crisis

State-local coordination in the three countries under consideration has increased since the 2015/16 migration crisis and after numerous deficits came to light (Thränhardt, 2020). While coordinative practices in the intergovernmental setting have been a common feature of IAR for decades in German cooperative federalism (although they functioned poorly during the crisis), administrations in Sweden and France had to expend more effort on implementing coordination across the various levels. As a result, all three countries tended either to converge towards more coordinated types of IAR (France, Sweden) or to enhance already-existing coordination patterns (Germany). Nevertheless, there continue to be significant differences between the three countries with regard to the degrees of (de)centralisation achieved.

Sweden—towards more centralisation and coordination: Starting from a conspicuously high degree of decentralisation and local autonomy, a number of significant steps towards centralisation were already taken before the 2015/16 migration crisis, especially in labour market integration as the focus of Swedish integration efforts (Lidén et al., 2015). In addition to the centralisation of the integration programme in 2010, further reforms were directed towards standardisation in the local implementation of nationally defined regulations during the refugee crisis, such as the Settlement Act of 2016. With the adoption of the Settlement Act, the distribution of recognised refugees across Sweden became centrally managed. Many municipalities had to establish the necessary organisational structures to comply with the new regulations (Lidén & Nyhlén, 2022). This legislation also obliged municipalities to organise housing for newcomers no later than two months after their favourable asylum decision, and for at least two years. These changes have increased central government influence on local integration management, and the 2016 Settlement Act can also be interpreted as a state intervention in the local sphere.

It is precisely these centralisation steps that have led to the development of the complex MLG structure. Thus, the integration programme involves a multitude of actors at the national (Swedish Public Employment Service and Swedish Social Insurance Agency) and local levels (municipalities). In addition to the integration programme, the Public Employment Service and the municipalities jointly implement programmes for subsidised employment, such as the ‘Extra Jobs’ programme. Regarding the distribution of refugees,² municipalities are bound to collaborate with the Swedish Migration Agency since they were obliged to accept a certain number of refugees in 2016. With regard to the increased efforts in intergovernmental coordination, the county administrative boards (Länsstyrelse) also deserve closer attention. As an important link between the municipal authorities on the one hand and the government and central authorities on the other, they regularly bring together all actors involved in migrant integration in the respective county, such as municipalities, the Swedish Public Employment Service, the Migration Agency, and civil society organisations. Their tasks are less operational in nature and focus primarily on creating an exchange platform for actors from different levels and sectors, institutionalising communication and identifying local needs in the integration field (Lidén & Nyhlén, 2022). Finally, almost all municipalities have concluded an agreement with the Swedish Public Employment Service to formalise their cooperative relations (Parusel, 2021). The centralisation trend was therefore paralleled by increasing intergovernmental coordination efforts, shifting away from an initially rather decoupled setting towards a more coordinated system.

The policy field of migrant integration carries a high potential for conflict since municipalities continue to carry a heavy weight in migrant integration, in particular the financial burden after the two-year integration programme. Almost all municipalities have set up their own labour market programmes to complement the public employment service, given that migrants who do not have a job after two years are entitled to income support from the municipalities.³ There are also long waiting times for language courses and the interviews have shown that some municipalities shirk their responsibility in housing recognised refugees by renting apartments in neighbouring municipalities. Conflictual relations also occur between the municipalities and the Migration Agency. Many local representatives deem the existing coordination mechanisms with this agency deficient as they lack timely information about the arrival of refugees, further exacerbating the tense situation in the housing market (Jedrzejewska & Spehar, 2020: 15 f.).

In sum, Sweden has moved closer to the ideal type of centralised-coordinated IAR. This applies in particular to labour market integration and the (two-year) management and financing of the establishment programme. However, it should be recalled that municipalities have remained important actors in migrant integration and have recently strongly reanimated their (labour market) integration activities—not least to avoid subsequent unemployment-related financial burdens.

France—informal decentralisation, increased coordination: In France, IAR were centralised in recent decades based on a clear vertical administrative organisation. In the traditionally decoupled setting, several cities alleged that the central and deconcentrated state administration had lacked communication and coordination during the migration crisis, sometimes accompanied by spectacular municipal protest actions.⁴ While we observe a cautious ‘localisation’ of integration-related activities, with many cities increasingly assuming a more important role in migrant management, the central government still closely steers and regulates local activities in this sector. Through contractual instruments such as city contracts under the Politique de la Ville⁵ or the territorial contract CTAIR (Contrat Territorial d’Acceuil et d’Intégration des Réfugiés),⁶ state authorities continue to dominate integration management. Many actions are initiated in a top-down mode, based on the contracts concluded between local government entities or inter-municipal bodies (especially the metropolis) and the departmental prefecture. Thus, local governments and inter-municipal bodies still lack formal competencies for addressing refugee policies. Instead, service provision remains primarily a central government task, only occasionally complemented by local government contributions. Nevertheless, local governments and inter-municipal bodies (metropolises) have increased in importance since the 2015/16 migration crisis. Cities often play a compensatory role, especially for vulnerable groups, since the state contributions are often not sufficiently target-group-specific. They may provide optional aid to those in precarious living situations via their municipal social centres (Centre Communal d’Action Sociale—CCAS), such as non-reimbursable financial assistance. Furthermore, an increasing number of cities and metropolises deliberately join city networks to approach the central state with a united voice and get more involved in formulating and implementing integration policies. These developments support the ‘local turn’ thesis and point towards a well-known pattern of administrative reform in France known as ‘informal decentralisation’.

With the introduction of the national integration strategy in 2018 (Stratégie Nationale pour l’Accueil et pour l’Intégration des Réfugiés), the degree of vertical coordination between the central state and local levels has increased significantly. Although new measures, such as territorial contracts, could not entirely eliminate conflicts between state actors (e.g., prefects, deconcentrated state agencies) and local actors, the intensity of intergovernmental coordination has significantly increased. To address deficiencies in the reception and integration of migrants, the 2018 integration strategy strengthened vertical collaborative relationships by employing public contracts. These contracts (CTAIR; see above) are voluntarily concluded between the prefecture and municipality or metropolis and include an annual financial envelope. In addition to incentivising local authorities to implement concrete integration actions, the contracts ensure that the state can exert influence on local policymaking through corresponding funding conditions and that policy implementation at the local level reflects the central policy framework (Huglo, 2014). Moreover, the prefects’ coordinating role was further strengthened. The new integration strategy had given rise to regional and departmental steering committees on asylum and integration policies led by the prefects. Prefects appoint regional and departmental coordinators for these matters and oversee working groups involving refugees to identify problems on the ground and to ensure that proposed measures are more appropriately tailored (Bonnotte & Sénimon, 2021: 164). The prefects act in a highly operational manner, especially in implementing the contract policy (CTAIR) and as an interface with civil society. In summary, France has moved from a centralised-decoupled type to an (informally) decentralised and coordinated setting of IAR. Yet taking the ‘starting conditions’ into account, the degree of (formal) centralisation in France remains considerable, particularly when compared to IAR in relation to migrant integration in Sweden and Germany.

Germany—reaffirming decentralisation and improving coordination: In Germany, the high level of immigration in 2015/16 served as a catalyst for increased local activity and engagement in migration and integration policies, based on an already highly decentralised setting and local experience in handling migration-related tasks. The nationwide distribution of refugees also confronted rural municipalities with the task of integration, resulting in the professionalisation of organisational structures and the revitalisation of municipal integration concepts in most county-free cities and counties (Schammann et al., 2020). While by the end of the 1990s, only an estimated 20% of all cities had formulated an integration plan, more recent figures from 2017 indicate that 69 of 79 large cities (87%) and 103 of 574 (18%) medium-sized cities did have such an action plan (Filsinger & Gesemann, 2018: 10). Furthermore, several municipalities have been politically advocating for a greater scope for action in migration and refugee policies (Schammann, 2020). Finally, numerous federal- and state-funding programmes have strengthened the municipalities’ role in integration policy.

The already-existing coordination patterns between local governments and Länder administrations were enhanced in response to various coordination failures during the refugee crisis. Migrant counselling services provide a prime example of vertical interdependence and highly coordinated IAR. The fragmented, overlapping, and sometimes opaque responsibilities and institutional structures require special intergovernmental coordination in the multi-level system. Here, the administrative district authorities (Bezirksregierung), which exist in some Länder at the meso level, adopt a coordinating function. For example, they provide overviews of counselling services and funding opportunities. In their coordinating and bundling function, they may be regarded as the counterparts of the prefects in France and the county administrative boards in Sweden. As this example demonstrates, inter-administrative coordination has been institutionalised and intensified as a result of the refugee crisis. Consequently, the decentralised-coordinated type of IAR has been reaffirmed and further strengthened in migrant integration in Germany.

Figure 5.2 below summarises our major findings regarding changes in the dependent variable of research (IAR).

Fig. 5.2

Shifts in IAR in the field of migrant integration. Source: own illustration

Discussion: Explaining IAR-Developments

The ‘External Pressure Hypothesis’

In all three countries under consideration, we observe that the sharp increase in the number of asylum applications in 2015 and thereafter served as a ‘catalyst’ (Schammann, 2020) and ‘critical juncture’ (Hernes, 2021) for strengthening both intergovernmental coordination and municipal involvement in migrant integration policies. This was not necessarily accompanied by a formal increase in local competencies, but rather by ‘informal decentralisation’ (as in France) or sometimes even, by contrast, a ‘national turn’ (Sweden). Municipalities in all three countries feel increasingly compelled to provide supplementary services to migrants due to insufficient or ineffective state services. In this way, they address specific groups, such as asylum seekers, who are not eligible for state measures. The growing employment activity of the cities can be explained by the fact that they seek to avoid the financial burden—that is, a dependence on social assistance—that may arise when migrants are not employable after participating in the state integration programme. With a growing sense of urgency, municipalities tend to establish their own structures—for example, in labour market integration—or seek new ways of collaborating with state agencies. In all three countries, local governments used their discretion and creatively complemented integration-related state contributions that had proved insufficient or ineffective, thus responding to the external pressure of (partially) failed state services in a crisis-driven policy field. Municipalities play a particularly important role for people without a valid residence permit, and they also differ from one another in the way they deal with this group. Lastly, financial support for integration measures from state agencies has significantly increased due to crisis-driven pressures, representing another striking commonality (and convergence) between the countries. Thus, the contractualisation in France exhibits similarities with the funding programmes of the German Länder but also with the newer labour market agreements between the Swedish Employment Service and the municipalities. All these agreements combine financial support and incentives with the obligation of cities to provide certain integration services.

Against this background, IAR have converged towards increased local engagement in migrant integration on the one hand and more intergovernmental coordination on the other (coordinated types of IAR). We also observe that, while France has moved towards increased decentralisation, Sweden has witnessed the opposite trajectory, which, from remote observation, indicates a convergent pattern (see Fig. 5.2). However, while IAR have shifted towards certain common characteristics, manifold differences and even divergent institutional trajectories were revealed upon closer examination of the three countries concerned. In the case of France, the central government continues to formally dominate the integration domain, while in Germany and Sweden, local governments remain key actors in integration services. As a result, the degree of formal (de)centralisation in the three countries varies significantly, as does the role of state actors and their interactions with local governments. Therefore, we must consider further endogenous factors when explaining the remaining differences and divergent paths of institutional development in the three countries.

The ‘Historical Path Dependence Hypothesis’

Our country comparison shows that many basic features and key elements of the intergovernmental systems have remained relatively stable over time, despite crises, thus confirming the assumption of path-dependent developments. In France, the characteristic features of the Napoleonic system have not disappeared, but partly intensified during the crisis. In Germany, too, the typical institutional features of highly intertwined coordination structures and administrative interdependencies within the peculiar model of cooperative federalism have proved to be fairly powerful, even under extreme pressure. Finally, in Sweden, the municipalities have remained important players in integration management and continue to be high scorers in local (fiscal) autonomy from an international comparative perspective (see Ladner et al., 2021)—the centralising trends and increasing state interventions notwithstanding.

There is no overall trend towards increased centralisation in migrant integration, which could have been expected from crisis literature and the ‘external pressure hypothesis’. Instead, the refugee crisis is addressed by intensified multi-level collaboration (Germany and Sweden) and informal decentralisation (France). Centralisation of migrant integration is not a crisis solution for all seasons, but only one possibility among others, depending on the country-specific contexts and political preferences (see below). This observation contrasts with the general convergence hypothesis and supports the assumption that historically inherited institutional legacies and endogenous forces of institution-building account for persisting country differences and even institutional divergence.

The ‘Actor Constellation Hypothesis’

Our interviews have revealed that the commitment of local executives (mayors) plays an important role in local governments’ engagement in migrant integration. In view of the political contentions surrounding the acceptance of migrants, characteristic of all countries accepting a significant number of migrants, this issue is becoming increasingly political and is influencing local provision for migrants. France provides a prime example of actor constellations and the couleur politique having the potential to initiate shifts in IAR, as local executives (mayors) massively challenged the centralist Republican integration model. This applies specifically to politically left-leaning cities in which a strong mobilisation was observed (Flamant, 2020). It also became evident that well-functioning coordination in the intergovernmental setting is more likely if the representatives of the various local government tiers share similar political convictions and preferences. The actor constellation hypothesis is also supported with regard to the German and the Swedish cases, yet based on a lower level of intergovernmental conflicts and contention. In Germany, it is not only local executives who influence IAR in migrant integration, but—even more significantly—the Länder governments, some of which exhibit more commitment in this policy field and have launched, in some cases (e.g., North-Rhine Westphalia, Hesse), significant funding programmes, while others have refrained from doing so. In Sweden, changes in the party-political constellations (e.g., the electoral success of the right-wing Sweden Democrats), as well as a general change in the country’s political and societal discourse about migrant integration and subsequent policy reforms—in labour market policy, for example—appear to be strong predictors of IAR. In addition, the party-political composition and policy preferences of local councils are crucial for the design of IAR and are among the drivers of the country’s ‘national turn’ away from its pronounced decentralised approach.

In summary, national and local actor constellations, interests, and political preferences account for persistent differences in integration management across countries and explain the variety in crisis responses and institutional solutions. They also explain deviations from historically inherited institutional paths and why certain institutions with deep roots have become issues of contention in the three countries. This observation contradicts the convergence thesis and challenges the assumption that institution-building is determined primarily by external pressure.

Conclusion

This chapter has addressed how intergovernmental systems in countries with different administrative traditions and local government systems have responded to similar pressures and to what extent this has changed relevant patterns of IAR. With migrant integration as our empirical field of investigation, we have adopted a policy-focused perspective on a substantive crisis-related policy issue, which has provided key insights into the dynamics of IGR and how institutional actors seek to manage a crisis in complex multi-layered systems. Based on an understanding of IGR as a process of coordination and conflict management across levels, we have concentrated particularly on IAR as a largely neglected area of comparative research so far. Our findings have revealed that the country-specific administrative cultures and traditions largely shape the IAR which have been established in the three countries in the aftermath of the refugee crisis.

Drawing on the distinction between centralised, decentralised, and multi-layered IGR processes as laid down in the conceptual framework of this book, we have elaborated a more refined typology with a particular focus on IAR. We have seen that, over the course of the crisis, countries have shifted in their predominant form of IAR towards more coordination and coupling, with Sweden moving from the decentralised-decoupled type towards the decentralised-coordinated type, France from the centralised-decoupled towards the decentralised-coordinated type and Germany reinforcing and re-confirming its characteristic decentralised-coordinated model.

Regarding the role of conflict in IGR and their potential to affect policy outcomes, migrant integration as a policy field has proven to bear a high potential for conflicts in all three countries. While this is on the one hand due to the sheer crisis-related pressure, urgency, the re-distributional character and the socially as well as politically and ideologically contested nature of the policy field, there are, on the other hand, also features related to the characteristics of public administration and IAR which fuel conflicts in IGR processes. This refers to the cross-cutting and multi-level nature of the task area and, therewith, a high need for coordination, the entanglement of levels, and the potential for inter-organisational conflicts. Furthermore, the malfunctioning of state agencies, a lack of resources at the local levels, opaque distribution of responsibilities, and funding issues are the main sources of conflict in the studied countries.

The 2015/16 migration crisis proved to be a critical juncture that led to increased municipal involvement in this area, without the formal competencies of municipalities being significantly expanded. Municipalities in Germany, Sweden, and France increasingly felt compelled to provide supplementary integration services during and after the migration crisis. Even though the state continues to dominate (France), or centralisation steps have been taken (Sweden), municipal activity has increased over time across all three countries. In addition to inadequate state offers, other factors were decisive for the design of the IAR. These include institutional changes (e.g., the closure of labour market agencies in Sweden and the associated service restrictions), as well as state support for local integration measures through corresponding funding programmes.

In all three countries, IAR have shifted towards increased coordination, thus mirroring a convergent path of institutional development to a certain degree. While the Swedish approach was initially marked by a largely decentralised and decoupled setting of IAR and shifted increasingly towards centralisation and coordination, France represents the opposite case. It increasingly ‘turned local’ in the sense of the typical French style ‘informal decentralisation’, accompanied by increased inter-administrative coordination between state actors and local/inter-municipal governments. Therefore, both France and Sweden exhibit certain convergent developmental patterns, if their respective ‘starting conditions’ (highly centralised vs. decentralised) are taken into account. Contrary to the widespread notion that federal states cannot respond to crises as quickly and in as coordinated a manner as unitary systems due to their institutional complexity, the existing type of decentralised-coordinated IAR in Germany was reaffirmed during the 2015/16 migration crisis. In other words, integration tasks were implemented in a vertically highly interconnected setting.

In conclusion, external pressure undoubtedly resulted in certain convergent developments in IAR (e.g., increased municipal engagement in migrant integration, intensified inter-administrative coordination, enhanced state subsidies and contract arrangements to incentivise local integration efforts). On the other hand, the historically different administrative traditions and the specific constellations of actors have been revealed as salient explanations for the remaining cross-country variation (e.g., state predominance vs. local government predominance) and within-country cleavages (e.g., integration proactive vs. passive local governments), as well as different institutional trajectories (e.g., centralisation vs. decentralisation).

Since the focus of this chapter has been primarily on the 2015/16 migration crisis, it remains to be scrutinised how the recent influx of refugees from Ukraine has influenced IAR in integration management. It will be interesting to see if the trajectories of IAR described in this contribution have prolonged, intensified, or changed.