Are small towns a battleground for migration governance? Negotiating refugee integration in small towns in the Netherlands

ABSTRACT

A surge in refugee arrivals post-2015 challenged small localities across Europe to respond to ‘refugee integration’ locally. Existing literature focuses little on governance dynamics in small localities, particularly regarding actors’ roles and relations. Based on fieldwork in four Dutch localities, we analyse the involvement of different actor types and explore if/how they develop collaborative relations. We show that small town governments engage in different forms of collaborative governance, from consolidated actor networks and intensive collaboration between local governments and CSOs to fragmented networks with more passive governments. However, we also find conflicts among actors due to power asymmetries and diverging interests and values. We identify several factors shaping interaction patterns, including localities’ size, political orientation, political leadership, residents’ attitudes and municipal decision-making. Despite outsourcing tasks, local governments remain pivotal in integration governance, giving them a crucial role in creating inclusive, participatory spaces, preventing actors’ alienation and designing efficient policy responses.

KEYWORDS:

• Local governance
• collaborative governance
• refugee integration
• battleground
• small towns
• migration

Introduction

Since asylum applications reached a peak in the EU in 2015 and 2016, the term ‘refugee crisis’ is often used at the European level to refer to (inter)national policy responses seemingly unable to cope with the complexity of migration (Scholten 2020). Not only have states struggled with implementing migration governance comprehensively and locally, but the perceived ‘crisis’ in migration governance has led to a growing role of small and medium-sized towns in refugee accommodation and integration, challenging local authorities’ responsiveness towards such policy issues.

Despite the growing role of small and medium-sized towns in integration governance, much of the literature in migration studies has focused on large cities, leaving a gap in our understanding regarding governance responses in medium-sized and small towns (Meijer, Zuzanna Popławska, and Szytniewski 2023; Pettrachin 2024; Schammann et al. 2021). Among existing studies in smaller places, only a few discuss particular governance processes as well as interactions between local state and non-state actors (see for notable exceptions Cabral and Swerts 2021; Haselbacher and Segarra 2022; Meijer, Zuzanna Popławska, and Szytniewski 2023; Miellet 2022; Semprebon, Marzorati, and Bonizzoni 2022).

Given that local governments have increasingly started relying on and collaborating with non-state actors to address complex policy issues such as migration more effectively (Wang and Ran 2023), it is pertinent to explore if/how small localities engage in collaborative governance processes and which factors shape patterns of interaction between state and non-state actors (Larruina, Boersma, and Ponzoni 2019). A core question concerns whether actors involved in local integration governance develop more collaborative or conflictual relations – and why. Since conflicts in collaborative governance arrangements may weaken or alienate (non-state) actors and impede decision-making and efficient policy implementation, it appears crucial to systematically analyse actor relations and underlying factors for collaboration and conflict (cf. Pettrachin 2024; Schiller 2018).

In our paper, we address these questions by examining how four smaller localities in the Netherlands have responded to the complex issue of refugee integration.¹ We analyse how local state and non-state actors engage in the integration process of recognised refugees who arrived after 2015 and the types of relations they develop. We aim to show how contextual factors may influence actor relations and lead to specific collaborative and/or conflictual patterns of interactions across localities. We adopt a local, comparative perspective to gain a nuanced understanding of ‘why and how cities and regions respond differently to similar challenges’ (Glorius et al. 2019, 20). The importance of a ‘local-to-local comparison’ (21) seems particularly pertinent in a decentralised unitary state such as the Netherlands, where the main tasks related to migration and – up until recently – integration policymaking lie in the hands of the national government. One thus may assume that local responses look relatively similar. Comparing different locality types embedded in the same national context promises in-depth insights into the relevance of different factors shaping local responses (cf.Meijer, Zuzanna Popławska, and Szytniewski 2023).

Drawing on a comparative case study, we emphasise the significance of actors’ local situatedness (Noordegraaf et al. 2019) which may result in different integration governance configurations, moving on a spectrum from a conflict-driven ‘battleground’ (Ambrosini 2021) to intensive cooperation (Schiller 2024). In our analysis of local actors and their relations, we thus consider the influence of contextual factors such as localities’ size, political orientation, or residents’ attitudes. We advance existing literature on integration governance responses in small and medium-sized towns in the Netherlands, showing that municipalities respond differently to the so-called ‘refugee crisis’. We further contribute to the collaborative governance literature in the field of migrant integration policymaking, by analysing constellations and interactions of local actors in four different localities. Gaining a better understanding of plural policymaking dynamics and factors shaping them is important because complex policy issues require complex responses, instead of a one-size-fits-all approach (Scholten 2020).

Conceptualising local migration governance

Migration is often described as a contested and equivocal policy issue (Noordegraaf et al. 2019) which cannot be solved by the state alone. This has resulted in a dispersion of authority and responsibility to (non-state) actors across local, regional and (inter)national levels (DiGaetano and Strom 2003). Scholars have increasingly shifted their attention to the local level, enhancing our general understanding of local migration governance. Yet, gaps persist in understanding migration governance dynamics in small and medium-sized towns, leading us to focus on local actors’ roles in such localities and examining emerging collaborative and conflictual relations.

Non-state actor involvement at the local level

Local governments increasingly organise the delivery of their public services ‘at arm’s length’ (Genugten, Marieke, and Voorn 2020, 1) by outsourcing tasks, collaborating with private or societal partners, or creating external government-owned agencies (cf. Nyseth and Ringholm 2008). In the context of migration, studies have shown that local migration governance is nowadays characterised by the involvement of local authorities and various non-public actors (Campomori and Ambrosini 2020; Schammann et al. 2021; Schiller et al. 2023). Local authorities carry out formal duties while outsourcing tasks to non-public actors which support refugees through formal and informal activities and services (Meijer, Zuzanna Popławska, and Szytniewski 2023). Local actors differ in their level of legal and structural autonomy, their degree of formality and professionalisation, as well as their service provision, activities and target group(s) (Campomori and Ambrosini 2020; Genugten, Marieke, and Voorn 2020). Some studies on small municipalities found that, compared to large cities, there are fewer actors involved, often forming dense networks, and occupying multiple roles (Semprebon, Marzorati, and Bonizzoni 2022).

As a starting point for examining actor dynamics in small localities, we identify what type of actors are involved and which roles they take. We adopt a broad definition of ‘non-public actor’, comprising ‘institutionalized’ actors (referred to as CSOs) as well as emerging civil society initiatives and residents. Regarding ‘institutionalized’ actors, we distinguish between ‘formal’ organisations that are formally (usually via a contractual agreement with local authorities) involved in the service provision for refugees and ‘informal’ actors which do not receive dedicated funding to implement specific tasks (including churches, faith-based organisations, or voluntary organisations). The latter’s services are often more flexible and comprise language support, legal advice, or facilitating exchange between residents and newcomers (Larruina, Boersma, and Ponzoni 2019). Importantly, informal actors can also apply for public subsidies, proactively taking up a role in integration governance structures. Finally, we conceive of residents (with and without migration background) as important actors in integration policymaking because they influence local governments’ responses, for example, by supporting or resisting refugees’ arrival and challenging local governments to adjust their policies (Ambrosini 2021).

Patterns of interaction between local actors

Besides highlighting local actors’ multiple roles, scholars have explored different ‘modes of governance’ (DiGaetano and Strom 2003, 363) defining relationships between non-public actors and local governments (Genugten, Marieke, and Voorn 2020; Nyseth and Ringholm 2008). As mentioned above, such relationships are often marked by collaboration. We understand collaboration as any joint activity between two or more actors to achieve a shared (policy) goal, deliver services and/or create public value (Prentice, Imperial, and Brudney 2019). Collaborations can be designed to be relatively permanent and long-term or temporary and ad hoc. However, the ‘ideal type’ of collaborative governance where independent actors work together ‘based on deliberative consensus and collective decision making to achieve shared goals that could not be otherwise fulfilled individually’ (Wang and Ran 2023, 1189) is in practice rarely achieved. Earlier studies on local migration governance highlighted that interactions can happen in more centralised/hierarchical or decentralised/horizontal/polycentric contexts (Schiller et al. 2023), leading to varying forms of cooperation, co-optation, or conflict (Schiller 2018). The battleground concept has been used to describe actor relations in asylum seeker reception in Italy, identifying a rather antagonistic relation between local actors (Ambrosini 2021; Campomori and Ambrosini 2020). In order to further unpack ‘local battlegrounds’ in other contexts, it is important to show if/how conflicts arise between different types of actors, which shapes they take and why they occur. Regarding the latter, scholars argued that groups with similar (political) values form coalitions against those with differing views on ‘integration’ and diverging values and interests (Hinger 2016; van Hulst and Yanow 2016). Power (asymmetries) also shape(s) interactions among actors, especially when (local) governments are involved (Glorius et al. 2019; Larruina, Boersma, and Ponzoni 2019; Schiller et al. 2023). Gaventa (2006) notes that power relations ‘shape the boundaries of participatory spaces, what is possible within them, and who may enter’ (26), leading to continuously changing forms of collaboration and tensions between actors. We thus consider actors’ interests, values and power (asymmetries) in analysing local patterns of interaction. Moreover, we pay attention to actors’ local situatedness to understand better why collaborative and conflictual dynamics evolve and possibly differ across localities.

Contextual factors

Scholars have identified various factors explaining plural local governance responses, including, localities’ size, political orientation, economic and demographic development, population composition, available resources and infrastructure, as well as local migration history (Glorius et al. 2019; Hinger 2016; Kos, Maussen, and Doomernik 2016; Schammann et al. 2021; Schneider 2022). Some of these factors may also account for differences in actor relations in different local contexts. First, localities’ size influences actor relations: small towns typically (used to) have little established arrival/integration infrastructure and are characterised by limited, rather homogeneous ‘associational ecologies’ (Semprebon, Marzorati, and Bonizzoni 2022, 12) with close and personalised relations between politicians, CSOs and residents (Haselbacher and Segarra 2022; Schneider 2022). Second, in their study on integration governance in small and medium-sized localities, Caponio and Pettrachin (2023) highlight the importance of local governments’ political orientation and presence of radical right parties (RRPs) in understanding (conflictual) dynamics between policy actors (cf. Pettrachin 2024). They find that collaborations with non-public organisations are less hierarchical and occur more frequently in municipalities with centre-left governments compared to centre-right ones. Third, residents’ attitude towards refugees appears to influence local governance dynamics, especially in the form of pro- or anti-immigrant movements. This seems particularly prevalent in small towns where the often less-diverse local population tends to hold more negative attitudes towards immigrants (Huijsmans et al. 2021; Kloosterman 2018), thereby shaping local governments’ responses (Glorius, Bürer, and Schneider 2021). Finally, active or passive political leadership (Haselbacher and Segarra 2022; Miellet 2022) and municipal decision-making shape actor relations. The latter refers to the central role of local governments in policymaking and -implementation (through their authority to enforce rules, control funds and delegate tasks), allowing them to significantly shape governance relations.

We formulate three expectations: First, various formal and informal actors are involved in local refugee integration and actors interact with each other. We expect the number of actors to be rather small and their relations tight. Second, because of the smaller number of actors and dense actor networks, we expect few(er) battles to emerge in the interactions between local actors in small localities. Third, we expect that local characteristics such as localities’ size, political orientation, residents’ attitudes, political leadership and municipal decision-making help explain variation across localities.

Methodology

We study refugee integration governance in four Dutch municipalities, including one medium-sized town, two small towns and one rural area. The Dutch case is interesting for its relatively high level of centralisation, with clear competences for municipalities and a more recent trend of task devolution. This makes for an interesting comparison with countries where the national level has taken a more laissez-faire approach, leaving it to localities themselves to define their roles and tasks. To allow for some generalisability of our findings for other small localities in the Netherlands, we aimed for high variation among the selected cases to allow for cross-comparison and identify potential explanatory factors for observed similarities and differences (Table 1). The four cases were selected from all Dutch municipalities with a population size under 250.000 involved in asylum-seeker and refugee reception after 2015 and still hosting an asylum seeker centre in 2021, ensuring that municipalities had previous experience with newcomers. We excluded exceptional cases that received significantly larger numbers of asylum seekers and are characterised by high politicisation and satellite localities connected to big cities. The selection was refined through theoretically relevant variables, including population size, share of residents with migration background before and after the arrival of post-2014 refugees, unemployment levels, demographic trends and political parties in government. Finally, to ensure regional variation, the cases are distributed across four provinces, namely the Western provinces South Holland and Utrecht, located in the densely populated and diverse metropolitan area ‘Randstad’, and the provinces Overijssel and Drenthe, located in the comparably more sparsely populated and less diverse East and North of the country, respectively (Figure 1). Due to the localities’ size and the often-limited number of actors, some respondents could have been easily identified, for example, because there was only one local official working on the topic of refugee integration. We therefore do not disclose the names of the localities to ensure respondents’ anonymity.

Figure 1. Overview of the selected cases in the Netherlands.

Table 1. Summary of cases’ key characteristics.

	Locality A Medium-sized town	Locality B Small town	Locality C Small town	Locality D Rural area
Province	Utrecht	South Holland	Overijssel	Drenthe
Size	140.000–170.000,^3^4	50.000–80.000	50.000–80.000	20.000–40.000 1 main town, 30 villages
% migration background (2021)	25% 16% ‘Non-Western’^5	12% 6% ‘Non-Western’	27% 16% ‘Non-Western’	9% 4% ‘Non-Western’
Number of recognised refugees	1200 in 2020^6	350	500	NA
Economic situation	Stable and growing Unemployment level < national average	Stable Unemployment level < national average	Unemployment level > national average High share of welfare recipients	Inner-municipal differences (poorer and better-off villages) Unemployment level = national average
Political orientation (2018–2022)	Progressive and conservative	Christian-conservative	Conservative-right, presence of RRP	Conservative/moderate with strong local party
Alderman for integration	Left-progressive	Christian-conservative	Conservative	Social-democratic

Our study is part of a European collaborative research project (Whole-COMM). Empirical data were collected by the first author between October 2021 and September 2022. Data collection comprised document analysis (including policy documents, newspaper articles and municipal websites) and semi-structured qualitative interviews with respondents at local, regional and national levels. Respondents were purposefully sampled based on their professional positions, e.g., as local officials working on integration or employees in CSOs supporting refugees. This resulted in a total number of 56 interviews which were evenly distributed across the four localities. In all municipalities, we interviewed the main stakeholders involved in local integration policymaking, namely members of local governments (aldermen), local officials working on integration, representatives from local housing corporations and from CSOs offering ‘social support’ to refugees as well as other non-public organisations. Additionally, the first author interviewed 15 actors at the regional and national levels. The acronym N - [locality A/B/C/D] - [interviewee number] is used to refer to respondents (see Jonitz, Schiller, and Scholten 2022, for interviewee details).

Due to the COVID-19 pandemic, most interviews between November 2021 and March 2022 were conducted online. The interviews comprised questions about the local distribution of responsibilities, actor’s relations, different forms of collaboration and conflict and local policy development. The interviews were transcribed and analysed with the software atlas.ti, resulting in deductively and inductively generated codes. Codes included key themes such as ‘perceived local characteristics’, ‘involved actors at the local level’, ‘forms of collaboration’ or ‘observed tensions’.

Due to the topic’s sensitive nature and dense actor networks in the localities, where most respondents knew each other on a first-name basis, some aspects related to existing conflicts surrounding refugee integration may not have been fully disclosed to the researcher. Moreover, our qualitative analysis of factors across four cases does not allow us to establish causal connections between factors and their actual ‘impact’ on actor relations. Instead, basing our analysis on respondents’ interpretation of their relations, we reflect on the potential role of factors from the viewpoint of those directly involved in the local governance of refugee integration.

Empirical findings

In the Netherlands, refugee integration ‘is an established and highly institutionalized process’ (Larruina, Boersma, and Ponzoni 2019, 55) in which municipalities have specific competences, derived from national legal frameworks. First, under the Housing Act (2014), municipalities must provide housing for refugees assigned to them by the national Central Agency for the Reception of Asylum Seekers. Second, under the Civic Integration Act (2013), municipalities are required to offer ‘social support’ to refugees while they follow their mandatory civic integration trajectory. In 2022, a new Civic Integration Act was implemented, giving municipalities a more central role in the governance of refugee integration. Third, under the Participation Act (2015), municipalities must assist welfare benefit recipients, including refugees, with their labour market (re-)integration.

Local actors and collaborations

We find that municipalities typically outsource responsibilities to other actors to carry out tasks related to ‘housing’, ‘social support’ and ‘employment’, on behalf of and in collaboration with local administrations (cf. Ambrosini 2021). We also identify informal actors supporting refugees without receiving dedicated funding by public bodies. Table 2 summarises the main actors that we identified in the four localities.

Table 2. Main local actors in integration policymaking.

Actor	Service	Locality A	Locality B	Locality C	Locality D
Formal actors	Social support	Local CSO (one-stop-shop)	Local (generic) welfare organisation	Local specialised CSO	National Dutch Council for Refugees
	Housing		Housing corporation	Housing corporations	Housing corporation
	Employment		Regional semi-public body	Municipal work-coaches and regional partners	Municipal dedicated client manager
Other (informal) actors	Informal language education, social activities	Volunteer network, 60–80 actors	Library	Welfare organisation, small CSO	National voluntary organisation (Humanitas), Library
Networks		Network Integration			Multifunctional neighbourhood centre

In locality A, the local government has a formal cooperation agreement in place with a well-established local CSO which has been assigned the task of ‘social support’ and also offers language courses, assistance with employment and finding accommodation, functioning as ‘one stop shop’ for refugees. The local government decided transferring the responsibility for (formal) ‘integration’ to one organisation only, creating a single contact point for refugees to receive comprehensive support, having a ‘better chance of it not going astray’ (N-A-1). With the new Civic Integration Act (2021), the CSO became a public body, changing its subsidy relation with the municipality into a horizontal collaboration (N-A-8, N-A-12). The local government also founded a ‘Network Integration’ in 2019, coordinated by the local welfare organisation, bringing together formal and informal actors, including churches, mosques, migrant-led organisations and a buddy project. The network represented a temporary exchange forum in which local actors were invited to provide input. Notably, respondents emphasised the diversity of informal actors supporting newcomers in town, with numbers ranging from 60 to 80. Finally, there is a group of 200 volunteers that has been mobilising support for refugees since 2015 and an alderman and mayor proactively shaping a positive narrative on migration (N-A-8). Locality A’s actor constellation is marked by a close, horizontal collaboration between one well-established CSO (now a public body) and the local government, with a (temporary) municipal network inviting other non-public actors to participate, albeit informally.

Contrary to locality A, the local government in locality B has contractual agreements with multiple actors to facilitate refugee ‘integration’, instead of bundling them under one roof. The main collaboration partner is a local welfare organisation implementing the task of ‘social support’. A member of the local government explains the strategic choice to collaborate with a ‘generic’ organisation that is locally embedded and offers its services to all residents because ‘if they organize a meal in a neighbourhood building, then that is not a meal in the context of integration, but a meal for the neighbourhood and simultaneously also an activity to integrate’. The local government has similar agreements with the local housing corporation (to provide housing for refugees) and with a regional semi-public implementing organisation (for labour market (re-)integration) (N-B-2, N-B-6). All three actors focus on broader target groups, refugees being one of them, reflecting the municipality’s integral approach to refugee integration. The municipal policy coordinator invites these actors to regular meetings; yet, beyond these meetings, there appears to be little formalised collaboration between them. A crucial informal organisation is the local library offering language support, practical help and information to newcomers (N-B-5). Locality B’s actor constellation is small and dense, but rather fragmented due to the clear task division. Actors work independently next to each other on their own area of expertise and have little to no overlapping responsibilities. Primarily relying on task delegation and coordination, the municipality itself had – until the new Civic Integration Act – a rather passive role.

In municipality C, the local government has delegated the task of ‘social support’ to a local CSO, founded in 2015 by volunteers (N-C-15). The CSO also offers informal language support to refugees, a service which is not funded by the municipality. The CSO thus occupies two roles – one as implementer of a legal task and formal contracted collaboration partner of the municipality and one as informal actor providing additional assistance to newcomers. The collaboration between the CSO and the local government appears transactional with the government providing funding and the CSO carrying out the task, with occasional possibilities to provide feedback. Like municipality B, the local government has a contractual agreement with housing corporations responsible for finding accommodation for refugees. On the topic of labour market (re-)integration, the municipality hired municipal ‘work-coaches’, while also collaborating with other municipalities and the public Employee Insurance Agency (N-C-5). Besides these formal organisations, there are a few other actors that provide services to newcomers, such as the local welfare organisation coordinating language activities in the library. Like locality B, the actor constellation in locality C involves few actors and is rather fragmented, with only little collaboration taking place between CSOs. The municipality has taken on a passive role, focusing primarily on task delegation, and leaving much of the responsibility with the local CSO contracted to provide social support to refugees.

Contrary to the other localities, the local government in locality D delegated the task of ‘social support’ to the national Dutch Council for Refugees. While their collaboration is formalised, exchange happens only occasionally, presumably because the Dutch Council for Refugees does not have a continuous presence in town. After 18 months, the task of assisting refugees is transferred to the local welfare organisation. Like localities B and C, a performance agreement with the local government defines the local housing corporation’s responsibility to find accommodation for refugees. Regarding employment, the municipality chose a traditional in-house service delivery through municipal staff (Genugten, Marieke, and Voorn 2020). They hired specialised ‘client managers’ to facilitate refugees’ labour market integration ‘more efficiently’ (N-D-9). Moreover, the municipality has an active alderman promoting a ‘social approach’ (N-D-15) to integration and seeking dialogue with ‘concerned’ residents. Informal actors include a national voluntary organisation offering language lessons to migrants, regardless of their status, and the local library providing language support to newcomers. Notably, these public and non-public actors, including a local language school, are located in a multifunctional neighbourhood centre, facilitating informal exchange and making services accessible for those living in the municipality’s main town. Locality D’s actor constellation is small, personalised and consolidated, marked by an intensive collaboration between local government, welfare organisation and local housing corporation, with the goal to ‘find joint solutions for arising issues’ (N-D-2). The municipal client managers and alderman proactively support refugees and seek to facilitate exchange between local actors and residents.

When analysing actor constellations across localities, we observe similarities and differences. In line with our first expectation, we see that ‘local associational ecologies’ (Semprebon, Marzorati, and Bonizzoni 2022, 9) are relatively small (especially in the three smaller towns B, C and D) and some actors perform several tasks at the same time. We see local governments collaborating closely and rather permanently with a small number of locally operating contracted actors such as housing corporations, welfare organisations and specialised CSOs, covering the formal ‘integration dimensions’ housing, social support and employment. Together, they constitute a relatively hierarchical and stable ‘service delivery network’ (Prentice, Imperial, and Brudney 2019) with local governments taking up an influential position due to their authority to delegate tasks and control funds (Schiller et al. 2023). Besides governments, local CSOs responsible for social support’ occupy a central role in local governance networks, functioning as first contact point for refugees. There are also various informal actors involved, providing additional support for refugees, including administrative support, informal language education or social network building (Larruina, Boersma, and Ponzoni 2019). Notably, in the smaller municipalities, the field of CSOs is significantly smaller and less diverse, and informal or emerging (volunteer) initiatives appear rather absent.

We can identify two patterns of collaboration. First, in localities B and C, actor networks are small and personalised, yet rather fragmented because local governments interact primarily ‘bilaterally’ and ‘transactionally’ with a limited number of – mostly contracted – local actors ‘with the main goal of delegating tasks or outsourcing services, particularly related to their specific policy priorities’ (Caponio and Pettrachin 2023, 13) Therefore, actors work rather independently of each other (except for some informal interactions) and are only occasionally brought together by local governments. Second, in localities A and D, local actor networks seem more consolidated and collaborations between public and non-public actors happen more frequently, also beyond formally assigned tasks. Local governments take on an active role in these collaborations (cf. Caponio and Pettrachin 2023), trying to promote participation and exchange between formal and informal actors (e.g., via a city-network for integration or a local neighbourhood house) (Gaventa 2006).

Conflictual interactions

Relations between local actors in the field of refugee integration are marked by collaboration, but also by conflictual dynamics (Campomori and Ambrosini 2020). The first identified pattern of conflict occurs between public and non-public actors (local authorities, CSOs and volunteer groups) that are (in)formally involved in the governance of refugee integration. Notably, this pattern seems to be significantly shaped by specific collaborative structures in place.

For instance, in locality A, the close collaboration between the municipality and the CSO serving as one-stop-shop led to conflicts because other non-public actors felt excluded from participating in formal policymaking processes. Due to its ‘monopoly’ position, the CSO was perceived as ‘gatekeeper’ by less established organisations which found it difficult to receive (financial) recognition by the local government. This led to frustration in the past, especially for actors adopting a different approach to helping newcomers. Similar tensions emerged within the ‘Network Integration’ due to differing ways of working, expectations and motivations as well as unequal distribution of resources and forms of inclusion among formal and informal actors (N-A-3, N-A-6).

Conflictual dynamics in locality A occurred primarily between the well-established CSO (representing the municipal approach) and informal actors, including volunteers, complementing the official service provision. Conflicts seem to stem from power imbalances and diverging approaches to ‘integration’ among local actors, with one CSO occupying a central position. The ways in which different actors negotiate integration policymaking are informed by their own interests, values, positionalities and frames of ‘integration’ (Campomori and Ambrosini 2020; van Hulst and Yanow 2016). Such tensions were not ignored but acknowledged and addressed. For example, the coordinator of the ‘Network Integration’ remembered acting as a ‘translator’ between formal and informal actors to foster more understanding. While not all tensions were resolved, discussions signalled strong civic engagement and led in some cases to increased communication and collaboration among actors, showing that conflicts related to power imbalances or diverging values do not necessarily turn into a ‘battleground, but can create participatory spaces and new alliances (Gaventa 2006; Wolf and Van Dooren 2018).

In locality C, conflicts emerged between the local government and the CSO offering ‘social support’ for refugees. This presents an interesting case because collaboration as well as conflict emerged between the same actors showing that relationships are dynamic, and collaboration can turn into conflict (and vice versa). Although the CSO acts as a formal collaboration partner for the municipality, it is also critical of the municipality’s approach. According to the CSO’s coordinator, the municipality is rather passive and restrictive, and people are often not given the chance to pursue their aspirations but are pushed to work in the low-paid sector on temporary employment contracts. Moreover, some political leaders adopted an anti-immigrant discourse, framing migration as an economic burden and ‘cultural threat’ (N-C-2) and disengaging with the topic by paying little attention to it (cf. Wolf and Van Dooren 2018), leaving much of the responsibility to the CSO. Consequently, conflicts are related to actors’ strongly contrasting values, interests and narratives about migration. As a response, the CSO set up a language programme with the help of 30 volunteers. Instead of an open conflict, tensions led to the creation of additional support structures (cf. Noordegraaf et al. 2019). It appears that the one-sided government approach resulted in the fragmentation between actors but also spurred civic engagement, balancing out the somewhat exclusionary municipal approach (Campomori and Ambrosini 2020).

In the small town in South Holland and the rural area in Drenthe (localities B and D), tensions arose primarily between local authorities and anti-immigrant residents. These dynamics have taken various forms, including public resistance against the arrival of refugees – also through violent acts against property (locality D) –, anti-immigrant comments on social media or individual complaints by residents towards political leaders. Such anti-immigrant movements are often met with solidarity movements welcoming refugees, including local governments taking a proactive stance. According to a member of the local government in locality B, there is both strong aversion towards migrants, supported by the presence of right-wing populists in the municipality, and a socially engaged movement welcoming migrants and refugees because ‘mercy and charity are key concepts’ in the traditionally ecclesiastical Christian town.

Respondents in localities B and D mentioned negative narratives about newcomers shaping residents’ perceptions, relating to the priority that is given to refugees by local housing corporations, but also to perceived cultural differences. Such anti-immigrant sentiments also demand a reaction from local governments. In locality B, one respondent explained that residents’ negative attitudes prevented the government from paying attention to the topic (N-B-1); others argued that local protests were one driver for the municipality to set up a communication campaign to proactively counter negative narratives (N-B-2). Similarly, the local authority in locality D did not adopt an anti-immigrant rhetoric but initiated neighbourhood dialogues to address citizens’ concerns. Conflictual dynamics seem to have led to a reaction from local authorities trying to address residents’ concerns, while designing strategies for refugee integration in their locality.

Comparing across localities, we observe various ‘battles’ between actors despite the rather small and dense actor networks, in contradiction with expectation two. The most pronounced tensions were visible in the medium-sized town where significantly more non-public actors are involved. Taking a comparative perspective, in two localities (A, C), frictions primarily emerged between state and non-state actors as well as between non-public actors differing in their formalisation, legal autonomy and (funding) relation to local governments. Reasons for tensions included conflicting values and interests, leading to disagreements over which approach works best, what type of services should be provided, for whom and by whom (Campomori and Ambrosini 2020; van Hulst and Yanow 2016). Some tensions were heightened due to power asymmetries within local governance networks, reflected in differences in budget, recognition and responsibility given to actors by local governments (Glorius et al. 2019; Schiller et al. 2023). Interestingly, we also observed that power asymmetries and the selective inclusion of actors in formal governance structures may lead to tensions between actors holding similar ideas and (political) values. Moreover, we find that local governments’ dominant’ discourses on migration and chosen strategy appear to be influential in shaping (conflictual) actor dynamics as they significantly shape power (im)balances between actors (cf. Schiller et al. 2023).

The other localities (B, D) were confronted with anti-immigrant sentiments among local populations expressing their concerns towards the arrival of refugees, creating challenges for local actors trying to navigate between residents’ concerns and newcomers’ needs (Glorius, Bürer, and Schneider 2021; Meijer, Zuzanna Popławska, and Szytniewski 2023). Here, conflictual dynamics emerged primarily among local populations and less between formally involved actors. Reasons for tensions included perceived cultural differences and diverging narratives about refugees between residents and governments.² Importantly, we do not want to overstate the existence of such ‘battles’ and argue that the four localities seem nonetheless characterised by rather collaborative modes of governance.

Factors explaining similarities and differences

Similarities across localities may be attributed to existing national regulations defining that all municipalities must provide housing and social support to refugees. However, local authorities have some leeway in the implementation of such tasks (Oomen et al. 2022), leading to local differences, for example, in the choice of collaboration partner (local or national), and the allocation of responsibility and budget.

While we observe collaborative and conflictual dynamics in all four municipalities, the intensity, regularity and ‘nature’ of existing collaborations and conflicts differ. There are various contextual factors explaining differences – and similarities – between the case studies. First, localities’ size seems to play an important role in shaping actor interactions. In line with existing literature on refugee integration in small towns, we find a limited number of actors in the two small towns (B, C) and the rural area (D). Relatedly, actor constellations are marked by dense networks, a high degree of personalisation and ‘short lines’ between public authorities and non-public actors, including residents (Haselbacher and Segarra 2022; Schammann et al. 2021; Schneider 2022; Semprebon, Marzorati, and Bonizzoni 2022). Contrary, in the medium-sized town (A), there is a significantly larger variety of actors involved, resulting in a more diverse landscape of service provision. It seems that this larger and more diversified network has resulted in complex patterns of interaction, with more actors claiming a ‘seat at the table’ and smaller actors feeling excluded, despite the local government’s intent to create spaces for participation (cf. Wang and Ran 2023).

Second, local governments’ political orientation and political leadership, especially that of the responsible alderman, seem to explain the emergence of specific actor interactions. In localities A and D, the aldermen responsible for integration take a proactive approach to integration, presumably because of their centre-left political orientation (N-A-8, N-D-11). This orientation seems to explain local governments’ active role, their close collaboration with CSOs, a fairly consolidated actor network and more participatory spaces (cf. Caponio and Pettrachin 2023) – but it does not prevent conflicts. The role of political orientation and leadership also becomes apparent in locality C, albeit in an opposite manner. According to a local official, integration is ‘not a topic to win elections’ (N-C-2), resulting in the de-prioritisation of the topic. The centre-right orientation, combined with the presence of an RRP, seems to (partially) account for the government’s rather passive role and transactional relation with the contracted CSO, which takes an inclusive approach and is more proactive in creating services on a needs-basis than the alderman (cf. Caponio and Pettrachin 2023).

Third, residents’ attitudes towards migration appear to influence actor relations, especially when expressed in the form of movements supporting or resisting the arrival of refugees. For example, in locality A, residents have a rather positive attitude towards migration, reflected, for instance, in an active volunteer group and a broad support for the municipality’s progressive approach. In locality B and D, residents visibly voiced their reservations towards newcomers, challenging the municipal approach and demanding governments to react to residents’ concerns (Larruina, Boersma, and Ponzoni 2019; Schammann et al. 2021). Notably, while resentments towards newcomers seem to be prevalent in the two localities with a rather ‘homogeneous’ local population (B and D; cf. Huijsmans et al. 2021), the local population in locality C similarly holds rather negative views, despite its long migration history and ‘diverse’ population. This presumably relates to a negative discourse on migration adopted by local political parties, centring on the ‘failed’ integration of migrant communities and the town’s limited capacity to bear the ‘economic burden’ (N-C-6).

Finally, our findings point to the importance of local governments’ decision-making and interpretation of the issue – reflected in the distribution of responsibilities and resources to CSOs for carrying out legal tasks – in explaining variations across localities: for example, in locality A, the municipal decision to put one CSO in charge has clear implications for the dynamics between the government, the CSO and other non-public actors – both in terms of participation and collaboration as well as emerging tensions. In locality B and C, municipal decisions to distribute tasks across actors have led to a fragmentation of provided services, possibly hindering collaborations between actors.

Conclusion

Our aim was to examine actors’ roles in local refugee integration governance and better understand collaborative and conflictual relations in small and medium-sized towns. In analysing responses to refugee integration in four Dutch localities, we find that the number of actors is limited in smaller localities, but nonetheless various actor types are involved in local governance processes. Local authorities outsource governmental tasks related to refugee integration to primarily local non- (or semi-)public actors, covering the ‘integration dimensions’ housing, social support and (to some extent) employment. Besides delegating and supervising tasks, public authorities also take on the role of ‘partners, sponsors and supporters’ (Nyseth and Ringholm 2008, 471) in local service delivery networks marked by the dispersion of responsibility, but also power asymmetries (Prentice, Imperial, and Brudney 2019; Schiller et al. 2023). Actors’ responsibilities, positions and influence vary, with some actors being responsible for multiple tasks and occupying more central positions than others. Next to formal organisations carrying out tasks for municipalities based on contractual agreements, various informal actors play an important role in offering additional support to newcomers. Other than expected, we found that despite the involvement of fewer actors and dense networks in small towns, actor relations are also marked by conflictual dynamics. Conflicts have often not led to a rift between actors, but to a questioning of municipal approaches and the creation of complementary services by (informal) non-public actors trying to balance out exclusionary dynamics derived from passive local governments and/or selective formal infrastructures. While some actors operate side-by-side municipalities ‘with their own agendas, missions and ambitions’ (Meijer, Zuzanna Popławska, and Szytniewski 2023, 18), others have formed coalitions with local authorities to respond to the task of refugee integration. Local integration governance in small localities thus is marked by collaborative ‘modes of governance’ as well as conflicts, showing that these dynamics (can) exist simultaneously and with differing degrees of intensity.

Overall, we identified two collaborative patterns: one more consolidated, with active governments and regular exchange between actors, and the other one more fragmented, with rather passive governments and limited interaction between (in)formal actors. We further observed two conflictual patterns, one between ‘institutionalized’ actors and the other between local governments and residents (holding anti-immigrant attitudes). Through a ‘local-to-local comparison’ (Glorius et al. 2019), we identified various factors which help understand these varying patterns. First, localities’ small size seems to relate to a limited and more homogeneous actor constellation, marked by proximity between state and non-state actors. A larger and more diversified actor network seems to result in more complex interactional patterns. Second, we identified political orientation and leadership as factors explaining municipalities’ activity (or passivity) regarding the topic of integration and its way of interacting with local actors, with centre-left governments engaging in closer collaborations and taking on more active roles (Caponio and Pettrachin 2023). Third, residents’ attitudes towards migration shape actor relations, often reflected in emerging tensions, for instance, when anti-immigrant movements encounter progressive governments. Finally, our analysis points to local governments’ (still) significant role in collaborative governance processes: they shape participatory spaces (Gaventa 2006), by setting up structures through their decision to involve certain stakeholders (and exclude others) and by framing ‘refugee integration’ in a particular way and responding to it accordingly. Their adopted strategy – and decision to share their power (or not) – significantly shapes local dynamics of conflict and collaboration.

Based on these insights, we formulate several takeaways for our theoretical understanding of collaborative local governance in small and medium-sized towns. First, even though the field of actors is smaller, local authorities are still challenged to choose with whom they collaborate with, and on what terms, to govern refugee integration in small localities. They engage in collaborative governance processes, reflecting a general trend in which governments increasingly organise their service delivery at ‘arm’s length’ (Genugten, Marieke, and Voorn 2020). Second, our analysis of factors explaining patterns of interaction provided nuanced insights into local integration governance configurations which are neither a strictly conflict-driven ‘battleground’ (Ambrosini 2021) nor characterised solely by intensive cooperation (Schiller 2024). We show that collaborative governance arrangements are also shaped by conflicts, mostly related to existing power asymmetries as well as differing frames, interests and capacities among different actor types (Scholten 2020). Notably, we observed that (political) alignment on an overall goal does not always prevent tensions, especially when actors do not receive adequate recognition and resources. Considering that conflicts may weaken and alienate (non-state) actors and impede effective policy responses (Pettrachin 2024; Schiller 2018), our findings highlight the importance of creating inclusive participatory spaces where such tensions can be negotiated, potentially leading to the formation of new coalitions. Given their authority and power, political leaders and municipal administrations have a key role to play here.

We underline the importance of exploring localities’ ‘situational responsiveness’ to better understand governance dynamics in small and medium-sized towns in the Netherlands. At the same time, also the Dutch national framework plays a certain role in stimulating some form of collaboration between local authorities and non-public actors as it defines clear responsibilities for municipalities. We thus propose that future research should explore how actor relations unfold in small localities in other national contexts while also assessing the relevance of different contextual factors in explaining differences and similarities across cases.

Ethics approval

We have received approval to conduct our research by the DPAS Research Ethics Review Committee of the Erasmus University Rotterdam. The respective approval number is ETH2021–0079.

Acknowledgments

The authors would like to thank the anonymous reviewers for their insightful and constructive feedback on an earlier version of this article.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the European Union’s Horizon 2020 research and innovation programme under grant agreement No 101004714.

Notes

1. We focus on integration governance targeting recognised refugees with a residence permit (henceforth: refugees) living in designated municipalities. We understand ‘integration’ as policy term referring to a bundle of national policies aimed at regulating the ‘integration’ of refugees, namely the Housing Act, Participation Act, and Civic Integration Act, with specific tasks for local governments. We analyse state and non-state actors taking on formal roles in carrying out these tasks or supporting refugees informally.

2. Importantly, this section discusses the most prevalent conflicts in the localities. Arguably, smaller forms of resistance also appeared in locality A and some tensions were also felt between non-public actors in locality B.

3. Data derived from the national statistical office Statistics Netherlands (CBS).

4. For anonymisation purposes, the exact number of residents is not disclosed.

5. Statistics Netherlands (CBS) defines a person with a migration background as a ‘person of whom at least one parent was born abroad’ and differentiates between persons with a western migration background and persons with a non-western migration background. The latter refers to persons ‘originating from a country in Africa, South America or Asia (excl. Indonesia and Japan) or from Turkey’ (CBS, 2024).

6. Data derived from (CBS, 2021).