Inter-Organisational Entanglements in Migrant Support Ecologies: Action and Collaboration Supporting Labour Market Integration

This paper examines inter-organisational behaviours in what we call ‘migrant support ecologies’ – shared physical and abstract spaces where multiple organisations work to help migrants access and transition within the labour market. Drawing on composite data generated through studies conducted in the UK and Brazil, we argue that actors and organisations in such environments operate in ‘common goal domains’, in which objectives are related but not necessarily integrated or coordinated, and they consequently adopt diverse interactional practices. We distinguish between four ideal types of migrant support organisation based on their activity scope and stakeholder focus before outlining how different organisations and their constituent actors engage in tactical and strategic coupling practices, reflecting shorter and more episodic interactions alongside complex, multithreaded ones. The findings show how different forms of cooperative arrangements may be pursued based on organisations’ capacities, focus and the types of value they seek to create for organisations, migrants and wider societies.


Introduction
A disparate range of organisations and their constituent actors influence migrants' transition into and through the labour market (Calò et al., 2021;Grosskopf et al., 2021;Siviş, 2021). Various informal networks, formal organisations and institutions including charities, branches of faith-based bodies, social enterprises and non-governmental organisations (NGOs) help to develop migrants' general employability and facilitate their access to work (Aerne & Bonoli, 2021;Bagavos & Kourachanis, 2021;İçduygu & Diker, 2017;Siviş, 2021). They provide direct support, for example, by securing voluntary work to build migrants' resumes, skills and confidence and by mediating their entry into employment (see, e.g. Dykstra-DeVette & Canary, 2019;Morano-Foadi et al., 2023). They also provide indirect support, advising on legal rights, skills development and peripheral administrative tasks, which facilitates migrants' labour market transition by removing barriers (ibid.).
Much of the research on supporting migrants' labour market entry and transition has focused on the activities of individual organisations and interactions between frontline service providers and their 'clients' (Dykstra-DeVette & Canary, 2019;Morano-Foadi et al., 2023;Nardon et al., 2021;Tomlinson & Egan, 2002). Many studies refer to collaborations between private, public and thirdsector organisations (TSOs) (Bagavos & Kourachanis, 2021;Calò et al., 2021;İçduygu & Diker, 2017;Siviş, 2021). However, they do not attempt to analyse in sufficient detail the nature of these interactions or collaborative arrangements to understand their forms or the contextual factors that shape their formation. Consequently, there is inadequate knowledge regarding inter-organisational arrangements, and there have been growing calls to consider the roles of wider networks of organisations, including the factors shaping their interactions (Finsterwalder, 2017;Hesse et al., 2019;Lee et al., 2020). This is in recognition of the multiple tensions that shape their activities and engagement, for example, the diversity of institutional logics followed by state, commercial and charitable organisational actors (Hesse et al., 2019;Maletzky de García, 2021). Richardson et al. (2020) urged researchers to adopt an 'ecosystems' perspective to examine how different organisations can and do cooperate to support refugees. Conceptions and invocations of ecosystems, generally in organisation and management studies (Adner, 2017;Baruch & Altman, 2016;Jacobides et al., 2018), specifically in service research (Chandler et al., 2019;Kuppelwieser & Finsterwalder, 2016) and in relation to supporting migrants and refugees (Finsterwalder, 2017;Finsterwalder et al., 2021;Kornberger et al., 2018), recognise that these may involve loose associations among organisations. Nevertheless, they continue to characterise support ecosystems as coordinated, cooperative arrangements among interdependent organisations and actors (Lusch & Vargo, 2014;Vargo et al., 2015). Ecosystems are thus applied in these studies as a normative model. However, adopting this framework uncritically risks overstating the extent to which these organisations are, in practice, aware of the presence and capabilities of other organisations, and cooperate in a systematic and coordinated manner to fulfil shared objectives. In short, ecosystems may represent laudable attempts to prescribe how things should be, but do not adequately capture how things actually are.
We recognise these tensions while responding to calls to examine how multiple organisations support migrants' labour market transition (Lee et al., 2020). Consequently, we seek to examine inter-organisational dynamics in what we call 'migrant support ecologies' -spaces where multiple organisations operate to create value for migrants to support their access to and transition within the labour market. The key research questions for this study are (1) 'what activities do organisations and actors perform to support migrants' labour market integration?'; (2) 'what type of interactive arrangements do organisations and actors engage in to support migrants' labour market integration?'; and (3) 'why do those activities and arrangements operate in particular contexts?' Moreover, in contrast to adopting ecosystems as a normative model and assuming there is mutual awareness, coordination and cooperation, the term support ecologies is thus introduced as an analytical framing to help conceptualise the support(ive) practices performed by organisations and actors individually, in relation to and sometimes in conjunction with others.
We thus advance existing knowledge by examining the coexistence and interaction between service providers and operators of organisations that support migrants' labour market integration. More specifically, we propose the notion of inter-organisational entanglements to conceptualise (a) contextual conditions and dynamics that influence service providers' actions and (b) specific factors that drive different forms of coexistence and interaction. This contributes to knowledge in two ways. Firstly, more specifically in relation to migration and integration, it helps to understand factors shaping the scope and nature of interaction and cooperation between service organisations and their constituent actors (e.g. employees, volunteers and affiliate members), which play a key role in supporting migrants. Secondly, more generally for organisation studies, studying migrant support services helps to conceptualise the dynamic, messy nature of organisational coexistence, which simultaneously involves resource competition, functional specialisation, pliable and selective cooperation and moral obligation that potentially shape individual and collective practices within inter-organisational arrangements.
It is important to clarify that the paper refers more generally to migrants, but the organisations we examine often concentrated much of their efforts and resources towards refugees. The reason for retaining the term migrant in this paper rather than using refugee exclusively is that the organisations being studied often engaged with a variety of migrants. They included economic and lifestyle migrants, alongside people who may have been coerced into migration, and those displaced by other forms of human or natural disaster, whose refugee status was contested (Havard, 2007).
Finally, it is necessary to stress at the outset that we adopt a supply-side focus, rather than attempting to include the demand perspective of migrants. We analyse the organisations and the constituent actors who provide services, and engage in value creating practices for, with and on behalf of migrants to support their transition into and through the labour market. This supply-side orientation enables us to (a) concentrate on examining why organisations and actors adopt different types of collaborative arrangements in migrant support ecologies and (b) distinguish between tactical and strategic forms of coupling, reflecting episodic and multithreaded engagement.

Conceptualising Migrant Support Ecologies
Central to the proposed conceptualisation of 'inter-organisational entanglements' is the ecological context in which organisations and their constituents operate (Adkins et al., 2007;Tsoukas & Dooley, 2011). This refers in part to the geographical location of organisations in relation to their 'clients' and other stakeholders, including organisations involved in competing or complementary service provision. However, ecology also refers to a more abstract geographical landscape, which includes other organisations and stakeholders involved in the same domain of activity or driven by similar goals, though not operating in close physical proximity (Bennouna et al., 2019;Garkisch et al., 2017). For commercial entities, this refers to companies engaged in analogous business ventures or functional specialisms. However, for social-oriented organisations, including charities, state-run service providers or social enterprises, this refers to entities that pursue similar moral or ideological goals, for example, concerning solidarity, empowerment or justice (Lawrence & Hardy, 1999). This also includes organisations driven by wider political agenda, particularly state actors seeking to govern and exercise power (Hesse et al., 2019;Maletzky de García, 2021).
Many contemporary writers invoke the notion of 'service ecosystems' to conceive and study interactional ecologies and their dynamics in service-related value creation (Chandler et al., 2019;Kuppelwieser & Finsterwalder, 2016;Vargo & Akaka, 2012). This extends to service research on migrants and refugees (Finsterwalder, 2017;Finsterwalder et al., 2021;Kornberger et al., 2018;Richardson et al., 2020). They conceptualise ecosystems as 'relatively self-contained, self-adjusting system[s] of resource-integrating actors connected by shared institutional logics and mutual value creation through service exchange' (Lusch & Vargo, 2014, p. 161). Resource integration is viewed as processes through which actors identify, assemble, attempt to gain control over and mobilise tangible and intangible resources to create multiple forms of value (Bruce et al., 2019). More importantly, within an ecosystems perspective, resource integration is seen to coincide with the institutionalisation of these ecosystem arrangements, reflecting increasingly formalised, coordinated and sustained arrangements (Vargo et al., 2015).
The invocation and adoption of ecosystems among service researchers are understandable, because these authors promote a 'transformative agenda', which seeks to design, drive and manage greater levels of interaction and cooperation among organisations to achieve positive social, moral, economic and health outcomes for migrants and refugees Finsterwalder et al., 2021;Nasr & Fisk, 2019). However, using the term 'service ecosystems' risks overstating several qualities including 'self-containment' (i.e. it is a stable, logically and clearly defined entity), 'connectivity', 'mutual value-creation' and 'shared institutional arrangements' suggesting awareness, goal and resource alignment and cooperation (Vargo & Lusch, 2016). Entanglement is thus used here as an evocative lexical device -a 'sensitising concept' (Blumer, 1954) -to stress that relations and interactions may take multiple, shifting forms across dynamic spaces, actors and practices.
Invoking the notion of 'entanglement' acknowledges that actors' coexistence and practices may be less ordered, interactive or integrated. More significantly, and perhaps unsettlingly for academics, adopting ecologies and entanglements reflects willingness in this study to accept the potential for mess. Following Law (2004), this position recognises that the world as an empirical context is 'vague, diffuse or unspecific, slippery, emotional, ephemeral, elusive or indistinct' (p.2). This study examines a dynamic amorphous phenomenon and therefore tries to avoid reducing it to a patterned, interconnected (eco)system with clear or fixed boundaries. In other words, it accepts the potential for indeterminacy and incompleteness in how it conceives a phenomenon and therefore how it accounts for its components, processes and outcomes empirically (Law, 2004;Tsoukas, 2017).
In this study, organisational ecology is approached as complex, unfolding, openended and ambiguous (Cilliers, 2000;Stacey et al., 2000;Tsoukas & Dooley, 2011). Organisations and actors may operate without full or even partial awareness of others in the same ecology, despite serving or engaging the same stakeholders or pursuing analogous objectives. This view does not reject that organisational actors may be aware of each other or come into contact, but it does not take as a core assumption they do, or that their interactivity substantially reflects interdependency. Interactions and therefore collaboration among organisational actors may be domain-, project-or activity-specific, accidental, loosely defined or governed, and constantly subject to coupling and decoupling. However, despite these ambiguous and potentially mobile forms of coexistence and interaction, these organisations may (a) create value and/or provide services for the same stakeholder community or communities and (b) pursue the same or at least similar goals.
Actors' activities and practices, informed by various factors including resource constraints, reflexivity, social idiosyncrasy and competing expectations regarding the value to be created, translate into organisational practices (Danış et al., 2019;Hesse et al., 2019;Maletzky de García, 2021). However, because actions and practices are subject to (re)interpretation and disruption, there is scope for unpredictability, serendipity and discontinuity. These ongoing tensions between continuity and change influence the scope and focus of inter-organisational interactions. The challenge for the current study is to examine how these dynamics operate in, and thus help to understand, inter-organisational entanglements among actors and entities seeking to create value with and for migrants and refugees.

Conceptualising Migrants' Labour Market Integration as Value Creation
It is important to acknowledge the role and use of value (creation) in this study. Previous work on inter-organisational arrangements has used alternative concepts such as individualised/common benefits (Khanna et al., 1998) and agreed goals (Castañer & Oliveira, 2020) in studying their dynamics. However, the emphasis in this study is on value, partly because this extends beyond benefits for organisations, but also because goals may not be singular, agreed upon or stable across a complex network of actors.
It is useful to view value in this context as being able to assume multiple forms, simultaneously, depending on how it is created and perceived (Heinonen et al., 2013;Lugosi & Ndiuini, 2022;Vargo et al., 2008;Zeithaml et al., 2020). Specifically, a social enterprise or service organisation working with migrants and refugees can continue to realise or extract direct and indirect economic value from these interactions (Harima & Freudenberg, 2020;Harima et al., 2021). For example, providing specialised services can be used to secure state funding, while supporting migrants to show corporate responsibility can also help create brand value for commercial entities. However, working with and for migrants in improving their psychological empowerment and financial independence by facilitating their labour market integration can also create transformative value for immigrants and wider society (Blocker & Barrios, 2015;Gross et al., 2021;Lugosi et al., 2023).
Empirically, the phenomenological and subjective nature of value makes it difficult to account for its existence as something objectified and singularised. Value is multidimensional and created 'in-use' (Heinonen et al., 2013;Zeithaml et al., 2020). Moreover, the aim here is not to describe what types of value different organisations sought to create. Nevertheless, conceptually, recognising the diversity of value creation (as process) and value (as outcomes) in migrant support ecologies helps to appreciate how and why organisations operating in these domains try to create and extract different forms of it. Moreover, acknowledging the multifaceted-ness of value also helps to understand why and how inter-organisational engagement in these contexts can function in pursuit of value creation.

Research Context and Approach
We synthesise data from two empirical projects conducted in the United Kingdom (UK) and Brazil over 4 years. The rationale for the locations is twofold. Firstly and foremost, the author team is based in these countries, and the data for the current paper were collected through two collaborative studies, which were built on longer, ongoing engagement in these field settings with various project stakeholders. Secondly, more generally, both countries have long, complex histories of migration that extend to the current period, and are therefore rich sources of data (Bloch, 2002;Lesser, 2013;Wejsa & Lesser, 2018;Yeo, 2020).
These were not designed to be comparative studies, and the aim is not to conduct comparative analysis. Nor did they seek to develop symmetrical data sets. Nevertheless, both studies examined the practices of diverse organisations involved in migrants' labour market transition, and consequently, there was conceptual convergence in the studies and in the data they generated. The first project focused primarily on refugees and support organisations and actors based in one county in the UK; the second project considered similar organisations and actors in São Paulo, Brazil. Both studies sought to identify and assess (1) practices that various individual and organisationally embedded actors engaged in to support migrants' or refugees' transition into and through the labour market; (2) different forms of value they created for diverse actors through those practices; (3) challenges they encountered; and (4) strategies and tactics that actors deployed in exercising their agency to address those challenges. Similar to Jung and Buhr's (2022) multisited research, studying organisations, actors and their practices in different contexts enabled us to appreciate the dynamics of analogous phenomena without conducting a comparative analysis.
The studies adopted a pragmatist position (Kelly & Cordeiro, 2020) in seeking to capture and interpret actors' behaviours, experiences and perceptions. Both studies followed a purposive sampling strategy that combined criterion and snowball approaches (Patton, 2015). The essential inclusion criteria were that the people, and their respective organisations, were involved in refugees' and migrants' labour market transition. However, this was expanded to include a wider set of actors, for example, entrepreneurs who were also engaged in value creation and capture, in cooperation with migrants and refugees, but not necessarily as service providers in a narrow sense. This was seen as essential because the emergent findings and literature showed that entrepreneurship was a key pathway to employment and economic empowerment for various migrants (de Lange et al., 2021;Lugosi et al., 2023). The aim was to identify 'intensive', data rich 'cases' (Patton, 2015) to better understand the phenomena of inter-organisational arrangements.
Importantly, in both studies, data collection and analysis were performed simultaneously rather than sequentially (Lofland, 2006). This informed the data collection and sampling, insofar as new respondents, organisations and issues could be incorporated into the study's sample and data collection because of their relevance and apparent interest. Moreover, the studies' scope and focus meant the findings they generated were assembled into a 'composite data' set, i.e. created for different projects, but having common features enabling the researchers to synthesise them in their analysis (see Lugosi et al., 2016Lugosi et al., , 2020 for studies adopting composite data strategies).

Data Collection
Reflecting previous work involving 'composite' data, evidence was collected using a mixture of observational methods, documentary analysis and semi-structured interviews (Kelly & Cordeiro, 2020;Smith et al., 2018;Torres et al., 2018). Since 2017, numerous visits have been made to charities and other NGOs, social businesses and commercial enterprises, involving largely non-participant observation, and an open, exploratory approach to data sampling (Adler & Adler, 1998). During visits, the researchers took images, focusing on organisational practices, facilities, routines, uses and users. This was complemented by opportunistically gathering other context-specific material, e.g. documents, and informal conversations with owners, operators/employees and clientele.
To expand the information collected during site visits and observations, material available online about organisations was also explored via desk research. This included descriptions of the organisations' histories, documentation regarding the scope and focus of their activities, alongside reports based on their research or that of related organisations. For commercial organisations, this included marketing and publicity material, and, where relevant, it considered online representations of organisations and their operators, available through popular media, including cultural commentaries in newspapers, and lifestyle magazines, social media platforms and blogs.
The scope and purpose of utilising these eclectic sources of data were not to attempt systematic content analysis; rather, assembling relevant fragments of data helped the researchers to gain context-sensitive understanding of organisations and their activities (see e.g. Smith et al., 2018 andTorres et al., 2018 for similar, multimethod studies). These disparate sources provided useful background information about the organisations and services, the migrants and refugees who founded, owned and in many cases operated them, alongside their employees, and their stakeholders, including partners, advocates and clientele.
Observations and background desk research on the organisations were complemented by three interactive workshops in Brazil with approximately 20 researchers and members of various migrant-support organisations attending each. The workshops, lasting about 1.5 h each, were led by two members of the team. They explored the challenges of labour market entry among migrants, the support and advocacy activities of diverse organisations and actors and their value creation activities. These workshops provided opportunities to present our preliminary findings based on our initial engagement with support organisations, gain feedback and capture another set of perspectives on services for migrants.
Finally, this work utilised a mixture of formal and informal semi-structured interviews. In the UK, interviews were conducted with representatives from two local charities, a social enterprise and a local education and training provider which supported migrants in their labour market transition, a representative from the local authority who managed a resettlement programme and an official of an international organisation supporting migrants, who collaborated with these organisations. These were selected, purposefully (Patton, 2015), because they were (a) the principal organisations in the geographic context providing support for migrants' labour market integration and (b) whose activities were therefore directly relevant to the study's focus. In Brazil, 16 people were interviewed, including migrant entrepreneurs, individuals involved with NGOs, charities and social businesses, as well as researchers and activists. Again, they were included because their insights were directly relevant to the study's focus.
The interviews with organisational stakeholders, which are the central focus of this paper, explored three main areas: firstly, the activities of the various organisations with which they were affiliated, and their roles within them, focusing on their past and present activities. Secondly, participants were asked about how they prepared migrants for work and how they supported their entry into and transition through the labour market. Finally, the interviews examined stakeholder perceptions of barriers and facilitators for migrants' labour market transition, which explored personal and micro-level factors, emerging in everyday practices and interactions, as well as meso-and macro-level ones, for example, at the organisational and interorganisational level, extending to societal and policy-related issues.
Interviews were augmented by countless other informal conversations during visits to organisations. As Pinsky (2015) argued, these types of 'incidental encounters' often provided further small insights. They were not analysed formally, but were captured in field notes, which informed subsequent interviews and the interpretation our findings.
Three members of the team were also involved in various formal and informal organisations supporting migrants and refugees, for example, as part of university outreach programmes, as volunteer advisors and advocates, or as members of support networks. As part of these entities, team members organised and participated in knowledge exchange events, and collaborated in several research projects, and continue to do so. In short, these projects were embedded in ongoing interactions. As Kelly and Cordeiro (2020) observed, this makes it difficult to disentangle interactions that could be defined explicitly and narrowly as data collection or research; nevertheless, this type of longitudinal engagement provided context-sensitive insights that helped to build richer and more credible understanding of these organisations and actors.

Data Analysis
Data were analysed over multiple cycles, during and after fieldwork, adopting a reflexive thematic approach (Braun et al., 2019), involving various configurations of parallel coding. Different members of the team analysed and interpreted the disparate data independently before sharing findings with other members. Data and researcher triangulation thus helped to make sense of an eclectic, rich and evolving composite data set. The analysis was inevitably informed by broad sensitising concepts stemming from the thematic areas of the interviews: roles and practices; factors facilitating and inhibiting labour market access; and preparatory and accessrelated support activities. Nevertheless, through a reflexive inductive approach, we identified specific data that were perceived to be novel and substantial in relation to the research questions. For example, what roles and related practices may have encompassed as they were enacted, e.g. coaching, representing, sponsoring [migrants] and role boundaries, which then helped us to differentiate between direct and indirect forms of intervention and interventions aimed at migrants or at organisational actors, including those who employed migrants or mediated their labour market transition.
Ongoing dialogue among the team regarding codes, sub-themes and meta-themes identified in the data was used to reorder the data in preparation for display and dissemination. However, as proponents of reflexive thematic approaches argue, analysis cannot be reduced to procedural tasks carried out at a single stage (Braun & Clarke, 2013). Rather, analysis and interpretation of the findings extended into the data presentation and writing process (ibid.). As part of this analytical writing, the first author re-examined the data from Brazil and the UK, using them to create a typology of migrant supporting organisational activities, based on activity scope and stakeholder focus. Through this ongoing interpretative process, it also became useful to present the data necessary to appreciate the dynamic nature of service provision, and the practices of specific actors and organisations, as a separate thematic area to help understand the migrant support ecology. Foregrounding these fundamental contextual issues then helped to identify the activities of different organisations, distinguishing between tactical, episodic forms of inter-organisational practices and strategic multithreaded forms.

The Evolving Ecology of Migrant Support
By way of context setting, this first section briefly sketches out three key sets of factors that shaped the dynamics of the migration support ecologies in the empirical settings: first, the diversity and evolving profile of arriving migrants; second, the changing migrant reception policies and funding available to support organisations; and third, the new service providers entering this sphere of activity and the new services they developed in response to emerging challenges. Appreciating the impacts of the three factors subsequently helps to understand how and why inter-organisational entanglements may have emerged in these support ecologies.
Brazil and the UK have both experienced substantial migration flows from disparate sources. Brazil has historically encountered diverse waves of economic and humanitarian migration. Migrants from Bolivia, Peru, Argentina, Colombia, Paraguay, China and Japan have constituted some of the largest groups in Brazil generally, and São Paulo specifically, during the twentieth and early twenty-first century (Baeninger et al., 2019). In recent years, flows have included humanitarian migrants whose mobility was driven by natural disasters, for example, from Haiti following the 2010 earthquake (Thomaz, 2013), and, more recently, those fleeing political and economic instability, including from Venezuela, Angola and Syria (Baeninger et al., 2019; Ministry of Justice and Public Security (MJPS)/Ministério da Justiça e Segurança Pública, 2017). These refugees and humanitarian migrants have complex needs because of forced displacement, so their labour market integration has required multiorganisational responses including from state institutions and TSOs (MJPS, 2017).
Britain has also experienced diverse migration flows, including extensive economic immigration from commonwealth countries during the mid-twentieth century onwards and particularly from accession countries following the 2004 enlargement of the European Union (Hansen, 2000;Somerville & Walsh, 2021). The UK received large waves of refugees from Europe during World War 2 (Bloch, 2002) and more recently from Iran, Iraq, Afghanistan, Syria, Eritrea, Somalia and Sudan (The Migration Observatory, 2021). Similar to Brazil, the vast majority of migrants concentrate in large cities such as London, although for refugees, the dispersal systems have resulted in a higher percentage being relocated to northern areas of the country (The Migration Observatory, 2022). The labour market integration of economic migrants, from the EU, for example, has often been managed primarily through commercial employment agencies (Forde et al., 2015) and informal social networks (Janta et al., 2011). However, for recent refugees, their complex needs, language and cultural differences, trauma and disrupted social capital has required support for labour market integration to be delivered by a more complex network of state, commercial and third-sector organisations, which have to address wider social, cultural and psychological challenges alongside transition into work (Mayblin & James, 2019;Morano-Foadi et al., 2021. The rising numbers of refugees in both countries prompted various policy responses at the state level, for example, Brazil granting Venezuelans the right to work and settle under a humanitarian reception scheme and the UK launching the Syrian Vulnerable Persons Resettlement Scheme (SVPRS) (Mavelli, 2017;Mayblin & James, 2019;Morano-Foadi et al., 2021;Moreira, 2014;Moreira & Baeninger, 2010). However, studies have shown that in both contexts, service provision was provided by a disparate range of actors, with variable levels of coordination and discontinuous funding available to service providers (see Calò et al., 2021;Morano-Foadi et al., 2021 for the UK, and Baeza, 2018;Ianni Segatto, 2021;Moreira, 2017 for Brazil).
Growing numbers of refugees have also driven various TSOs to innovate, thus expanding their service portfolios, or to extend existing services to refugees (Blunt, 2021;Mayblin & James, 2019;Morano-Foadi et al., 2023). For example, this included organisations whose primary focus was initially on supporting homeless people or women escaping from abusive relationships, but whose client base expanded to include refugees and other types of migrants.
The complexities regarding refugees and the changing policy and service landscape of reception are foregrounded here to stress that these represent dynamic phenomena, involving heterogeneous people. Invoking the term 'migrant support ecologies' rather than the common parlance of 'ecosystems' as organised and coordinated networks of action to comprehend the situation is particularly apt. This was highlighted by the UK County Council Official coordinating the SVPRS programme: The Home Office then liaises with the local authority and they really want one point of contact. So it's a funded scheme, so obviously we get all our funding through the Home Office and we can use, local authority's saying how, if you like, the freedom to..., to then see how they're going to meet what's called a statement of requirements. That's like our sort of contract with the Home Office for getting that funding. But how we do that, there is a certain amount of freedom. So some people have kept that in house, but we've decided to commission services. So I commission services to meet the needs of our newly arrived families. We procure the housing, so we have that role. Then we commission services to support them according to what's required of us for the first year. After the first year we have considerable amounts of freedom then.
Most of the programmes supporting refugees into work were delivered by TSOs. Even Home Office guidance on the financial aspects of the resettlement programme was rather limited. Devolved governance, with restricted funding and reliance on a disparate range of charitable, community and commercial/social enterprise-type organisations, risked creating a fragmented support landscape (Ianni Segatto, 2021; Morano-Foadi et al., 2021). These pressures forced local organisations and actors to compensate accordingly in terms of their decisions to provide different types of support, improvising their activities and occasionally attempting to mobilise other organisations' resources through collaborative arrangements (Morano-Foadi et al., 2023).
The fragmented characterisation of the support ecology was identified by other stakeholders in our empirical findings. One of our interviewees, Amanda, 1 who worked in a specialist charity supporting refugees and asylum seekers' access to the labour market, observed: [The Syrian Resettlement Programme] is a very decentralised programme... I am not aware of what guidance there is from the centre on this matter [access to jobs for refugees] ... not even ... promotion of English, I think, there isn't enough in terms of provision. And you're depending a lot on local supply which often just isn't there. So actually, if you look at the refugees [one of the first families to arrive in the county] they're being taught by volunteers. … There aren't really English programmes as far as I am aware.
An important theme in the observations from various practitioners was the evolving dynamics of the situation within the service supply and client demand. Similarly, one of the members of a Brazilian faith-based organisation noted that their clientele was constantly evolving, with new migrant and refugee cohorts presenting new questions and challenges. For example, this included conveying cultural sensitivities about self-presentation in recruitment and work settings. Despite being part of a large, well-established, international organisation, the representative referred to their work as being like an ongoing 'experiment', which was continually changing. The evolving profile of new migrant cohorts and the challenges presented by their cultural, religious and linguistic characteristics has been recognised elsewhere (Moreira, 2017). The unfolding indeterminacy of the situation drove organisational actors to reflect on their clients' needs and on how they could best be addressed, either through their services, but also those of others operating in the same ecology to support these client segments.
For example, Franco, one the operators of 'Presença', a Brazilian social enterprise observed: 'In the beginning of [this enterprise] we were meant to focus on communication to fight prejudice. When people came [seeking help with other issues, including work] we referred them to other NGOs.' A moment later, in reference to a specific incident where they referred refugees to a charity to support an application, Franco commented: 'Then, Maria [Presença's co-owner] got on the phone, tried to know what happened after they were referred. But nothing happened.' Maria continued: Then I got crazy. That company offered 14 jobs and I sent [the refugees] to [a charity] and nothing happened. Why? Because I think [the employer] they have a lot of online channels and [the charity] couldn't take care of that offer in the way that I would like to. Then, this day I decided to start working from 5 to 10 doing that.

Inter-Organisational Entanglements in Migrant Support…
These types of incidents, stemming in part from the eclectic needs of their clientele but also from the highly variable nature of existing migrant support service provision, drove organisations such as Presença to innovate and diversify their portfolio of activities. In the case of this social enterprise, innovation also extended to the essential nature of the company. As Maria commented: "We created an NGO because companies asked 'are we an NGO or social project?'. [The prospective organisation] want[ed] to donate but needed this status of NGO." Maria noted that companies who donated money to charitable organisations could use this to reduce their tax liabilities. Importantly, these evolving intra-and inter-organisational activities were responses to the fragmentation of support and indeterminacy of the support ecology. They are not cited in this analysis simply to valorise the activities of these support organisations; rather, they also help to stress that adaptive behaviours were driven by the necessities of the ecologies' dynamics and the persistence of resource constraints.
Contextual dynamics and pressures also meant that various service organisations and actors entered this ecological space, bringing diverse resources, sensibilities and capacities, which shaped their engagement with various stakeholders, and the outcomes they sought to achieve. The next section explores their activities and their stakeholder focus, providing a general typology to help understand their roles in the ecology, and it considers how their activities and focus potentially shaped interorganisational arrangements.

Organisational Activity Scope and Stakeholder Focus
It is useful to begin by distinguishing between four ideal types of migrant support organisation, based on the scope of their activities (diversified versus niche) and their clientele (broad, thus accommodating versus narrow and specialist), which are summarised in Fig. 1. Specifically, there were a number of organisations engaging in multiple activities, including assorted services that were available to a broad range of clientele. 'Diversified accommodator' organisations operated employment, advocacy, training and education services; they provided counselling and health services, and advised in other areas including housing, and in some cases, enterprise. These services could be available to a range of clientele including the homeless, people with learning difficulties, those with drug dependencies, people who have experienced violence, ex-military and of course migrants and refugees. These were often large, heavily resourced and tied to larger institutions such as religious and philanthropic entities.
'Diversified specialists' also engaged in a variety of activities and provided multiple services, but targeted particular audience niches. Examples of these were charities who worked specifically with migrants and refugees, tailoring their different functions to their clients' needs. These provided a range of counselling, advisory and support services while promoting migrants' welfare, development and empowerment.
'Niche accommodators' provided a narrow range of activities, but for a variety of audiences. Examples of these were state-run and privately operated employment agencies that serviced the general public alongside migrants and refugees who had distinct profiles and needs. The main priority was to connect employers with potential employees, but they did not provide training or education. Finally, 'niche specialists' were organisations that concentrated on particular areas, e.g. food and culture related activities, or operating services for or with migrants. These included running job mediation services, organising media and advocacy campaigns, alongside providing training and finance to support enterprise activities.
Different organisations and actors operating within migrant support ecologies arguably shared what we call a 'common goal domain'. This term is invoked to stress that it is different to a 'shared goal', implying something agreed, which Castañer and Oliveira (2020) suggested was a central component of inter-organisational arrangements. The common goal domain could be summarised, broadly, as supporting migrants' psychological, economic and social empowerment, and independence through facilitating their transition into and across the labour market. However, given the diversity of organisational scope and stakeholder focus, outlined above, organisations adopted different value-creating (and value-extracting) strategies, involving different forms of engagement, which are outlined below.

Tactical Coupling and Episodic Engagement
Tactical coupling reflects attempts to create and exploit instrumental, often shortterm, domain-specific relations to pursue a particular value-related objective. This was evident in activities of several niche specialist (social) enterprises in Brazil.

Inter-Organisational Entanglements in Migrant Support…
For example, 'Presença' tried to challenge representations and perceptions of, and thus attitudes to, refugees among employers and the general public. They engaged in multiple forms of what they called 'provocations'. This initially involved recording hundreds of video stories, which were used to create media campaigns to challenge xenophobia, and refugees' narratives were also published as a series of books. They created a digital system that allowed refugees to record short videos and create visual CVs, which employers could then browse. They also contacted organisations encouraging them to utilise their technologies and, once one did, leveraged this further by contacting rival companies to inform them that their competitors were already participating. To support these activities, they also designed 'a stamp [a visual certification], for companies that hire refugees'. This could then be used by organisations to create value in public relations and social responsibility credentials.
Other examples of tactical coupling and short-term, episodic engagement in Brazil included 'Mordida', which showcased refugee chefs as part of pop-up events, while the owner-operator of 'Gosto' engaged in advocacy work alongside promoting food-related cultural heritage. These organisations and their constituent actors utilised their public relations expertise and mobilised technological platforms in media campaigns that sought to valorise refugees, transform perceptions of them and facilitate positive encounters (see Lugosi & Allis, 2019).
Importantly, tactical coupling also included referral practices through which all four types of organisation could utilise the skills and resources of other actors or organisations, though not all did. For example, employment agencies, which represented niche accommodators, providing a limited set of services to a wider variety of clients, did not necessarily have the specialist skills or resources to provide language training, legal or housing advice, coaching, mentoring or counselling services. This was also the case for niche specialists such as social enterprises or niche accommodator migrant and asylum charities. Referral was often essential for service provision, especially for specialist fields, and other forms of value creation that required market knowledge. However, such short-term engagement could also be initiated by various actors operating in the support ecology. For example, a social business representative observed: When we get a refugee who need their qualifications validated, we recommend them to [another specialist service provider], who have a partnership with [another agency], who receive money [from the state]. … If they arrive and they don't speak Portuguese, we refer them to some projects, some partners, NGOs. Nowadays if they have some capacity, people share with each other. Yesterday [another organisation] had a course about Photoshop; the other day it was painting walls. All the NGOs, they recommend some refugees to be part of that course.
These did not necessarily involve cross-organisational coordination of actions. Cooperative activities were often domain-specific, reflecting episodic engagements and involving limited interaction between actors. They functioned partly because other actors had specialist knowledge and dedicated resource capacities. However, inter-organisational interaction required sufficient knowledge of these actors operating in the same support ecology. Arguably, these interactions required embedded social capital, where actors from different organisations knew of each other, operated within a common goal domain, and maintained a certain level of trust that enabled client referral to operate (see Morano-Foadi et al., 2023). These thus had the potential to morph into stronger, extended, multithreaded, inter-organisational arrangements, as discussed below.

Strategic Coupling and Multithreaded Engagement
Diversified accommodators, for example, 'Portão Aberto', a large, church-affiliated organisation in Brazil, provided an assortment of support services directly to a range of clientele, including migrants. However, in the UK context, diverse accommodator organisations such as the local authorities of a city council charged with supporting refugees as part of a formal resettlement programme represented even larger and more complex entities, whose responsibilities extended beyond this stakeholder group. The council became a key 'fiscal agent' (Sandfort & Milward, 2008), commissioning services, for example, from educational providers and social enterprises, to create and run dedicated programmes for language learning, skills development and employability support.
Strategic coupling was also evident among niche specialists, such as 'Prosperar', a social enterprise that negotiated directly with Uber Eats and AirBnB, facilitating migrants' access to markets and consumers through their platforms, through which they could promote, distribute and monetise foodstuffs and create cultural experiences. They also established a working relationship with Google, which consequently provided advertising credits that could be used by migrant entrepreneurs to improve their search engine visibility. 'Prosperar' also launched a catering business, utilising refugees' knowledge and skills, and established relations with corporations who engaged their services, often on a repeat basis.
One of the UK-based social enterprises built and nurtured an ongoing partnership with a large nationwide food retailer, through which they could deliver their work-based ESOL (English for Speakers of Other Languages) programme in service skills. This programme provided refugees with work placements, and also involved longer-term work coaching for the refugees and support for the organisation to ensure positive and mutually beneficial cooperation. At the core of all these interactions was the role of advocacy-driven intermediation, with the social enterprise actors helping to understand the goals, motivations and resources of the different stakeholders; calibrate expectations of the different parties; reduce psychological fears and negate risks; address individual challenges; and evaluate practices and their outcomes.
Representatives from this enterprise sought to create longer-term relationships. These often began as speculative discussions, but aimed to facilitate long-term, mutually beneficial partnerships. For example, Kate, who worked for a UK social enterprise 'Thrivelab', reflected on creating these pathways: And so I said to the [employer] -would you consider some part-time child-friendly hours for Syrian women, because you know, that would be a good baby step for them to come in. And they said -what are you talking about? We haven't even thought of this. … So I'm planting a seed. … In [the region] there's an Association of Care Providers, who are the umbrella body for the sector. So I said to [the Association] would you be interested in developing a route for Syrian women to come into the care sector? And they were like -we haven't even thought of that. So everywhere that I'm going, I am saying to employers -is this something you would consider?
It is also useful to note that these speculative interactions, aimed at developing partnerships, were informed by the embeddedness of the social enterprise actors in relationships with other migrant supporting organisations in this context. Specifically, the idea of approaching care sector representative came from another partner organisation: I have heard through [a local asylum charity], that some of the Syrian women there would like to go into the health sector and they would like to train as nurses. And the way into the nursing sector could be to go through the care sector. … Which seems like a very good route to me. (Kate, Thrivelab, social enterprise) In sum, these practices represented longer-term, multithreaded, cooperative arrangements, enacted by creating linkages between multiple stakeholders. However, these arrangements may have involved limited direct coordination of partner activities. Rather, the emphasis was on creating value from multiple stakeholders' specific resource bases: i.e. technological infrastructures and other forms of social and cultural capital, including specialist, domain-specific expertise and cultural knowledge. These organisations operated within a common goal domain, but often sought to extract particular forms of value. Commercial organisations may have pursued narrow economic interests or public relations gains that translated into 'brand value', which ultimately represented monetizeable assets. However, for philanthropic, religious or governmental agencies, value may have taken the form of perceived inclusivity or social justice for their clientele. Importantly, those different types of value could coexist, so the same activity enabled multiple forms of value to be co-created.
Within such multithreaded engagement, involving strategic coupling, there were examples of attempts to create a network primarily for the purpose of knowledge exchange. This was evident in the UK context, where the head of the 'Thrivelab' social enterprise created a refugee employment network among disparate service providers including representatives from charities, state bodies and other social enterprises. The network was a loosely defined entity that communicated through email lists, periodic meetings and occasional events. The membership changed partly as new actors or organisations were identified to be providing related services or engaging in activities that aligned with the interests of those of existing network members. Members also withdrew from the network, often because they stopped providing migrant-oriented services. Some organisations only worked with refugees on a narrow set of specific projects, which had limited funding and were therefore time and resource constrained; others withdrew services as part of strategic organisational changes or restructuring. This network did not attempt to coordinate goals or functions among actors; rather, it provided a space where actors operating in the migrant support ecology became aware of the existence, scope and function of other actors and organisations. It thus acted as a facilitator of multithreaded engagement. Participants distributed information regarding specific events and used the forum to build social capital, which could be the basis of future cooperation. It was evident that some of the members had engaged more actively in coordinated actions, for example, applying jointly for project funding and collaborative service delivery. However, members also applied separately to central government funds for specific projects, so were effectively competitors operating in the same space.

Discussion and Conclusion
Previous work advocating greater integration of services supporting migrants, underpinned by stronger coordination and cooperation, has used ecosystems to conceptualise how these networks could be designed and governed (Finsterwalder, 2017;Finsterwalder et al., 2021;Kornberger et al., 2018;Richardson et al., 2020). These perspectives recognise that systems are adaptive and subject to change (Janzen et al., 2021); nevertheless, they continue to foreground the importance of integration and interdependency among organisations as fundamental features of support-service-provision networks. These are undoubtedly laudable aspirations, and constructive alliances have been shown to be effective in addressing the disparate needs and interests of migrants and members of receiving populations (Danış & Nazlı, 2019;Janzen et al., 2021). There may also be systemic incentives that drive organisations to seek collaborative interactions; states and NGOs who fund integration programmes regularly promote collaborative arrangements among stakeholders to deliver services (Calò et al., 2021;Ianni Segatto, 2021). However, it is also important to recognise that service provision may continue to be fragmented, primarily because it comes from a diverse range of organisations, whose actions are not coordinated (Morano-Foadi et al., 2021). Discontinuities in the scope, level and comprehensiveness of service provision may also be caused by a range of additional factors including the evolving flows and profiles of migrants, shifting social and political attitudes towards migration and its governance, as well as the resource constraints faced by states and organisations (Mayblin & James, 2019). In many contexts, these pressures have driven multiple organisations and actors, from across state, commercial and third sectors to enter and operate within migrant reception and support ecologies (Baeza, 2018;Janzen et al., 2020Janzen et al., , 2021. Drawing on data from Brazil and the UK, we have proposed a conception of migrant support, based on ecology, which recognises the potential for concurrent action, without assuming the presence of interaction, coordination and interdependency. Moreover, even when there is mutual awareness and interactions, we have sought to conceptualise why those may take different forms. The data were used to outline and explore the dynamics of inter-organisational interactions among disparate actors and organisations engaged in supporting migrants' transition into the labour market. We proposed the term 'migrant support ecologies' to help conceptualise the spatiality of their coexistence. Following Tsoukas and Dooley (2011), and others (e.g. Bennouna et al., 2019), invoking ecology recognises that organisations and actors may operate in common physical and abstract (social, political and moral) landscapes, which should not be thought of as self-contained, stable or statically bounded entities. Moreover, adopting ecology serves to stress that organisations and actors can coexist and operate in shared or at least related spheres of action, without assuming awareness or interaction.
Furthermore, we introduced the term 'common goal domain' to emphasise that these actors and organisations can have a broad overarching ambition to empower and support migrants in their labour market integration, but do this through diverse value creation practices. Following Law (2004), who urged researchers to acknowledge the fundamental messiness of social and organisational practices, we adopted the term entanglements as a conceptual device to stress that relations between organisational actors can be fragmented and discontinuous, varying in scope and scale. The conceptual terms introduced in this paper thus offer credible alternatives to those implied by ecosystems insofar as they recognise and accommodate the messiness of inter-organisational practices.
Moreover, beyond acknowledging fluidity and discontinuity, we distinguished between organisational types based on their scope and stakeholder focus before outlining how organisations and their constituent actors may engage in tactical and strategic coupling practices, reflecting shorter and more episodic interactions alongside complex, multidimensional ones. We thus argued that understanding the dynamics of organisations' activity scope and stakeholder focus can help to appreciate why different forms of engagement may occur. This complements and extends Hesse et al. (2019) and Maletzky de García's (2021) work examining how institutional logics shape the dynamics of collaboration in supporting migrants. More specifically, we proposed a unique conception of organisational ideal types, which can be applied to conceive and evaluate inter-organisational entanglements in other support ecologies.
The findings of the current study showed how different forms of inter-organisational arrangements may be pursued based on organisations' capacities and stakeholder focus, alongside the multidimensional types of value that they seek to create for organisations, migrants and wider societies. Importantly, recognising how the distributed nature of transformative practice operates within the common goal domain helps to conceptualise how a disparate network of actors and organisations continually support migrants' labour market transition, despite the potential for ambiguities and ongoing changes in coordination (i.e. shared goal setting) and cooperation (i.e. intentional actions that help realise those goals).
Arguably, the conceptualisations from this study are transferable beyond these UK and Brazilian contexts. Distributed networks of actors and organisations are likely to operate in a variety of international scenarios. The interrelated notions of ecologies and common goal domains, which acknowledge the possibility that diverse organisations and their constituent actors engage similar stakeholders, pursue related value-laden goals, while seeking disparate value creation outcomes, provide useful ways to conceptualise the dynamics of such arrangements. This is not a normative framework prescribing how things should function; rather, as a critical analytical perspective, it helps to understand how and why things function in practice.

Implication for Practice
Conceiving migrant support ecologies as dynamic, evolving inter-organisational arrangements, without assumptions regarding collaboration, or even full, mutual awareness between organisations, can help practitioners to approach the challenges they face in terms of information management, as opposed to coordination or integration (Goerzen, 2018). Specifically, in shifting landscapes of service providers, provision and demand, attempts at coordination of disparate organisations or actors by any one service provider operating in the support ecology may be unfeasible because it is resource-intensive. Individual service organisations may therefore invest in the management of information regarding complementary or substitute service providers, for example, among TSOs or community actors. This may primarily involve 'market scanning' (Olsen & Sallis, 2006): attempts to identify what types of organisations are operating in a close geographical and/or domain-specific (service) ecology, what types of resources and capacities they deploy and what types of value creation they are pursuing in relation to the 'clients' or stakeholders. This knowledge can be used to direct clients to seek support from these alternative providers, for example, through referrals. Moreover, the effective capture, management and, if necessary, dissemination of this information can thus inform the decision of organisations operating in this ecology to engage in tactical or strategic coupling and thus episodic or multithreaded engagement.
If we continue to acknowledge that, in practice, support for migrants' labour market integration may operate in fragmented migrant support ecology scenarios as outlined in this paper, strategic investment from states or international NGOs may be focused towards developing diversified or niche specialist organisations' information management capabilities, for example, information concerning occupation-specific training, job mediation services or volunteer opportunities provided by other organisations. Investment can help build their capacity to manage information about the service ecology to inform subsequent decisions to pursue different forms of organisational coupling that are directly relevant for facilitating migrants' labour market integration.
However, it is also useful to consider how this type of capacity building in market scanning and knowledge management capabilities, which facilitate inter-organisational coupling, might shift from individual service providers to third parties. In migrant support ecologies, individual service providers may not know or understand which other organisations provide complementary activities, or they may see others operating in this ecological space as competitors for limited resources. Consequently, it might be more feasible for a third-party organisation to manage knowledge about service providers and provision in the support ecology. A key challenge for this is the effective translation of information regarding how strategic or tactical coupling among individual service providers can become sources of mutual value creation in how they support migrants' labour market integration. Universities have been proposed to perform these intermediating roles insofar as they are able to traverse organisational boundaries (Jubilut, 2020). Moreover, these translational, intermediary responsibilities may extend beyond information management towards active coordination and resource integration, where inter-organisational arrangements may increasingly adopt the principles of normative ecosystem models. However, it is important to appreciate that, in practice, the scope and form of inter-organisational entanglements may continue to be shaped by the complex dynamics of migrant support ecologies.

Limitations and Implications for Future Research
This paper utilised composite data generated across two countries over 4 years. However, future studies could pursue longitudinal methodological strategies to better understand how inter-organisational arrangements evolve over time and under different social and political conditions, for example, during (inter)national crises, changes in governments or policy shifts. Expanding the sample and adopting systematic comparative approaches could also enable researchers to capture and assess cooperative behaviours, including how different organisations and actors seek to create value. Furthermore, this study's purposive sampling of organisations, based on their relevance to the research questions and the possibilities for access meant that other organisations and their constituent actors were not included. Expanding the inclusion criteria, adopting systematic stratified approaches and setting quotas for organisational types, for example, according to size and activities, could help to build a wider evidence base.
Future research may expand the sample of organisational actors to be studied, again adopting alternative sampling approaches and inclusion criteria to capture a wider set of perspectives regarding interactions, collaborative activities and value creation practices. The current study's composite data also included clients (encompassing refugees and other types of humanitarian and non-humanitarian migrants); however, it was not possible to explore their perceptions, behaviours or experiences in this paper. Future research can also incorporate their views, and those of other stakeholders, for example, local community actors, to understand the dynamics of collaborative behaviours and value creation activities, and to evaluate their implications for migrants' integration more generally and for labour market integration specifically.
The findings of this study do not point to causal relations, suggesting that one interactional or cooperative approach is more effective than another for certain types of organisations. Nor do the findings assess the utility of interactions and their outcomes regarding value creation. Tactical and thus episodic, instrumental interactions, and more complex, extended, strategic engagement are performed -they are experiential, emotive and thus subjective. A key challenge for future research is to integrate this paper's conceptualisations in evaluative or diagnostic frameworks to assess the quality and substance of (inter-)organisational engagement. How and why they are adopted, and become embedded in networks, and why they are disrupted or reconfigured remain significant questions for future enquiry. Future work can also seek to assess whether certain inter-organisational coupling strategies are more effective for pursuing specific types of objectives or value creation outcomes.
Furthermore, the dynamic perspective developed in this paper suggests that an interpretative rather than positivist approach seems more suitable for trying to understand the nature of such interactions and their consequences, particularly in assessing the potential for value creation. Importantly, the entanglements proposed here serve as sensitising concepts for future empirical studies of analogous interorganisational arrangements in service ecologies where diverse types of organisations, including social and commercial enterprises, state institutions and TSOs operate.