Relational perspective on collaboration barriers in public service ecosystems

Public services, such as social and healthcare services, rely strongly on the collaboration of diverse actors, agencies and institutions at varying governmental levels. This collaboration is focal in creating well-functioning service paths and, more widely, value-creating service systems (Osborne et al., 2022). Relational failures in collaboration may cause an inability to respond to the needs of the public, thus causing harm to both citizens and the system. To exemplify, different understandings of value may lead to conflicts and more inefficient service paths (e.g. Burn and Needham, 2025; Engen et al., 2021; Rossi and Tuurnas, 2021). Collaboration is also affected by wider societal developments in the institutional environment, such as reforms, legislation, financial and budgetary questions and demographic changes.

In recent years, the ecosystem concept has gained increasing attention in public management research. Public service ecosystems are used to understand the systemic and relational nature of public services (cf. Osborne et al., 2022). Existing research on public service ecosystems has focused on viewpoints such as managerial practices and processes (Osborne et al., 2021; Petrescu, 2019), organizational governance (Wirtz and Müller, 2023), value creation (Ojasalo and Kauppinen, 2024; Osborne et al., 2022) and disruptions in ecosystems (Eriksson et al., 2021). The emerging concept of value co-destruction highlights how misaligned interactions may undermine service outcomes (Engen et al., 2021). Recent studies have also focused on examining different actors in public service ecosystems, such as professionals (Maijala et al., 2025), civil society organizations (Tuurnas et al., 2025) and citizens (Kurkela et al., 2025).

In this article, to stress the importance of interaction and human relations for the outcomes of the service processes taking place in ecosystem settings, the relational aspect is highlighted. The article focuses on the interconnectedness of social relationships and networks in public administration horizontally and vertically across ecosystem levels. Therefore, we ask the following question: How are barriers to interorganizational collaboration framed in the analytical context of the public service ecosystem?

To clarify the main concepts of this study, the ecosystem concept is employed to draw attention to the interconnectedness of actions and operational frameworks across micro, meso and macro levels – from service encounters to organizational rules and legal frameworks. In this study, collaboration is understood as intentional, coordinated and purposive interactions oriented towards joint problem-solving or service outcomes (cf. Osborne et al., 2022). Collaboration is therefore conceptualized as a specific subset of the broader category of interaction – ranging from incidental encounters to co-production – making it analytically relevant for the study of public service ecosystems. While this study uses collaboration as its primary analytical concept, it also recognizes informal, incidental and emergent interactions as part of the wider service ecosystem context within which collaboration is shaped (cf. Burn and Needham, 2025).

To operationalize the relational approach in the context of the ecosystem, we employ frame analysis to scrutinize the social and healthcare ecosystem in the Finnish context, which has been characterized in recent years by wide-ranging reforms to social and healthcare services and employment services and has a tradition of broad responsibilities and universalism as a Nordic welfare society (cf. e.g. Keskimäki et al., 2019; Nordin and Johanson, 2025; Tynkkynen et al., 2022). Specifically, the large qualitative dataset (56 informants and 38 interview situations) is from the Social Insurance Institution (Kela) and three wellbeing services counties (WSCs). Kela is Finland's national social security agency, which provides financial support for health, unemployment, families, students, housing and pensions, ensuring basic economic security for residents. Meanwhile, the WSCs – established in 2023 – are responsible for healthcare, social welfare and rescue services in 21 regions. Their collaboration is particularly visible in multi-actor service processes, such as social assistance and rehabilitation. For example, Kela provides the basic part of social assistance, while supplementary and preventive support – such as housing costs or special needs – comes from WSCs.

Thus, this collaboration is essential for value creation in the Finnish health and care ecosystem because these institutions share responsibility for a wide variety of joint processes, influencing the quality of life of Finnish citizens as service users. Moreover, the functioning of these organizations' service processes is highly dependent on their potential to navigate and collaborate in the service ecosystem.

By combining a relational ecosystem perspective with frame analysis, this study contributes to the public management literature in two ways. First, it demonstrates how collaboration challenges are interpreted and constructed across the macro-, meso- and micro-levels of the service ecosystem. While the (public) service ecosystem concept has gained increasing attention among public administration and management scholars, there remains a need for empirical studies that apply it explicitly as an analytical framework (cf. Trischler et al., 2023), particularly by incorporating multiple ecosystem levels within a single analysis.

Second, the study advances relational approaches by showing how meanings and framings shape the functioning of social and healthcare public service ecosystems. Although relationality lies at the core of ecosystem thinking, empirical research that foregrounds this perspective is still developing (e.g. Osborne et al., 2022). By focusing on how actors frame collaboration challenges, this study offers insight into the interpretive processes through which relational dynamics are enacted in public service ecosystems. While a relational ecosystem perspective has been adopted by Burn and Needham (2025), there is still room to further develop this approach, especially by mobilizing the ecosystem lens explicitly as a core analytical framework rather than as a background concept.

As multiple interdependent actors are commonly involved in service provision, it is important to observe the dimensions of public services from a broader perspective than just the views of internal management or externally perceivable results (Engen et al., 2021; Osborne, 2006; Osborne et al., 2022). A public service ecosystem perspective offers an illustrative framework for observing service provision, in which the interaction and collaboration of different actors is expected. As a concept, the service ecosystem expands views beyond linear and network-like service processes and can show value being co-created by the collaboration of different actors (e.g. Osborne et al., 2020; Petrescu, 2019). Service ecosystem thinking draws from service research and service-dominant logic. As Vargo and Lusch (2016, pp. 10–11) described, a service ecosystem is “a relatively self-contained, self-adjusting system of resource-integrating actors connected by shared institutional arrangements and mutual value creation through service exchange”. However, the ecosystemic nature of services is at the heart of modern public administration and management research, which highlights collaborative governance, value co-creation and the central role of citizens in their diverse roles defining service value (Osborne et al., 2021, 2022; Trischler et al., 2023). An ecosystemic lens has been seen as useful for understanding the complexities of social care and healthcare (e.g. Burn and Needham, 2025; Leite and Hodgkinson, 2023).

The three basic levels of the ecosystem include the institutional macro level, the service operational meso level and the individual micro level (Trischler et al., 2023). Osborne et al. (2022) added a fourth ecosystemic level, the sub-micro level, which consists of individual beliefs. The public service ecosystem is further illustrated by metaphors, introducing the institutional level of the ecosystem as “atmosphere”, the service level as habitat, the individual level as population and the level of beliefs as subsoil (Osborne et al., 2022, p. 638).

In this paper, we connect the ecosystem approach with relationality, an emerging approach in the PAM field (Bartels et al., 2025; Wilson et al., 2024). In this context, the metaphor of ecosystems inspires consideration of the roles of social capital, relationships, power dynamics, resource flows and organizational and institutional culture. However, relationality can be supported by other conceptual and theoretical discussions that draw attention to the interconnectedness of organizations in service delivery, interactions and human relations. Relationality can be used as an approach for (re)thinking governance systems. It goes beyond describing administrative activities or referrals to applying specific methodological approaches that focus on networks. In fact, relationality exists, at least loosely, in multiple current academic discussions in the field of public administration and management. That is, the public sector can be seen as a plural system of actors with “inevitable degrees of interdependencies” (cf. Bartels and Turnbull, 2020, p. 1326).

To understand the relationality of the public administration domain more thoroughly, Wilson et al. (2024) set out a relational research agenda for public services. Specifically, they criticize the heavy reliance on quantitative measures and statistics in defining and managing public service quality. As several critical studies related to performance management have noted, this approach leads to framing what constitutes good public service (e.g. Diefenbach, 2009). Against this backdrop, rather than focusing on technocratic measurement models, Wilson et al. (2024) suggest an approach based on negotiating conventions and socially constructing phenomena, allowing knowledge and measurements to be interpreted in the context of social relationships and community insights. This approach emphasizes collaborative learning and continuous improvement, suggesting that measurement should be a dynamic, evolving process rather than a rigid, fixed system. Naturally, dynamic approaches do not fit easily into models and systems and may therefore be discarded as too complex and messy to be taken seriously by administrations (e.g. Gyllenhammar et al., 2023). Similarly, instead of “transactionalization”, the relational public service agenda pays more attention to the diverse range of interactions and intersections that service coordination must address instead of focusing on single citizen-consumer dichotomies (Wilson et al., 2024, pp. 555–556).

Analytically, the relational approach is outlined by three commitments (Bartels and Turnbull, 2020). First, it emphasizes an ontological focus on networks of interactions and relationships rather than on individuals or institutions. Second, it adopts an epistemological orientation aimed at explaining the emergent properties of co-creation processes. Finally, the relational approach employs diverse methodologies to study how social networks and practices are carried out. This study is aligned with these approaches, focusing its analytical interest on relations and interaction in service ecosystem settings, highlighting the emergence of collaborative processes and using the specific methodology of frame analysis to detect the driving and hindering elements of collaboration in terms of interaction and the role of relations in interorganizational context.

The collaboration between WSCs and Kela forms the empirical context of this study. As a result of historical and broad reforms, WSCs (established in 2023) are responsible for providing social and healthcare services, as well as rescue services. Finland is divided into 21 self-governing WSCs (Ministry of Social Affairs and Health, 2025). Previously, until the end of 2022, municipalities (altogether 309) were responsible for organising social and healthcare services. However, they were able to form joint municipal authorities to provide these services in collaboration with other municipalities, or they could outsource the provision to other municipalities or private service providers (Ministry of Social Affairs and Health, 2019). Due to the enormous reform, WSCs form a new institutional level, which is democratically led and acts regionally.

Kela, as a nationwide public organization, is responsible for providing social security, which means support in different life situations, such as unemployment, illness, disability, old age, birth of a child and loss of a guardian. Kela is responsible for, for example, social assistance, child benefits, unemployment benefits, parental allowance, some parts of pensions, reimbursements for medicine expenses, financial aid for students and rehabilitation subsidies (Kela, 2025).

Because of Kela's wide responsibilities, it can be said that in some parts of their lives, every Finnish citizen has some kind of connection to Kela's services. In addition, Kela and WSCs share a wide variety of joint processes. We acknowledge that WSCs and Kela do not form the entire Finnish social and healthcare ecosystem, but they are highly focal institutions in that ecosystem, and the functioning of their collaboration affects the whole ecosystem.

The study data consist of semi-structured interviews conducted in Kela and three WSCs gathered in 2023 and 2024. Twelve interview situations were conducted as group interviews and 26 as individual interviews. Altogether, 56 informants were interviewed (transcribed in approximately 749 pages), of which 26 represented Kela and 30 represented WSCs (see Table 1).^[1]

The informants worked in top management, middle management or supervisory positions or employee positions. In the interviews, the researchers and informants discussed several topics connected to collaboration and co-creation with diverse actors, the roles of different actors and creation of value. One of the most focal themes was the collaboration between Kela and WSCs. All the authors took part in conducting the interviews.

To understand the relational perspective on the hindrances and pitfalls of interorganizational collaboration in ecosystem settings, this study utilizes frame analysis. While definitions of frame analysis vary, the central idea in analysing frames is to address the question of what is now occurring of examination in each situation (Entman, 1993; Goffman, 1974). Instead of reflecting reality as it is, the data illuminate a window through which one part of reality can be seen (Linström and Marais, 2012).

Framing involves organizing pieces of reality and information (Linström and Marais, 2012) as well as selecting and positioning certain perspectives at the centre of the data (Entman, 1993). Thus, frames are ways to define or interpret events in the environment (de Vreese, 2005; Goffman, 1974). The aim of framing is to enhance the significance and memorability of the chosen perspective. Presenting the same content in diverse ways (i.e. through various frames) can significantly influence how specific information is understood and how subsequent actions occur (Entman, 1993). Frames thus reveal themes that can be examined from multiple perspectives, and their presentation affects how people interpret each situation (Linström and Marais, 2012).

The analysis of frames highlights the power of communication to direct attention to certain interpretations of a situation or phenomenon instead of facts (Linström and Marais, 2012; de Vreese, 2005; Linström and Marais, 2012). It is essential that the frames are connected to the larger research context (Koenig, 2006).

Our research utilizes the qualitative approach to frame analysis, which focuses on the content of expressions and delves specifically into how things are presented, and allows more nuanced ways of framing (Linström and Marais, 2012). With an inductive approach, we rely on frames to emerge from the data. Although in an inductive approach frameworks result from the research, our analysis also takes the ecosystemic levels into account.

During the data gathering phase, major notions were shared in joint discussions and in a diary. Before coding, these measures built a joint understanding of recurrent viewpoints in the data, guiding us to utilize the service ecosystem as a perspective for data analysis. After the data gathering and transcription, the quotations (see Table 2) concerning and interpreting hindering elements of collaboration were coded by four authors using NVivo software. These codes were connected into one NVivo file and divided into subcategories. The subcategories were then observed from an ecosystemic perspective. At this point, the questions of what wider themes emerge throughout all the ecosystemic levels and what they describe guided the analysis. The aim was to find linkages between different ecosystemic levels. These linkages or interpretations among the ecosystemic levels form the four frames recognized in the study. These frames are (1) insufficient knowledge and flow of information, (2) institutional and organizational restrictions and scarcity, (3) immature organizations due to reform (4) and differing modes of action.

The first frame arises from interpretations concerning insufficiencies in joint knowledge formation. In many cases, the hindrances and challenges are connected to an insufficient flow of information between partnering organizations and a lack of knowledge concerning the operations and processes of partners.

The macro-level viewpoint considers the norms of the service system. The regulation of information flow is seen as an issue that affects organizational operations. Notwithstanding a willingness and desire to cooperate with partner organizations, legislation concerning data protection is seen as one of the key issues and juridical challenges that hinder everyday interactions among professionals, simultaneously burdening citizens through the responsibilities of delivering information between organizations and professionals.

The legislation builds a framework for information sharing, causing one-way information flows and not enabling feedback on processes. The legislation on data protection poses further challenges to the everyday collaboration of service professionals within the same service processes. From the micro-level perspective, questions concerning the flow of information also emerge in other forms; the complicatedness of contacting fellow professionals in partner organizations and a lack of communication possibilities become underlined, as the legislation is seen to “trip up the work”, “prevent cooperation, because there is no such permission” and cause an “authorization circus” or even fear among service professionals. One of the objectives mentioned is “to speak freely” with important partners.

Data privacy is linked to the challenge of “one-way flow of information” between organizations, referring to service professionals not knowing about the progress in service processes in other organizations or whether they have reached the service user to begin with. This may inflict stress on the service user, who must then act as “a transmitter of information.” The flow of information is linked, for example, to the challenges of formal and informal forums and meetings among partners, also spanning digital means. Nationwide information systems are seen as “trustworthy and well-functioning” but simultaneously rigid. Finding the correct persons to interact with on specific issues concerning joint service users is often described as difficult, especially regarding wide-ranging or newer organizations (WSCs).

The data also point to the importance and lack of adequate “comprehension concerning partners” in the service system and the wholeness of partners in it. This refers to the flow of information concerning organizational preconditions and modus operandi and is seen to influence the ability to support service users in joint service processes. Regarding the everyday work of professionals, the system is described as wide and fragmented, which causes extra challenges in aiding service users in finding appropriate services or benefits.

The organizational understanding of the everyday lives of service users is relevant but lacking in some cases. Aside from feedback and recruiting experts with experience, the data point out the importance of valuing interaction and addressing feedback to where they are due instead of collecting them fragmentedly between service users and professionals. The data include examples in which comprehension of the service user's situation is not sufficient (i.e. a service user has not been examined by doctors, for example, about learning difficulties and therefore lacks diagnosis and services). This is connected to insufficient understanding of partner organizations' preconditions, inflicting situations in which service users are not served as expected.

The second frame that emerged in our analysis concerns institutional restrictions that are connected to bureaucratic silos between organizations and scarcity of resources. These aspects are interpreted as heavily affecting the possibility of collaborating in a meaningful way.

Concerning the scarcity of resources, macro-level political decisions regarding budgetary cuts influence the everyday collaboration between organizations and service professionals. Budget reductions in the field of social and healthcare services, in addition to already tight resources, cause organizational scarcity, referred to in the data, for example, as “a funding problem”, “a severe shortage of resources”, “a resource deficit” and a situation in which a WSC “has completely run out of money”.

These factors may cause a stronger need for prioritizing organizational activities, giving fewer possibilities for developing collaboration, because the basic services need to be secured. Lack of time and collaboration as an “under-resourced” activity were widely used arguments regarding the lack of collaboration. Simultaneously, some insights viewed the lack of adequate resources as grounds for developing collaborative activities.

The scarcity of resources in one partner institution can influence work within others, causing extra work or shaping its allocation and, eventually, institutional roles, as the following example illustrates:

Regarding scarce resources, institutional settings may even cause situations in which other institutions are seen as “competitors”. In some notions, steering by ministries may cause organizational silos, both inside an organization and between organizations. These silos may, for example, hinder interactions between social insurance (Kela) and social work (WSC), and a siloed environment does not support cooperation.

Moreover, at the micro level, “bureaucratic” organizations and processual rigidity translate to “complicated forms” for citizens to fill in and rigid service processes but also organizational “facelessness”, which may bother service users and the actors in partnering organizations. Bureaucratic restrictions and scarcity also affect the micro level. For example, sparse resources emerge in service professionals' work in various ways. Insights in the data highlight the lack of allocated time or “bandwidth” for professionals to cooperate with partners. In addition to budgetary issues, these include, for example, the COVID crisis and “staffing shortages”. In the context of limited resources and bureaucratic restrictions, professionals also need to navigate – and help service users navigate – the service system.

The third frame that crosscuts all three ecosystemic levels includes interpretations of systemic change due to the ongoing implementation of national reforms, leading to the “immaturity and incompleteness” of WSCs, which affect collaboration, development activities and strategic clarity. This frame emerges as a discussion and challenges recent reforms concerning the social and healthcare services that have taken place in Finland. This historically “enormous change”, implemented in 2023, caused groundbreaking changes in organizations and shaped the everyday work of professionals. This reform also reshaped the entire context of collaboration in social and healthcare services.

Because of the reform and new institutional arrangements, the incompleteness of the WSCs is a recurring viewpoint in the data. The new counties bear the heavy task of integrating former organizational cultures and modes of action, which were previously led by municipalities. In addition, the terms of organizational development and completeness of WSCs appear to be in divergent “point of departures” and phases. Some WSCs are considered to be “late adapters”; therefore, the counties' capabilities for collaboration and building alliances are not similar. As the organizations are still “taking shape”, “in a start-up phase”, “still in search phase” or “in a construction phase”, the incompleteness of one organization reflects heavily on the others in the service ecosystem.

Other reforms are mentioned in the data as something that has already affected collaboration or will affect it in the future. For example, the reform concerning employment services (implemented in 2025) shifts the responsibility of organizing employment services to municipalities, stressing the whole service system and its cooperation relations.

Organizational and even systemic incompleteness are wide issues affecting and stressing the everyday work of service professionals, for example, in situations where “looking for work tasks” is still “in the search phase”. As organizations go through enormous changes, the relations of cooperation also change. One central cause, as recognized in the data, is “staff turnover”, which may be a consequence of reforms but also of other issues, such as overall heavy workloads and dissatisfaction in the social and healthcare system. As professionals change organizations or units, the so-called tacit knowledge vanishes. This may also burden professionals in partnering organizations handling the service user's processes:

Reforms have also affected such useful joint projects and experiments that originated from municipalities and aimed to improve joint processes and collaboration with various actors, including Kela.

The fourth main frame, cross-cutting all ecosystemic levels, discusses the question of actors' differing modes of action. From a macro-level perspective, this interpretation emerges as a variance in societal tasks and responsibilities. Both Kela and WSCs are described as “strong institutions” with shared responsibilities regarding welfare services, as WSCs are responsible for healthcare and social service tasks, and Kela is responsible for social security tasks. Simultaneously, the data indicate unclarities in the division of labour between the institutions, also affecting the organizational day-to-day life. This appears particularly in questions of social assistance, with Kela being responsible for the “basic part” of social assistance, while WSCs provide the supplementary parts of it.

The differing institutional logics that guide organizational actions emerge from the viewpoints of localness, local knowledge and “local democracy”, often described as the guiding logic of WSCs. In contrast, Kela's role is often described using vocabulary that emphasizes a centralized role (e.g. “nationwide”). As these differing logics guide the organizational solutions, the challenge is the “consolidation” of these logics.

Funding in both institutions is also noted to have different logics, despite their high dependence on government financing:

The complexity of integrating local and nationwide modes of action emerges as real-life challenges between organizations. At the meso level, the different organizational structures hinder cooperation from the nationwide Kela viewpoint, as WSCs act locally, utilizing different solutions and having varying working cultures.

The unclear, ill-suited and unorganized processes or collaborative structures are also interpreted as collaboration challenges. This emerges as a lack of “systematic”, “rational” or “easy-access” means for collaboration, often also seen as “person-related”. Despite the possible strategic guidance, the well-functioning implementation of cooperation may be impossible because cooperation is not “being prioritized” or due to resource-related issues, such as working hours not being specifically allocated for cooperative actions.

From the viewpoint of professionals working at the service interface, one challenge is a lack of clarity, such as an appointed “contact person” or “match” in partner organizations. An unclear labour allocation among professionals may cause tensions. For example, some notions mention service county professionals acting as “interpreters” to assist and guide service users in “their rights”, “reading” and “clarifying” decisions made by Kela.

In addition to different organizations, the interplay between differing professional backgrounds and cultures may also cause tensions, and dissenting expectations concerning the role of professionals emerge in the data. Social work professionals in WSCs note a key challenge of “lacking discretionary power” in questions of social assistance among professionals in Kela. This translates to an inadequate understanding of the “overall situation” and “individual needs” of service users.

In this study, we examined the barriers to interorganizational collaboration in Finnish social and healthcare, viewed analytically as a public service ecosystem. As outcomes, we identified four frames: insufficient level of knowledge and flow of information, institutional and organizational restrictions and scarcity, immature organizations due to reform and differing modes of action (see Table 3). These frames build a picture of the barriers that emerge throughout the levels of the health and care ecosystem, underscoring the relational nature of the ecosystem (cf. Jepperson and Meyer, 2011).

The first frame, insufficient comprehension and flow of information, includes interpretations of hindrances concerning the flow of interaction and a lack of comprehension of the wholeness of the service system but also of service users' needs. The second frame of institutional and organizational restrictions and scarcity implies limited resources, as well as institutional silos and rigidity. The third frame, systemic change and incompleteness, includes discussion around broad reforms and the institutional incompleteness caused by them. Finally, the fourth frame, differing roles and modes of action, discusses the differing institutional tasks and professional culture, as well as unstructured modes of collaboration.

The results highlight the relational nature of reform as a social phenomenon that shapes, and even shakes up, the whole institutional context in which collaboration is undertaken (cf. Wilson et al., 2024). In our results, Finnish social and healthcare services reform is the explanatory element of all frames. The reform affects the barriers, for example, through the flow of information and comprehensions concerning the systemic wholeness but also through the understanding of the situations of partners, the role and tasks of institutions and the resources and financing of the overall service system.

While barriers to collaboration, or at least many of them, would exist without reforms, one could say that reforms cause an extra vortex. Interpretations connect many of the hindrances to reforms as a direct result or as a side effect of a system that is undergoing continuous change and, in many ways, is incomplete. This highlights the emergent nature of the relational aspect of the public service ecosystem, in which reform emerges and characterizes the relations and collaboration processes of diverse institutions, whether they have been a direct target of reform or not (cf. Bartels and Turnbull, 2020). This also becomes apparent in the case of Kela, which has not been under reform but has been forced to adjust its governance and interactions, as reforms have heavily affected the partner organizations.

Reforms not only alter formal structures but also destabilize trust and role expectations among actors (Pollitt and Bouckaert, 2017). This can easily lead to situations in which organizations retreat to protect their own domains, entailing managing control and collaboration in tandem rather than as alternatives to one another (Evans et al., 2024). Indeed, reforms greatly affect the coordination of interactions and connections among diverse actors beyond individual relationships within service systems (cf. Wilson et al., 2024).

From these viewpoints, the focus of the results has been on relational ontology. Instead of individual relationships, we focus on the interactions and interdependence of differing institutions acting in the social and healthcare ecosystem (Bartels and Turnbull, 2020). This focus has also emerged at the micro level in our results, exploring everyday possibilities for collaboration in social and health services instead of singular service encounters between individuals.

In addition to governmental structures, the function of collaboration relies on interdependence among actors (Elston et al., 2023). From a managerial perspective, this implies that collaboration must be assessed strategically (Bryson et al., 2006). A central managerial task is, therefore, to create conditions that enable systematic boundary-spanning collaboration and reduce dependence on individual actors (van Meerkerk and Edelenbos, 2018). To ensure that partnerships do not remain merely symbolic or rhetorical (Dickinson and Glasby, 2010), strengthening collaboration requires granting sufficient discretionary space to the boundary-spanning actors in aligning the interests of various stakeholders (cf. Williams, 2002).

Based on the findings of this study, three implications for the management of collaboration can be identified. First, the focus of management cannot be limited to the internal functioning of individual organizations but must extend to collaborative capacity emerging at organizational interfaces. Since key barriers relate to information flows, unclear responsibilities and differing institutional logics, the problem does not reside within single organizations but in the relationships between them.

Second, in reform contexts, the challenges in leadership extend beyond coordination, reaching instead to the management of uncertainty. Under such conditions, a key task of management is to maintain continuity amid ongoing incompleteness. This requires the maintenance of functional collaborative relationships and the continuous evaluation and institutionalization of emerging practices generated during the reform process.

Third, the findings suggest that the key issue is not simply increasing collaboration but targeting it appropriately. Not all problems or objectives inherently require intensive interorganizational collaboration. Building collaboration in situations in which genuine interdependence is lacking can consume resources without corresponding benefits and may even reduce efficiency. Instead, collaboration should be managed selectively – only where interdependence makes it necessary and where it is likely to generate genuine value.

Building on these managerial implications, we propose that future studies investigate relational perspectives on public service ecosystems through a comparative perspective across different politico-administrative systems. This study acknowledges that there is likely to be differences across various countries and cultures concerning the proximity of ecosystemic actors.

With a relational lens on the public service ecosystem, and with an empirical focus on interorganizational collaboration, this study contributes to the emerging theoretical discussion of relationality in public services, denoting that collaboration is shaped by systemic structures and evolving institutional contexts rather than just interpersonal dynamics. The study introduces questions of knowledge and information, institutional and organizational restrictions and scarcity, maturity of organizations and differing organizational and professional modes of action as key pressure points for relational governance.

As an additional theoretical implication, this study deepens the understanding of reforms and service integration as challenges to relationality. While shaping relations, reforms may create tensions at the meso level of the public service ecosystem as organizations reshape and negotiate their roles and responsibilities. These tensions may also emerge at the micro level as experiencing uncertainty and collaboration difficulties. However, practices at the micro level may challenge interpretations concerning the complications in reforms, for example, in establishing partnerships or well-functioning venues for interaction.

In this study, the identified barriers appear not as isolated but relationally embedded; they emerge from the interactions and interdependencies between institutions (in this case, WSCs and Kela) rather than from individual actors alone. Finally, this study has also been able to identify reforms as changing the institutional landscape, acting as a background vortex, intensifying existing barriers and reshaping institutional relationships. This whirl of events affects organizations not directly targeted by a reform, underlining the interdependent nature of public service ecosystems even further.

It is worth noting that this study has some limitations. First, the four organizations observed here are an important part of the Finnish public service ecosystem but do not form it. It is possible that there are variations between different organizations in how collaboration is seen. For example, geography and local culture and traditions of collaboration may influence the relations of different actors. In addition, the collaboration might look different when observing, for example, third sector or private service providers. Second, due to 56 interviewees, this study relies on rich data. Informants representing different positions, from top and middle management to employee positions, provide us with a wide understanding from different perspectives. However, it cannot grasp the profession-related questions of collaboration, which would also be a key question to study.