From ‘Master and Servants’ to ‘Equal Partners’ in a Health Community: Transforming local mental health services by action research and social innovation

Research context, procedures and participants

This study was conducted at the conclusion of a four-year longitudinal AR project (2020–2023), ‘Leading for Innovation and Collaboration in a Health Community for Mental Health, Aiming to Increase the Value of the Services for the Service Users’. The project aimed to create integrated health and welfare services for joint service users across their units and organisations, defining themselves as a Health Community. Service users were local mental health patients. The 20 participating leaders represented 20 different units within the four local public health and welfare organisations: two municipalities, a mental health hospital and the Norwegian Labour and Welfare Administration (NAV), hence they represented the service providers. For further description of the context, see Senneseth and Berg (2025, p. 72).

To answer the aforementioned research question, qualitative data were generated through evaluations of and reflections on the entire four-year process among the participants. All participants in the leader network were invited to participate in this study and were recruited because they were key decision-makers or frontline leaders from the local service providers. Representing the central public health and welfare system, they had a mandate to make necessary changes proposed by national policies. Altogether, 16 of the 20 invited leaders participated in this study. Before presenting the data and analysis, we further outline the research context by describing the AR project.

The action research project

The ‘actions’ in the AR project were pragmatically chosen collaboratively by the action researcher and the participating leaders by examining the best methods available to address participants’ real-life questions and challenges. These initiatives/actions aligned with the social innovation definition presented above as change in social relations, involving new ways of doing, organising, knowing and framing (Avelino et al., 2019) Thus, these actions are characterised as social innovation initiatives in this article, as presented in Table 1. These actions are not, per se, the data for this study, but are essential to understanding this study’s context and answering the research question.

The portfolio of social innovation initiatives implemented in the AR project enabled the use of various methods new to the context (see Table 1), conducted abductively (Gehman et al., 2018) and systematically integrating emerging knowledge needs with innovative approaches. Networks and temporal organisations (Lundin & Söderholm, 1995), such as service design projects (Patrício et al., 2020), enabled flexible collaboration and learning with opportunities for collective action described as central for transformations, such as realigning normative, institutional and cultural elements; exploring knowledge complementarities and facilitating collective sense-making (Vink & Koskela-Huotari, 2022).

Table 1 was designed to aid reader comprehension (Cloutier & Ravasi, 2021) of the research context by describing the portfolio of social innovation initiatives implemented. Each row describes a social innovation initiative based on four central dimensions in the columns. The first column describes the question the leaders asked and the required knowledge creation. The second column describes the participants, who were recruited strategically to provide epistemically humble (Skelton et al., 2024) and trustworthy knowledge to answer or handle the question in the first column. The third column contains descriptions of the actions taken to generate knowledge and these actions’ outcomes. The fourth column describes the new relational spaces created for collaborative learning and/or governance purposes in each specific social innovation project or process.

Data and analysis

Reflexivity is vital in AR projects, and systemic reflexivity is pivotal in creating transformative capabilities (Moore et al., 2018; Wittmayer et al., 2019). In this study, we zoom in on two such co-reflexive processes at a particular moment in the larger AR project through the participating leader’s evaluations and co-reflections after four years of a process implementing social innovation initiatives together. The specific data developed for this article were generated in two contexts: First, a one-and-a-half-hour focus group interview (Hummelvoll, 2008) was conducted with the four key decision-makers representing the four organisations. Second, a one-day evaluation meeting was conducted in the leader network, providing 4 hours of recorded reflections and discussions. Altogether, 16 leaders, decision-makers and frontline leaders from different units within the four organisations participated in the meeting. The focus group interview and evaluation meeting were audio-recorded and transcribed verbatim. In addition to generating data to answer this article’s research question, these methods created arenas for co-reflection and learning between the participants and a foundation for decisions concerning further collective actions, which are central ingredients in the iterations of AR processes.

Through reflexive thematic analysis (Braun & Clarke, 2022) of the focus group interview and evaluation meeting transcripts, we identified five key themes of processes and outcomes described by the leaders. Initially, the first author became familiarised with the data by reading and listening to the interviews, aiming to grasp the essence of the leaders’ descriptions and stories. During this phase, words, sentences and ideas for potential themes were noted. Next, the first author reread the transcripts to delve deeper into how the leaders described their experiences, focussing on new social relations’ specific contributions and processes concerning new ways of doing, organising, framing and knowing; and descriptions of transformations. Post-it notes then were used to create main theme groups, presented in this article’s findings section. Throughout the analytical process, the themes were discussed continuously with the second author.

Reflections on the action researcher’s role

Action researchers can adopt various roles (Wittmayer & Schäpke, 2014). During the four years of the AR project, the first author became deeply embedded in the case. What can be described as deep knowledge and thick descriptions of the context were obtained by coordinating the portfolio of social innovations, conducting surveys, longitudinally participating in and/or facilitating various formal meetings and participatory processes, conducting informal conversations and multistage focus group interviews, and serving as the project leader in one of the service design projects described in Table 1. The various encounters allowed the action researcher to assume the role of a change agent, engaging as a transformative action researcher (Horlings et al., 2020). As a PhD student in innovation, the first author functioned as a knowledge broker between the `outsider position’ from the academic realm and an ‘insider perspective` from the real-life field of innovation and as an experienced health professional (psychologist).

However, deep involvement can blur awareness of one’s power dynamics; thus, precautions were taken during organisational research with close collaborators, for example, communicating when data were collected for citation purposes to avoid participants feeling monitored. Transparency was another key ethical consideration, with most knowledge cocreated in groups and written reports readily accessible to participants. This way, we also avoided delays typical of traditional research, and the participants could utilise the knowledge immediately.

Conducting AR is described as risking entanglement in ‘dilemmas of power inequalities and complex relational dynamics’ (Bartels & Friedman, 2022, p. 101). As an action researcher, the first author reflected on own stance regarding power dynamics and values embracing equality and justice. This allowed continuous awareness and discussions with participants concerning the process, procedures and what is experienced as equal partners. This helped us develop literacy and consciousness concerning the blind spots existing concerning just participation. Thus, listening to all participants actively and humbly was an aim, including awareness of power balances between the researcher and participants and between participating actors. Furthermore, one’s biases and experiences can influence understanding and interpretations of participants’ stories. Thus, the second author’s critical view has been important, both in the research process and in the analysis and writing of this article.

Research ethics

The Norwegian Centre for Research (NSD, now renamed Sikt. Project number: 464812) approved this study, and we adhered to ethical guidelines throughout. Participation in interviews and surveys was voluntary, with written informed consent obtained from all participants. To maintain anonymity, participants are referred to below as leaders without linking them to their specific organisations, unless it is particularly important for explaining findings (e.g., concerning power relations).

Transforming power relations from ‘master and servant’ to ‘equal partners’

The leaders described how new ways of collaborating in projects and cogoverning in networks and projects transformed their relationships, including changes in their actions and behaviours, which led to transformations in power dynamics between the hospital and municipalities’ representatives over the four-year period. Previously, municipal representatives had viewed hospital representatives as domineering. Municipal leaders described earlier meetings as arenas in which ‘the hospital was just telling the municipality how it should be’ without room for dialogue. Thus, collaboration ‘was not experienced as being on equal footing’. Their past experiences with asymmetrical power relations aligned with a well-known phenomenon described in research from the Norwegian health context (Vik, 2018). A shift towards a partnership on equal footing was experienced during the past four years, initiated by a new hospital director’s behaviour. A municipal leader described how the new way of ‘doing’ collaboration was based on dialogues and exploration of the problem at hand together:

We began discussing the challenges related to collaboration. No one had predetermined decisions about those challenges; we explored them together. This approach was entirely new compared to how the hospital had handled things before.

The new hospital director described a strategic decision to move away from conducting decision-making on behalf of other organisations: ‘I was determined that this had to change. It was a strategic choice. I do not get involved in (what) is not my business’. Self-steering of each organisation’s participation in joint initiatives and decisions concerning whether to make changes in own organisations were established as the new norm for collaboration: ‘This is how we do it now’, one municipal leader said. As the new approach evolved into new routine practices between actors over time, it indicated an institutional change (Scott, 2014) representing a shift in power relations between parties.

Power can be used to exercise ‘power over’ others or to share ‘power with’ others (Avelino, 2021). In this study, we found that the hospital director initially held a dominant position and had power to make changes. The director then shifted towards equal participation among self-steered actors. It is not the organisation, but rather individuals within an organisation and the organisational culture that act. Therefore, how individuals within organisations with dominant power positions choose to exercise their power can impact maintenance or transformation of relationships and power dynamics between actors significantly. This finding demonstrates that one leader can be pivotal in transforming power positions between organisations of hospitals and municipalities by changing their actions.

However, avoidance of exercising power dominance is not enough to create transformation. To create the dynamics of exercising ‘power with’, all relevant actors’ in the Health Community had to be included, come forward, and take on responsibility for own service provision and make necessary changes. For example, by acknowledging how they contributed to harmful practices and changing these practices. The changes in the actors’ mundane practices described actions that we understand as changes in how the actors conduct services for service users, and how they do collaboration between the organisations, aligning with descriptions of new ways of ‘doing’ collaboration as part of a social innovation process, in which the outcome was a transformation in the power balance between the actors. The social innovation portfolio conducted with AR processes provided a transparent and systematic method to include and enable this participation, and allowed the actors to participate on their own terms, aligning with descriptions of epistemic and procedural justice (Skelton et al., 2024). Hence, the awareness and sensitivity to these central AR principles contributed to the transformation from asymmetric to equal partners.

Transforming relational ignorance into relational awareness

Participation in the new relational spaces provided by the social innovation initiatives and the processes of dialogues, cocreation of joint experimentation/iteration of new services, and co-reflections for learning permitted participants to get to know each other and the situated contexts in new ways. What had been relational ignorance between the actors – that is, lack of knowledge about who the other service providers were, which services they provided for the joint service users, and interdependencies and dynamics between the actors – shifted towards relational awareness. This new knowledge and awareness enabled the actors to understand the Health Community’s `parts’ and ‘whole’. As one leader explained it: ‘The way we sat together, from different organisations, was more important than what we did’. As they learned about the many service providers involved, each other’s roles and the variety of services provided, they realised they had been ignorant about each other. One leader burst out to another participant, ‘I did not know you, and I didn’t even know your unit existed!’ By focussing solely on their domains, they had been ignorant of the other actors and their joint responsibilities across organisational borders. One leader stated: ‘The focus was “this is my (domain and)” responsibility’. Through social innovation initiatives, participants identified relational, structural and institutional barriers. They also revealed the blind spots on how their organisation might hinder others’ efforts to serve joint service users: ‘It was constructive; we examined the pain threshold for collaboration and became aware of our sphere of interaction’, one leader stated.

Service users’ needs and experiences while encountering services were described as meaningful and legitimate focal points for all actors, with one leader describing this as the core mandate: ‘The service users’ experience concerns the core of our mandate and task’. This way, systemic awareness developed concerning how the services provided as ‘parts’ and as a ‘whole’ affected service users. This concerned that most actors did not know what services the others provided; thus, when they referred service users to other units/organisations, they created wrong impressions about what the service user could expect from other service providers. This destroyed service users’ trust in the health and welfare system and caused frustration between actors, and everybody felt they wasted time and energy sorting out misunderstandings. Further, new insight concerning how ‘revolving door’ patients were harmed in encounters with different service providers was identified; such as being treated disrespectfully or not getting access to the services they needed when they needed them. This way, practices that caused lack of service integration and resource alignment were identified concretely. This included the high costs, both economically and in human suffering, of either providing harmful services or faulty due to negligence. This new cocreated insight transformed the framing of existing practices from legitimate to illegitimate in a systemic perspective of the health community as a whole, even if practices in each silo organisation were according to policy and routines.

The leaders from all organisations described how these projects enhanced the understanding that all epistemic perspectives were needed to comprehend the ‘value-in-use’ (Vink et al., 2021) for service users to enable the organisation’s ability to align resources. Furthermore, as the understanding that fidelity to sectoral policies did not guarantee safety and value for service users in a systemic perspective, a new narrative emerged; that services delivered according to routines could still contribute to harm to the service users.

These descriptions align with how contextualising and recontextualising (Meier & Dopson, 2019) can foster awareness of the systemic perspective (Senge, 2006) through dialogues and reflexive practices. This, in turn, enhances the mentalisation capacity (Fonagy et al., 2011) of all actors regarding the situated and evolving context as they experience it. We understand these processes as continuous sensemaking (Patrício et al., 2019), enabling participants to flexible move the understanding between the silos and the systemic whole (Hajjar et al., 2024), thus seeing ‘the parts’ and ‘the whole’ together.

We view sensitivity to ensuring participation in these reflexive processes as contributions to the emergence of a collective consciousness (Kinder et al., 2021) towards understanding and taking collective responsibility for services provided (or not), including taking on responsibility for own actions contributing to value co-destructive practices (Järvi et al., 2018) in need for dismantling or exnovation (Hartley & Knell, 2022).

Transforming capability to self-steer own unit in a systemic whole

The empowering shifts in power dynamics and relational awareness of the systemic whole elicited a capability to take on the responsibility to make changes in own units where they were mandated. Thus, new ways of organising between the actors emerged based on self-steering and self-organising changes in own unit towards what they understood was best for the systemic ‘whole’ of the Health Community and their joint service users.

It was crucial for leaders to make strategic decisions about who they wanted to include for participation in the different processes. The leaders needed to ensure that the participants had maximum fitness for the purpose. Sometimes, this could be a person’s professional background. Other times, it could be a person’s collaborative skills, local ‘know-how’ or an informal leadership position that was desirable. Being able to select their participants during all phases of the social innovation initiatives fostered a great sense of ownership: ‘It was new for us to select who we wanted, and this gave us stronger ownership’. This self-steering principle was experienced as a flexible, expedient and respectful approach, enabling actors to address their challenges. The approach also honoured each organisation’s capabilities when external pressures on organisations during the COVID-19 pandemic constantly shifted and impacted organisations differently over time. As one leader noted:

It’s a respectful way of collaborating because we have different laws, policies and organisational contexts. None of us have stable economic situations, and we all face changing challenges.

All the leaders emphasised the importance of stable platforms for collaboration and governance, enabling them and their organisations to comprehend the complex situated landscape. This understanding allowed them to decide how and when to engage.

Collaborating with external organisations also legitimised internal changes. One leader explained how external perspectives helped increase self-awareness of own organisations impact on others, and to dismantle silo barriers in own organisations: ‘It is about enhancing our self-awareness, understanding how we actually work and taking a step back to dismantle these walls we have built around ourselves’. That was also the case when it came to handling their own contributions to the harmful practices identified. The new understanding called for a transformation of own practices, but these had to be aligned with the other actors’ practices. Making changes within one single ‘part’ of the silo systems would not provide sufficient change. They recognised that they all needed to act and handle failures if the `whole` of the services in the Health Community should be integrated and provide value for service users.

We understood the leader’s descriptions of the processes as central features of AR involving a multi-learning environment with joint planning, experimentation and reflexivity to steer practice towards the aim (Patrício et al., 2019; Wittmayer et al., 2019). The outcome was increased capability for all actors to participate in improving services for their joint users through collective governance (Kinder et al., 2021; Senneseth & Berg, 2025).

Transforming the organisational culture from ‘us’ and ‘them’ to ‘we’

The combination of joint projects and the self-steering and self-organising norm within the systemic whole transformed the organisational and collaborative culture. The participants described the changes as ongoing through various encounters in networks and projects, rather than any single project. One leader emphasised that the organisational change in culture occurred between meetings: ‘The portfolio focusses on what we do outside of the meetings; the meetings are about building a shared understanding, developing a common knowledge base and discussing our challenges’.

The leaders reflected on their past tendency to speak negatively about other organisations to service users. ‘It’s a shame, but that’s what we did’, one leader admitted, expressing regret. She described how she now felt genuine pain when hearing negative remarks about collaborators: ‘It really hurts’, she said, placing her hand over her heart. This shift in perspective stemmed from a growing familiarity with other organisations and their challenges. ‘I know that you do…and we do the best we can’, she acknowledged, emphasising that understanding others’ efforts transformed her view and enhanced empathy. Earlier, they had narratives about each other as competitors, fighting over resources and suspecting that others would trick them into delivering services. Now they understood that all actors were short of resources and faced similar pressures. ‘We now understand that the pressure is on all of us, and we understand what it means for each of us’. While acknowledging that transformations take time, one leader expressed hope by observing changes in their frontline staff: ‘It is so obvious that they now have a different understanding and attitude (…) beneficial for our joint service users’.

Participants from all organisations described transformations as a shift from an organisational culture of ‘us and them’ to a ‘we’ culture, acknowledging that no single solution or quick fix exists. A ‘we’ culture is crucial for systemic understanding and collective actions (Kinder et al., 2021; Senge, 2006). The new culture for handling value failure together was developed collectively, breaking with earlier blaming and scapegoating by trying to find ‘the fault’ and ‘the guilty’. Leaders took on collective responsibility for the failure, anticipating that increased demand for changes would be impossible to handle alone in the silos for the future. We understand this as transformation in the narrative on governing to meet future challenges (Wittmayer et al., 2019) towards collective actions and collective goals. The leaders further understood the need for personal engagement to develop their roles (Senneseth & Berg, 2025) and services to exercise ‘power with’ each other (Avelino, 2021) for the future and to conduct collective leadership (Kinder et al., 2021).

Transforming services and harmful practices

The findings indicate that the actors transformed the services they collectively identified as harmful and expensive, thereby realigning practices towards service integration and safety for service users and resource alignment between the organisations. The harmful practices concerned how service providers mislead service users by creating wrong impressions about what they could expect from another service provider, and concerning ‘revolving door’ patients as they were harmed in the encounters with several service providers.

The new understanding called for a transformation of each actor’s own practice, which needed to be aligned with all other actors’ practices. Making changes within one single ‘part’ of the silo systems would not provide sufficient change. As one leader noted, ‘We just cannot do that anymore’. The leaders recognised they all needed to act and handle failures where they were mandated, if the `whole` of the services were to be integrated and provide value for service users. Hence, the focus shifted away from pressuring others to change, towards taking on responsibility to improve one’s own practices, aiming to secure patient safety for service users. This aligns with descriptions of a psychologically safe environment (Edmondson & Lei, 2014) between the actors, creating institutional trust (Finch et al., 2023) and a fundament for addressing value co-destructive practices and iterating new solutions (Hartley & Knell, 2022) together.

Due to the detailed picture of the (lack of) alignment of resources and service integration, and its consequences for service users, each actor could see both the parts and the whole of the Health Community, framing the problem as shared, yet acknowledging their own contributions. Recognising their contributions to the problem and learning about others’ willingness to make changes empowered each actor to alter their practices and service delivery. However, they had not yet reached a state of providing no harm to service users. ‘That is what we are trying to experiment with now’. Hence, the actors had to endure the uncertainty about whether the new solutions would create safe services for the users.