Balancing between collaboration and autonomy: leadership insights from a Swedish preschool research-practice partnership Open Access

In decentralized educational systems, school leaders work at the intersection of autonomy, control and collaboration. Leaders are expected to exercise professional judgment with a high degree of independence, without being tightly constrained by external oversight, while still being accountable for outcomes. This autonomy and the authority to make key decisions (Ingersoll, 2009) increasingly coexist with demands for inclusive and collaborative decision-making involving local stakeholders (Jarl et al., 2024). As a result, school leaders must continuously balance their professional judgments with external accountability pressures, field-level expectations and the need to engage meaningfully with community needs.

In Sweden, recent reforms of New Public Management (NPM) and New Public Governance (NPG) have further reshaped leadership roles in education by expanding the autonomy of municipalities. School leaders are now embedded in multifaceted governance networks where political mandates, professional responsibilities and stakeholder demands intersect. The expectations on school leaders are particularly acute in preschools, partly due to the ambiguous mission outlined in the national curriculum (Sundström, 2024) and partly due to the variation in local staffing (Berg, 2022). Together, these forces make school leadership increasingly complex, requiring a balance between professional discretion and local accountability.

Research-practice partnerships (RPPs) offer one structured approach to support educational leaders in such settings. Defined as long-term collaborations between researchers and practitioners (Farrell et al., 2021), RPPs aim to improve education through shared inquiry and sustained engagement. They advance their goals by engaging diverse sources of expertise, designing structures for participation and joint work and supporting the application of research towards local objectives (Farrell et al., 2021; Weddle et al., 2021).

For the Swedish preschool context, RPPs may be especially helpful given the tensions between (1) the preschool’s ambiguous mission (Sundström, 2024), with an increased attention to the mission of teaching and (2) its varying local conditions, often with a lack of certified preschool teachers (Berg, 2022). These conditions create a context where leaders must respond to both policy-level ambiguity and on-the-ground constraints unique to Swedish preschool settings. Several features of the RPP model lend themselves to this setting, including explicit attention to power dynamics across a diverse range of partners and the design of collaborative infrastructure to support evidence-informed decision-making.

At the same time, the potential for supportive mechanisms via an RPP must be realized from the ground up, which means success is not guaranteed. Although existing RPP research suggests positive outcomes at the elementary and secondary education levels, there is limited attention to early childhood settings (e.g. Brotman et al., 2021; Datnow et al., 2023; Schindler et al., 2017). Moreover, most studies are based in the United States of America, where accountability systems, governance structures and leadership expectations differ significantly from those in Sweden (Sjölund et al., 2023a, b; Poekert et al., 2020). In contrast, unique features of the educational systems in Norway and Sweden, such as high decentralization, high trust and low accountability, might influence the potential for RPPs to succeed. In Sweden, in particular, the relatively weak regulatory frameworks that shape teachers’ work (Ryve and Hemmi, 2019; Wermke, 2011; Wermke and Forsberg, 2017) further contribute to a decentralized system. These features situate municipal preschool leaders in central, context-sensitive leadership roles (Nordholm et al., 2023), highlighting the need to explore how collaborative approaches to research and development, like RPPs, might function in such environments.

This study aims to explore if and how the RPP approach might support preschool leaders in navigating the new power dynamics of collaborative decision-making introduced by New Public reforms and, if so, how it helps. Specifically, we examine the decision-making processes within a Swedish RPP focused on preschool improvement. The work makes several contributions. First, we examine how RPPs can serve as a framework within which educational leaders draw on research and researchers as they navigate the multiple demands for autonomy and collaboration. Second, we demonstrate the dynamic process of negotiation over time and bring forward the idea of “collaborative autonomy” as the balance that this RPP developed in their work together. Finally, by exploring the application of the RPP model in a non-USA context, this study contributes to a broader conversation about how and whether RPPs can be adapted to different countries and policy environments.

Below, we outline the governance context of Sweden’s decentralized education system, with a focus on the leadership challenges faced in preschool settings. We then introduce RPPs as a promising model for supporting collaborative decision-making and educational improvement in this context.

Swedish schools operate within a decentralized educational system, where local municipalities hold significant authority over decisions regarding governance and administration. This structure allows for a high degree of local autonomy, enabling municipality and individual school leaders to make decisions for their local school communities. Historically, Swedish principals and educators have had professional discretion in implementing curricular materials, designing pedagogical approaches and making decisions in classrooms (Wermke and Höstfält, 2014). This autonomy reflects a broader educational philosophy that values professional judgment and locally adapted solutions. According to the law (SFS, 2010:800), it is the principal’s responsibility to lead and coordinate the educational work in a preschool.

Despite this flexibility, schools face growing accountability pressures (Nordholm et al., 2023). Municipalities are responsible for monitoring school performance in order to ensure that local schools adhere to national guidelines. To align with the demands of greater administrative oversight and accountability at the regional level, school leaders and teachers are now expected to document and evaluate their work more systematically. However, while these measures aim to ensure consistency across the system, they may limit the professional discretion that teachers and principals have traditionally enjoyed. As a result, school-level autonomy now becomes a contested space.

In this context, autonomy in school leadership extends beyond local decision-making authority to include control over core educational functions, such as personnel, teaching and resource allocation (Hanushek et al., 2013; Cheng et al., 2016; Ingersoll, 1996, 2009; Wermke et al., 2022). This autonomy becomes especially important in collaborative efforts, such as RPPs, where the use of research can inadvertently undermine practitioner independence if not carefully managed (Sjölund et al., 2022; Cooper et al., 2020).

To better understand how leaders navigate these dynamics, the concept of distributed leadership may provide a valuable lens. Unlike hierarchical models of leadership, a distributed view emphasizes how leadership emerges through interactions and shared responsibilities (Spillane and Camburn, 2006; Harris and DeFlaminis, 2016). It acknowledges that those outside of formal leadership positions can enact effective leadership practice and is most impactful when intentionally aligned with clear purposes. Research suggests that well-supported distributed leadership enhances trust, team effectiveness and instructional capacity (Hargreaves and Fink, 2006). In collaborative efforts such as RPPs, it can help balance authority and expertise among diverse actors while supporting sustainable change.

These contemporary conditions can be viewed through the lens of historical reform. For much of the 20th century, Swedish education followed a bureaucratic model, in which central authorities made policy and civil servants implemented it with fidelity. In the 1980s, however, this approach began to shift toward NPM, which introduced private sector efficiency logics and performance measurement into education (Jarl et al., 2024). More recently, NPG has gained ground, promoting networked collaboration among schools, municipalities, civil society and the private sector (Peters and Pierre, 2017). These historical trends have led to principals’ being embedded in regional and local governance structures while also being responsive to top-down expectations for results.

Overall, school leaders, including those in preschools, must navigate overlapping and sometimes contradictory demands. They are expected to exercise professional autonomy, comply with external accountability systems, participate in collaborative initiatives and ensure the delivery of high-quality instruction (Jarl et al., 2024). These tensions define the complexity of educational leadership in Sweden’s decentralized system and create the conditions under which collaborative models, such as RPPs, must operate.

Building on the broader governance context in Sweden, we now turn to the specific challenges faced by preschool leaders in Sweden. Preschools play a central role in children’s cognitive, emotional and social development, starting as early as one year old (Persson, 2015). Guided by the Education Act (SFS, 2010:800) and the national curriculum (Swedish National Agency of Education, 1998/2018), preschool educators are tasked with promoting equity and delivering high-quality educational experiences.

However, expansion and decentralization have contributed to uneven conditions across preschools. Local factors, including teacher qualifications, class size and resource availability, shape variation in educational opportunities in preschools. A key tension arises from growing expectations for structured teaching, which are not always matched by adequate staffing, leading to a shortage of certified preschool teachers (Berg, 2022; Frankenberg et al., 2018; Sheridan et al., 2019). In addition, the increasing diversity of the student population adds further complexity to delivering consistent, inclusive and pedagogically sound education (Sheridan, 2009).

These pressures underscore the need for sustained, practice-relevant professional development to support preschool leaders and educators in addressing local needs (i.e. mathematics, science, digital literacy and sustainable development), especially in areas where preschools may lack resources or existing expertise (Elliott et al., 2020; Sheridan et al., 2011). Cross-organizational collaborations, including university-community partnerships and early learning center affiliations, have been shown to enhance teacher preparation and improve educational quality (Lees and Kennedy, 2017). Key success factors for these collaborations include field-based learning, shared goals, and research-driven preschool development, with frameworks such as the “third space concept helping to bridge theory and practice (Lees and Kennedy, 2017; Del Grosso et al., 2014; Mörk, 2022).

Given these persistent tensions that Swedish preschool educators and leaders face, we see a growing need for collaborative models that can support context-sensitive leadership and improvement. We next consider how RPPs, as one such model, might address the unique demands facing Swedish preschool leaders.

RPPs have experienced a period of strong growth in the USA (Arce-Trigatti et al., 2018), and the concept has gained traction internationally, including within the Nordic context (e.g. Fjørtoft and Sandvik, 2021; Prøitz and Rye, 2023; Sjölund and Lindvall, 2024), where Nordic scholars have identified RPPs as a way to frame collaborative efforts between education researchers and practitioners (e.g. Fjørtoft and Sandvik, 2021; Korhonen et al., 2024; Sjölund and Lindvall, 2024). RPPs are distinguished through five core principles, as suggested by a field-driven definition from 2021 (Farrell et al., 2021). In general, RPPs are theorized to (1) have an orientation towards long-term collaboration, (2) work toward educational improvement or equitable transformation, (3) feature engagement with research as a leading activity, (4) be intentionally organized to bring together a variety of expertise and (5) employ strategies to shift power relations in research endeavors to ensure equitable representation across partners. Research on RPPs highlights their potential to enhance practitioners’ access to research, strengthen research skills and improve student outcomes (Coburn and Penuel, 2016; Farrell et al., 2018), making them a promising model for reform efforts (Datnow, 2020; Bassok et al., 2021), especially in challenging contexts (Popa et al., 2023).

In terms of early childhood and preschool settings, RPP examples continue to emerge, emphasizing the co-creation of knowledge among researchers, practitioners and occasionally parents; supporting ongoing professional learning and adapting research to local contexts (Brotman et al., 2021; Datnow et al., 2023; Schindler et al., 2017). With strong leadership and stakeholder engagement, RPPs can foster the conditions needed to implement and scale evidence-based practices while also contributing to professional development through reflective and practice-oriented learning (Datnow et al., 2023). RPPs may thus be a promising model for improving preschool education, particularly given their focus on inclusive participation and attention to power dynamics within research in service of improvement (Farrell et al., 2021).

Despite their potential, RPPs face common challenges. For instance, one issue lies in the cultural and epistemological differences between researchers and practitioners, which can contribute to misaligned expectations or communication breakdowns (Sjölund and Lindvall, 2024; Farrell et al., 2019). These tensions may have roots in traditional power dynamics typically found in research, where practitioners may feel like passive recipients rather than active co-creators of knowledge. Further complicating the work are practical barriers, such as syncing up schedules, navigating different institutional demands and meeting the pace of work, often in practice (Donovan et al., 2021).

At the same time, through explicit attention to the health and development of the partnership itself (Farrell et al., 2021; Henrick et al., 2023), RPPs constitute a model where, in practice, initial structures such as organizational hierarchies and decision-making structures can productively be contested and renegotiated as the RPP evolves (Anderson, 2023; Coburn et al., 2008). Additionally, participants in RPPs tend to continuously renegotiate their roles and relationships to adapt to the evolving partnership dynamics and needs (Farrell et al., 2019; Sjölund, 2023; Sjölund et al., 2023a, b). RPPs have the potential to balance priorities and power differences through considering factors such as early and ongoing engagement of partners, adequate representation of diverse perspectives, the presence of trusting relationships (Meyer et al., 2023) and partners honoring one another’s needs and goals with genuine concern in a way that reinforces their inherent value or worth (Riedy and Penuel, 2024). Given the described practical usefulness of the RPP model above and that the RPP principles pay explicit attention to, for instance, how to (1) intentionally organize partnerships to bring together a variety of expertise and (2) employ strategies to shift power relations in research endeavors, we hypothesize that RPPs could be quite a useful model for the Swedish preschool context described earlier.

This study therefore explores whether the RPP model supports preschool education leaders in navigating the new dynamics of collaborative decision-making introduced by New Public reforms and, if so, how it helps. We ask: How do preschool leaders and researchers navigate decision-making processes within an RPP in the context of Sweden’s decentralized education system?

We begin by describing the context of an RPP focused on preschool within the Swedish context, followed by an outline of the data collection and analysis, with a focus on an approach to analyzing power and equity in RPPs via decision-making.

This study (Figure 1) examines a collaborative research and preschool improvement program that integrated principles of sustainable development into preschool settings. The program involved three main entities: educational practitioners (including preschool teachers, principals and municipal office employees), researchers and a nonprofit, independent research and improvement institute. The institute was funded by both municipal and private school authorities who supported both practice-oriented research and improvement initiatives. The program spanned multiple cities and municipalities across Sweden, involving 300 practitioners, 3 researchers and an institute representative.

At the beginning of the RPP, the partners agreed on a program and a research plan as a baseline for their collaborative efforts. The program plan outlined roles and responsibilities for the participants, where local process leaders, appointed by school authorities, would plan and lead development work. At the same time, principals would form and lead development groups in their preschools. A program board (including representatives from the municipal office, researchers and the institute) would oversee the program, making decisions on the plan and budget. The group met regularly, engaged in seminars and drew on an electronic project platform for communication and documentation. Institute project leaders served as coordinators who provided support and supervision through the division of labor and active leadership throughout the program. Broadly, researchers aimed to develop theoretical and practical knowledge regarding preschool staff’s ability to organize and lead educational activities for sustainability through focus groups, documentation and recordings of educational sessions. The research aimed to explore the scientific basis and proven practices in sustainability education, as well as the communication and participation of children and leadership practices at various organizational levels.

The program aligned with the five core principles of RPPs outlined in Farrell et al. (2021). First, although it operated within a time-limited framework (four years), the program fostered long-term collaboration, with partners emphasizing ongoing engagement beyond the funding period to support sustainable development education in preschools. Second, by centering their work on improving education for sustainable development, the team aligned with RPPs’ goal of educational improvement or transformation. Improvement through research inquiry was central to this partnership, with initial phases involving scoping surveys and observations, followed by iterative research and improvement activities. The program intentionally integrated diverse expertise, a hallmark of RPPs. Finally, the program actively worked to shift power dynamics, with leadership roles designed to balance the influence of researchers and practitioners, such as a preschool municipal leader chairing the board and a neutral facilitator prompting discussions on roles and responsibilities.

The first author acted as an “external” researcher for the program and did not actively participate in shaping it. However, to build trust, when attending meetings and improvement seminars, the first author introduced himself and responded to questions as they were asked. The first author also socialized during breaks to further build relationships with the program participants. Co-authors #2 and #3 were not involved in the program or data collection but contributed to the analysis and writing of this article.

The data for this study comprise video-recorded observations and transcripts from two types of meetings within the program, collected between autumn 2022 and spring 2023. In total, the first author gathered approximately 33 h of recorded meetings: 15 h of board meetings and 18 h of process manager meetings. The program board consisted of 9 preschool leaders (occupational roles included municipal or private school authority head of preschool department), one institute representative and one researcher. The process manager group consisted of fifteen preschool leaders (with occupational roles primarily as municipal preschool development leaders), one institute representative and two researchers. Most of the data were collected during the COVID-19 pandemic, meaning that most meetings were conducted digitally via Zoom. These video recordings offer a rare opportunity to gain in-depth insights into the interactions between researchers and practitioners within the context of RPPs (Coburn and Penuel, 2016). To maintain anonymity, as far as possible, we do not reveal the program’s name or the organizations involved. Additionally, in the data, we use generic terms like “researcher” and “process manager.”

To explore leadership in the Swedish collaborative context, we draw on the work of Ingersoll (1996, 2009), who suggests that power within an organization resides with those who control the most critical decisions. To answer our research question, we iteratively read and analyzed transcripts across the four themes to examine how power and authority were negotiated within the existing decision-making system. Fairclough’s (2003) concept of antonyms (conflicting meanings) has guided us in how participants conceptualize power dynamics by exploring linguistic contrasts between terms that express opposing ideas about control and autonomy. For example, concerning empowerment versus disempowerment, participants discussed local improvement work, using language that emphasizes autonomy: “There’s no right or wrong in this, I think, so if we feel this way locally, we act based on our context […]” Here, autonomy in local decision-making contrasts with feelings of being “researched on” (as highlighted by the institute representatives in a subsequent meeting), which reflects disempowerment or a lack of control over research decisions. These contrasting expressions reveal tensions between perceived autonomy in practice and constraints imposed by the research process.

By the end of our iterative process, we identified broad key areas for decisions: (1) autonomy in development and research, (2) research partners’ involvement in process manager discussions, (3) practice partners’ involvement in research and (4) negotiating collaborative autonomy.

A central issue for the Swedish RPP focused on supporting preschool leaders and educators was navigating the balance between autonomy, control and collaboration. While there was general agreement that programmatic decisions should be made jointly, decision-making regarding research and local development work was more contested. In this RPP, researchers and practitioners tended to work independently with their data, using it for different purposes. Researchers used it for analysis and knowledge production, while practitioners used it for local improvement efforts. However, as collaboration deepened, questions arose about the extent to which researchers should be involved in practice and vice versa.

The first theme indicated a consensus amongst partners that decisions related to research and local development work should be made autonomously, rather than jointly, as is typically done for overall program activities. Both researchers and practitioners describe a division of power regarding research and development, which surfaced during a board meeting discussing documentation from the use of an observation protocol. As one researcher argued in the meeting:

In this quote, the researcher contrasted the different kinds of data used for development and research. In their eyes, researchers work with the aggregated research data, while practitioners work with the unprocessed development data. Implicit, then, was that researchers and practitioners would have more autonomy on how to use “their” data. This quote highlights the idea that researchers “pull out” with the research data to conduct analysis, and practitioners “pull out” with their data to inform their development work. While they continued to hold meetings and improvement seminars, informing and discussing with each other, researchers and practitioners were not practically involved to a great extent with each other’s work. This pattern was articulated both by another researcher and a preschool leader in the same meeting:

Again, these two quotes establish the distinction between data used for research on the one hand and local improvement work on the other. Hence, while regional leaders did receive data, knowledge and some guidance from researchers, there was limited involvement in each other’s research or improvement work. In sum, it is emphasized in this first part of the narrative that researchers and practitioners do not feel a need to participate in deciding over each other’s more practical work.

Over time, however, the boundaries between collaboration, autonomy and control shifted. A second theme emerged from one of the process manager meetings regarding the involvement of researchers in process manager group discussions. Before the meeting described below, the research group had limited presence in the process leaders’ conversations, instead of gathering in their own group. However, this dynamic had now begun to change, with an increased openness to dialog and co-creation.

During the meeting, the previous division of responsibilities was problematized, and the researchers themselves expressed a need for greater insight and more precise feedback from the local contexts:

Similarly, process leaders sought a more active presence from the researchers, not just as knowledge providers in professional development but also as critical discussion partners:

Furthermore, educators noted the importance of researchers not only collecting data but also engaging in the joint learning process:

This discussion marks a clear shift away from the previous division of roles toward a more integrated and interactive relationship between researchers and process leaders. Here, the negotiations can be seen as questioning the last consensus on development work and research. The particular focus here was on educational practice, where the researchers wanted to gain a deeper understanding of local practices. This preference was supported by the process managers who similarly conveyed that they wanted researchers to be more involved in challenging their local practices.

Also at this time, a third theme emerged, based on discussions regarding the analysis of data collected from conversations with children. At this point, the team renegotiated the involvement of practice partners in research. This was part of the program that followed the use of the observation protocols discussed in the first theme. For context, the following quote is extracted from the institute representative presenting results from a survey evaluation of the last improvement seminar with preschool teachers.

The institute representative’s comment about “building knowledge together” reflects a desire for greater influence over the research focus, particularly a stronger emphasis on content. While the teachers did not necessarily seek to conduct the research themselves, they wanted a say in shaping what was being studied.

The following quote emerged during the discussion after the institute representative’s presentation, where preschool leaders expressed that it had not been clear from the outset that researchers intended to analyze how teachers conducted the conversations. Their initial expectation was that the analysis would focus on the content of the conversations:

From this discussion, practitioners (both preschool leaders and teachers) want to be involved in deciding, or at the very least informed of, the process and focus of research. The above quotes convey an evolution from information exchange between two separate communities towards greater integration.

In the third stage, we saw an explicit negotiation around the level of involvement of researchers in preschool development work. The conversation in the process leader group reflected ongoing negotiation, where the desire for shared work across research and practice lines became more explicit and evident. Instead of researchers positioned as knowledge providers and practitioners as implementers, participants from both groups discussed the level and nature of guidance they sought from one another. While autonomy remained central, there was an expressed preference for structured input from researchers that does not compromise local decision-making:

This quote illustrates the shift from viewing research as external to practice to recognizing that it can support local development and improvement. The process manager went on to explain that popular science articles, initially framed as a dissemination tool, were instead seen as potential catalysts for professional learning:

The group did not seek a one-size-fits-all directive. Instead, they argued for structured flexibility, where a common framework provided direction while allowing practitioners to align the work with their contexts:

The discussion moved towards a new agreement, where researchers would focus on transformative working methods (i.e. providing tools and processes for local adaptation) rather than prescribing specific themes or dictating content:

Throughout these phases, we see that involvement was not predefined or settled but continually shaped through dialog and negotiation. While researchers still maintained an analytical perspective, their role was no longer separate from the developmental process. Instead, knowledge production and local change efforts became intertwined and negotiated through ongoing interaction.

Our study aimed to investigate how negotiations across autonomy, control and collaboration in a decentralized educational system influence decision-making processes and governance structures in a preschool-based RPP and whether and to what extent the RPP model was supportive. Our findings suggest that these negotiations were dynamic and evolving, initially emphasizing autonomy but later shifting towards what we define as “collaborative autonomy.” Below, we discuss three key findings of school leadership in a broader perspective as related to the Nordic welfare state context and NPG in general and the preschool context in particular.

The RPP program’s established meeting structure, led by the institute, can be understood as a response to the evolving challenges of decentralized preschool leadership within the Nordic welfare state. As discussed earlier, the shift from the traditional bureaucratic model to NPM and, more recently, to NPG altered the landscape of educational leadership. Educators and researchers navigated their RPP efforts within the context of NPG, with its emphasis on collaborative decision-making among various stakeholders. An alignment with RPP principles is evident in the renegotiation of the decision-making structure, representing a deliberate shift towards more distributed leadership (Spillane and Camburn, 2006). Given the challenges Swedish preschools face with an ambiguous mission (Sundström, 2024) and the complexity of teaching themes such as sustainable development (Elliott et al., 2020), finding a practical approach to collaborative decision-making is crucial for improving preschool practices in these areas.

Our first finding, that both researchers and practitioners initially favored autonomy over shared decision-making, reflects the initial structure of decision-making. This theme mirrors the broader challenges faced by school leaders in navigating the decentralized governance landscape, where the distribution of power and influence required them to balance independence and collective action. The initial push for autonomy among both researchers and practitioners can be seen as a strategic adaptation to the decentralized management systems in education, where local actors are increasingly expected to take ownership of their decisions (Jarl et al., 2024).

Our second finding, that researchers later sought greater insight into local contexts while practitioners wanted structured but non-prescriptive guidance, demonstrated how tensions arose and began reshaping governance structures in the partnership. The negotiation that occurs in the following two moments – (1) involvement of researchers in process leader discussions and (2) practice partners’ involvement in research – highlights the complexities involved in maintaining the balance between collaboration and autonomy. The relative autonomy was renegotiated when both process managers and researchers expressed a need for closer communication and exchange of ideas and information at process manager meetings. Researchers sought more information on local contexts, and process managers wanted to be challenged in their thinking based on the researchers’ expertise. This is reasonable when faced with leading improvement in a complex area, such as sustainability in preschool (Elliott et al., 2020; Sheridan et al., 2011), when educators have access to experts in the field who can help inform and challenge local practices. Hence, the RPP participants renegotiated the leadership structure of the collaboration in relation to improvement work to better leverage the expertise of researchers.

Secondly, the practitioners’ frustration with the researchers’ analysis reflected the broader challenges inherent in decentralized governance systems, where a lack of coordination can lead to misunderstandings and conflicts (Nordholm et al., 2023). Therefore, while the RPP’s governance structure is designed to support collaborative decision-making, the practical implementation of these principles can still encounter significant obstacles, not least when balancing power and participation. In this theme, researchers and practitioners initiated a renegotiation of research practices to incorporate the practice perspective better.

Finally, the shift towards “collaborative autonomy” in the final phase of their work illustrates the negotiated balance between autonomy and shared leadership in decision-making. Researchers and process leaders explicitly negotiated governance, focusing on themes for further development work, which portrayed the improved collaborative management that emerged throughout the negotiation of their work over time. The “new” consensus among participants in the RPP was a stance of involvement and a kind of mutual and collaborative decision-making structure. While researchers were still responsible for research and practitioners were still responsible for development, a negotiation on the level of involvement occurred, leading to new information sharing and a critical challenging of perspectives. We describe the newly established roles and responsibilities as “collaborative autonomy,” defined as maintaining autonomy for researchers and practitioners in collaborative efforts while at the same time utilizing the differences in expertise through extensive information exchange and critical challenging of perspectives.

This is also reflected in the findings of this study, which suggest that while Swedish RPPs might aim to uphold collaborative ideals (Prøitz and Rye, 2023), the actual practice of this RPP initially reflects a more conventional division of roles. Researchers and practitioners, despite their collaboration, operate somewhat independently, particularly in the research phases. This division highlights how the adaptation of RPP principles is shaped by the specific features of the Swedish context, characterized by high levels of autonomy, reinforcing that the degree of collaboration between researchers and practitioners is inherently context dependent. It underscores how RPPs are subject to local interpretation based on varying goals and conditions (Meyer et al., 2023). The independent lines drawn between researcher and practitioner shifted during their engagement, reflecting a deepened understanding and appreciation for each partner’s contribution to the work, which Riedy and Penuel (2024) described as an important factor in establishing trust and navigating power differences in RPPs.

For Swedish RPPs to fully embrace the principles of involvement and equal power and perhaps alleviate the need for such an extensive renegotiation (Farrell et al., 2021), a shift towards greater practical engagement of practitioners and researchers at the outset of the research process may be necessary (Meyer et al., 2023). This shift would require addressing the historical traditions of educational research in Sweden and confronting the existing power dynamics that continue to shape research-practitioner relationships. Moving forward, in some cases, Swedish RPPs may consider adopting a more integrative approach to ensure that practitioners are valued for their knowledge and actively involved in shaping the research process from start to finish (Meyer et al., 2023). Similarly, it could ensure that researchers might be more involved and informed in local development work. The concept of “collaborative autonomy” may help facilitate conversations aimed at these ends.

This study illustrates how RPPs can serve as a model for navigating the complex environment of Swedish preschools, which operate within a decentralized system where leaders must balance professional autonomy, accountability demands and collaboration with various groups. Our findings indicate that while initial interactions in the RPP reflected a preference for autonomy and traditional role divisions, ongoing negotiations within the partnership gradually led to a more integrated form of collaboration. This evolving dynamic – what we term collaborative autonomy – captures how researchers and practitioners can maintain their respective roles while actively engaging with one another to support educational improvement through research inquiry.

These findings reflect the broader shifts in Sweden’s educational governance, where decentralization and NPG have shaped the realities of leadership. Preschool leaders, positioned at the intersection of multiple expectations, must navigate tensions between autonomy, collaboration and control. The RPP examined here mirrors these tensions and offers insight into how structured collaboration can support leadership in such contexts. However, the study also reveals that achieving this balance presents challenges. The desire for professional independence can run counter to the need for shared accountability, and misunderstandings about roles or expectations can strain collaboration. The evolution of the partnership necessitated deliberate efforts to build trust, promote transparency and create an environment for reciprocal learning. These findings suggest that while collaborative autonomy can strengthen RPPs, its success depends on sustained dialogue, clearly negotiated responsibilities and a willingness to adapt.

Further research could explore how the concept of collaborative autonomy evolves across different educational and governance contexts. While this study demonstrates that collaborative autonomy can facilitate more balanced and participatory decision-making in a Swedish preschool RPP, such a model may not be suitable for all partnerships. In systems with more centralized authority or rigid accountability frameworks, the space for negotiated autonomy may be more limited, and/or the dynamics between researchers and practitioners may follow different logics. Understanding these evolving dynamics could offer critical implications for the design and governance of future RPPs.

In conclusion, this study contributes to the growing literature on RPPs by showing how decision-making in preschool partnerships evolves through ongoing negotiation of autonomy, control and collaboration. The shift from independent roles to a more integrated approach highlights the need for RPP models that are flexible and contextually grounded. Collaborative autonomy, as both a conceptual and practical tool, offers a way to navigate these tensions, enabling RPPs to meet the demands of decentralized systems while supporting educational improvement and leadership development.