At the Interplay Between Needs and Expectations: Regional Perspectives on Being an Intermediary Support Structure in Knowledge Governance

Introduction: During the last decade, structures for ‘knowledge governance’ as a means of strengthening the quality of social services has been used by the Swedish government and national authorities. In this process, regional collaboration and support structures (RCSSs) have been identified as a key asset as an intermediary link between national and local actors. The aim of this study was to explore the perspectives of these regional actors on expectations associated with being an intermediary support structure in knowledge governance processes in social services in Sweden. Methods: The study uses an abductive, qualitative approach, mainly with open questions in an empirical-driven manner (inductively) and partly in a theory-driven manner (deductively). Interviews with representatives from the RCSSs were used for data collection. Conventional content analysis with elements of directed content analysis of the data material was used. Results: A variety of expectations on RCSSs was reported from the national and local levels. According to the respondents, the feeling of inadequacy in relation to the perceived expectations was recurrent because the expectations were hard to meet. They also experienced role ambiguity and that their mission was unclear. Conclusions: The findings show that RCSSs seem to handle the situation by considering, assessing, and prioritising different actors’ needs and expectations in relation to their own perception of the intermediary mission. The findings also indicate a lack of a collective sense-making process with other key actors on different levels in the knowledge governance system. To reach the core purpose of knowledge governance, it is necessary to establish a common understanding of the relationship between expectations and prerequisites for facilitation between national, regional, and local actors.


Introduction
Extensive national and local efforts are currently being made in Sweden to improve the conditions for social services to provide a practice based on best available knowledge.Structures for governance based on knowledge as a means for improvements and quality insurance have been developed and used by the Swedish Government since the late 1990s.In the Government bill (Ds 2014:9), knowledge governance has been defined as 'a system or processes aimed at providing evidence-based practice where quality-assured knowledge is spread and used, while eliminating non-evidence-based or harmful methods' (our translation).It has also been stated that knowledge governance consists of all strategic decisions and activities concerning (1) the implementation of methods and work practices that are based on the best available knowledge, and (2) the development and maintenance of the necessary structures and processes for knowledge development, for example, follow-ups, improvement efforts, research, knowledge valuation, development of guidelines, knowledge spread, and so on (SALAR, 2017).
This study is part of a project with the overall aim of advancing understanding of the prerequisites for effective knowledge governance processes.The case setting is Sweden and ongoing efforts in knowledge governance for social services.
Regional Collaboration and Support Structures: Intermediaries Facilitating Knowledge Governance Processes Knowledge governance involves actors at national, regional, and local levels, and their associations and surrounding networks.At the regional level, the so-called regional collaboration and support structures (RCSSs) are viewed as key actors, expected to take on an intermediary role in knowledge governance processes (Hjelte et al., 2022).They are assumed to be the link between national and local levels in knowledge governance processes and to be facilitators of the development and implementation of an evidence-based social services practice through collaborative relationships (Denvall & Johansson, 2012;Hjelte et al., 2022).It has been concluded previously that human service organisations, such as social services, may benefit substantially from the strategic involvement of boundary-spanning intermediary organisations with the purpose of contributing to the utilisation of research findings to achieve an evidence-based practice (Graaf et al., 2017).
An intermediate organisation has been defined as an organisation that functions in the midst of the users and producers of knowledge (Clayton et al., 2018;Smedlund, 2006), and they have been stated to exist between the government and the private sector (Vonortas, 2002).In the literature, organisations or actors considered to act as intermediaries are very diverse, and the concept of an intermediary has been used interchangeably between contexts (Mignon & Kanda, 2018).Mignon and Kanda (2018) compared different intermediaries based on differences in terms of their source of funding, the scope of their intermediation, and their target recipients.They concluded that these differences have an impact on intermediaries' short-term or long-term orientation, actor-level or system-level focus, and demand-side or supply-side target.Intermediary organisations in social services have been described previously as taking on various roles and functions to help bridge knowledge-practice gaps by facilitating the implementation of evidence to practice (Franks & Bory, 2017).
The role and function of organisations that link resource systems with user systems, for example, the RCSSs, have been stated to be important but still understudied issues (Proctor et al., 2019), and research on organised facilitation of system-wide planned change and development from the intermediary position is not commonplace.Facilitating, in a broad sense, has been said to mean providing opportunities to other people by educating, gathering and distributing resources, influencing regulations, developing local rules, and creating 'spaces' for others to act.This includes establishing, nurturing, adjusting, and altering connections between different actors (Hakkarainen & Hyysalo, 2016).Hakkarainen and Hyysalo (2016) concluded that the nature of intermediation consists of a range of tasks, not only concerning engagements that may be typically thought of as 'facilitation' .Brokering contacts and interactions between different actors, as well as facilitating their work and learning, are highlighted.The authors further concluded that the complexity of intermediary work, including the fact that challenges that may occur are hard to predict, demands high capability of intermediaries to adjust their role and actions to changing circumstances in context with multiple stakeholders.It has also been suggested that it is necessary to consider variation, use a wide range of methods, work longitudinally and with sensitivity and understanding of change processes (Hupe, 2011).Proctor et al. (2019) studied 119 intermediary organisations with the aim of disseminating evidence-based interventions in child behavioural change.They reported that there is no uniform approach used by intermediary organisations when purveying or facilitating implementation of evidence-based practice and that intermediaries seem to use a variety of combined strategies in such efforts.
Challenges that may impede intermediary organisations have been suggested to include inadequate funding, insufficient readiness for change in the adopter context, and a lack of an adequate timeline for implementation (Franks & Bory, 2017).Westerlund et al. (2015) concluded that it is crucial to clarify the division of roles and responsibilities between managers, facilitators, and others in planned change processes.They suggest that role ambiguity and unclear expectations may substantially impede facilitation efforts.
Although research on regional actors' intermediary role as facilitators in knowledge governance in social services is relatively scarce, some challenges and concerns regarding the prerequisites for RCSSs to function and act in such a role have been identified (Hanberger et al., 2011).In a feasibility study examining the conditions for regional structures in support of knowledge development during 2010-2012, the authors stated that representatives from RCSSs perceived the initiative as having an unclear starting point and imprecise aims and objectives.In addition, RCSS representatives experienced vague communication from national level, problems with coordination between authorities, businesses, new and existing networks, and sometimes inadequate anchoring of the RCSS role and mission in relation to professional groups at the local municipal level (Hanberger et al., 2011).Johansson (2019) studied the implementation of evidence-based practice in social services in Sweden between 2009 and 2016.Findings from these studies indicated that this effort had a clear top-down perspective.The authorities at national level with primary responsibility for the implementation were the decision-makers, and the RCSSs were financed through annual agreements with national level actors.During this period, the RCSSs were expected to implement evidence-based methods, but, in practice, they were not only mediators of evidence-based practice between the national and local levels but also served as transformers of ideas by adapting initiatives from national level to suit the adopter organisations (municipalities) at local level (Johansson, 2019).
However, in recent years, the governance ideals for knowledge governance of social services in Sweden seem to have changed at the national level.The idea of governance has changed from a clear top-down perspective towards network-based governance.This change also seems to have influenced expectations on RCSSs from expectations closely related to the implementation of evidence-based methods towards increased expectations on an intermediate function based on dialogue and collaboration with stakeholders at different levels and with different tasks (Hjelte et al., 2022).The influence of network-based forms of governance is not exclusive in Sweden; it can be seen as a more general emerging change in ideas about governance in the public sector, including social work and healthcare (Adriaenssens et al., 2019;Frahm & Martin, 2009;Osborne, 2006).Such a change in governance seems to involve both opportunities and risks.According to Sørensen and Torfing (2017), network-based governance might be a way to achieve effective public regulations and enhance democratic legitimacy.At the same time, it also implies the existence of power and influence over decisions being formed in continuous negotiations between mutually dependent actors at different levels (Kooiman & Jentoft, 2009;Maron & Benish, 2021).Therefore, this kind of governance has often led to unclear demarcations between various areas of responsibility.In addition, the power and influence of different actors may also risk becoming invisible because political intentions are created through interactions between a number of actors rather than through formalised structures (Frahm & Martin, 2009;Johansson, 2012).
RCSSs have been identified as a key asset in knowledge governance in social services as an intermediary link between national and local actors and are described as crucial to the impact of knowledge governance at local level.However, the question of what this really means and requires in practice is unclear.Thus, the question of how RCSSs perceive, make sense of, and manage expectations from other actors in the knowledge governance system on their intermediate role needs further attention.
Accordingly, the aim of this study was to explore the perspectives of representatives from RCSSs on expectations associated with being an intermediary support structure in knowledge governance processes in social services in Sweden.The research questions in focus are: (1) Regarding areas of responsibilities and actions, what expectations do RCSS representatives perceive as being interlinked with the role of an intermediary support structure between national and local level in knowledge governance?(2) (How) do existing perceived expectations on RCSSs affect their ability to act as intermediaries in the knowledge governance process?(3) What strategies do RCSSs apply to manage expectations in their intermediary role?
Making Sense of the Intermediary Role: An ongoing Construction Through Communication It has been stated that 'sense-making' determines behaviour, that is, that a situation and its circumstances need to be comprehended explicitly in words to inform action (Weick, 1995).The conditions under which the RCSS operates imply that the intermediate role is formed in dialogue and interaction between mutually dependent actors at different levels in the knowledge governance system.Sense-making has been described as the process through which people work to understand issues or events that are novel, ambiguous, confusing, or in some other way violate expectations (Maitlis & Christianson, 2014).Weick (1995) considered organisations as open systems that influence and are influenced by their environment.To create stability and predictability, the organisation and its individuals strive to create a shared understanding among important actors about circumstances and events that affect them (Weick, 1995).The need for sense-making arises when the organisation experiences that collective actions with the environment become ambiguous or confusing (Weick et al., 2005).This may happen when the expectations from important actors in the environment (image) are not in line with the organisation members' own ideas about who they are and what their mission is (identity).Thus, there is a need to understand the situation at hand and agree on an interpretation (a narrative) that seems reasonable, which can include an explanation for the ambiguous situation but also how to resolve it (action).Weick (1995) states that it is important to keep in mind that sense-making is an ongoing process, and the identity is continuously affected by the interaction with the environment.Therefore, the narrative can be constantly reconsidered and changed, depending on which expectations are directed towards the organisation (Weick et al., 2005).

Study Setting: The Organisation of Knowledge Governance in Sweden
The development of knowledge governance has been ongoing since 2010.Until 2015, annual agreements (Platform agreements) formed the basis of the knowledge governance system.These agreements were concluded jointly by the government (Ministry of Health and Social Affairs) and the Swedish Association of Local Authorities and Regions (SALAR) as representatives of the municipalities.RCSSs were initially funded mainly by the state to build and strengthen the regional level in the task of being an intermediary link between local and national levels and to support the implementation of knowledge and evidence.The government operated through the National Board of Health and Welfare (NBHW), which, for example, was responsible for providing treatment guidelines (Johansson, 2019).SALAR's role and responsibility was mainly to coordinate nationally initiated initiatives and to provide funds to the regional level.
At the time of the study (2018), there had been changes in the organisation of the knowledge governance system.NBHW was given a more prominent role by serving as a hub for knowledge governance.Together with SALAR, they were commissioned to develop a model for how knowledge development should be organised, which resulted in a voluntary agreement at the national level (the Partnership), and each county was given the opportunity to join (SALAR, 2016).The Partnership includes representatives from SALAR, NBHW, and RCSS (as the counties' formal representatives), and they collaborate with relevant actors in the knowledge governance system.
In most counties, RCSSs are part of a larger organisation and support structure and have existed for decades.The RCSSs may be found in different types of organisational contexts, often directed and funded by the municipalities in each county or by other local and regional authorities and organisations such as regions and universities.Thus, RCSSs have a wide scope and mission and are considered as a support structure for the municipality's common issues.However, this study focuses solely on the role and mission of RCSSs in relation to knowledge governance in social services.RCSSs employ anywhere from a few people up to around 20 who often have experience in social work, management or from strategic assignments.Some have a PhD and work at a Research and Development (R&D) unit.
Each RCSS has joined the Partnership on behalf of the municipalities in each county.Unlike the time of the Platform agreements, RCSSs receive no state funds contributing to knowledge governance issues (SALAR, 2016).This means that SALAR, state authorities, and RCSSs do not have any major opportunities to control what the municipalities choose to prioritise in their activities.In line with this, the purpose of the Partnership is to contribute to increased collaboration and dialogue between actors at local, national, and regional levels, and by coordinating initiatives, contribute to needs-adapted knowledge development and effective knowledge support (SALAR, 2016).

Study Design and Data Collection
The study uses an abductive approach and qualitative methods for data collection and analysis (Graneheim et al., 2017).The research questions have been approached mainly with open questions in an empirically driven manner (inductively).Interviews with representatives from the RCSSs were used for data collection.Conventional content analysis with elements of directed content analysis of the data material was used.All RCSSs in Sweden were approached, and representatives from 14 of the 21 regions in Sweden were willing to participate in the study.The material covers approximately 235 of the country's 290 municipalities.The participating RCSSs vary in size and are from different organisational contexts with different premises.
Nineteen group interviews were conducted.The interview questions focused on main themes such as knowledge governance, evidence-based practice, the Partnership, and RCSSs.Contact persons from each participating RCSS were asked to invite people they considered would contribute to the study.83 respondents participated in the interviews, of which around 15 were representatives from municipalities (usually heads of administration who were involved in knowledge governance) and four respondents represented healthcare in the region or a university.The interviews were conducted between October 2018 and April 2019.The interviews lasted between 30 and 120 minutes (average 80 minutes).All interviews were conducted by researchers from the research team.All interviews were recorded and transcribed verbatim.

Data analysis
Data were analysed using a combination of conventional and directed content analysis, whereby codes and categories were mainly derived from data inductively and then, in a final step of the analysis, sorted into categories based on theory (Hsieh & Shannon, 2005).Data analysis was carried out in the following steps (Elo et al., 2014): (1) Each author read the material in its entirety.
(2) All data related to the aim and research questions of the study were highlighted.
(3) The material was coded and recoded in relation to the research questions (Graneheim et al., 2017).(4) Categories and subcategories emerged from the coded material (Table 1).
(5) Condensed and categorised data on ideas on (1) the expectations placed on the RCSSs, (2) the consequences of those expectations on their ability to act as intermediaries in knowledge governance, and (3) the strategies used to manage expectations were interpreted and deductively organised in a template based on Weick's theoretical framework on sense-making (Weick, 1995;Weick et al., 2005).

Methodological Considerations
In qualitative research, the term trustworthiness has been suggested to define aspects of validity and reliability (Elo et al., 2014).Graneheim and Lundman (2004) suggest that the concepts of credibility, dependability, and transferability may be used to further define aspects of trustworthiness.
The results are based on self-reported data from interviews with information obtained from representatives from different RCSSs.Self-reported data entail a risk of bias, which has to be considered in relation to the trustworthiness of the study findings.First, people tend to report what is socially desirable, thus the findings may be biased due to so-called socialdesirability bias (Fisher & Katz, 2000).Second, data were collected by interviews, which entails a potential source of bias due to measurement errors (Gillespie & Chaboyer, 2013).
The abductive approach, combining mainly inductive analysis with partly deductive coding, was considered the most appropriate.In accordance with Elo et al. (2014), data collection and analysis were guided by the research questions; the analysis aimed at (a) condensing raw textual data into a brief, summary format; (b) establishing clear links between the research objectives and the summary findings derived from the raw data; and (c) in the final step of coding, the data were interpreted deductively in relation to both the research questions and Weick's theoretical framework on sense-making (Weick et al., 2005).Thus, results are organised and described by the content of the categories and in relation to theory.
Credibility refers to how well methods of data collection and analysis address the intended focus of the study (Polit & Beck, 2004).The question of how well interviews captured all relevant aspects of the RCSSs' experiences remains unanswered.An attempt to increase credibility was made by collaborating on coding and categorisation and by frequently discussing the process of analysis and the emerging findings.The similarity of issues raised by respondents from the different RCSSs strengthens the trustworthiness of the results as a representation of the perspectives of RCSSs in Sweden.Despite the effort to collect rich data, it must be recognised that the empirical base is limited.The results may be comparable with other similar studies on intermediaries and the facilitating role, but transferability of the results to other contexts has to be considered with caution because of the long tradition of municipal autonomy with independence and freedom of action in Sweden, not least within social services (Swedish Agency for Public Management, 2011).

Ethical Considerations
All procedures used in this work comply with the ethical standards of the Swedish Ethical Review Act (SFS 2003:460) on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008.At the time of the interviews, the respondents signed a consent form for participation in the study.The respondents were also informed that the material would be treated confidentially and stored in such a way that unauthorised persons could not access it.They were also informed that participation was voluntary and that they could withdraw at any time.

Expectations on RCSSs as Intermediaries
The Image of the RCSSs Interviews with representatives showed that there seems to be a variety of expectations with regard to RCSSs as intermediaries in the knowledge governance of Swedish social services.These expectations exist at both national and local levels.Several respondents stated that RCSSs were to 'disseminate knowledge and information' initiated by national authorities.This concerned both the NBHW, who expected RCSSs to disseminate research findings and guidelines, and other national authorities such as SALAR, who expected them to disseminate information to the municipalities.
It is under construction [knowledge governance], and so far, it is pretty much from the top [national level].There are several authorities that are part of the Partnership.After all, they receive their regulatory letters [from the Swedish government] and want to step down in their mission through the Partnership.
In addition, many respondents experienced that RCSSs were also expected to act as a support to the local level, and external monitoring was assumed to be a central issue.Respondents from one RCSS believed that one of their main tasks was to stay updated and spread new knowledge relevant to social services because they experienced that most municipalities could not perform such external monitoring themselves.However, according to the empirical material, the greatest expectation regarding RCSSs providing support to the local level concerned enabling of development.According to the respondents, this could include a variety of issues.Representatives of one RCSS assumed that an important aspect was to inspire local levels to initiate development efforts together with the RCSS.Respondents from several RCSSs experienced that adapting initiatives from national authorities so that they could be transformed into the local social service practice was a central task.In addition, the support could also be about drawing attention to initiatives that came from the local level.One respondent described different kinds of support to the local level as follows: I think that RCSSs have a big part in it, helping to get an overview of things, helping to support implementation, helping to capture what the municipalities themselves say they need knowledge about.
Another important expectation, according to several respondents from different RCSSs, was that RCSSs would act as a mouthpiece for the local level in relation to the national level.According to representatives from one RCSS, this was the idea behind the development of the Partnership.Instead of knowledge governance being a top-down phenomenon, the respondent perceived that the aim was to create a dialogical governance through the Partnership.Several respondents from different RCSSs expressed that an important function as a mouthpiece was to pick up and transfer local needs to the national authorities.In addition, several respondents considered that their participation in the Partnership was a way for the local level to convey their interests and have an influence over decisions made at national level.
Based on the data, it also seems that there was an expectation that RCSSs would function as some kind of regional node in knowledge governance.According to representatives of several RCSSs, there was an expectation from the local level that they would provide a meeting area for representatives of different municipalities.In addition, several respondents from different RCSSs considered that they were expected to coordinate support and development in the municipalities in the region by monitoring the situation and bringing together municipalities with similar needs for support and development.

Impact of Perceived Expectations on the Role as Intermediaries
The Ambiguous Situation: A Discrepancy Between 'Image' and 'Identity' RCSSs expressed feelings of inadequacy due to a perceived discrepancy between expectations and the possibility to fulfil these.One such discrepancy concerned the identity, that is, what their role and mission comprises.In relation to their role as a mouthpiece, several respondents perceived that the national level had unrealistic expectations.One respondent expressed it as follows: There is an unreasonable expectation that we should keep track and that we can sort out, for example, what is highest on the agenda in X municipalities [the number of municipalities in the region] in the social services area.How should I be able to judge this?It is completely impossible.
According to several respondents, there was also an expectation of immediacy where the RCSS representatives were expected to be able to express the municipalities' views directly in network meetings with different actors at the national level.The respondents expressed that they could not work that way.Because they were representatives of the local level, they considered there was a need to discuss the issues with the municipalities before conveying a message to the national actors.Another perceived discrepancy concerned 'how much' they were supposed to do.A number of respondents considered that many RCSSs did not have adequate resources to handle the relatively large number of assignments that came from national level, although the content of the assignments was often reasonable, based on RCSS's intermediary role in knowledge governance.

Explanations for the Discrepancies
The respondents presented several potential explanations for the discrepancies.Representatives of one RCSS believed that a contributing factor to the excessive expectations was that national authorities did not limit their expectations to areas that the RCSS had chosen to prioritise.Representatives of another RCSS believed that the number of assignments was a consequence of the authorities at national level not always being coordinated; all the authorities involved wanted RCSSs to meet their specific needs.In addition, representatives of several RCSSs argued that a contributing cause of the excessive expectations was that actors at the national level did not always understand the conditions under which RCSSs operate or the resources available to them.One respondent described it as: We are well aware of this in the X region.Given the resources we have, we are expected to do much more than we are able to.I attend every RCSS [meeting] and just feel that 'we have not done this, we have not done that, we have not been able to think about this' … Representatives of several RCSSs believed that national authorities sometimes did not have sufficient awareness that some RCSSs had dual roles.As mentioned earlier, many RCSSs were incorporated as part of pre-existing regional structures, for example, R&D units.They were expected to be part of the knowledge governance and at the same time work on local and regional assignments as an R&D unit.According to representatives of one of the RCSSs, expectations from national actors did not take these conditions adequately into account.Respondents also stated that the municipalities themselves might also experience difficulties adopting all national initiatives, especially the smaller municipalities.
But I can feel that out in the municipalities, we have lost this with evidence-based practice.… The smaller municipalities tend to lose it.… [They] are flooded with a lot of different things that are put on the municipalities.

Action: The Resolution to the Ambiguous Situation
The RCSSs seemed to have used different strategies to manage the ambiguous situation with what they perceived as unreasonable expectations from actors at the national level.However, the strategies did not seem to aim at resolving the ambiguous situation.Instead, they were directed at actors at the local level to meet their needs and conditions, which the RCSSs perceived as their main mission.One strategy was to prioritise.The RCSSs appear to have used different arguments when prioritising what they should and should not do.One argument mentioned by several respondents was what they saw as their main mission.According to several respondents, their main task was to support the municipalities based on needs expressed at the local level.As mentioned earlier, several RCSSs were located in pre-existing structures such as R&D units.Representatives of these RCSSs stated that the RCSS did not stem from knowledge governance, but instead from R&D units that were created based on local support needs.The work done within the framework of R&D units seems to have been given priority when it was necessary.One respondent described that their original mission in the R&D setup was not to support national authorities but to support municipalities.Another respondent described it as follows: The RCSS becomes some kind of plus menu instead of it being the basic menu.
Another line of argument related to who the respondents saw as their main employer, that is, the actual 'owners' of RCSS efforts.Some respondents referred to 'the Board' or 'the R&D Council' , which consisted of representatives of the municipalities in the region.In most cases, the issue of actual 'ownership' was simply associated with the funders of RCSS activities.According to the respondents, this also influenced their priorities.One representative of an RCSS expressed it this way: We work here completely on behalf of the municipalities.… SALAR and the NBHW cannot come and tell us what to do.They were able to do that when we were paid for by state funds.
A second strategy was to act as a gatekeeper in relation to the local level.The respondents stated that meeting expectations from the national level was not only perceived as difficult by the RCSSs but also by some of the municipalities.Therefore, RCSSs sometimes acted as a gatekeeper who filtered what was considered urgent and relevant for the municipalities.According to several respondents, the main purpose was to safeguard the municipalities because they were not always considered able to adopt and cope with all the initiatives from the national level.Therefore, this strategy was primarily aimed at limiting the number of initiatives that reached the municipalities within the framework of knowledge governance.Representatives of one RCSS described it as follows: Recently, I have begun to highlight this … that the local level is not able to cope, does not have resources to deal with all issues.And you forget that, the more spin you get on that superstructure [national level].And then we lose … we cannot run faster than local level if we want results.
A third strategy was to sort and package information and initiatives from national authorities.Expectations from national authorities were excessive and not always coordinated.This seems to have resulted in a need to package information and initiatives in a way that the local level could handle.One respondent stated that most of the time was spent managing the national actors' initiatives because most were not structured and coordinated.Another respondent stated that it was a central task for RCSSs to package the information and initiatives in a good way so that they became manageable for the local level.According to one respondent, the municipalities expressed that they needed support from RCSSs to sort information and initiatives from the national level: I have asked the municipalities: What do you want?What do you want us to do … should a researcher be employed here [at RCSS]?No, they say.We want you to help us sort through the noise.

Discussion
This study explored how RCSSs make sense of and manage expectations in their intermediary role in knowledge governance in social services in Sweden.The findings in this study confirm previous research that the intermediary function of RCSSs is supposed to be based on dialogue and collaboration with stakeholders at different levels and with different tasks in the knowledge governance system (Hjelte et al., 2022).In line with previous studies, our results also indicate that there seems to be a variety of expectations on RCSSs.The role of intermediary organisations has previously been associated with a broad range of facilitating activities (Hakkarainen & Hyysalo, 2016;Proctor et al., 2019), and previous studies on RCSSs confirm the diversity of expectations.Among other things, they have been described as key actors in knowledge transfer and knowledge development in social service practice (Eliasson, 2014;Hjelte et al., 2022).In addition, the findings confirm previously identified barriers, such as inadequate (and sometimes unclear) funding, unclear mission and expectations and role ambiguity, that may impede intermediary facilitators engaged in system-wide efforts (Graaf et al., 2017;Westerlund et al., 2015).
However, the findings from this study also shed light on the processes by which RCSSs make sense of and manage expectations in their intermediary role.Such sense-making processes seem to emerge at the intersection between the RCSSs and other actors in the knowledge governance system when there are discrepancies between their own perception of their mission (identity) and expectations placed on them by other actors (image).By relating expectations from the environment to their own perception of their identity, RCSSs seem to create a plausible narrative of the disparities, which not only contains how expectations differ from their own perception but also explanations for why such disparities occur (cf.Weick et al., 2005).
What can also be seen is that the solution to such discrepancies is not necessarily that intermediate actors, in line with Weick et al. 's (2005) assumptions, strive to establish a shared understanding of their identity together with actors in their environment in order to create stability and predictability.According to the authors, organised systems that build on coalitions and negotiations between different actors, such as the knowledge governance system, deal with uncertainty (e.g., role ambiguity or unclear missions) by creating a coherent interpretation (a narrative) of the situation that may inform the system for action.The locus for this process is realised through coordination and distribution of information among interdependent actors (Weick, 1995;Weick et al., 2005).In this study, however, RCSSs in their intermediary role, seem to handle the ambiguity through a balancing act whereby they consider, assess, and prioritise different actors' needs and expectations in relation to their own perception of the intermediary mission rather than through a collective sense-making process with other actors on different levels in the knowledge governance system.By prioritising the local level needs and expectations, and by developing strategies to handle perceived unreasonable expectations from the national level, knowledge governance appears in some respects to be a relatively loosely coupled system (cf.Weick, 1976).Thus, RCSSs appear to be able to maintain their identity without being the subject of dialogue and negotiations with other actors in the knowledge governance system.
The use of different strategies may not only be interpreted as a way for RCSSs to handle perceived unreasonable expectations placed on their role as an intermediary link.It may also be interpreted as an expression of discretion and distribution of power and influence in the knowledge governance system.According to the findings, RCSSs were perceived as having power and influence over decisions at the national level in their role as a mouthpiece for the local level.In addition to this more formal influence that the role of mouthpiece entails, the discretion that RCSSs have in terms of facilitation at the local level could also be regarded as an opportunity for power and influence.Power and influence in organisational systems is expressed in actions that form what people accept or reject, and occur through, for example, control over permitted or disallowed cues or actions (Weick et al., 2005).In terms of control over permitted cues or actions, the strategies used by RCSSs to transform, structure, and prioritise among initiatives from national level may be interpreted as an expression of the distribution of power and influence over which issues should be acted upon, transferred, and implemented into social services practice and in what order.This in turn suggests that RCSSs may not only be seen as transporters and transformers of knowledge on behalf of initiatives from the national level (Johansson, 2019) but also gatekeepers regarding what knowledge should be considered most important to develop and implement within the framework of knowledge governance of the Swedish social service.This distribution of power and influence does not necessarily have to be detrimental to actors at the local level if their perceived needs are met.However, it can be disadvantageous for actors at the national level, who may have limited insight and influence on how evidence-based practice is implemented at the local level.In this sense, parts of the implementation might become some kind of 'black box' for actors at the national level.They are aware of their own input, but they are not aware of or in control over how and to what extent knowledge is transferred to the social service practice.

Implications for Practice
The purpose of knowledge governance in social services in Sweden has been to contribute to needs-adapted development and effective knowledge support for social services practice through a coherent system that is characterised by collaboration and dialogue between actors at different levels.The expectations on RCSSs have been to function as an intermediary link, encompassing both a bottom-up and a top-down flow.The results from this study imply that, in some respects, knowledge governance appears to be a relatively loosely coupled system.This does not necessarily have to limit RCSSs' opportunities to meet perceived needs in social services practice.However, if the aim is to create a coherent and transparent system where actors at different levels have similar opportunities to influence what knowledge should be developed and implemented within the framework of knowledge governance, it seems to be essential to establish a common understanding of the role of RCSSs as intermediaries.This highlights the need for a collective sense-making process in order to reach a common understanding regarding expectations, and the conditions for facilitation, but also what should be prioritised between actors at different levels in the knowledge governance system.
This study encompasses the views and experiences of RCSS representatives regarding the role of RCSSs as an intermediary link in knowledge governance in social services in Sweden.To gain a fuller understanding of this matter, further research covering perspectives from local level actors as well as national key actors is needed.A question of particular interest concerns local level actors' ideas on how knowledge governance should be organised and implemented and what facilitation activities they believe would enhance the implementation of knowledge/evidence and proven experience in social services practice.

Conclusions
The findings in this study shows that, from the perspectives of participating representatives from the RCSSs, expectations on the intermediary role and mission are perceived as unclear and hard to fulfil, rendering feelings of inadequacy.The RCSSs manage role ambiguity by considering, assessing, and prioritising different actors' needs based on how they perceive and define their intermediary mission.Thus, the RCSSs are able to maintain their identity without engaging in dialogue and negotiations with other actors in the knowledge governance system.A collective sense-making process is missing, and knowledge governance appears to be a relatively loosely coupled system where RCSSs are acting as gatekeepers in relation to the local level and where actors at national level may have limited insight into, and influence over, how evidence-based practice is implemented at local level.This in turn, implies that knowledge-based initiatives from the national level are not always made accessible to local level settings and, hence, are not considered for social service practice.Therefore, a common understanding of the relationship between expectations and prerequisites for facilitation and a consensus opinion regarding the RCSS role and mission, between national, regional, and local levels seem to be essential.

Table 1 .
Example of a category with subcategories and codes of expectations on regional collaboration and support structures