Big ideas, soft governance: managing large-scale school improvement at the national agency level in Sweden

Abstract

Purpose

Using a large-scale school improvement program in Sweden as a case, this article aims to explore the state governance of a large-scale school improvement program in Sweden and how officials at the state agency level made sense of the reform ideas and operationalized them in policy actions.

Design/methodology/approach

Data were integrated from Swedish Government Official Reports and formal directives from the Ministry of Education. Officials of the Swedish National Agency for Education (SNAE) were also interviewed. Data were analyzed to identify how regulatory rules, professional norms and cultural–cognitive beliefs shaped SNAE's design of the program.

Findings

The article shows how different types of governance (i.e. regulatory rules, professional norms and cultural–cognitive beliefs) set the direction for managing large-scale school improvement. In particular, in the studied case, the lack of clear regulatory directives enabled sensemaking processes clearly influenced by normative ideas and cultural–cognitive beliefs.

Research limitations/implications

The findings are mostly presented from the perspective of managers, so further study is required to attain a broader understanding of the state agency level's role and function.

Practical implications

By illustrating the strengths of understanding various dimensions of educational governance, the findings are highly relevant to both policymakers and educational managers at different levels of school systems.

Originality/value

The article offers a valuable perspective on large-scale school improvement and educational governance by focusing on a level that has hitherto received little attention.

Citation

Nordholm, D. and Adolfsson, C.-H. (2024), "Big ideas, soft governance: managing large-scale school improvement at the national agency level in Sweden", International Journal of Educational Management, Vol. 38 No. 1, pp. 302-316. https://doi.org/10.1108/IJEM-04-2023-0219

Introduction

Reform policies targeting whole systems and overall strategies have become increasingly prominent over time, illustrating a movement toward more nationwide and large-scale approaches to school improvement (see, e.g. Hopkins et al., 2014). A crucial aim of such large-scale and systemic approaches is to activate and involve multiple levels of education systems (i.e. the national, district and school levels) so that they work in concert (see, e.g. Fullan, 2010). Besides the changed conditions for policymakers and school actors due to this changed school landscape, these policy shifts also have consequences for school improvement research. That is, when studying large-scale school improvement reforms, one must pay attention to the different levels of education systems and to the relationships among them. As noted by Hopkins et al. (2014, p. 269), “we can only learn about system change by studying systems, their components, and the interactions among their components.”

Reviewing the research on school management shows that leading work has supplied important details on effective school leadership and on the relationship between school management and student results (see, e.g. Bendikson et al., 2012; Hallinger, 2005; Leithwood et al., 2020). Also, when it comes to local education authorities (LEAs), earlier research offers important insights into how successful LEAs operate and support school managers in various contexts (see, e.g. Anderson and Daly, 2013; Leithwood, 2010; Leithwood et al., 2019; Liljenberg et al., 2022; Nordholm, 2016; Rorrer et al., 2008). However, less attention has been paid to the level above the LEA level, i.e. the level of the national school agency (i.e. the Swedish National Agency for Education –SNAE). That is, educational policy and state policy directives must often, in a first step, be interpreted and operationalized by a national school agency in more concrete school reforms and programs before they reach LEAs and school actors in the local schools. Yet, we have limited knowledge of how officials at the national school agency level make sense of and operationalize state intentions and directives for national school improvement programs before they are implemented at the LEA level and in local schools.

Previous research has also shown that the implementation of educational policy is far from a straightforward process; rather, it is dependent on contextual factors at various levels of education systems. For example, by studying educational leadership and policy-making mechanisms in three US states, Seashore Louis et al. (2008) showed that each state had very different means with which to pursue educational policies, concluding that political culture is a mediating factor for states' responses to increasing demands for leadership and accountability initiatives. Similarly, Harris and Jones (2018) pointed out that policymakers worldwide remain preoccupied with identifying the “right” policies rather than considering the conditions and contextual factors that could contribute to effective policy in practice. As also argued by Brauckmann et al. (2023), using Sweden as a case, it is essential for educational leaders to develop a profound understanding of contextual factors in times of new public management. By recognizing these and other works, this article is positioned in a tradition in which sensemaking activities at different levels of the education system are essential for analyzing and understanding the underlying ideas and results of educational policy (cf. Coburn, 2005; Spillane and Anderson, 2014; Spillane et al., 2002).

In this paper, a large-scale national school improvement program in Sweden, Collaboration for the Best School Possible (CBSP), is taken as an empirical case for studying these processes. In Sweden, the responsibility for education is roughly divided between the central government and its agencies, which are responsible for ensuring the overall quality of education, and the LEAs or their independent counterparts, which are responsible for providing education in preschools and schools. Most public schools are organized by the municipalities, each having a school board and one or several superintendents. Swedish school boards consist of appointed politicians, representing the political parties of the municipal council such that the school board has the same political composition as does the municipal council. Notably, Sweden has 290 municipalities, so there is great variety in local school governance and municipal school organization (Johansson et al., 2016).

The CBSP program started in 2015 and stands out as a striking example of a large-scale school improvement reform (Adolfsson, forthcoming). For example, at the beginning of 2022, the program involved school actors from approximately 250 schools, 50 process leaders at SNAE and 130 educators from Swedish universities. In the first step of CBSP, the LEAs, principals and teachers analyze the schools' results jointly with SNAE, leading to a number of efforts in response. Then, when the concrete efforts start, the universities get involved by providing their academic expertise. The aim is to offer a combination of researchers with doctoral degrees, who can contribute theoretical and scientific perspectives, and university teachers with extensive practical experience, for example, as school leaders and/or teachers. A primary aim of CBSP is to develop a capacity for sustainable school development at the local municipal level in collaboration with SNAE and the universities and, in so doing, avoid a top–down approach. The financial resources are close to SEK 100 million (about USD 10 million) annually (Swedish National Agency for Education, 2019). A considerable proportion of this funding is distributed to LEAs and schools during their participation in the project—normally about two years. Despite this large-scale national investment in terms of the number of involved schools and economic resources, previous work (Author, forthcoming) has indicated a lack of clear state directives and guidance to SNAE as to how this school improvement reform should be designed and implemented. This raises important questions concerning the links and relationships between the government and SNAE levels.

Theoretically, this article is based on a neo-institutional point of departure (e.g. Scott, 2001, 2008) and on previous work examining sensemaking in organizations (e.g. Weick, 1995, 2001). This theoretical framework provides a valuable perspective on how reform intentions of different types are articulated at the state level and, in a next step, are interpreted and operationalized at the level between government politicians and LEAs.

Using CBSP in Sweden as a case, this article explores the state governance of a large-scale school improvement program and how officials at the state agency level make sense of the reform ideas and operationalize them in strategies and policy actions.

The article starts with a description of the theoretical point of departure with a particular focus on the relationship between institutions and social organizations, in addition to presenting the concept of sensemaking in organizations. The article then introduces the material and the data analysis, before the empirical findings are presented in three subsections. Finally, the article provides a discussion and conclusions based on these findings.

Theoretical points of departure

In analytically examining and understanding different aspects of the governance and management logics linked to the implementation of CBSP, we drew inspiration from neo-institutional theory. Focusing on the relationships between institutions and social organizations makes it possible to distinguish underlying assumptions, mechanisms and indicators in the governance. In the current case, the Swedish government, as an institution, has long had an ambition to improve the school results nationwide. Following the work of (Scott, 2014), institutional structures are constituted by three “pillars,” i.e. regulatory rules, professional norms and cultural–cognitive beliefs. Accordingly, these dimensions—individually and combined—can be seen as constituting different forms of governance and management logics in relation to different social organizations, in this case SNAE and, in the next step, the LEAs and schools.

The regulatory aspect refers to the laws and rules that together regulate and constrain organizational behavior. Laws and rules are, in this context, associated with regulatory activities, such as rule setting, directives, inspection and sanctions, intended to influence future behavior.

The normative pillar, in turn, comprises aspects that constrain and empower behavior through values and norms. Scott (2008) explained that values are to be understood as notions of the preferred and desirable—that is, various forms of acknowledged social agreements as to the expectations and conventions that direct members of organizations to act in desirable ways. Such pressure may arise from norms, expectations and attitudes linked to how members of organizations should work. In the absence of legal obligations, the normative element primarily operates through soft regulations with no formal legal sanctions attached. However, actions and outcomes can be compared and assessed in light of norms and values. Regarding norms, Scott (2008, pp. 54–55) clarified that they “specify how things should be done; they define legitimate means to pursue valued ends.” Hence, the normative pillar constrains organizational behavior but simultaneously enables social actions that balance responsibilities, privileges, duties, licenses and mandates.

Regarding the third pillar, cultural–cognitive aspects are shared conceptions of social reality that serve as a platform for sensemaking. According to Scott (2001, 2008), these thoughts and identifications are often deeply embedded in the life of social organizations. Concerning their legitimacy, Scott emphasized the power of cultural–cognitive factors because they primarily rest on preconscious and taken-for-granted understandings. However, even if shared cultural understandings tend to work in a homogenizing or isomorphic direction (see, e.g. DiMaggio and Powell, 1983), Scott also pointed out that cultural–cognitive conceptions often vary and, consequently, could have heterogenizing effects as well.

Given these different types of governance and management logics, when studying and attempting to understand how policy and reform processes arise and play out in and between different levels of a school system, the sensemaking concept is essential (see, e.g. Weick, 1995). Studies based on this concept have produced essential insights into how organizational change occurs, by providing a lens through which to understand human agency in relation to organizational conditions (see, e.g. Park and Datnow, 2009; Spillane et al., 2002). A sensemaking perspective on reform and policy implementation implies placing school professionals and their experiences, knowledge, beliefs and context at the center of the analysis. The implementation of a large-scale school improvement program like CBSP, in which several organizational levels are involved, will accordingly be framed by how school professionals at these different organizational levels make sense of the improvement program in relation to the contextual conditions (Weick, 1995).

Consequently, to understand governance and implementation processes in relation to a large-scale school program like CBSP, it is important not just to analyze the formal aspects of this regulation. Rather, it is also crucial to consider the soft forms of the governance logics and mechanisms that were more and less salient in this CBSP program and the ensuing sensemaking activities at SNAE. This implies moving beyond a more traditional implementation and governance analysis of educational policy and reform. The three pillars elaborated on by Scott provide valuable concepts for such an analysis by offering a clear direction while still permitting an exploratory analysis. In the following section, the material and data analysis are presented in detail.

Material and data analysis

To deepen our understanding of the governance and management of CBSP, we analyzed Swedish Government Official Reports, formal directives and Ministry of Education decisions. A request was also sent to SNAE to access documents detailing how the CBSP process had been formalized and operationalized over time. After dialogue with SNAE, 16 documents of different types were sent to the research group. The first type of document (n = 12) comprised internal working material (e.g. drafts, meeting notes, plans, figures and images). These documents offered background on how CBSP had been established and developed over the years. The documents also gave the research group valuable input in formulating suitable interview questions. The second type of document (n = 4) was characterized as more formalized and public and could be described as the formalized results of internal work conducted by SNAE.

In addition, a sample of six (n = 6) SNAE officials was interviewed. Two of these interviewees were heads of the current implementing unit at SNAE and four were directors of studies leading the four working teams organizing and implementing CBSP. The interviewees had extensive insight into and long experience of working with CBSP in designing the collaborations and working processes between SNAE, participating municipalities or independent school organizers, schools and universities. Based on their key positions at SNAE, they also had many years of experience in interpreting the intentions underlying the program and in translating these intentions into concrete school improvement programs and strategies. The interviewees were informed that the study was guided by the ethical guidelines set by the Swedish Research Council (2011) concerning voluntariness, anonymity, how results were to be published and so forth.

Both researchers have worked in CBSP in different municipalities and therefore have had discussions with several interviewees over the years. This probably had a positive impact in terms of gaining access to the field and collecting data. It is worth noting, however, that the researchers have never been involved in CBSP project design as such or in project governance, which is the main focus of this analysis.

The sampled documents and the transcribed interviews were imported into software for qualitative data analysis (NVivo 20). In an initial step, all data were organized and analyzed based on the three pillars (i.e. regulatory, normative and cultural–cognitive) identified by Scott (2001, 2008). More specifically, regarding the regulatory pillar, the analysis explored laws, rules and formal directives that might regulate and constrain the state agency officers' sensemaking efforts. The analysis also paid attention to regulatory activities that might influence sensemaking activities. Concerning the normative pillar, the data analysis explored ideas and notions beyond regulatory governance, identifying what was preferred and/or desirable at the state and SNAE levels (see also Weick, 1995, 2001; Weick et al., 2005). Thus, the analysis paid attention to extracts indicating the values and norms that constitute the basis of dominant prescriptions, obligations, evaluations, goals and means—i.e. “the desirable” or how things should be done. Regarding the third pillar, the cultural–cognitive pillar, the data analysis paid attention to the shared conceptions and frameworks of social reality that serve as a platform for individual and shared sensemaking activities (Weick, 1995). The analysis strove to detect thoughts, identifications and assumptions embedded in SNAE's “life” and work processes that affect the sensemaking activities. As noted above, such shared cultural understandings could work in both homogenizing/isomorphic and heterogeneous directions.

Altogether, despite this somewhat deductive point of departure, the design of the analysis permitted more exploratory elements as well. More specifically, although the regulatory, normative and cultural–cognitive pillars set the direction of the data analysis, the contents of these pillars and the sensemaking activities they generated became empirical matters to investigate. In addition, the current research design enabled important differences among and/or nuances in the documents and interviewees to be revealed.

Results

The results presentation is structured according to the three pillars identified by Scott (2001, 2008). Based on these pillars, the analysis focuses both on the state-level governance and on how officials at SNAE made sense of these intentions and managed and operationalized them in strategies and actions.

The regulatory dimension of CBSP

As noted above, the state intentions underlying CBSP are revealed in various policy documents. The reasons for the initiation of the program and the government's intentions for the reform can be found in document U2015/3357/S:

Despite the requirements of the Education Act, national and international reports and knowledge measurements indicate that equality in the school system is not maintained. Not least, the supervision and the inspections carried out by the School Inspectorate show this. Many principals and schools are requesting support for continued development following the School Inspectorate’s decision … The increased differences in school results require powerful measures at national and municipal levels if equality in the school system is to be maintained.

In addition to this rather normative assumption regarding the relationship between measuring and equality, the documents reveal how the state level has gradually recognized that the negative developments in many Swedish schools must be reversed, and this is where SNAE comes into the picture (see, e.g. U2019/03786/S):

The government believes that long-term planning and continuity is needed in the national support and development efforts to strengthen equality and provide support to the school organizers in this work. Against the background presented above, the Swedish National Agency for Education should be assigned to continue [to] carry out efforts in dialogue with school organizers with the purpose of raising schools' results and increasing equality within and between school units. The efforts should concern schools with poor results … and/or are assessed to have difficult conditions to improve their results on their own.

Given this background, SNAE was assigned the task of designing CBSP. That is, based on the regulatory guidance provided, SNAE started to make sense of the state intentions and operationalized them in implementation work. In that sense, sensemaking activities tend to evolve when something triggers or disrupts the previous way of working (cf. Dutton and Dukerich, 1991; Weick, 1995). One significant example of this sensemaking work is the state expectation that the large-scale improvement work should be based on dialogue between SNAE, school organizers and individual participating schools (Swedish National Agency for Education, 2019 Dnr 2015:778).

According to the assignment, SNAE must work in dialogue with school organizers to carry out targeted efforts with the aim of raising knowledge results and increasing equivalence. The dialogue is one of the prerequisites for making the efforts adapted to the needs of the school units and becoming sustainable in the long term. Dialogues on school improvement must always be based on national goals, requirements and guidelines. Evaluations from previous work show that the dialogue between the agency and the school organizer and school unit must be based on trust, openness and clarity.

This regulatory governance, as described by the four SNAE officials, set the direction for how they understood and further operationalized the program. In the interviews, they stressed the importance of working dialogically, and of the fact that the number of LEAs and schools that should be invited to participate in the program is determined at the state level, as described by Official 1 at SNAE:

The government’s decree clearly sets out that we must work in dialogue. … We are not some control unit that comes out, but it really is a form of dialogue throughout … And then money controls us, but so far we have received a lot of money, and that’s the signal that this is still important. … After all, there’s less freedom and there should ideally be fifteen [school organizers and schools] every term, and that’s quite a lot to deal with.

The process of selecting which LEAs and schools will be invited to participate in CBSP is based on the outcomes of the Swedish School Inspectorate's nationwide inspection. Accordingly, LEAs and schools that have received substantial criticism from the Inspectorate are offered support from SNAE to address the identified deficits. It is not obligatory to participate in CBSP, which means that an LEA or school can refuse the support offered. At the same time, the more indirect regulatory pressure from the Inspectorate and SNAE is described as very strong, and the deficits identified by the Inspectorate must be remediated by the current LEA and schools if they are not to risk sanctions from the Inspectorate and, especially, a poor reputation on the school market. Considering this, it becomes quite “improper” to reject participation in CBSP. This was confirmed by the SNAE officials, who said that it was very unusual for an LEA or school to actually refuse to participate in the national improvement program.

This indirect regulatory pressure was also described as something that could legitimize and reinforce the program in relation to, for example, the school teachers. That is, CBSP is not an ordinary local school improvement effort, as two national school agencies stand behind it, giving the program considerable weight and legitimacy.

To sum up this first part of the “Results” section, the analysis reveals that despite little regulatory control of the aim, content, and methods of CBSP, the regulatory governance dimension was nevertheless evident in the SNAE officials' understanding and operationalizing of the national CBSP policy. Accordingly, the regulatory dimension worked somewhat more indirectly in relation to the LEAs and schools. However, and as shown in the following sections, normative and cultural–cognitive elements were also integrated in this regulatory dimension, affecting the SNAE officials' understanding and actions in relation to the CBSP policy.

The normative dimension of CBSP

Due to the relative absence of detailed directives and decisions concerning CBSP, with few details on how the program should be designed and on what roles and functions various actors should have in the CBSP process, the SNAE officials wanted extensive sensemaking concerning how the program should be understood and operationalized. In the interviews, one SNAE official described this as follows: “I think this is an interesting question, since we have never had clear and strong directives from the department, and we do not have them today either,” and there is nobody who says “now you should do it like this” (Official 2). Another official (Official 4) looked back on the initial CBSP implementation process and described it in a similar way:

In the beginning it was … you had to build everything from scratch, step by step. It was like the “wild west”. We received a political order that we had to interpret, handle, and solve … There was quite extensive work we had to do at the beginning, and over time we have refined the CBSP process, especially the initial analyzing phase.

Due to the lack of strong directives, this implied that the operationalization and implementation of the program had to be based on other logics and mechanisms than regulatory ideas. Accordingly, more normative aspects of management and control, based on values and norms, constituted an important basis for SNAE's work in relation to the involved LEAs and schools. A clear example of such normative aspects was the dialogue approach. In internal SNAE policy documents (Dnr, 2015:778), the Agency described this in greater detail in terms of a “promoting, supporting, and listening role” throughout the improvement process. In addition, in these dialogues, SNAE was to provide advice and support in matters related to the conditions for school improvement, for example, in terms of providing the LEAs or schools with support material and research linked to their identified improvement areas. That is, the dialogue approach became a softer way of exerting influence and pressure, compared with more traditional regulatory methods.

An example in which dialogue constitutes a crucial platform in the working process is in the initial analysis phase. In this first step of the CBSP process, actors from the LEAs and schools analyzed different forms of data to identify several improvement areas. This analysis process followed a formal structure and was led by a SNAE official. An important principle guiding this process was that the improvement areas should not be identified top–down, but through a cooperative process. This in turn was supposed to reinforce the legitimacy of the ensuing improvement efforts. As to whether the analysis process mostly resulted in consensus between the school actors and SNAE concerning the improvement needs, one SNAE official (Official 3) commented as follows:

I think we almost always end up in some form of consensus. Through the process, we can support the LEA and the principals to see this and that. On the occasions when this does not happen, it is often because a principal finds it difficult to expose his or her deficiencies or lacks of competence.

Consequently, the dialogue method seems to have become normative, a legitimate way to attain a shared understanding between SNAE, the LEAs and the principals concerning what can be considered areas of adequate improvement. Even regarding the content and focus of CBSP, the directives were quite general. However, one issue identified in the directives as a cornerstone was the schools' and LEAs' “capacity to monitor, analyze, and initiate improvement efforts with the aim of developing the schools” (U2019/01553/S). More precisely, even though the initial analysis process is described as a strategy for identifying and defining the schools' and LEAs' improvement needs, the normative pressure in the analysis process is described as strong. This resulted in almost all CBSP programs, on a general level, consisting of similar improvement efforts and features. One example of such an improvement effort that almost all LEAs and schools implement is strengthening the schools' and LEAs' quality assurance system. SNAE's internal policy document (Dnr, 2015:778) describes this as follows:

To improve schools' results and the equality within and between schools, the schools and LEAs must maintain their capacity to improve themselves. A well-functioning quality assurance system will comprise the solution.

This interpretation of the state directives and the normative pressure from SNAE concerning the content and focus of CBSP was confirmed and justified by an interviewee (Official 3) from SNAE:

You can absolutely see this as a form of pressure, that we have such a strong focus on the quality assurance system, that we see this as the key to attaining successful school improvement. … However, we rely on research, and I feel confident in this focus.

The above quotation also points to another crucial normative aspect of the governance and implementation of CBSP, namely, that it should rest on a scientific basis. The interviewee, however, did not further develop what this would actually mean or whether there is research that problematizes or even questions such an assumption. In the formal directives and decisions from the Ministry of Education, the scientific basis and proven experience are mentioned as follows in document U2019/01553/S:

The improvement efforts that are initiated and supported by SNAE should rest on a scientific basis and proven experience. The efforts should be based on research concerning successful factors linked to good teaching.

How this directive was made sense of and operationalized by SNAE officials become, above all, visible in the design of CBSP. After the initial analysis phase, an improvement phase follows in which the LEAs and schools work with a university for almost two years to address the identified improvement needs. Accordingly, adding a university to the CBSP process became a way to realize the normative directives. In addition, involving the universities expresses the normative assumption that the knowledge possessed by SNAE and its officials does not correspond to the requirements for scientific knowledge and proven experience required in CBSP. It can finally also be understood as an effective way to legitimize CBSP as such and as a way to mobilize normative pressure on the LEAs and schools, as explained by Official 5:

It is about how the program must be scientifically based, and it must rest on the latest research … In this case, we can’t work with a consultant, in other cases that can be good but now we must quality assure the program and justify it very carefully.

In sum, it is clear that normative strategies and actions constitute an important part of the management and operationalization of CBSP. This is evident in how CBSP is designed in the form of dialogical working methods, the analysis phase, the scientifically based improvement efforts and so forth. The fact that both a national school agency, i.e. SNAE, and universities are involved in CBSP gives the program legitimacy, on a general level, and constitutes normative pressure in relation to the LEAs and schools.

The cultural–cognitive dimension of CBSP

As detailed above, cultural–cognitive aspects comprise shared conceptions and a shared, taken-for-granted understanding within an organization of central aspects of a specific phenomenon, in this case, CBSP. This in turn serves as an important platform for organizational sensemaking that works in both homogenizing and heterogenizing directions (see, e.g. Scott, 2008, 2014). Accordingly, despite the regulatory and normative directives provided to SNAE, sensemaking activities rooted in the understandings of specific organizations were still required by the officials when managing and designing CBSP. This third and final section explores these types of sensemaking processes.

An essential element of both the regulatory and normative governance logics is, as shown above, that CBSP should be based on dialogue and on collaboration between the participants, i.e. SNAE and the participating universities. However, despite these shared points of departure, a more nuanced image emerges if we look at CBSP chronologically – i.e. from the first contact of SNAE with the LEAs and schools and the initial analysis until the final evaluation report, about two to three years later, as described here by Official 2:

There is a certain degree of autonomy in the work that should be done, but we also have an “internal operating logic” within SNAE saying that this is what we need to work on and achieve.

In more detail, the analyzed documents summarizing SNAE's experience of the initial steps of the program (see, e.g. Document Dnr, 2015:778) reveal sensemaking processes leading up to the cultural–cognitive understanding that the LEAs and individual schools often lack the capacity to analyze, plan, and implement improvement initiatives. Similarly, the interviewed officials described how this cultural–cognitive understanding had been developed based on the considerable number of meetings and the work they had undertaken with the participating school actors. In that sense, they detailed how they made sense of their assignment and how they formulated ideas about how a dialogical analysis model should actually be designed and implemented within CBSP, and about the difficulties linked to this process, as reflected on by Official 3:

When I started three and a half years ago … I didn’t understand. So should we ask them [i.e. the LEAs and schools] what processes and structures they have? And why do they look the way they do? … They will try to reinvent the wheel, but how can they know what they don’t know? … And often they realize this themselves, or they discover along the way in the analysis that they need help at all levels.

Another cornerstone of CBSP, also discussed above, holds that the program should focus on and be organized around the LEAs' and schools' quality assurance system. However, from a cultural–cognitive perspective, and given that sensemaking activities tend to emerge when there is no obvious way to act and think in an organization, the quality assurance system becomes a way to reduce complexity within an organization in terms of forming and sharing a homogenizing idea linking schools' results and improvement processes. In addition, the sensemaking processes evoke not only a shared cultural–cognitive understanding of which the quality assurance system is a cornerstone but also ideas as to what a desirable quality assurance system should be, as articulated in document Dnr, 2015:778:

In order for the schools' results to improve, and to increase the equality within and between schools, the selected school units and LEAs need to develop an ability to self-renew. An effective quality assurance system with a qualified analysis forms the basis of such an ability.

The cultural–cognitive idea that the quality assurance work is the key to improving results and increasing equality was reinforced in the interviews. In more detail, the interviewed officials confirmed this cultural–cognitive notion, clarifying the connection between the quality assurance system, schools' results and the work toward increased equality, here explained by Official 4:

There is an assumption regarding the quality assurance system and its importance and connection to student outcomes. It has probably not happened that participants [in CBSP] say that everything is going well in the systematic quality work, but for the students it is not going so well. I don’t think that has ever happened.

The above quotation illustrates how the respondent understands the relationship between the schools' quality assurance activities and the quality of teaching. More specifically, the quotation illustrates the cultural–cognitive understanding that without a functioning quality assurance system, on an overall level, the teaching quality in the classroom cannot be good either.

Over time, homogenizing strategies and processes have also been developed from the sensemaking processes concerning how the SNAE officials should work within each CBSP project, regarding process and content. The interviewees explained, for instance, that there were some fundamental concepts that all SNAE teams referred to in their work with the LEAs and schools, as a way to develop a common language linked to the improvement processes. For instance, the expression “putting the fingers in the right jam jars” refers to the importance of clarifying obligations and responsibilities between functions and organizational levels. Other concepts that SNAE recurrently used and implemented in describing different levels within a school improvement process were “structure, process and individual.”

Another key idea highlighted in the policy material and the interviews concerns the school development modules developed by SNAE. These modules could be found on the SNAE website and comprise extensive materials linked to different school improvement areas. They were described as scientifically based and should therefore work as solid school improvement tools and provide a mental “map” for all improvement efforts and for everyone involved in CBSP, including the universities. Accordingly, these modules, and their content, were seen as quality guarantors of CBSP and as exerting homogenizing pressure on the officials' and school actors' sensemaking processes:

[We have] great faith in the school improvement modules developed by SNAE. And then we have prescribed, at SNAE, that we must always consider our scientifically based modules and justify why we should not use them, or why you at the universities should not relate your work to the modules. (SNAE, Official 3)

However, and as noted by Scott (2008, 2014), there are also examples of heterogenizing sensemaking processes in the form of different understandings and interpretations among the SNAE officials of the central ideas of CBSP. One example of this concerns the conditions and consequences linked to the management of CBSP. The fact that the four SNAE teams have considerable autonomy and room for action concerning how CBSP should be carried out was described as both strength and challenge, as elaborated on by Official 4:

The first problem is that we are four teams and these four teams do not cooperate within the processes in any sense. Because we don’t have time to do that. … [Our manager] gives us enormous room for action within the framework of the process. We do have the process map, but it doesn’t explain how we should manage our efforts.

The above quotation highlights an important and interesting aspect, namely, that even though the four teams have considerable autonomy, they still did not manage (or prioritize) inter-team meetings. These conditions underpin heterogenizing sensemaking processes in the form of considerable scope for interpretation among the SNAE officials linked to CBSP. One example concerns the pre-understanding and the entrance into the analysis phase of CBSP. Some officials were clearer in emphasizing the importance of dialogue and of building shared knowledge, while others, as indicated above, had the somewhat different mindset that LEAs and the participating schools could not reasonably be expected to have the required answers and knowledge. Such an approach to school improvement, more characterized by a top–down approach, is illustrated here by Official 1:

Yes, now we are starting with a new municipality that asked if we could read their quality reports, and we usually do that. Then we look at them and then we think that here we need to “explode” some things. … And they use models and matrices, of course, if they fit. If they ask us, we take a look at them, but they rarely do.

Summing up this third and final part of the results presentation, the analysis showed how the regulatory and normative ideas were made sense of in a particular organization based on cultural–cognitive understandings mostly working in a homogenizing direction. However, and importantly, the analysis also showed how cultural–cognitive ideas operated in a heterogenizing direction based on the histories of the four teams and on the sensemaking activities that have taken place over time in CBSP.

Discussion and conclusions

By using CBSP in Sweden as a case, this article explored the state governance of large-scale school improvement and how officials at a national agency of education (i.e. SNAE) made sense of and operationalized this reform in a concrete school improvement program. The relationship between the state decision level and the national school agency level is a hitherto rarely explored relationship in the field of educational management. Accordingly, this article contributes both theoretical and empirical insights into school management and educational reform linked to systemic and large-scale school improvement approaches (see, e.g. Fullan, 2010; Hopkins et al., 2014). In this final part of the article, the main findings will be discussed in light of the article's theoretical concepts and a sample of previous research within the field.

In more detail, by applying key elements of neo-institutional theory, in terms of different dimensions of governance, school management and the sensemaking concept (Scott, 2001, 2008, 2014), this multilayered analysis offers a deeper understanding of governing and managing effective school improvement. This became evident in the implementation of CBSP, in which there was a clear lack of detailed directives from the government department. However, whether this absence of regulatory governance, offering extensive room for action by SNAE officials, should be seen as problematic or preferable is beyond the scope of this article. This article instead demonstrates the importance of having capable leadership if considerable room for action is left to educational professionals, even at the national school agency level. This study has also highlighted the importance of including other dimensions of governance and management besides the traditional regulatory dimension. In fact, given the size and objectives of CBSP, one possible conclusion is that there is an outer limit regarding the use of regulatory directives—a finding that seems transferable to other contexts and improvement programs. Therefore, attention should also be paid to the normative and cultural–cognitive aspects of governance and management to understand the potential and complexity of educational reform and large-scale school improvement. In this regard, the absence of expanded regulatory ideas in the current case demonstrates to educational managers, at different levels of educational systems, the importance of understanding the strengths of normative and cultural–cognitive ideas, often deeply rooted in specific organizations. As shown above, there are always particular “histories” and “logics” in all organizations, regardless of the number of detailed policy directives and intentions. From this follows the importance of considering the professional actors', in this case the SNAE officials', sensemaking process in relation to the state directives. Accordingly, the officials' interpretations and understandings of the current reform, especially when the regulatory dimension of the governance is less detailed, will greatly affect the implementation process and outcome.

These results also evoke important questions linked to specific elements of managing large-scale improvement work, for example, questions linked to distributed versus shared perspectives on educational leadership and school improvement (see, e.g. Brown et al., 2020; Harris, 2011; Harris and Spillane, 2008; Spillane and Healey, 2010). The present results show that the SNAE officials were granted considerable autonomy and trust in interpreting and operationalizing CBSP. However, at the same time, there was no clear infrastructure or supportive framework at the national agency to help managers and officials shoulder this huge and complex assignment. Another condition that reasonably made a trust-based and more distributed working approach more difficult was that CBSP grew gradually and significantly. This implies that the officials have been forced to “lay the rails while driving the train,” which arguably also made it more difficult to distribute roles and assignments and thus lead more collectively. In addition, this made it difficult to take a more comprehensive approach to the improvement program regarding strategies and activities within the program's different phases.

Linked to this discussion, the results also reveal that there were few arenas for knowledge sharing and shared sensemaking activities at SNAE. From a learning organization perspective, which, for example, emphasizes the importance of developing professional learning communities at various levels of education systems (see, e.g. Horton and Martin, 2013; Lomos et al., 2011; Stoll et al., 2006), this is problematic. Obviously, there was a lot to learn and make sense of that could bolster the internal improvement capacity at SNAE. However, the analysis showed that there was no clear idea of how these sizable learning opportunities could contribute to organizational learning and thus become natural parts of SNAE that could reinforce its ongoing work (cf. Senge, 2006). Because the cultural–cognitive dimension comprised such an important part of the implementation process, the lack of such collective processes resulted in the emergence of several heterogenizing ideas among officials at the national agency linked to CBSP. Arguably, building structures and routines that enable such a learning organization is especially important when the regulatory governance dimension is weak.

In conclusion, previous research has offered important results concerning effective school leadership (see, e.g. Bendikson et al., 2012; Hallinger, 2005; Leithwood et al., 2020) and how successful LEAs operate and support school managers in various contexts (see, e.g. Anderson and Daly, 2013; Lee et al., 2012; Liljenberg et al., 2022; Nordholm, 2016). Given that less attention has been paid to the level above the LEA level, i.e. to the national school agency (i.e. SNAE) level, this article makes an important contribution by showing how different types of governance (i.e. regulatory rules, professional norms, and cultural–cognitive beliefs) could set the direction for managing large-scale school improvement at the state school agency level. In particular, in the current case, the lack of clear regulatory directives permitted sensemaking processes clearly influenced by normative ideas and cultural–cognitive beliefs. In that sense, this article echoes and extends previous work on sensemaking in educational research (see, e.g. Coburn, 2005; Spillane and Anderson, 2014; Spillane et al., 2002). It also echoes previous research showing the importance of the context in which policy is to be translated and implemented (cf. Seashore Louis et al., 2008). Lastly, the article demonstrates the complexity of educational reform and large-scale school improvement in which various actors not only must make sense of directives but also must transform them into concrete routines (cf. Harris and Jones, 2018). This, as noted by Brauckmann et al. (2023), demands a profound understanding of contextual leadership factors, evidently not only among principals and superintendents but also at the national agency level.

Regarding the limitations of this work, the present findings are mostly presented from the perspective of SNAE managers and the sample of interviewees is restricted. Therefore, even though the interviewees possessed essential knowledge of CBSP and the link to the state level offering a unique perspective, further studies using, for example, questionnaires and/or focus group interviews could include a larger number of officials involved in the program and attain a deeper understanding of the state-agency level's role and function in interpreting and operationalizing this large-scale school improvement reform.