Swedish school goals are set at the national level by parliament and the government. The local school board is responsible for assuring an appropriate and effective organisation whereas principals and teachers are responsible at the school level for the  students’ learning environment. The hierarchical rationalistic idea is that each level in this multi-level system (the governing chain) takes responsibility for what is stated in the legal governing documents (Nihlfors & Johansson, 2013). The common goal is to promote the  development and education of all students and to encourage  a lifelong desire to learn (Skolverket, 2011). A school´s outcome is ultimately symbolized by the student’s graduation certificate. 

          Of great importance in Sweden is the concept of pedagogical leadership whose  main target is to help students achive their goals. This concept embraces different types of leadership, termed in international discourse as ‘transactional leadership’, ‘transformational leadership’, ‘instructional leadership’, etc. (Burns, 1978). Pedagogical leadership can be exerted directly or indirectly (Törnsén & Ärlestig, 2014). All levels in the multi-level school system exert such leadership (Uljens, Sundqvist & Smeds-Nylund, 2016). In such a rationalistic systemic thinking, the upper level provides the lower level with the best conditions for its work. A prerequisite for pursuing a coordinated pedagogical leadership is that actors in the multi-level system have a common understanding of existing goals. If the actors are not linked by common beliefs and values (Robinson, 2017), the base for pedagogical leadership is weak.

          An effective multi-level governing chain presupposes good functionality at each level and between levels. However, this is not always the case;  research shows challenges to governing a multi-level-system(Johansson & Ärlestig, 2020; Ståhlkrantz & Rapp, 2020). This is often expressed as difficulties in governing loosely-coupled systems (Boyd & Crowson, 2002; Fuzarelli, 2002; Weick, 1976). One reason for couplings can be found in the existing asymmetric relation of powers (Rapp, Aktas & Ståhlkrantz, 2020). An obvious example is that  the knowledge advantage of  teachers (as agents) determines the teaching contents. The higher level (the principal) cannot take for granted that agents act in line with the principal´s interests (Ferris, 1992; Jensen & Meckling, 1976).

          One way to strengthen the cooperation in the chain is to create  ‘communities of practice‘ (Wenger, 1998)  Coldren & Spillane (2007) also talk about ‘communities of practice’ and argue that by acting as boundary spanners actors will affect each other’s work. Through boundary-spanning the gaps can be sealed and a more tight connection can be established in the multi-level system.

          There are not many Nordic studies that simultaneously deal with pedagogical leadership, multi-level school system and students’ outcome. One exception is Uljens, Sundqvist & Smeds-Nylund (2016) who have studied the non-hierarchical Finnish school system. Their  results show, among other things, the importance of cooperation, trust and a culture aiming at coherence (Fullan, 2015) in the chain.         

          A purpose of this study is to identify, formulate and present appropriate efforts to influence cooperation in the governance chain and find out how the chains’ connections can be tightened. In  turn this will increase  understanding about  how boundary-spanning activities can be developed to strengthen the connections in the chain. Furthermore, this study will increase our knowledge about the conditions required for the governing  opportunities of  the chains to affect students’ learning. The questions are: 1) How is pedagogical leadership exerted on different levels of the governing chain? 2) What role can boundary spanning play for students knowledge development?         

          The study has been conducted in a smaller town/urban area (SKR, 2016) in the south of Sweden. Results from the study will be compared with outcomes from Uljens, Sundqvist & Smed-Nylund (2016).   

Using the case-study approach (Yin, 2009), our study focused on municipal actors in the local governance chain of the comprehensive school (students 6-16 years). Its empirical material consists of interviews with a local school board chair, vice chair, superintendent, vice superintendent, four principals and nine teachers. Interviews were semi-structured and based on open-ended questions (Cohen, Manion & Morrison, 2018). The goal was to make dominant discourses visible. The starting point of these interviews was a basic questionnaire, one that allowed the interviewer to probe deeper into responses of special interest. Informants were asked to describe how the local school actors work and the expectations the informants had of other levels in the governance chain. Interviews took 45-70 minutes, were recorded and transcribed in full. Afterwards each text was processed and analysed with the goal of improving the understanding of the phenomena. In the initial analysis steps, the texts were read briefly and, based on the study´s aim, comments of special interests were identified. Subsequently, the texts were reread several times and the relevant comments were categorized. Proper ethical guidelines have been observed throughout the research process (Vetenskapsrådet, 2017). Before the interviews all informants were informed about the study´s aim. At any time informants could terminate their participations in the study. The implementation of this study presented us with several challenges regarding informant confidentiality. For example, even though the name of the town is not mentioned, the municipality might still be identified and thus it could be possible to identify some of the informants. Those informants whose identities might be revealed (the chair, the superintendent) were told about this confidentiality issue and asked whether they had objections to being part of the study – none did. To further mitigate the effect of a potential confidentiality breach, the chair and the superintendent were provide with printed texts of their interviews and given the possibility to comment on them. This step had the double effect of 1) making sure there were no inaccuracies in the transcribed text and 2) since these informants could be recognised, assuring that this would not create problems for those individuals.. One informant did ask for some adjustments, which were an editorial nature and did not change the content in the original text. An additional benefit of the informants’ review of their own text is that the validity of this data have been strengthened.

The Swedish school system is hierarchically built and is based on the rationalistic idea that what is stated in governing documents will be realised in schools. Outcomes are measured by student grades. Student graduation certificates symbolize the governing chains’ pedagogical leadership. The analysis of this study´s empirical material indicates that actors in the governance chain mainly exercise indirect pedagogical leadership, for example, by governing through documents and following up on outcomes (transactional and transformational leadership). A direct pedagogical leadership is exerted mainly by teachers meeting with students (instructional leadership). Furthermore, study results reveal poor understanding of a common goal, complicating conditions for boundary spanning. Even with consensus between the school board and the superintendent, challenges exist in the transfer (the boundary room) between the superintendent and principals, and even in the boundary room between principal and teachers. Formal power (legality) to act as boundary spanner and intervene is reserved for the principal (see principal-agent theory). At the same time the agent can either cooperate (legitimate) or counteract (not legitimate) a boundary spanning. Boundary spanning assumes that principal and agent have intentions to cooperate, but a weak cooperation will decrease the opportunities to affect students’ knowledge development. If the Swedish hierarchical governance chain adapts the Finnish non-hierarchical model (pedagogical leadership cooperation, coherence and trust), possibilities of working towards increased outcomes will be strengthened. That will also tighten the loosely-coupled chain. By striving towards a more non-hierarchical system, principals and agents on a common platform can cooperate toward a common goal, that is, promoting increased learning by all students. This can strengthen both pedagogical leadership and trust in the governing chain. The boundary spanning focus must continuously be directed towards the chains most important level – the teachers! It is in the meeting between teacher and student that outcomes improve.

Burns, J. M. (1978). Leadership. New York: Harper Torchbooks. Boyd, W. L. & Crowson, R. L. (2002). The quest for a new hierarchy in education: From loose coupling back to tight? Journal of Educational Administration 40(6) 521-534 Coldren, A., & Spillane, J. (2007). Making connections to teaching practice: The role of boundary practices in instructional leadership. Educational Policy, 21(2), 369–393. Cohen, L., Manion, L., & Morrison, K. (2018). Research methods in education (8th ed.). Abingdon: Routledge. Ferris, J., M. (1992). School-Based Decision Making: A Principal-Agent Perspective. Educational Evaluation and Policy Analysis 14,(4), 333-346 Fullan, M. (2015). Coherence: The Right Drives in Action for Schools, Districts, and Systems. Corwin Press Fuzarelli, L. D. (2002). Tightly coupled policy in loosely coupled systems: institutional capacity and organizational change. Journal of Educational Administration 40(6), 561-575. Jensen, M. C. and Meckling, W. H. (1976). Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure. Journal of Financial Economics 3(4), 305-360. Johansson, O. & Ärlestig, H. (2020). Skolans styrning – om styrkedjan och dess mellanrum. I Johansson, O. & Svedberg, L. (red.) Att leda mot skolans mål. Vägval och möjligheter. Gleerups. Nihlfors, E., & Johansson, O. (2013). Rektor - en stark länk i styrningen av skolan (1. uppl.). Stockholm: SNS förlag. Rapp, S., Aktas, V. & Ståhlkrantz, K. (2020). Schoolboards' expectations of the superintendent – a Swedish national survey. Educational Review. https://doi.org/10.1080/00131911.2020.1837740 Robinson, V. (2017). Reduce Change to Increase Improvement. Corwin: Thousand Oaks Skolverket (2011). Läroplan för grundskolan, förskoleklassen och fritidshemmet 2011 - Lgr 11. Stockholm: Skolverket. SKR (2016). Kommungruppsindelning 2017. Omarbetning av Sveriges Kommuners och Landstings kommungruppsindelning. Stockholm: Sveriges Kommuner och Landsting. Ståhlkrantz, K., Rapp, S. (2020). Superintendents as boundary spanners - facilitating improvement of teaching and learning. Research in Educational Administration & Leadership REAL. 5. 376-415. Törnsén, M. & Ärlestig, H. (2014). Om ledarskap och pedagogiskt ledarskap. I Törnsén, M., & Ärlestig, H. (Red.). Ledarskap i centrum : Om rektor och förskolechef (s. 17-29). Gleerups. Uljens, M., Sundqvist, R., & Smeds-Nylund, A-S. (2016&. Educational leadership for sustained multi-level school development in Finland. Nordic studies in Education 36(2) 103-124. Vetenskapsrådet (2017). God forskningssed. https://www.vr.se/download/18.2412c5311624176023d25b05/1529480532631/God-forskningssed_VR_2017.pdf Weick, K. E. (1976). Educational organizations as loosely coupled systems. Administrative Science Quarterly, 21, 1-19. doi:10.2307/2391875 Wenger, E. (1998). Communities of Practice. Learning, Meaning and Identity. Cambridge University Press. Yin, R. K. (2009). Case study research: Design and methods (4th Ed.). Thousand Oaks, CA: Sage.