# Leading from the middle: its nature, origins and importance

Andy Hargreaves (Lynch School of Education and Human Development, Boston College, Chestnut Hill, Massachusetts, USA)
Dennis Shirley (Lynch School of Education and Human Development, Boston College, Chestnut Hill, Massachusetts, USA)

ISSN: 2056-9548

Article publication date: 15 January 2020

Issue publication date: 22 January 2020

3035

## Abstract

### Purpose

The purpose of this paper is to cover a 10-year period in ten of Ontario’s 72 school districts on the nature, origins and importance of “leading from the middle” (LfM) within and across the districts.

### Design/methodology/approach

The research uses a self-selected but also representative sample of ten Ontario school districts. It undertook three-day site visits in each of the districts, transcribed all the interviews and compiled an analysis into detailed case studies.

### Findings

LfM is defined by a philosophy, structure and culture that promotes collaboration, initiative and responsiveness to the needs of each district along with collective responsibility for all students’ success.

### Research limitations/implications

To be sustainable in Ontario, LfM needs support and resourcing from the top. The current environment of economic austerity therefore threatens sustainability. Globally, examples of LfM are emerging in at least three other systems. The analysis does not have identical questions or respondents in phases 1 and 2. Ontario’s version of LfM may differ from others. The collaborative design may downplay criticisms of LfM.

### Practical implications

LfM provides a clear design for leading in complex times. Compared to top-down leadership the whole system can address the whole of students’ learning and well-being. LfM is suited to systems and cultures that support local democracy, community responsiveness and professional empowerment and engagement.

### Originality/value

LfM is an inclusive, democratic and professionally empowering and responsive process that differs from other middle level strategies which treat the middle merely as a way of connecting the top and bottom to get government policies implemented more efficiently and coherently.

## Citation

Hargreaves, A. and Shirley, D. (2020), "Leading from the middle: its nature, origins and importance", Journal of Professional Capital and Community, Vol. 5 No. 1, pp. 92-114. https://doi.org/10.1108/JPCC-06-2019-0013

## Publisher

:

Emerald Publishing Limited

## Introduction

One key characteristic of high-performing educational systems is a strong teaching profession (Hargreaves and Shirley, 2009, 2012; OECD, 2011; Tucker, 2019). Ontario, in Canada, has been repeatedly identified as an exemplary high performer, not just because of its overall achievement scores, but also due to its high levels of equity in a context of considerable cultural and linguistic diversity (Campbell et al., 2017; Fullan and Rincón-Gallardo, 2016; Mehta and Schwartz, 2011). One of the features of Ontario and other Canadian provinces that also perform well is a teaching profession that has high status, that is well rewarded, and that experiences a positive environment of professional development, collaboration and support (Campbell et al., 2017; Fullan and Quinn, 2015).

A robust teaching profession is evident in three components of professional capital. First, human capital refers to the capacities of highly qualified teachers who have been trained in universities with strong connections to practice. Second, decisional capital entails the professional judgment developed over time through practice, mentoring and coaching. Finally, social capital describes how teachers plan, reflect and support each other collegially (Hargreaves and Fullan, 2012).

An important part of social capital among educators is how leaders collaborate to support and develop their teachers and each other in understanding and undertaking improvement and innovation together (Ainscow, 2015; Bryk and Schneider, 2002). After teaching, school leadership is the second most important in-school influence on student achievement (Leithwood, 2018). Like teachers, school leaders also benefit from opportunities to build their social capital by collaborating with colleagues across schools and systems (Evans and Stone-Johnson, 2011).

This paper reports the results of research covering a ten-year period in ten of Ontario’s 72 school districts on the nature, origins and importance for system coherence on the one hand, or collective professional responsibility to learners and communities on the other, of what Ontario educators themselves have termed “Leading from the Middle” (LfM) within and across their districts. It distinguishes LfM from top-down and bottom-up leadership. It also makes a defining distinction between the new concept of LfM, and alternate conceptions of leading in the middle (LiM) (Barber and Day, 2014; Schleicher, 2015).

Conclusions and implications are drawn about the sustainability of LfM as a leadership and change strategy within Ontario – especially in light of a sudden shift in political control and direction within the province – and about the spread of LfM beyond Ontario to other systems as the start of a small global movement.

Bad leaders can be strong leaders – people with relentless energy, determination and strong will. This is why researchers who advance seemingly neutral frameworks supporting “strong leadership” need to be cautious (Leithwood, 2013). Strong leadership can be evil as well as effective. It can become strongman leadership (Bauman, 2016). The manifestation of strong leadership as a set of vices rather than virtues is most obvious in the behavior of political demagogues.

While some kinds of strong leadership in education can be expressed through genuine displays of courage and integrity (Shirley and MacDonald, 2016; Shirley, 2017), strong leadership as “strongman” leadership, whether practiced by men or women, can also manifest itself in everyday ways. These include school principals who mistreat or silence their teachers through stonewalling them, not responding to their repeated requests, or isolating them from their colleagues (e.g. Blase and Blase, 2002).

In psychiatric terms, many unethical bad leaders are what Kets de Vries (2006) identifies as narcissists, sociopaths and control-freaks of organizational change. They are common in the high-threat environments of turnaround schools and low-performing districts in the USA and England. In these countries the pressures for improved achievement scores are relentless and intense. Punitive measures to secure rapid improvement in such environments have been commonplace (Daly et al., 2011; Duke, 2016; Mintrop, 2004; Hill et al., 2016).

Not all top-down leadership is bad, though. In the early 2000s, for example, Ontario tried to learn from and mitigate the punitive excesses of its Anglo-Saxon global counterparts. Following the management theories of Kotter (1996), a guiding coalition that included professional leaders and other partners focused on a small number of priorities related to literacy, math and high school graduation rates (Fullan et al., 2006). The Ministry of Education provided schools with extensive professional development and financial support for implementation. The government gave teachers on-the-ground support in the form of mentors and coaches whose skills were closely matched with the needs of particular schools. Literacy attainment improved significantly and sustainably, and achievement gaps narrowed for a number of groups, including second language learners and children with learning disabilities (Fullan and Rincon-Gallardo, 2016).

This strategy yielded measurable success where the educational change goals were simple and few. But, as Bentley (2010) points out, the strategy is weakly suited to complex and uncertain environments where innovation has to be encouraged, and multiple initiatives addressing different constituencies are being pursued simultaneously. Indeed, with Joanne Quinn, Michael Fullan has subsequently recognized the limitations of this model that he previously promoted, and, following Hargreaves and Shirley (2009, 2012), and along with Johnson et al. (2015), now argues that complex systems need interlocking coherence rather than merely linear alignment of strategies and reform elements (Fullan and Quinn, 2015). More widely, Elmore (2016, p. 531) has similarly revised his own earlier and influential exposition of the linear nature of “getting to scale” (Elmore, 1996). This approach, he now argues, was “either very superficial or downright wrong” (Elmore, 2016, p. 531) because it was naïve about the complexity of change.

Top-down change in education can work when the purpose is straightforward, the results are easily measured, and there is public confidence in schools. Such an approach is anachronistic, however, when the goals of school systems are increasingly complex (Payne, 2008; Sarason, 1990). Policy makers around the world are now calling for the development of innovation, creativity, critical thinking, inclusion, well-being and senses of belonging. These goals are impossible to implement straight from the top because leaders at the top cannot possibly know everything across such a broad span of student learning.

In contrast to top-down approaches, some reformers are now advocating for greater autonomy for schools and teachers, increased freedom for curriculum design, and personalized uses of technology (Horn and Staker, 2015; Mehta and Fine, 2019; Salokangas and Ainscow, 2017). But the previous history of bottom-up innovation and school autonomy is not impressive. In the 1960s and 1970s, innovative ideas often didn’t spread beyond a few isolated classrooms or schools (Gross et al., 1971; Loucks and Hall, 1977). When innovations did spread, their implementation was often superficial (Anderson, 2010; Cohen, 1990). Teachers often used new methods that they did not fully understand, or even when they did understand them, these methods were isolated experiments that withered away without adequate and continuing support.

Bottom-up innovations, wherever they are, often have value in their own right (Chung, 2011). A government or system does not have to adopt an idea or strategy across all its schools for an innovation to have widespread impact. Howard Gardner, for example, has never attempted to implement or directly influence any system to implement his original and innovative ideas wholesale, but in his 2009 book on Multiple Intelligences Around the World, 42 authors from 15 countries on five continents write about the multifarious ways in which they have applied his ideas (Chen et al., 2009). Fink’s (1999) study of declining innovation in a trailblazing high school found that leaders who had spent some of their most formative professional years in the school, subsequently drew on and adapted much of what they learned when they assumed large-scale system leadership positions later in their careers (see also Fink and Brayman, 2006). However, without explicit strategies to expand the impact of local innovations in real time, there is no reason to believe that reforms based on bottom-up change will fare any better than top-down ones in having positive system-wide impact in the present, as well as over longer periods of time in the future.

One of the consequences of putting too much emphasis on control at the top, even when it is combined with freedom or individual school autonomy at the bottom, is that it has bypassed the middle of the system. This has led to incoherence across entire systems. Without engagement of the middle, the top lacks first-hand knowledge of what is happening in the schools, and the bottom lacks colleagues who can help to communicate what is transpiring in other schools undertaking similar reforms.

The idea of introducing a middle level, middle tier, or simply middle leadership in school systems has therefore become increasingly attractive. This is especially true where there has been a decline in the importance of intermediary local control of public education. In England and Sweden, and in association with the charter school movement in the USA, for example, districts, local authorities, or municipalities, as they are variously called, have been by-passed, weakened or eliminated altogether. Districts were completely abolished in New Zealand in 1989 (Wylie, 1997). In Canada, Nova Scotia’s nine school districts have been eliminated following a government commissioned review (Glaze, 2018). In England, local authorities have also had their powers seriously curtailed, even though this has led to no demonstrable overall improvement in student outcomes (Ladd and Fiske, 2016).

Where strong direction from central government exists in combination with marketplace competition among individual schools, attention to developing some kind of middle level leadership has sought to plug the policy implementation gap. In this respect, Michael Barber and his colleagues have argued for a new middle tier to replace the traditional middle stratum of democratically elected school districts or local authorities. In a report for the Massachusetts Business Alliance in Education about the future of education in the commonwealth, they referred to earlier work with Barber’s colleagues at McKinsey and Company where they identified “a critical role for what they called the “middle tier.” The first responsibility of the middle tier, they asserted, was “to provide targeted support to schools and monitor compliance.” Second, they stated, the middle should “facilitate communication between schools and the center.” Third, the middle should “encourage inter-school collaboration.” Finally, the middle tier should “moderate community resistance to change by making the case for a different future” (Barber and Day, 2014, p. 45; our italics). The idea of a “middle tier” here appears to be part of an effort to minimize democratic control and the public’s possible opposition to government policies. Here, the middle serves as a buffer that will insulate policy makers from parents or community activists who could question or challenge their leadership.

Other cases for middle leadership emphasize the idea of coherence in school systems. Michael Fullan (2015, p. 1) has described LfM as:

[…] a deliberate strategy that increases the capacity and internal coherence of the middle as it becomes a more effective partner upward to the state and downward to its schools and communities, in pursuit of greater system performance […]. This approach is powerful because it mobilizes the middle (districts and/or networks of schools), thus developing widespread capacity

(Our emphases in italics).

Like Fullan, Andreas Schleicher (2015) has also argued for a stronger role for the “meso” or middle level of change in school systems. Here again, the middle is seen to play a role by helping to implement changes from the top, and to move around ideas and strategies that are percolating up from beneath. The middle improves efficiency and performance by breaking down the walls of miscommunication and misunderstanding that can flourish in large organizations like school systems. However, it does not appear to have much driving force or momentum, or a clear identity of its own.

For others, middle level leadership is about a set of roles that have career pathways, formal responsibilities, and additional remuneration. For example, Jonathon Supovitz (2014, p. 28) refers to how system leaders in England describe middle leaders as those “who worked closely with teachers, who could more closely foster and monitor individual teachers’ improvement” in order to enhance performance. These middle leaders give teachers the sense that they are being listened to. One potential danger of being in the middle, like this, though, is that it can lead to what Alma Harris and Michelle Jones (2017, p. 214) describe as tensions “between expectations that the middle leader role had a whole-school focus versus […] loyalty to their department, and […] between a growing culture of line management within a hierarchical framework versus a professional rhetoric of collegiality.”

In all these cases, middle level or middle tier leadership is a role or function that creates coherence, increased efficiency and enhanced performance in a complex system. This idea applies especially in instances where democratically controlled districts have been weakened in order to remove restrictions to market competition and centralized control. LiM in such settings is about creating more efficient and, sometimes, more competitive systems, not building better democratic and professional communities.

Middle leadership has surfaced, therefore, as a solution to systems where local control has been diminished. But what does and should LfM look like in societies and communities that resist the marketization and privatization of education and that retain an interest in strong democratic control? Some societies value an attachment to the power of communities, towns and cities, not just as middle levels that connect other people’s policies and purposes, but also as democratic creators and drivers of educational purposes for themselves. What can it look like to lead from the middle in these jurisdictions, where districts are strong and valued, rather than weak and vulnerable?

The province of Ontario, in Canada, is an example of a system that, over the past decade and more, has achieved high performance and rapid improvement in a context of great diversity by bringing together teachers, leaders and schools within and across its 72 districts to pursue collective responsibility for all students’ success. We have had the opportunity to investigate, interpret and also support this system of what its educators call LfM over the past decade.

## Context and research design

The research reported in this paper was funded to document collaborative work with a Consortium of ten Ontario school districts as a follow-up to a 2008–2011 evaluation of a government-funded project entitled “Essential for Some, Good for All” (ESGA). This prior project focused on improving the academic results of students with special needs in all 72 of Ontario’s school districts (Hargreaves and Braun, 2012). The Consortium was facilitated by the Council of Ontario Directors of Education (CODE).

One key outcome of the study was identification of a new strategy described by Ontario’s educators as “LfM” (Hargreaves and Braun, 2012; Hargreaves et al., 2018). This term was used to describe situations where districts developed their own strategies to promote inclusion of all students in ways that suited the diversities of their own communities, then networked and circulated these strategies across the districts. This circulation of knowledge occurred in an environment of transparent collective responsibility for participation and results.

The follow-up research to the ESGA project reported more extensively in this paper was developed in collaboration with the ten CODE Consortium districts at their renewed invitation to assist further progress in their continuing efforts to “lead from the middle” through a range of selected district-by-district projects focused on implementing new government policies and priorities that had been advanced in 2014 (see below), after the completion of the first research study. This second study was requested in order to articulate the theories of action undergirding the Consortium’s projects. A research team was tasked with gathering perceptions of the projects’ strengths and limitations, connecting findings to existing leadership literature, and providing feedback to participating districts and to other districts outside of the Consortium.

The study reported here had four research goals that were developed with due consideration for the Ministry of Education’s new priorities as expressed in a report entitled Achieving Excellence (Ontario Ministry of Education, 2014). The first goal, and the one most relevant to this paper, was “to understand and articulate the model and theory of action undergirding the CODE Consortium’s current LfM projects, so they can be communicated clearly to project participants and diffused more effectively to other jurisdictions” (Hargreaves et al., 2018, p. 12). Second, the study sought to “gather perceptions of the projects’ strengths and weaknesses.” Third, the research team would “connect these findings to the existing evidence base of leadership models” detailed previously in two reports (Hargreaves and Braun, 2012; Hargreaves and O’Connor, 2017). Fourth, the study would “share the interim and final results” to support “improvement planning.”

In the first project, a research team had worked with a self-selected but also representative sample of ten Ontario school districts to interpret and explain the core principles and practices that had been adopted in ESGA. It conducted a survey of self-reported perceptions and practices related to ESGA among a sample of school principals, teachers and special education support staff in nine of the ten districts. It also undertook three-day site visits, using rotating research teams to strengthen cross-case validity and reliability in each of the districts, transcribed all the interviews, and compiled an analysis of them into detailed case studies of approximately 10,000 words each as a basis for subsequent analysis. An 11th case was also added of interviews with senior policy makers and leaders in CODE and the provincial Ministry of Education.

The methodology used in each of the two studies has been described in two reports (Hargreaves and Braun, 2012; Hargreaves et al., 2018). A summary of the first research study is provided in this section as a basis for contextual understanding of and comparison with the second period of collaborative research with the school districts that commenced in 2014. The analysis in the first study identified a four-pronged strategy for ESGA. First, $25m of government funding to develop an inclusive education strategy was distributed equally across all districts, large and small, to build commitment among a critical mass of school district leaders, since three-quarters of Ontario’s school districts are small or medium in size. Second, within the province’s philosophical framework for inclusive education, each district designed and developed its own inclusive education strategy to respond to the unique diversities of its own community – be this Indigenous Canadian, Franco-Ontarian, immigrant and refugee, Catholic, Old Order Mennonite, poor working class, etc. Third, a small coordinating and facilitating team of retired superintendents was appointed to connect the districts with each other, circulate knowledge among them, and cultivate senses of shared responsibility across them for all students’ success. Last, districts were required to report the results of their efforts to each other in face-to-face annual meetings. These interrelated elements of the CODE strategy were described by one of the coordinating and facilitating team in our 2009 interviews as “LfM.” That is where our research team’s engagement with and articulation of this leadership and change strategy originated. Over time, following our first study and the inclusion and dissemination of our results in our book (Hargreaves and Shirley, 2012), the concept of LfM began to enter the mainstream vocabulary of Ministry policy makers and school district administrators more widely. It also awakened the curiosity of ten of the 72 districts (nine of whom had participated in the study), who wanted to continue learning from one another. Together, this new iteration of the CODE Consortium membership and the research team refined the understanding of LfM to encompass seven principles that informed their new separate and combined change projects and drove subsequent research. These are represented in Figure 1. The principles were described as follows: 1. Responsiveness to diversity: districts and their schools work with others to generate solutions that respond to distinctive local needs and diversities through practices like differentiated instruction and universal design for learning. LfM projects engage with students’ distinctive identities and develop cooperation to better support students among special education support teachers, curriculum consultants and regular classroom teachers. 2. Responsibility: districts take collective responsibility for all students’ success by working in professional learning communities. They examine student data and bring together teachers with special education consultants, speech pathologists and mental health specialists, to devise strategies to support all of the students in the districts who have struggled with their learning. A culture of “my students” becomes a culture of “our students.” 3. Initiative: LfM is about fewer initiatives and more initiative. It is about educators themselves seizing the initiative together to acknowledge and respond to challenges in their schools and communities and to develop strategies to address them. 4. Integration: districts seek to integrate their efforts with government priorities wherever possible, by linking to literacy reforms or efforts at the time to close achievement gaps, for example. 5. Transparency: districts act together to establish transparency of participation and results regarding their progress in improving learning, establishing well-being, and building identity. They share their strategies and results with each other through the networks of their relationships and at public sessions that display their projects and their impact. 6. Humility: no district sees itself as superior to all the others. Each district demonstrates curiosity to learn from the rest. All districts commit to learning from other systems elsewhere. 7. Design: districts work together to ensure that the six prior principles are put into place through deliberate designs, and then disseminated throughout their schools and systems. At Consortium meetings, in the second project, the districts worked together with the research team to use these seven principles of LfM to organize the next phase of shared investigation and project development. The example in Figure 2 was used for professional development purposes in a CODE Consortium meeting in 2015 in Toronto. Consortium participants filled in the slides in advance of the meeting so that members from all of the other districts could see what kinds of progress they were making – as well as challenges they were facing – with their LfM projects, in regard to identity. The slide in Figure 3 on student and teacher engagement exemplifies how these collaborative activities, conducted in a spirit of shared inquiry, then connected with the districts’ projects as developed within the LfM framework. As the research developed in collaboration with the districts, it became clear that, compared to the period when ESGA was being implemented, there was less orchestrated effort or investment on the part of the Ontario Ministry of Education to support a coordinated and continuing strategy of LfM across the districts. Specific resource allocations for leadership across and among the districts were discontinued. The key administrator (the Assistant Deputy Minister) in the province who had become the system’s champion for LfM was recruited to be Director of one of Canada’s largest school districts. In effect, the new project continued the cross-district strategy, but now only across the ten participating Consortium districts themselves, not among all 72 of Ontario’s districts. As a strategy of province-wide coherence and cohesion, LfM was no longer center stage. Still, as an idea and a strategy, LfM remained very much alive within the participating districts themselves. So, as part of our overall interviewing strategy and protocol, we asked participants to describe their own understandings of LfM as they experienced it at the time and in the past. ## Methodology Our five-member research team conducted initial visits to CODE Consortium meetings in 2015 to begin collaboration and develop our research design alongside members of the CODE Consortium as well as the Ontario Ministry of Education. In May 2016, our team conducted site visits to all ten of the participating districts across the province. At least two team members visited each district. Team membership was mixed and rotated in order to enhance cross-validation of interpretation. Team members conducted interview-based mini-case studies over one to two days with each school district. We interviewed 222 educators, selected project leaders, and project coordinators at the district and Ministry level. Interviews lasted approximately one hour each and were conducted in private locations in each district office or school building. Interviews were audio recorded with the knowledge and informed consent of the interviewees. Further interviews were conducted with a small sample of administrators, teachers and support staff from a range of elementary and secondary schools. These were chosen in consultation with district leaders who participated in the CODE Consortium. Separate interviews were conducted with CODE Consortium leaders and a group of senior Ministry of Education staff. Semi-structured interview protocols were developed collaboratively with members of the CODE Consortium. Questions were constructed to elicit information on the design of their improvement projects, as well as to obtain evidence of educators’ perceptions regarding the implementation and the impact of LfM and LfM projects. Questions were piloted during the initial collaborative meetings in the Spring of 2016 and then implemented during the May 2016 data collection period. The sample was of the same nine districts that had participated in the earlier study of ESGA, with a substitution of one Franco-Ontarian district by another. This sample is representative of districts across the province in terms of geographical spread, urban and rural distribution, religious and non-religious schools, and standardized test scores. The districts volunteered to participate in the research and to provide funding through CODE. Participants were recruited through purposive sampling, assisted by the CODE leaders. This form of recruitment was selected because the CODE leaders have no power or authority over participants. No conflict of interest, coercion, or undue influence was involved. CODE is a non-governmental, third party entity. It provides services with which participants choose to engage. Members of the research team were trained to apply the informed consent procedure in accordance with current standards for research ethics. All participants signed informed consent forms before the interview commenced, once they decided that they were willing to participate in the study. The team analyzed interviews using the constant comparative method (Glaser and Strauss, 1967) to identify salient themes within each of the ten districts, including ones concerning LfM. This was followed by member-checking with participants at meetings of the CODE Consortium. After initial coding, the team wrote individual case studies of 5,000–10,000 words each for internal use only, describing findings from each individual district, based on the themes and also on the emerging narratives inherent to each district. Each district was assigned a case number and the research team created a consistent citation format to keep track of all the supporting quotes to document exactly when and where they originated. For example, O1_Principal#1_5.14 describes the first principal interview conducted in the first (alphabetically organized) school district on May 14, 2016. This procedure was adopted to ensure accurate retrieval of data and to ensure that the data came from a wide range of participants and districts. For ease of readability, we simply refer to educators’ roles when participants are quoted in this paper. A timeline of policy developments in Ontario since 2002, and of the initiation and completion of the two phases of our research related to ESGA and LfM is presented in Figure 4. ## Findings According to a Superintendent of Curriculum in one of the districts, LfM “kind of morphed into different things” over time. By 2014–2018, even though LfM had less funding and no governmental support aside from encouragement and attendance of senior Ministry staff at Consortium meetings, it had assumed greater depth and complexity in relation to students, teaching and learning. It had become an embedded set of practices that varied in their structures and design yet were clearly recognizable by Consortium participants through member checking with diverse constituents across all ten of the districts. These had philosophical, structural and cultural components. ### A philosophy of the heart of practice Consortium educators viewed LfM as their concept, something that they had created and sustained even in the absence of government support. It was their initiative – not anything that came out of the Ministry or any other branch of the government – and they had held onto it and developed it in a spirit of genuine collegiality in spite of a year of labor unrest and other distractions. A director in one district explained: I think what we’re trying to do, and to me the essence of Leading from the Middle, is making real your vision. It’s about moving those ideas into concrete practice and making a positive difference, for all your students and for all your staff so that everybody just loves their learning environment. I don’t see it [LfM] as connected to personnel. I see that the whole notion of Leading from the Middle is a concept or a philosophy. [It is] this idea of wanting to get as close to the action as you possibly can. In this view, LfM is not a location such as a middle tier. Instead, it means getting close to the teaching and learning that is at the heart of the profession. In the words of a mental health consultant, “Sometimes the middle is the students, sometimes the middle’s the teacher. It depends on where learning is happening, or where the learning is.” A Superintendent in a (publicly funded) Catholic district explained: “Leading from the Middle really speaks to us through our Catholic social teachings and that notion of subsidiarity, that the work and the change and the impact of that change will happen at the ground root.” Educators stated that LfM, in its essence, is about supporting students with all their diverse identities. One district superintendent said that “Leading from the Middle really forces us to look beyond those categories, those roles, to see how we can serve others to really uplift them.” LfM as a vision or philosophy of practice that stayed close to students was evident when Consortium teachers identified “students of wonder” and worked with diverse teams of colleagues to study these students’ assets as well as their challenges. It was also manifest when students learned about the lives of missing Indigenous women, studied inequities in water quality across communities, or brought a Syrian refugee family to their community after studying the refugee crisis in their social studies classes. Teachers stated that LfM projects took concrete shape in classrooms and schools when they strove to develop students’ skills of self-advocacy in writing their own Individual Education Plans, when students were engaged in inquiring into and representing their own mental health issues, and when they assessed student’s work by sitting beside them in processes of pedagogical documentation (a kind of assessment for learning) rather than using standardized tests to make data-driven interventions. These cheek-by-jowl forms of engaging students in reflecting on their own learning were also evident when forums were organized for all students to share their ideas about what could be done to improve school climate and when apps were designed that enabled students to inform educators when students are concerned about other students’ lack of well-being. In all these ways, educators said that LfM was about placing students, their learning and their well-being at the heart of and close to teachers’ practice. This student-centered view of LfM provoked deeper reflection among educators about their beliefs, relationships and strategies. One district leader asked: What are our values? What are our customs? What are our beliefs? We know that it starts with the beliefs and it results in a chain of events, a chain of thoughts, relationships that develop, connections that are made and actions that are planned and actions that are implemented. That becomes the work that we do collectively and that becomes the best that we achieve. The power of these beliefs and how they engaged educators with their students was expressed in a meeting of the senior administrative team in one district where participants represented their understanding of LfM by writing on birch-bark slices – an iconic symbol of Ontario. For a mathematics coach, what mattered most was “getting back to that reflective piece” of always “asking effective questions to push your thinking.” These slices had statements on them such as “growth: a ripple effect,” and “because it grows out from the middle.” This activity revealed how LfM was experienced by the senior administrative team not as a mechanistic or bureaucratic phenomenon. It was rather as an organic activity that “grows and spreads from an idea.” It involves “teacher-student voice” and trust to “let it grow, let it flourish.” ### A structure of interdisciplinary teams and committees One director of an urban school district regarded LfM as a theory of leadership that entailed placing people “in learning teams and supporting them through their inquiries.” “When I think of Leading from the Middle,” the Superintendent of Leadership and Learning in the same district said, “an organization has to put some structures in place and identify what the function of that structure is.” In her district, consultants and coaches worked with teachers to improve mathematics learning. In another district, a Mathematics Task Force distributed surveys to ask students what they liked about mathematics, what they found frustrating and difficult, and what kinds of supports they themselves would most like to receive. Even though these kinds of transformations involved a measure of reorganization and information gathering by the traditional educational bureaucracy, there was much more to all this than simply operating as an intermediary middle-tier that implements government policies from the top down. Recalling one of the seven original principles of LfM in which districts and their personnel take initiative rather than simply implementing other people’s initiatives, one superintendent observed: “I don’t see Leading from the Middle as the Ministry is at the top and boards (districts) are in the middle, and schools are at the bottom. I see that the Ministry lays out the game plan for everybody, but the action orientation rests with the boards (districts).” From this point of view, a diverse array of individuals at all different levels could lead from the middle. One director stated that these included “district staff,” “system leaders,” “school leaders,” and others as well. In a district with a large population of students with Fetal Alcohol Spectrum Disorder, mental health professionals provided by community-based non-profit agencies were considered to be part of their LfM team because their services were viewed as absolutely necessary. LfM was seen by Ontario’s educators as activating everyone in their schools and systems by rethinking and redesigning the structures of interdisciplinary diverse and locally responsive teams. One elementary school principal stated that LfM could be understood as the “expertise that comes in and helps our team problem solve, and helps to build our knowledge and capacity and mindset.” For a district director, LfM “means advocacy; it means people influencing the direction of the organization from the inside, in the grassroots, as opposed to top-down.” Here, LfM did not mean building up a bureaucracy with excessive guidelines and constraints, least of all one that was insulated from legitimate public concerns. Rather, it meant “flattening the organization, so it’s more that we’re all Leading from the Middle. We’re taking away the hierarchy and protocols and leading all together from the middle. We’re all sitting at a table here. You have a voice at the table.” With LfM, the structures of interdisciplinary teams were designed to evolve so they would become more responsive, inclusive, and participatory for all. ### A culture of collaborative professionalism Consortium boards (districts) experimented with LfM not just as a philosophy or a structure, but also as a method of group work. One superintendent stated: “Our Leading from the Middle methodology is about the circle as a resolution.” “The circle itself is very much an asset-oriented stance” to promote dialogue. “It’s about the fact that everybody came [to one meeting] with a little cheat sheet of a few things they wanted to say, but when it comes together, it actually becomes the conversation.” These conversations, she said, “speak to how the whole Leading from the Middle is a whole idea of trust. I’m trusting you to know what’s really important for your students, for your staff.” One elementary school principal underlined these perspectives when stating, conversely, that “there is a negative aspect that comes out of Leading from the Middle when people are feeling from above them that they’re being held back or there are constraints.” A Superintendent of Special Education echoed this view: “I’ve always believed that Leadership from the Middle is all inclusive”: Somebody raised a point [at a Consortium meeting] and said we need to start thinking about our teachers and how they are leaders from the middle as well and it just made me really think differently about the whole concept. It shifted my paradigm a little bit. I thought “She’s right!” We have to have that voice. That voice has to be heard! Teachers are an important voice in the Leading from the Middle concept. Eventually, LfM overlapped with a culture of collaborative professionalism (Hargreaves and O’Connor, 2017, 2018; Campbell, 2016; Fullan and Hargreaves, 2016; Ontario Ministry of Education, 2016) where teachers and other educators work together for all students’ success and well-being in ways that are assertive and show initiative. One district, for example, offered modest amounts of funding up to$1,000 for teachers to undertake innovations together. A district leader who designed the program said:

I’m trusting you to know what’s really important for your students. You have a great idea. Let’s see what we can do with it. Let’s see how we can make it grow. Really, it subsidizes a lot of the stuff that’s already happening but it’s amazing what can be done with that amount of money.

“They feel connected, they feel comfortable to try something that’s specific to their school culture that might be an initiative that might take off and have lift to it,” one superintendent said. “Sometimes,” a colleague added, “it starts off locally, and it just spreads from one school to another school, and part of the grant is that idea of sharing. Sharing so that we can all take it and rework it to fit our own communities.” In other words, knowledge is distributed rapidly and effectively across schools, enhancing teachers’ creativity and professionalism:

At the end of the innovation grants, we come in and we share them. We share them with superintendents, we share them with the Director so people have an idea. We share them with other administrators and various contacts throughout the district, so you’ll get people saying, “Oh you did that as innovation grant? That’s great. Can we meet? Can I set you up with some teachers?” The teachers who are coming forward are coming forward knowing that they might be asked to be a leader to lead this initiative going forward.

LfM, in this respect, promotes a culture of collaboration, sharing, initiative and responsiveness to the needs of the students, educators and communities in Ontario’s schools. It is founded on a philosophy that prioritizes democracy and community that is made possible by clear structural design decisions. As it evolves, it becomes a culture of collaborative professionalism in itself. This entails new ways of doing things together and acting independently. Ultimately, it drives learning and teaching forward with an ethic of care for all students.

Most efforts at educational change entail tensions and contradictions. LfM is no exception. Examples of these tensions and contradictions were evident among project coordinators at the ministerial level and also among some teachers. Two issues were prominent. On the one hand, there were concerns about the persistence, sustainability, coherence and measurable impact of LfM without top-down support in terms of funding, system priority and accountability. On the other hand, the continuation of a 20 year-old, standardized, large-scale assessment system (known as the EQAO) alongside more recent efforts to innovate pedagogically and support students’ well-being, pulled efforts at LfM in different and arguably contradictory directions.

One project coordinator in the Ministry appreciated the concept of LfM but expressed concern that the different boards were “inadvertently perhaps working at cross purposes at times, just by not understanding, or seeing their coherence.” Another colleague in the Ministry agreed. “The only worry I would have,” this colleague said, “would be if the stuff gets lost. Knowledge mobilization is the big question.”

As we have shown, in the earlier phase of LfM, there were deliberate and explicit strategies, accompanied by targeted funding to share and circulate (or mobilize) practices and improvements across the 72 districts. By the second phase of work within the Consortium, however, apart from the work of this project and the Consortium itself, coherence across the districts in LfM was assigned no specific priority by leaders in the Ministry at the top. As a result, this project leader observed, important opportunities for learning across the districts were lost.

Some project coordinators in the Ministry went further in articulating their concerns about lack of continuing coherence and argued for greater accountability. One project leader regretted that “we haven’t had specific measures” for LfM. Another stated that “if it’s not monitored, it’s just a ‘feel-good.” Moreover, he added, he didn’t want to be merely “at the provincial average for EQAO. Our kids deserve the best we can bring them every day,” but there needed to be better ways to demonstrate that kind of success.

Teachers’ concerns about LfM were directed less at the loss of learning opportunities across districts and more at the inconsistencies between the province’s strategies for innovation and for large-scale assessment, respectively. In districts where the directors encouraged teachers to be innovative, there was enthusiasm for LfM. It was viewed as a way for teachers to let students “have the lead” in designing inquiry-based projects that would help them “to be critical thinkers and good citizens.” While one teacher conceded that “I’ve really been struggling with that a little bit” because it entailed learning new ways of teaching, he also acknowledged that “the results have been remarkable in terms of the quality of work. It has shot through the roof!”

The difficulty for teachers resides in the contradiction between the possibilities of LfM for pursuing a philosophy of getting closer to the student through team-based structures and cultures of collaborative professionalism, on the one hand, and the province’s large-scale testing system on the other (Hargreaves, in press).

“I do have EQAO pending as a grade-three teacher,” one commented. She wondered whether the character education and critical thinking skills she was teaching “would come through when [students are] presented with a paper-pencil test for three days in a row. There’s a complete disconnect.” Another teacher stated: “I feel like EQAO is preparing students for a very antiquated version of education.” “I’d keep my own kids home [from taking the test] if I could,” she added. Another teacher who had been able to transfer out of teaching grade 6, in which the EQAO was administered, to grade 5, observed: “It’s a big difference. If we didn’t do math today, it doesn’t matter. We’ll catch up with it. The kids are engaged.”

Meanwhile, administrators in some districts endeavored to use LfM to improve lagging EQAO results in mathematics. While these administrators argued that testing “helped drive standards” and “helped with accountability,” teachers described the test as “hugely detrimental” to students in general and especially “unfair” for those with disabilities, different mother tongues, and backgrounds of poverty. When asked whether the test should be discontinued, one teacher said, “I don’t think I could think of a teacher that would say, ‘Well, no. We need to keep it. It’s so useful and great.”

In the second period of LfM, therefore, tensions and contradictions were evident in concerns related to lack of funding and continuing focus for districts to work together around innovation and well-being while long-standing pressures to raise the bar and narrow the gaps in achievement scores as measured by standardized tests remained. Ontario continued to wrestle with these tensions and contradictions as it tried to push through its innovation and well-being strategies via collaborative processes of LfM that brought leaders closer to learning and learners alike. In this respect, its strategy differed from other systems where top down reform and high stakes testing led many districts to abandon ambitions for innovation altogether (Adamson et al., 2016; Baird and Elliott, 2018; Koretz, 2017).

## Discussion: leading from the middle; not in the middle

The research reported here took place in two stages, although it was not planned as such in advance. The first stage of ESGA research surfaced an idea and strategy of LfM involving districts working together in a deliberately designed way to meet the needs of all their own and each other’s students more effectively. LfM was seen at this time as an alternative to top-down change and leadership. It enabled and empowered districts to drive change based on their own needs and expertise rather than simply implementing changes imposed from the top.

In the second stage, educators’ perceptions of LfM and its importance led to a review of other research, literature and policies in other systems regarding middle level or middle tier leadership. This review generated a process of theory building that was based upon new interview data elucidating what LfM meant to Ontario’s educators in the ten Consortium districts. This was then combined with theory testing of how middle level or middle tier leadership was being interpreted and presented in other research, literature and policy systems.

In the first phase of our research and of Ontario Liberal Government education policy until 2012, LfM existed alongside more conventional strategies of LiM. The province’s top-down, tri-level policy priority in this period was focused on driving up results and narrowing achievement gaps in literacy and numeracy on the EQAO, and on using the middle tier of school district leadership to implement this top-down priority with urgency and fidelity through a process of LiM.

This top-down strategy, combined with LiM approaches to implementation produced sustainably steady improvements in achievement on the EQAO. Even so, some aspects of bad leadership (not as extreme as cruel and callous ones) were also evident at the top and in the middle in the pressure to meet performance targets, in the form of teaching to the standardized achievement test, and concentrating undue attention on students scoring just below the point of measured proficiency (Hargreaves, in press).

Within and alongside this approach to top-down reform, we witnessed the parallel emergence of LfM strategies among special education administrators and superintendents, using top-level guidance and support to steer inclusive education reform as a stimulus. These LfM strategies were more explicitly focused on developing all students’ achievement and success, especially among students who were most vulnerable, by taking collective responsibility for measurable improvements in the students’ learning results.

From 2014, after Ontario’s new premier, Kathleen Wynne, issued Achieving Excellence with its emphasis on promoting broad excellence, well-being, equity understood as inclusion, and continued public confidence, school and system leaders started to view LfM as a way to help them move closer to all their students and support their learning and development. It evolved away from mainly being a mediator of top-down drives for measurable success in literacy and mathematics on EQAO, as it was during the first phase of our research. LfM was now expanded so that it was philosophically and ethically as well as strategically connected to addressing genuine concerns for all students, and to building authentic collaboration with other educators across the system through the Consortium to achieve this end.

In this new phase, LfM gave less evidence of being a cynical or manipulative approach to middle-level implementation. The system opened up and began providing opportunities for in-depth study of students and their learning. Its proponents showed less evidence of using the calculative methods to post achievement gains that had some presence in the earlier phase of policy and research.

In Ontario, LfM became much deeper than the kind of LiM or middle tier leadership that has been advanced in the wider literature. The essential differences between the two approaches are summarized in Figure 5.

In market systems, or ones focused on narrowly defined priorities such as literacy and mathematics, LiM has been about improving measured student achievement. It has ultimately been about ensuring fidelity with top-down central government priorities. Initially conceived as a method to improve the efficiency of market-driven systems under central government control, where local democracy has been weakened or eliminated, the systemic principles of LiM to improve performance and coherence have spread to and influenced many other countries. LiM in this sense is now common in England, the USA and some Australian states, for example.

LfM provides a clear point of contrast to LiM. As understood by Ontario’s educators, it is more explicit and direct about the purpose, power and agency of a new kind of leadership. LfM is not against stronger performance or improved coherence, although these things may result from its introduction. It differs in that it approaches performance and coherence in a way that is inclusive and empowering, and that has the capacity and commitment to engage with people through all parts of the school system. In this way, LfM, as practiced in Ontario, has been developing into a powerful way of strengthening local communities and local agency. It addresses the particular needs of schools and communities rather than only promoting the capacity of abstract systems to undertake self-improvement.

LfM, then, came to regard those in the middle not just as a mediating layer that connects the bottom to the top, but as encouraging and addressing the heart and soul of leadership at its core. Our interviews across the ten districts revealed educators’ persistent concern for and initiative taken on behalf of Ontario’s children, their learning, and their well-being. LfM was, in this sense, about much more than just incremental adjustments among levels to raise performance or about developing more coherent systems. It endeavored to support those educators who are closest to the work of teaching and learning in classrooms and schools so they can collaborate together purposefully, responsibly and transparently. It affirmed the value of educators having the local and also collective authority, expertise, and confidence to develop new strategies to educate the children they interact with every day and genuinely know better than a distant administration at the top.

## Implications of LFM in Ontario

Just a few weeks after the research team’s report was published in April 2018, Ontarians elected a populist government under Progressive Conservative Premier Doug Ford. The new government is fulfilling its campaign promises to reduce taxes and to balance the budget by eliminating 3,475 teaching positions in four years’ time along with many other forms of support, such as coaches and consultants, that the Ministry of Education had been able to provide to districts and schools over many years (Gray and Alphonso, 2019). An immediate government travel ban for public educators and district leaders, on the pretext of needing to implement austerity measures, directly inhibited the capacity for face-to-face collaboration that is essential for LfM and that was evident in the CODE Consortium. A 10–20 percent increase in secondary school class sizes will not only make classrooms more difficult for teachers to manage, but will also increase educators’ overall workload, prompting massive protests from students, parents, and teacher unions (Martin, 2019). Meanwhile, the framework for student well-being, championed by the former Premier, has been quietly removed from the Ministry website. LfM will have to struggle to endure in this unfamiliar but not unprecedented setting of a populist policy environment in Ontario.

This kind of “bad leadership” from the top that is driven more by agendas to reduce public expenditure than by improving the quality of teaching and learning, could be partially counteracted if districts could, in the spirit of LfM, find new ways to work collaboratively to share their initiatives together within this new policy environment – through online platforms, for example. Educators have at times proven to be remarkably resilient in the face of unfavorable policies, as we have documented in relation to change networks in California, Texas, and England that demonstrated measurable gains in student learning against the odds despite discouraging policy environments (Hargreaves and Shirley, 2009, 2012; Shirley 1997, 2002).

But it is not only a resurgent populism that has prompted the decline of LFM as a change strategy. LfM is not self-sustaining. It must be nurtured and supported continuously by those at the top. It requires a structure of support, resources and direction, as well as an embedded culture of shared habits and beliefs. With more than \$25m of allocated government funding, ESGA galvanized all 72 districts and their leaders to develop an inclusive strategy for students with special educational needs, that supported other students too, and that yielded definitive gains in equity in ways that drew attention from around the world. However, the data we have collected from these ten districts and from policy leaders reveal that the original collaboration among the 72 districts that was so evident at the time of ESGA has largely faded away. In Ontario, as elsewhere, the idea that cross-district collaborations can be funded as one-time projects and then sustained after the funding has been withdrawn is erroneous.

Within and beyond this populist phase, the challenge for Ontario is not to leave Ministry funding streams and assigned roles as they are, at the top, while then seeking additional temporary resources to fund limited cross-district initiatives below. Instead, restructuring must move elements of funding and responsibility from the center of the Ministry to the districts and to affiliated professional organizations with proven track records of success at initiating and spreading change. A system cannot lead from the middle haphazardly or temporarily. It must be resourced and organized to lead from the middle sustainably.

## A global movement of LFM?

If the immediate prospects for LfM in Ontario seem not very promising, the work of the Consortium of school districts in this decade, and the associated contribution of our research with the Consortium, has supported the beginnings of an international movement of educators and researchers who are producing their own definitions and signature practices affiliated with LfM.

In Scotland, following a country review by the OECD (2015) in which one of us participated, the core principles of LfM have been adopted as a national strategy to get the 32 local authorities (school districts) to work together in the form of six Regional Improvement Collaboratives (RICs) (Scottish Government, 2018). The RICs have been established to share resources and ideas and to take collective responsibility for improved equity and for reducing bad variation of quality across districts.

In California, where funding has been redirected to school district control in line with former Governor Brown’s advocacy for subsidiarity, LfM has been advanced as a way to break down cross-district isolation and competition in an inequitable system (Fullan and Rincón-Gallardo, 2017; Torlakson, 2015).

In the Canadian province of Nova Scotia, one of us is co-directing a three-year developmental evaluation of the province’s evolving inclusive education strategy, led by its Deputy Minister, Cathy Montreuil (a former Assistant Deputy Minister in Ontario when we were conducting the project reported here). Because Nova Scotia had previously decided to abolish its school districts and to separate principals from the teacher’s union (Glaze, 2018), a key part of our collaborative work with the province is to rebuild a culture and structure of collaborative professionalism and LfM across the system.

These are just three examples of how other systems are attending to, exploring and implementing their own versions of LfM and the core ideas underpinning it.

## Limitations

There are three primary limitations to this research. First, although this research has provided a rare opportunity to gather and analyze data on leadership and policy implementation in a high performing system over more than a decade, the research questions and respondents across the two phases of the research are not identical or strictly comparable in any controlled sense, although nine of the ten districts are identical across the two studies. Statements of progress, change and development regarding the ideas and interpretation of LfM are therefore made by inference rather than direct extrapolation.

Second, while the collaborative aspects of this research design enhance the validity and authenticity of the data given the relationships of high trust and co-creation through which the data and findings were produced, they also risk minimizing criticisms of the LfM strategy for which participants could feel at least partly responsible. Member checking enabled participants to read drafts of the research report and to rectify errors and to make suggestions. While this level of engagement has the advantage of clearing up misstatements and resolving contested interpretations, it also may have led the final interpretations to be less rather than more critical of LfM in Ontario.

Finally, the identification, analysis and interpretation of LfM in this account is confined to the one system of Ontario, even though other systems are now experimenting with versions of this approach. This means that inferences about the interpretation or implementation of LfM in other systems should only be made with caution in the absence of comparable research conducted in those systems. Future research should clarify how educators in these other systems understand LfM in their own contexts. Even when similar terminology is used, it could be that the meaning of the terms in action is distinctive to each system and setting. Future research is necessary to clarify these matters of comparative definition and practical enactment as LfM gains momentum as an international movement.

## Conclusion

In democratic and inclusive educational policy environments of increasing urgency and complexity, such as the one that characterized Ontario over the course of our research, LfM offers a promising option to achieve greater quality and equity by having leaders collaborate closely with each other to improve teaching, learning and well-being. LfM, in this respect, provides evidence of an integrated philosophy, structure, and culture of change that has advanced aspects of shared, democratic, or distributed leadership that have been advocated for many years but that have been difficult to bring about in practice on a systemic scale. LfM represents a significant advance over LiM. The first is a strategy that engages and empowers a strong profession to serve and support learning and well-being for all students within and across local and diverse communities. The second is an abstract and systemic method to increase efficiency and coherence in any kind of context, including ones that have abandoned or avoided community democracy, and that are less committed to developing strong teaching professions capable of exercising collective professional judgment.

## Figures

### Figure 1

Seven principles of leading from the middle

### Figure 2

Powerpoint slide enabling transparency of activities on the subject of identity across the consortium

### Figure 3

Sample slide exhibiting transparency of one district’s activities

### Figure 4

A timeline of ESGA and LFM projects in Ontario

## References

Adamson, F., Astrand, B. and Darling-Hammond, L. (Eds) (2016), Global Education Reform: How Privatization and Public Investment Influence Education Outcomes, Routledge, New York, NY.

Ainscow, M. (2015), Towards Self-Improving School Systems: Lessons from a City Challenge, Routledge, London.

Anderson, S.E. (2010), “Moving change: evolutionary perspectives on educational change”, in Hargreaves, A., Fullan, M., Lieberman, A. and Datnow, A. (Eds), The Second International Handbook of Educational Change, Springer, Dordrecht, pp. 65-84.

Baird, J. and Elliott, V. (Eds) (2018), “Special issue: the consequences of metrics in education: gaming, malpractice, and cheating”, Oxford Review of Education, Vol. 44 No. 5, p. 2018.

Barber, M. and Day, S. (2014), The New Opportunity to Lead: A Vision for Education in Massachusetts in the Next 20 Years, Massachusetts Business Alliance for Education, Boston, MA.

Bauman, Z. (2016), Strangers at Our Door, Polity Press, Oxford.

Bentley, T. (2010), “Innovation and diffusion as a theory of change”, in Hargreaves, A., Lieberman, A., Fullan, M. and Hopkins, D. (Eds), The Second International Handbook of Educational Change, Springer, Dordrecht, pp. 29-46.

Blase, J. and Blase, J. (2002), Breaking the Silence: Overcoming the Problem of Principal Mistreatment of Teachers, Corwin Press, Thousand Oaks, CA.

Bryk, A. and Schneider, B. (2002), Trust in Schools: A Key Resource for Improvement, Russell Sage, New York, NY.

Campbell, C. (2016), “Collaborative professionalism: of, by and for Catholic school leaders”, Principal Connections, Vol. 20 No. 1, pp. 6-7.

Campbell, C., Zeichner, K., Lieberman, A. and Osmond-Johnson, P. (2017), Empowered Educators in Canada: How High-Performing Systems Shape Teaching Quality, John Wiley & Sons, Marblehead, MA.

Chen, J., Moran, S. and Gardner, H. (Eds) (2009), Multiple Intelligences Around the World, Wiley, San Francisco, CA.

Chung, B.P. (2011), “Changing school from inside out: a qualitative case study of a self-initiated change over time in South Korea”, unpublished dissertation, Lynch School of Education, Boston College, Chestnut Hill, MA.

Cohen, D. (1990), “A revolution in one classroom: the case of Mrs Oublier”, Educational Evaluation and Policy Analysis, Vol. 12 No. 1, pp. 311-329.

Daly, A.J., Der-Martirosian, C., Ong-Dean, C., Park, V. and Wishard-Guerra, A. (2011), “Leading under sanction: principals’ perceptions of threat rigidity, efficacy, and leadership in underperforming schools”, Leadership and Policy in Schools, Vol. 10 No. 2, pp. 171-206.

Duke, D.L. (2016), The Children Left Behind: America’s Struggle to Improve its Lowest Performing Schools, Rowan & Littlefield, Lanham, MD.

Elmore, R. (1996), “Getting to scale with good educational practice”, Harvard Educational Review, Vol. 66 No. 1, pp. 1-27.

Elmore, R. (2016), “‘Getting to scale…’ it seemed like a good idea at the time”, Journal of Educational Change, Vol. 17 No. 4, pp. 529-537.

Evans, M. and Stone-Johnson, C. (2011), “Internal leadership challenges of network participation”, International Journal of Leadership in Education, Vol. 13 No. 2, pp. 203-220.

Fink, D. (1999), “The attrition of change: a study of change and continuity”, School Effectiveness and School Improvement, Vol. 10 No. 3, pp. 269-295.

Fink, D. and Brayman, C. (2006), “School leadership succession and the challenges of change”, Educational Administration Quarterly, Vol. 42 No. 1, pp. 62-89.

Fullan, M. (2015), “Leadership from the Middle”, Education Canada, Vol. 55 No. 4, pp. 22-26.

Fullan, M. and Hargreaves, A. (2016), Bringing the Profession Back in: Call to Action, Learning Forward, Oxford, OH, available at: https://learningforward.org/docs/default-source/pdf/bringing-the-profession-back-in.pdf

Fullan, M. and Quinn, J. (2015), Coherence: The Right Drivers in Action for Schools, Districts, and Systems, Corwin, Thousand Oaks, CA.

Fullan, M. and Rincon-Gallardo (2016), “Leadership from the middle”, Education Canada, Vol. 55 No. 4, pp. 22-26.

Fullan, M. and Rincón-Gallardo, S. (2016), “Developing high quality public education in Canada: the case of Ontario”, in Adamson, F., Astrand, B. and Darling-Hammond, L. (Eds), Global Education Reform: How Privatization and Public Investment Influence Education Outcomes, Routledge, New York, NY.

Fullan, M. and Rincón-Gallardo, S. (2017), “California’s golden opportunity: taking stock: leadership from the middle”, available at: https://michaelfullan.ca/wp-content/uploads/2017/09/17_Californias-Golden-Opportunity-Taking-Stock-FinalAug31.pdf

Fullan, M., Hill, P. and Crévola, C. (2006), Breakthrough, Corwin, Thousand Oaks, CA.

Glaser, B.G. and Strauss, A.L. (1967), The Discovery of Grounded Theory: Strategies for Qualitative Theory, Aldine Transaction, New Brunswick, NJ.

Glaze, A. (2018), “Raise the bar: a coherent and responsive education administrative system for Nova Scotia”, available at: www.ednet.ns.ca/adminreview

Goodwin, D.K. (2016), Team of Rivals: The Political Genius of Abraham Lincoln, Simon and Schuster, New York, NY.

Gray, J. and Alphonso, C. (2019), “Ford government to cut over 3,400 teaching positions over four years”, The Globe and Mail, available at: www.theglobeandmail.com/canada/article-ford-government-warns-public-sector-it-is-considering-wage-hike-caps/

Gross, N., Giacquinta, J.B. and Bernstein, M. (1971), Implementing Organizational Innovations: A Sociological Analysis of Planned Educational Change, Basic Books, New York, NY.

Hargreaves, A. (in press), “Large-scale assessments and their side-effects: the case of mid-stakes tests in Ontario”, Journal of Educational Change.

Hargreaves, A. and Braun, H. (2012), “Leading for all: a research report of the development, design, implementation and impact of Ontario’s ‘essential for some, good for all’ initiative”, Council of Ontario Directors of Education, Toronto.

Hargreaves, A. and Fullan, M. (2012), Professional Capital: Transforming Teaching in Every School, Teachers College Press, New York, NY.

Hargreaves, A. and O’Connor, M. (2017), “Collaborative professionalism”, Monograph prepared for the World Innovation Summit for Education, Qatar Foundation, Doha, Qatar, available at: www.wise-qatar.org/2017-wise-research-collaborative-professionalism

Hargreaves, A. and Shirley, D. (2009), The Fourth Way: The Inspiring Future for Educational Change, Corwin, Thousand Oaks, CA.

Hargreaves, A. and Shirley, D. (2012), The Global Fourth Way: The Quest for Educational Excellence, Corwin, Thousand Oaks, CA.

Hargreaves, A. and O’Connor, M. (2018), Collaborative Professionalism: When Teaching Together Means Learning for All, Corwin, Thousand Oaks, CA.

Hargreaves, A., Shirley, D., Wangia, S., Bacon, C. and D’Angelo, M. (2018), “Leading from the middle: spreading learning, well-being, and identity across Ontario”, Toronto, Ontario, CODE Consortium.

Harris, A. (2003), “Teacher leadership as distributed leadership: heresy, fantasy or possibility?”, School Leadership & Management, Vol. 23 No. 3, pp. 313-324.

Harris, A. and Jones, M. (2017), “Middle leaders matter: reflections, recognition, and renaissance”, School Leadership and Management, Vol. 37 No. 1, pp. 213-216.

Hill, A., Mellon, L., Goddard, J. and Laker, B. (2016), “How to turn around a failing school”, Harvard Business Review, August, available at: https://hbr.org/product/how-to-turn-around-a-failing-school/H02Z5P-PDF-ENG (accessed May 31, 2017).

Horn, M.B. and Staker, H. (2015), Blended: Using Disruptive Innovation to Improve Schools, Jossey-Bass, San Francisco, CA.

Johnson, S.M., Marietta, G., Higgins, M.C., Mapp, K.L. and Grossman, A. (2015), Achieving Coherence in District Improvement: Managing the Relationship Between the Central Office and Schools, Harvard Education Press, Cambridge, MA.

Kellerman, B. (2004), Bad Leadership: What it is, How it Happens, Why it Matters, Harvard Business School Press, Cambridge, MA.

Kets de Vries, M. (2006), The Leader on the Couch: A Clinical Approach to Changing People and Organizations, Wiley, San Francisco, CA.

Koretz, D. (2017), The Testing Charade: Pretending to Make Schools Better, University of Chicago Press, Chicago, IL.

Ladd, H.F. and Fiske, E.B. (2016), “England confronts the limits of school autonomy”, Working Paper No. 232, National Center for the Study of Privatization in Education (NCSPE) Teachers College, Columbia University, New York, NY.

Leithwood, K. (2013), “Strong districts and their leadership: a paper commissioned by The Council for Ontario Directors of Education and the Institute for Educational Leadership”, available at: www.ontariodirectors.ca/downloads/Strong%20Districts-2.pdf

Leithwood, K. (2018), Leadership Development on a Large Scale: Lessons for Long-Term Success, Corwin, Thousand Oaks, CA.

Lieberman, A., Campbell, C. and Yashkina, A. (2017), Teacher Learning and Leadership: of, by and for Teachers, Routledge/Taylor & Francis, London.

Loucks, S.F. and Hall, G.E. (1977), “Assessing and facilitating the implementation of innovations: a new approach”, Educational Technology, Vol. 17 No. 2, pp. 18-21.

Martin, T. (2019), “Throngs of people protest at Ontario legislature against Ford government’s changes to education system”, The Globe and Mail, available at: www.theglobeandmail.com/canada/article-unions-to-rally-at-queens-park-to-protest-ford-governments-education/

Mehta, J. and Fine, S. (2019), In Search of Deep Learning: The Quest to Remake the American High School, Harvard University Press, Cambridge, MA.

Mehta, J.D. and Schwartz, R.B. (2011), “Canada: looks a lot like us but gets much better results”, in Tucker, M.S. (Ed.), Surpassing Shanghai: An Agenda for American Education Built on the World’s Leading Systems, Harvard Education Press, Cambridge, MA.

Mintrop, H. (2004), Schools on Probation: How Accountability Works (And Doesn’t Work), Teachers College Press, New York, NY.

OECD (2011), Strong Performers and Successful Reformers in Education: Lessons from PISA for the United States, OECD, Paris.

OECD (2015), “Improving schools in Scotland: an OECD perspective”, available at: www.oecd.org/education/school/Improving-Schools-in-Scotland-An-OECD-Perspective.pdf

Ontario Ministry of Education (2014), “Achieving excellence: a renewed vision for education in Ontario”, available at: www.edu.gov.on.ca/eng/about/great.html

Ontario Ministry of Education (2016), “Collaborative professionalism”, Policy/Program Memorandum No. 159, Toronto, Ontario, available at: www.edu.gov.on.ca/extra/eng/ppm/ppm159.pdf (accessed May 31, 2017).

Payne, C. (2008), So Much Reform, So Little Change: The Persistence of Failure in Urban Schools, Harvard Education Press, Cambridge, MA.

Povey, D. and McInerny, L. (2019), The Leadership Factor: The Seven Characteristics of Exceptional Leaders, John Catt Educational, Woodbridge.

Salokangas, M. and Ainscow, M. (2017), Inside the Autonomous School: Making Sense of a Global Educational Trend, Routledge, New York, NY.

Sarason, S. (1990), The Predictable Failure of Educational Reform, Jossey-Bass, San Francisco, CA.

Schleicher, A. (2015), Schools for 21st-Century Learners: Strong Leaders, Confident Teachers, Innovative Approaches, OECD, Paris.

Scottish Government (2018), 2018 National Improvement Framework and Improvement Plan for Scottish Education: Achieving Excellence and Equity, Scottish Government, Edinburgh.

Shirley, D. (1997), Community Organizing for Urban School Reform, University of Texas Press, Austin, TX.

Shirley, D. (2002), Valley Interfaith and School Reform: Organizing for Power in South Texas, University of Texas Press, Austin, TX.

Shirley, D. (2017), The New Imperatives of Educational Change: Achievement with Integrity, Routledge, New York, NY.

Shirley, D. and MacDonald, E.A. (2016), Teachers College Press, New York, NY.

Spillane, J. (2006), Distributed Leadership, Jossey-Bass, San Francisco, CA.

Supovitz, J. (2014), Building a Lattice for School Leadership: The Top-to-Bottom Rethinking of The Mindful Teacher, Leadership Development in England and What it Might Mean for American Education, Consortium for Policy Research in Education, Philadelphia, PA.

Torlakson, T. (2015), A Blueprint for Great Schools: Version 2.0, California Department of Education, Sacramento, CA.

Tucker, M. (2019), Leading High-Performance School Systems: Learning from the World’s Best, ASCD, Alexandria, VA.

Wylie, C. (1997), Self-Managing Schools Seven Years On: What Have We Learnt?, New Zealand Council for Educational Research, Wellington.

California Department of Education (2016), “Our journey together on the California way”, available at: www.cde.ca.gov/eo/in/bp/bp2intro.aspevery

Crossley, D. (Ed.) (2013), Sustainable School Transformation, Bloomsbury Press, London.

Fullan, M. (2019), Nuance: Why Some Leaders Succeed and Others Fail, Corwin, Thousand Oaks, CA.

Hargreaves, A. and Ainscow, M. (2015), “The top and bottom of leadership and change”, Phi Delta Kappan, Vol. 97 No. 3, pp. 42-48.

Hargreaves, A., Boyle, A. and Harris, A. (2014), Uplifting Leadership: How Organizations, Teams, and Communities Raise Performance, John Wiley & Sons, Marblehead, MA.

Harris, A. and Jones, M. (2019), “Teacher leadership and educational change”, School Leadership and Management, Vol. 39 No. 2, pp. 123-126.

Munby, S. (2019), Imperfect Leadership, Crown House Publishing, London.

New Zealand Ministry of Education (2012), Leading from the Middle: Educational Leadership for Middle and Senior Leaders, Learning Media, Wellington.

Spillane, J.P. (2012), Distributed Leadership, John Wiley & Sons, Wilmington, DE.

## Corresponding author

Andy Hargreaves can be contacted at: hargrean@bc.edu