Spanners and enhancers: The role of teacher leaders in school evolution

Abstract

This study explores how the New York City Department of Education attempted to impact instruction and school quality through Teacher Career Pathways, a teacher leadership program. We apply Complexity Theory for Teacher Leadership (CTTL) to examine how teacher leaders may contribute to the improvement of stakeholder experiences and school quality. Using mixed methods, we compare schools with formal teacher leaders appointed into roles with targeted responsibilities and positional authority to lead to schools without formal teacher leaders. Diverging from prior literature, we find that schools with formal teacher leaders show improvement in school quality that matches targeted program areas not mirrored in schools without teacher leaders, suggesting formality may provide an important role in school change. In line with CTTL, we suggest that the effectiveness of formal teacher leaders may be attributed to their functioning as boundary spanners and network enhancers, reinforcing and amplifying changes over time.

Keywords

Teacher leadership complexity theory school quality mixed methods school improvement middle leaders teacher leaders

Introduction

Decades of education research have found that schools are remarkably recalcitrant as institutions, with the core of instruction and the dynamics between teachers, students, and school and system administrator remaining largely unchanged (Bryk et al., 1998). To move away from “tinkering” around the edges of the system and reach deeper into classrooms, teacher-driven approaches to organizing school improvement have been promoted. One approach is teacher leadership, which has since the 1980s experienced waves of interest and evolution among school leaders, policymakers, advocacy organizations, researchers, and private foundations.

In response to calls from the field for more large-scale, empirical research on teacher leadership (e.g., Harris et al., 2019; Nguyen et al., 2020), this paper is the first of three devoted to examining the impact of teacher leadership in New York City schools. This study explores how the New York City Department of Education (NYCDOE) attempted to impact instruction and school quality and institute and deepens distributed leadership practices through Teacher Career Pathways (TCP), a teacher leadership program. Here, we apply mixed methods to better understand how teacher leaders may contribute to the improvement of stakeholder experiences and school quality, as measured by survey data and the School Quality Review (SQR).

Sensitive to limited empirical work that explicitly focuses on theoretical dimensions of teacher leadership (Nguyen et al. 2020), we engage and expand the Complexity Theory for Teacher Leadership (CTTL) framework to build a more robust conceptual knowledge of teacher leadership. We theorize that strategic distribution of leadership to teacher leaders enables them to become boundary spanners, working across hierarchies and structures within schools. Through boundary spanning, teacher leaders become network enhancers—strengthening the connections between school leaders and teachers that are required for communication, decision-making, and individual and organizational evolution. By conceptualizing teacher leaders as boundary spanners and network enhancers, we can consider change at the individual and school levels as early evidence of potential communication feedback loops that reinforce and amplify changes over time, which could move schools from recalcitrance to resilience.

Teacher leadership and TCP

Teacher leadership refers to “the use of teachers’ expertise about teaching and learning to improve the culture and instruction in schools” such that teacher and student learning is enhanced (York-Barr and Duke, 2004: 261). While there still exists a lack of consensus around defining teacher leadership (Pan et al., 2023), this research focuses on formal roles that blend teaching with instructional leadership and are “titled, ranked, differently structured, differently authorized, and differently paid” (Smith, 2019: 1). This contrasts with informal roles, where teachers may be encouraged to lead initiatives, professional learning, departments, or grade levels in their schools, but without any formally defined responsibilities, recognition, or compensation. Practically speaking, the formal roles are in alignment with the types of roles in TCP's program design. Theoretically speaking, it aligns with some literature suggesting that for teacher leaders to effectuate change in their school communities, it is necessary for them to be recognized within the hierarchy of schools (Smith, 2019).

Teacher leaders refer to “hybrid teacher leaders,” defined by Margolis (2012) as “a teacher whose official schedule includes both teaching K-12 students and leading teachers in some capacity” (292). Margolis's definition and its emphasis on hybridity is most suitable here given our examination of teacher leader roles where the leadership work is situated within the day-to-day work of teachers in school settings.

Noteworthy for international readers may be the similarity between formal teacher leaders and middle leaders. Middle leaders have been identified as having a “hermaphroditic role” (Siskin, 1993), operating as a teacher and a leader. They are deemed “key brokers within organizations” (Earley and Bubb, 2004) to support teachers’ professional learning and influence change. Middle leaders both lead and teach (Grootenboer et al., 2023), typically in formal and remunerated roles, embodying the same hybridity as the TCP case. Additionally, middle leaders, like teacher leaders, influence school quality and effectiveness by focusing on managing relationships, communicating effectively, and leading teams (De Nobile, 2017).

While the concepts of teacher leaders and middle leaders are not necessarily the same (Bennett et al., 2007; Lipscombe et al., 2023), for this paper, where teacher leaders focus on pedagogical expertise and the development of instructional knowledge and practice among peers (Taylor et al., 2011), the work of middle leaders is highly relevant. We consider middle leaders and teacher leaders as interchangeable, though we primarily use the term “teacher leaders” to match the NYCDOE language. We exclude teacher leaders who primarily serve in supervisory or administrative roles that emphasize managing or evaluating teachers, given the explicit emphasis of TCP on cultivating nonevaluative teacher leadership. Note that our focus on teacher leaders is not meant to overstate their power within systems but to serve as a starting point for identifying structural levels within the core educational system where the interaction of individuals between and within levels may influence outcomes (Lemke & Sabelli, 2008).

Examining teacher leadership as embodied in TCP allows us to make a unique contribution to the teacher leadership literature in three regards. First, few studies examine formal teacher leadership where teacher leaders stay in the classroom, in addition to taking on leadership duties across the school. Most studies examine informal teacher leadership, wherein teacher leaders organically assume leadership responsibilities (e.g., Meirink et al., 2019; Lee and Ip, 2023). Second, where studies take up formal teacher leadership, often the teacher leaders have left the classroom and are functioning in solely a leadership capacity (e.g., Supovitz & Comstock, 2023; Yost et al., 2009). On these two counts, our analysis of TCP is novel in that we apply mixed methods to examine a formal teacher leadership program where teacher leaders remain in the classroom while serving as leaders. Finally, though our data and methodology lend themselves to a traditional program evaluation, our application of CTTL extends our work. We highlight how the data collected routinely by administrative agencies like school districts might be extended to scholarly research.

The CTTL framework

Complexity theory suggests that a system is a collection of interconnected and interacting parts that are defined by shared boundaries and properties, and together function as a whole (Morrison, 2002). These interactions are intensely integrated with each other and involve myriad individuals and variables so that the outcomes of the system cannot be predicted by a linear equation; the behavior of the system is not explained by individual components but as a function of their interaction (Morrison, 2002; Snyder, 2013). These relationships between individuals and components of the system mean that feedback between them shapes the evolution of the system, resulting in iterative and emergent change that expresses itself on multiple time scales and levels simultaneously (Snyder, 2013; Mital et al., 2014). At the same time, research on teacher leaders indicates that context influences teachers’ development as leaders and a failure to acknowledge and address the complexity in teachers’ contexts hinders their development as they work to develop as leaders (King and Holland, 2022). Merging this knowledge of teacher context and complexity theory, CTTL suggests that teacher leaders play an essential role in connecting the various components and individuals in complex systems and initiating the feedback loops that instigate and sustain system-wide change. For our purposes, we suggest that it is through (1) boundary spanning and (2) network enhancement that teacher leaders send information across structural levels and individuals and target it where it needs to go most.

Schools embody a web of nested structures, meaning that within the complex systems of the school, there are other complex systems which give rise to new patterns of activities and rules of behavior throughout the whole (Davis and Sumara, 2008; Resnick and Scherrer, 2012). Figure 1 illustrates a school as multiple nested structures nested within the broader community. Porous divisions between complex systems and nested structures within them mean that policies, procedures, and processes at any one level can promote or inhibit actions on one or multiple levels. This means that the exchanges between individuals in each sphere of the system may inform exchanges across all of them. The bidirectional exchange between members within and individuals across levels enables teacher leaders to act as “boundary spanners”—individuals who can operate at the edge of an organization or a level within it and act as exchange agents between their location and other parts of the system (Keszey, 2018; Buchanan et al., 2023; Snoek et al., 2017).

Stemming from teacher leaders’ work as boundary spanners is their role as “network enhancers”—individuals who are critical to creating, maintaining, and strengthening communication channels between individuals and groups. These network enhancers foster “clusters of communication relationships” that combine to form organizational networks and, in doing so, support vital organizational work, including channeling the flow of information; connecting people with common interests; establishing common interpretations of information; enhancing the social influence of individuals and groups; and enabling an exchange of resources (Munsayac, n.d.: 3). This can be likened to networks in the physical sciences. Most relevant here is the idea of complex root networks in trees. McElrone et al. (2013) observe two sorts of networks: one within trees, where complex networks of roots absorb water that eventually travels up the tree through long tubes, and one between trees, where special fungi connect trees, establishing a symbiotic relationship that increases the absorptive surface area of the root system across the “networked” trees. In the case of teacher leadership, we can think of root networks as teacher leaders operating as the roots, sending nutrients (e.g., quick and accessible information on teaching and learning) across the school via their connections to teachers, school leaders, and students.

Figure 2 illustrates how the work of teacher leaders positions them to create and enhance networks through their connections to a vast array of individuals throughout the school and operating at different levels of the hierarchy. In this way, teacher leaders have the potential to make educational systems more efficient by connecting ideas, practices, and resources through frequent interactions across complex systems that create the conditions for high-functioning work environments (Zadok et al., 2024).

Applying this to TCP, we would expect to see that within each level and across them, systems are living, learning, and attending to their members (Resnick and Scherrer, 2012; Sands, Margolis and Murphy, 2022). Specifically, we expect to see growth at the edges—the boundaries of the system—where teacher leaders were supposed to target their efforts. Where teacher leaders were intentionally positioned to share ideas, information, and practices across the school hierarchy, we expect to see positive changes for individuals and the school community (Keszey, 2018; Davis and Sumara, 2008). The examination of program documents and survey and SQR results should provide a better understanding of how stakeholders and school communities might be evolving in the ways complexity theory suggests due to teacher leadership.

Data and methods

In this paper, we investigate three questions:

What are the program intentions and scope of work for teacher leaders?

How do stakeholders in schools with teacher leaders feel about organizational processes and impact over time?

Does school quality improve in schools participating in TCP in ways that differ from schools that do not participate?

Our goal is to understand the conditions that might facilitate boundary spanning and network enhancement and subsequently may explain our results. Data for this study come from three sources: TCP program data, TCP annual surveys, and the SQR.

Teacher Career Pathways program data and method

The primary data source used to establish the sample for the analyses was TCP school program participation data, which included yearly data on schools with qualified and staffed teacher leaders. These data were maintained by the NYCDOE's Office of Teacher Leadership, which managed TCP. It was used to establish the three main test groups for analysis: schools that had staffed teacher leaders; schools that had qualified teacher leaders but never staffed them; and schools that did not have any eligible teacher leaders. A school with qualified teacher leaders was a school where at least one teacher passed the district's application process to qualify as one of three kinds of teacher leaders and was thus a “qualified teacher leader.” A school with staffed, or appointed, teacher leaders was a school where the principal chose to appoint at least one qualified teacher leader into one of those three formal teacher leadership roles. Principals appointing a teacher leader were responsible for paying the salary addition to compensate them and agreed to cover the release time for teacher leaders to perform their duties. The district outlined a broad scope of responsibilities and activities for each kind of role, commensurate with its level and pay. How the role was implemented and what specific activities teacher leaders engaged in, either individually or in teams, was decided by the school.

The second set of program data we used consisted of documents and reports detailing TCP, including its mission, goals, strategy, operations, and activities. We consulted a variety of publicly available documents, including federal grant proposals, reports and program evaluations, program newsletters, role descriptions, and web pages. We applied the principles of qualitative document analysis to examining these documents, reviewing the content to ascertain meaning, deepen our understanding of the program and its components, and develop empirical knowledge (Bowen, 2009). In particular, we focused on “node identification”—identifying the roles and responsibilities of teacher leaders, how they were supposed to carry out their responsibilities, and who they were supposed to engage in their work. The purpose of this review was to provide direction in applying our framework and analyzing the results from our quantitative data analysis. A limitation of the quantitative data used in this study is the inability to explicitly link the experiences of individuals in schools to changes observed schoolwide. However, an examination of program documents provides us with insight into how TCP was conceptualized specifically to amplify the results of individual-level work to the school-level work.

Teacher Career Pathways survey data and method

Survey data came from the annual TCP program evaluation administered to principals staffing at least one teacher leader; appointed teacher leaders; and teachers in schools with appointed teacher leaders. Participation in the survey was optional and confidential. Table 1 shows the response rates for 2016 to 2020,¹ the years analyzed for this study. While the teacher response rates are notably low, the sheer number of respondents given the size of the district suggests that they may provide a useful guidepost.

Table 1.

Spring survey responses rates for all stakeholder groups.

	Principals		Teacher leaders		Teachers
Year	# Respondents	% Response	# Respondents	% Response	# Respondents	% Response
2016	176	44%	616	58%	2986	Unknown*
2017	245	49%	1269	59%	4839	24%
2018	345	61%	1264	65%	4932	21%
2019	376	57%	1503	62%	8289	30%
2020	425	61%	1522	80%	8080	29%

*Note: For 2016, TCP staff were missing records that prevented identification of the number of teachers who received the survey.

To conduct our analyses, we converted Likert-scale items into numerical values. For questions enquiring about agreement, a 6-point attitudinal scale was used. “Not applicable” responses, which appeared as a response option starting in 2019, were dropped from the analysis.

Questions consistently asked all five years and aligned to our theoretical framework were categorized into constructs. The teacher and teacher leader surveys had four survey constructs: distributed leadership, school culture, support and collaboration, and impact. The principal survey had three constructs: distributed leadership, impact, and strategy.² We tested the reliability of the questions in these constructs within years and across all years of the analytic window using Cronbach's alpha.³ After determining reliable constructs, we created an index for each one that involved establishing 2016 as a threshold year and standardizing every variable to the threshold, averaging the standardized variables together, and then standardizing that average again. We analyzed the survey data using descriptive and inferential statistics. Independent sample t-tests were used to compare 2016 and 2020 scores to each other. Satterthwaite's approximation for unequal variances was used to accommodate for the unequal sample sizes and unequal variances.⁴

School quality review data and method

To assess school quality, we used publicly available SQR results. The SQR was a process run by the NYCDOE Office of School Quality that examined the organization of schools to support teacher and student learning (NYCDOE, 2022). Schools participated in an SQR every three years. A rubric was used to rate schools on 10 indicators that fell under three categories: Instructional Core, School Culture, and Systems for Improvement. We examined five indicators for which we had data for all years: Curriculum, Pedagogy, Assessment, Higher Expectations, and Teacher Teams & Leadership Development. Ratings for each indicator fell into four categories: Underdeveloped; Developing; Proficient; and Well Developed.

Following preliminary testing, we compared schools with qualified teacher leaders only to schools with no teacher leaders, and schools appointing teacher leaders at least once compared to schools with no teacher leaders.⁵ To determine a pre- and post-TCP implementation score for treatment schools, all schools with qualified or staffed teacher leaders were assigned to a cohort based on the first year they had qualified or staffed teacher leaders (e.g., schools with qualified teacher leaders for the first time in 2014 were assigned to cohort 1, schools with staffed teacher leaders for the first time in 2015 were assigned to cohort 2, etc.). The cohort year was then used to determine the baseline score for an indicator—the score a school earned for an indicator prior to having teacher leaders—and the end score—the most recent score a school received for an indicator that was during or after the first year of having teacher leaders. For comparison schools, schools were assigned to a cohort based on the earliest available pair of scores. All schools without a pair were dropped from the analyses.

We applied the method used in the survey results to standardize scores for each indicator. We conducted two paired sample t-tests for three groups of schools: qualified but never staffed teacher leaders, staffed teacher leaders, and no teacher leaders to determine if there was a statistically significant difference in means between the baseline and end scores.

Sample

Table 2 presents select characteristics, including descriptive statistics, on schools in the three subgroups: schools that only ever had qualified teacher leaders, schools that had staffed teacher leaders for at least one year, and schools that never had any teacher leaders. I include data for all years from 2014 to 2021. With sizable samples, we observe small differences across many school characteristics between schools that had qualified teacher leaders but never appointed them, schools that staffed teacher leaders, and schools that had no teacher leaders ever. Notably, schools staffing teacher leaders had a higher percentage of middle schools and high schools than the other groups. In addition, the student population served is markedly different. Schools staffing teacher leaders had a higher percentage of males, Black and Hispanic students, students with disabilities, English language learners, and students from lower socioeconomic status. Regarding student achievement, the standardized mean scale scores for all grades in ELA and math were lower in schools staffing teacher leaders compared to other subgroups. The student population and test scores suggest that schools opting-in to participate in TCP were serving subgroups of students that had a greater need for services and additional support to improve educational outcomes. The high standard deviations in some cases, particularly with respect to test scores and funding, indicate that significant variability is driven from differences between schools.

Table 2.

Select characteristics of NYCDOE schools by group, 2014–2021.

	Qualified teacher leaders only (n = 1463)		Staffed teacher leaders ever (n = 7892)		No teacher leaders ever (n = 3709)
Characteristic	Mean	Standard deviation	Mean	Standard deviation	Mean	Standard deviation
Total enrollment	735.308	670.579	583.486	514.725	585.207	360.252
Per pupil funding	$8585.93	$23,336.76	$7703.11	$4822.39	$7164.72	$6572.16
Female	0.490	0.082	0.478	0.108	0.482	0.059
Black students	0.262	0.257	0.326	0.252	0.248	0.274
Hispanic students	0.362	0.229	0.466	0.250	0.371	0.249
Students with disabilities	0.217	0.170	0.241	0.163	0.220	0.163
English language learners	0.125	0.159	0.149	0.151	0.139	0.131
Free/ reduced price lunch or eligible for Human Resources Administration benefits	0.702	0.248	0.817	0.171	0.737	0.241
ELA, all grades (3–8)	0.387	1.046	−0.261	0.895	0.323	1.020
Math, all grades (3–8)	0.337	1.030	−0.268	0.910	0.352	0.994

Note: All variables except for school level have missing data. Per pupil funding is calculated using the Fair Student Funding amount from available School Allocation Memorandum divided by total enrollment. All test scores are standardized by subject. No test scores are available from SY 2019–20 and SY 2020–21.

Findings on TCP program intentions

Analysis of program documents and reports indicate that TCP placed an intentional focus on multiple boundary-spanning roles for teacher leaders. This occurred at three main nodes: (1) As a link between school administrators and teachers, as well as between teachers themselves, with the intention to improve pedagogy, student learning, and instructional leadership; (2) As a liaison between school administrators and TCP program managers, guiding principles on how to use teacher leaders and program assets locally; and (3) As a facilitator between management (district officials) and employees (the teachers union), to improve communication and broaden buy-in regarding the program as a whole. Further analysis of program policy documents indicates that teacher leaders were positioned as network enhancers in these boundary-spanning roles.

Linking school administrators and teachers and teachers with each other

In the initial 2012 application to receive Teacher Incentive Fund grant monies, the TCP theory of action for impacting change describes supporting teachers “learning from one another” as well as building “school-level capacity through … instructional leadership.” The program design focused on enhancing networks between teachers and to assist principals in their role as instructional leaders. School administrators were encouraged to lean on TCP teacher leaders to guide conversations and “ask the right questions” that will ultimately improve pedagogy (K-12 Dive, June 2019). The TCP website (April, 2024) maintains this program intention by asserting an approach where teacher leaders “amplify the voice and needs of students, caregivers, and colleagues…” Targeting the role to improve and “de-noise” conversations, creating a common language grounded in classroom practice, and amplifying teacher and student voice illustrate an explicit program focus on enhancing instructional leadership networks and cultivating channels of communication between individuals to improve teaching and student learning. Put simply, TCP aims to build connections between teacher leaders and other stakeholders to increase knowledge of teaching and learning needs in schools.

Liaison between decision-makers to improve the program

Learning from previous unsuccessful attempts to implement teacher-leader programs, TCP placed a greater emphasis on formalizing structures for program effectiveness, coherence, and accountability. Part of these efforts included 17 teacher team leaders who worked full-time on the program “liaising with school administrators to discuss their vision for the teacher leader roles and advise them on how to support teacher leaders in their schools” (K-12 Dive, June 2019). Current website descriptions of roles continue to emphasize this theme asserting that TCP teacher leaders “work closely with colleagues, school leaders, and district leaders [to] strategize the evolution of pedagogy across systems by constructing enduring structures” (April, 2024). Teacher Career Pathways program design emphasized teacher leaders not only working within professional learning structures but also creating new ones, focused on coevolving growth in teaching and growth of the larger school system at the same time. Program policy documents highlight teacher leaders collaborating with administrators to create higher quality and more frequent conversations about teaching and student learning, deepening root connections within and across schools.

Facilitating communication between labor and management

Teacher leaders were not only positioned to span boundaries within the school setting but also between “management” and “labor” in steering the program as a whole. Teacher Career Pathways was designed to be jointly managed by the teachers’ union and district officials to increase buy-in, communication, and better decision-making (K-12 Dive, June 2019). Placing teachers and teacher leaders at the boundaries where impactful decisions were being made led to several structures being built into the program to enhance school and district networks. This includes Teacher Leader in Action! Mini-Conferences, which “focus on authentic practice and create opportunities for teacher leaders to cross-pollinate ideas and leadership practices across the city” (TCP Newsletter, Winter 2024). Teacher leaders within TCP were intended to “create ripple effects” contributing to the collective growth and overall trajectory of NYC Public Schools by using resources more efficiently through more frequent interactions and information sharing.

Overall, these findings indicate that the efforts of teacher leaders were targeted at the boundaries of nested structures within schools and wider systems. By working in different capacities and across levels of school hierarchies, teacher leaders were intended to influence a host of different relationships and expand networks needed for the exchange of critical information. We subsequently examine the relational dynamics and stakeholder experiences to gauge whether teacher leaders had a demonstrable impact on relationships at the boundaries.

Findings on the relational dynamics and stakeholder experiences in TCP schools

The survey responses organized by construct were highly reliable, with reliability of items ranging from fair ( $\propto$ = .776 for school culture on the teacher survey) to excellent ( $\propto$ = .947 for impact, also on the teacher survey). Note that these alphas represent the score for all years 2016–2020. The average response on distributed leadership increased steadily over time across all stakeholder groups (Table 3), indicating a substantive change in the distribution of power and authority in school decision-making and a gradual increase in the creation of networks enabling more frequent interactions between individuals across the school. Teacher leader and principal average responses improved by nearly the same amount over the five-year period (differences in means of 0.424 and 0.461, respectively, p < .001), indicating that both groups believed teacher leaders were having more conversations with school leaders, included in decision-making with school leaders about the allocation of resources and selection of instructional goals and curricular choices. Average responses for teachers improved a large amount from 2016 to 2020 (difference in means of 0.597, p < .001), indicating a belief that teacher leaders had an increased voice in decision-making.

Table 3.

Comparing average responses, by Survey Index & Stakeholder groups.

Stakeholder group by index	M	SD	t
Distributed leadership
Teachers (n = 10,808)
2016	−0.309	1.079	27.105***
2020	0.288	0.864
Teacher leaders (n = 1826)
2016	−0.171	1.041	8.855***
2020	0.254	0.806
Principals (n = 592)
2016	−0.214	1.175	4.735***
2020	0.247	0.806
School culture
Teachers (n = 11,060)
2016	−0.341	1.094	27.629***
2020	0.275	0.875
Teacher leaders(n = 1826)
2016	−0.155	1.078	6.756***
2020	0.181	0.845
Support
Teachers (n = 11,066)
2016	−0.039	0.955	7.884***
2020	0.129	1.098
Teacher leaders (n = 1826)
2016	−0.071	1.002	4.594***
2020	0.152	0.930
Impact
Teachers (n = 5050)
2016	−0.289	1.122	15.473***
2020	0.243	0.870
Teacher leaders (n = 1826)
2016	−0.155	1.096	7.233***
2020	0.206	0.807
Principals (n = 594)
2016	−0.122	1.132	4.312***
2020	0.280	0.749
Strategy
Principals (n = 592)
2016	−0.157	1.091	4.338***
2020	0.244	0.841

Note: For teachers, support is calculated based on how many kinds of support they received from a teacher leader. For teacher leaders, support was calculated based on how frequently they provided different kinds of support to teachers. Samples for each year are not balanced. *p < .05, **p < .01, ***p < .001.

Teachers and teacher leaders both reported improvement in school culture and greater engagement in support and collaboration activities with each other. Teacher average responses on school culture questions increased substantially from 2016 to 2020 (difference in means of 0.615, p < .001). While not as high, teacher leaders also felt more positive about school culture (differences in means 0.336, p < .001). These changes indicate that teachers and teacher leaders felt more valued by their colleagues and school leaders and were increasingly committed to teaching in their school community. Both teachers and teacher leaders engaged in more support and collaboration activities, though these changes were small. This meant that teachers received a few more different kinds of support from teacher leaders (e.g., resources to improve their instructional practice, collaboration to design an assessment and/or curriculum, or the ability to observe a teacher leader as they modeled a lesson). At the same time, teacher leaders increased the frequency with which they provided these supports and engaged in collaborative activities with teachers (e.g., acted as a liaison between teachers and school leaders, reviewed and analyzed student work with colleagues, or led a professional development session for peers).

Given the improvements in distributed leadership, school culture, and support and collaboration, we would expect to see an increase in beliefs of impact and teacher leadership as a school improvement strategy for principals. All three stakeholder groups report increasing agreement that teacher leadership had a positive impact on the school community (differences in means ranging from 0.361 for teacher leaders to 0.531 for teachers, p < .001), noting improvements in, for example, student learning, teacher instructional practices, and teacher classroom management, and better teacher insight on their strengths and opportunities for growth. Principals reported that teacher leadership was critical to their strategy for school improvement, including a strategy to attract new teachers and retain current staff, making the system more coherent and resistant to environmental challenges (difference in means 0.401, p < .001). These results suggest a self-sustaining feedback loop: By improving different aspects of school life for adult stakeholders, teacher leaders appear to become a critical part of the school improvement effort, creating a feedback loop that further cements the need for teacher leadership.

Findings on organizational change and school improvement

The SQR analysis provides an opportunity to gauge if changing the experiences of stakeholders through specific avenues influences school quality. It also provides the first opportunity to examine the impact of having teacher leaders in formalized roles. Staffing teacher leaders had a small impact on school quality, supporting CTTL that by changing the relational dynamics between stakeholders, the wider school community can benefit (Table 4). Schools with staffed teacher leaders experienced statistically significant positive change across the five areas of the SQR. All three areas related to the instructional core (Curriculum, Pedagogy, and Assessment) improved, with differences in means ranging from 0.228 (p < .001) for Curriculum to 0.091 (p < .05) for Assessment. This means that schools improved in their ability to establish a rigorous, engaging, and coherent standards-aligned curriculum, enact instruction that produces high-quality student work, and align assessment practices to the curriculum, and use assessments to inform instruction (NYCDOE, 2020).

Table 4.

Differences in changes on school quality review indicators by staffed teacher leaders & time period.

		Qualified teacher leaders (n = 116)			Staffed teacher leaders (n = 638)			No teacher leaders (n = 557)
SQR indicator		M	SD	t	M	SD	t	M	SD	t
1.1 Curriculum	Before	0.361	0.801	−0.616	−0.022	0.939	6.093***	−0.021	1.130	2.887**
	After	0.314	0.983		0.206	0.797		0.118	0.896
1.2 Pedagogy	Before	0.243	0.898	−0.249	−0.154	0.952	4.334***	0.192	1.055	−1.346
	After	0.220	0.774		0.015	0.769		0.128	0.878
2.2 Assessment	Before	0.323	0.928	−2.172*	−0.112	0.980	2.208*	0.130	1.016	−0.1074
	After	0.128	0.868		−0.021	0.859		0.125	0.936
3.4 High Expectations	Before	0.386	0.828	0.761	−0.007	0.973	8.276***	−0.092	1.064	8.992***
	After	0.450	0.744		0.318	0.808		0.331	0.840
4.2 Teacher Teams & Leadership Development	Before	0.363	0.897	−3.103**	−0.015	0.961	3.172**	0.003	1.064	1.395
	After	0.084	0.917		0.124	0.909		0.071	0.948

Note: This table shows results from 2014 to 2019 for schools with only qualified teacher leaders, schools with staffed teacher leaders, and schools that never had staffed or qualified teacher leader schools during that time frame. *p < .05, ** p < .01, *** p < .001.

Schools with staffed teacher leaders also experienced a moderate increase in Higher Expectations (difference in means of 0.326, p < .001) and a small increase in Teacher Teams & Leadership Development (difference in means of 0.139, p < .01). Improvement in these areas indicates that schools established a culture of learning that clearly communicated and fostered higher expectations of staff, families, and students. At the same time, schools bolstered their ability to engage teacher teams in collaborative practices and apply an inquiry approach to improve classroom practices (NYCDOE, 2020).

It's important to note that schools without any staffed teacher leaders did experience statistically significant positive change in the average score for Curriculum and High Expectations, with differences in means of 0.139 (p < .01) and 0.423 (p < .001), respectively. The improvement of non-TCP schools in their ability to establish a strong, standards-aligned curriculum and a culture of higher expectations may be indicative of some of the system-wide capacity-building initiatives that the NYCDOE undertook to bolster school quality and student achievement. However, schools with staffed teacher leaders improved by a greater amount in the case of curriculum, which suggests that teacher leaders may have improved the ability of stakeholders to build their capacity to integrate such instructional improvements into their school's practices and processes. In the case of high expectations, other reasons, like the differences in student demographics, may be attributable to the slightly greater improvement for non-TCP schools.

Whereas staffing teacher leaders was beneficial to schools, having qualified but not appointed teacher leaders yielded no improvement. Qualified teacher leaders alone were insufficient to incite positive changes in school quality. Indeed, it appears that in the areas of Assessment and Teacher Teams & Leadership Development, schools with only qualified teacher leaders showed statistically significant decreases, while schools without any teacher leaders improved in the areas of Curriculum and Higher Expectations—areas associated with the instructional core and school culture. Put differently, schools with only qualified teacher leaders—those teachers who were qualified by the NYCDOE to be teacher leaders but not appointed into the role—appear to decline in their ability to align assessments to curriculum and pedagogy, and to support the development of collaborative teacher teams aimed at improving instruction. This suggests that without individuals positioned and empowered by administrators to work in such a way, the system itself might stagnate or decline. We return to this idea in our next section.

The compounding effects of teacher empowerment in decision-making

The complementarity of the program documents, survey results, and SQR highlights ways in which teacher leadership might be deployed at the boundaries of relationships and functional work areas to reinvent relationships within schools and expand the capacity of the schools to evolve, like that of middle leaders (Zhang et al., 2022). Bridging the program intentions to the relational experiences of stakeholders to the school quality, we find that teacher leaders were most effective at boundary spanning and enhancing networks when they had a mandate to act. The more explicit the mandate in program documents, the more direct the change. For example, TCP teacher leaders were directed to work with teachers to improve their instructional practice. The transformative potential through the enactment of this intention appears in the complementary experiences of teachers and teacher leaders with respect to the support and collaboration activities—where teachers report teacher leaders providing more kinds of support activities, and teacher leaders report engaging with peers in such activities more frequently. The increases over time in, for example, distributed leadership, collaboration, and culture across stakeholder groups would suggest that individuals processed information and subsequently made decisions about what resources were needed and how to receive/direct them.

That the changes in relational dynamics and school quality appear to increase over time in targeted areas suggests that teacher leadership may have a compounding effect on results. In other words, the changes appear to build on each other, with each year of improvement leading to additional improvement in the following year, and so forth. These compounded effects may provide evidence of feedback loops in schools (Margolis, 2020; Poekert et al., 2016; Sands et al., 2022; Warr and Mishra, 2023). Recalling Figure 2, feedback loops should enable the exchange of information across the system and incorporation of new information into decision-making to enable compounded change at each level (Poekert et al., 2016). The whole community could be evolving together, with the teacher leaders serving as the mediators and facilitators of this growth (Thien et al., 2024; Santagata et al., 2023). In other words, the autonomy granted teacher leaders to work across boundaries might mediate the effect of distributed leadership on instructional quality, teacher and school leader innovation, and school culture (Hsieh et al., 2024).

The lack of improvement in schools that had qualified teacher leaders but did not staff them suggests that in complex systems, the ability to span boundaries and subsequently enhance networks needed to disperse, process, and apply new information may require some level of authority, autonomy, and responsibility. This may explain the pervasiveness of middle leader roles in many countries (e.g., Malaysia in Thien et al., 2024, England in Day et al., 2016; Israel in Shaked, 2023). Our findings echo the relational aspect of middle leaders’ role highlighted in the review of literature by De Nobile (2017) where their managing relationships, leading teams, and communicating effectively featured very strongly. Thien et al.'s work finds that middle leaders’ focus on learning-centered leadership has a significant effect on teacher professional learning. Specifically, middle leaders’ emphasis on teacher learning can create environments where teachers are cultivated to engage more in professional learning. Zhang et al. (2022) observed middle leaders in China acting as “boundary brokers and knowledge engineers,” facilitating both teachers’ long-term learning and school organizational change (458).

Of course, teacher leadership does not necessarily replace the existing hierarchy. In this case, the power of teacher leaders—who have been given positional authority to lead by the district and their principals—is contingent on how the school leader chooses to engage them (Day et al., 2016). This is reflected in our own survey results on the principals’ decision to include teacher leaders in more decision-making arenas over time. It is evident the principal support is a key enabler of successful teacher leadership (De Nobile, 2017).

The challenges of boundary spanning across the school hierarchy notwithstanding, we nevertheless advocate that school and district administrators should devolve or license leadership to teacher leaders to support schools in fulfilling their goals with some parameters. First, it appears that appointment into formal roles with recognition, responsibility, and positional authority may be needed to support broader school reform. Without that formal position, teacher leaders may lack the ability to shift systems, processes, and practices needed to effectuate improvement in student learning. Second, our finding that teacher leaders were most effective where the scope of their role and responsibilities and the intended outcomes aligned implies that administrators, in collaboration with teacher leaders, should be intentional in how the roles are constructed and implemented. Finally, while further statistical testing can bring the compounded effects into clearer relief, for now, it appears that system and school administrators should consider a longer-term investment in teacher leadership in order to reap more extensive benefits.

Assumptions and limitations

The unique program characteristics of TCP present several assumptions and limitations. Chief among them is that participation in TCP is not mandatory, which could impact the survey results. Relatedly, the survey data are potentially biased. Even though all survey takers were informed that the survey was confidential and anonymous, respondents might have been motivated to misrepresent the impact of teacher leadership on their school community.

Selection biases could be at work when it comes to which schools participate in TCP and it is not clear whether this would tilt the results in an upward or downward direction. The survey results pose a challenge with respect to selection bias as those who feel most positive about teacher leadership or their school may feel more inclined to respond. Additionally, schools serving students with greater educational needs tended to staff teacher leaders, which may have further biased the sample.

Finally, the school quality indicators pose a challenge as there could be significant lead and lag time between the baseline and end scores, calling into question whether any observed changes are from TCP or other factors. As this is a descriptive analysis that attempts to establish whether there is a relationship between having teacher leaders and changes in school quality, this limitation is not disqualifying but just a consideration when interpreting our findings.

Conclusion

Proponents of teacher leadership assert that teacher leaders, like middle leaders, play a critical role in moderating the relationship between teachers and school leaders, acting as a liaison, helping create enthusiasm and support for instructional change initiatives by sharing information, fostering engagement in school processes, and breeding trust (Margolis, 2020; Nguyen et al., 2017). The findings in this paper extend such earlier findings, providing evidence that changing the relational dynamics between stakeholders and improving school quality are intimately connected, and that teacher leaders, with the right support from senior leadership throughout the system, can play a significant role in that disruption. Different from earlier work, our paper contributes to the theory-building that is much needed in the realms of teacher and middle leadership (Nguyen et al., 2020; De Nobile, 2017). Looking ahead, our intimation that the relationship between teacher leaders’ work, stakeholder relational dynamics, and school quality results appear to substantiate the existence of feedback loops is ripe for future research. Namely, researchers seeking to build CTTL might pursue process tracing in schools with teacher leaders to better understand when, how, and under what conditions teacher leaders initiate and sustain feedback loops and the impact of their actions. Such future research can build on our work in better elucidating how adaptive systems like schools engage in a give-and-take between continual learning and stabilizing those changes to improve and maintain school quality.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Sara Sands Francis

Fiona King

Notes

Author biographies

Sara Sands Francis is an Instructional Associate Professor of Public Administration and the Associate Director of the Elizabeth D. Rockwell Center on Ethics & Leadership at the University of Houston Hobby School of Public Affairs. Her career has spanned the private, public, and nonprofit sectors. As a researcher, her work focuses on the intersection of K-12 education policy, political science, and public administration, including teacher leadership and complexity theory.

Jason Margolis is a professor of education and EdD in Educational Leadership Program Director at St Bonaventure University. His career as a teacher and researcher has focused on the intersection between teacher professional development, school change, and complexity theory-inspired approaches to teaching and learning.

Fiona King is an associate professor and Chair of the EdD programme at the Institute of Education, Dublin City University (DCU). Her career as a teacher and researcher has focused on teacher professional learning, leadership for inclusion, teacher leadership, and social justice leadership.

References

Bennett

Woods

Wise

and Newton W (2007) Understandings of middle leadership in secondary schools: A review of empirical research. School Leadership & Management 27(5): 453–470.

Bowen

(2009) Document analysis as a qualitative research method. Qualitative Research Journal 9(2): 27–40.

Bryk

Sebring

Kerbow

, Rollow S and Easton JQ (1998) Charting Chicago School Reform: Democratic Localism as a Lever for Change. Boulder: Westview Press.

Buchanan

Mills

Edward

, Mathieu E, Snyder M, Weitman M, Goodsell C and Thurman K (2023) Teacher leadership collaborative: Boundary-crossing spaces for teacher empowerment. Professional Development in Education 49(6): 1152–1166.

Davis

Sumara

(2008) Complexity and Education: Inquiries into Learning, Teaching, and Research. New York: Routledge.

Day

Sammons

(2016) The impact of leadership on student outcomes: How successful school leaders use transformational and instructional strategies to make a difference. Educational Administration Quarterly 52(2): 221–258.

De Nobile

(2017) Towards a theoretical model of middle leadership in schools. School Leadership & Management 38(4): 395–416.

Earley

Bubb

(2004) Leading and Managing Continuing Professional Development: Developing People, Developing Schools. London: Paul Chapman.

Grootenboer

Tindall-Ford

Edwards-Groves

and Attard C (2023) Establishing an evidence-base for supporting middle leadership practice development in schools. School Leadership & Management 43(5): 454–472.

10.

Harris

Jones

Ismail

and Nguyen D (2019) Middle leaders and middle leadership in schools: Exploring the knowledge base (2003–2017). School Leadership & Management 39(3-4): 255–277.

11.

Hsieh

Song

(2024) Analyzing the relationship between distributed leadership and instructional quality in Taiwan: The mediating roles of teacher autonomy and innovation. Educational Management Administration & Leadership: 1–19.

12.

Keszey

(2018) Boundary spanners’ knowledge sharing for innovation success in turbulent times. Journal of Knowledge Management 22(5): 1061–1081.

13.

King

Holland

(2022) A transformative professional learning meta-model to support leadership learning and growth of early career teachers. International Journal of Leadership in Education: 1–25.

14.

Lee

DHL

NKK

(2023) The influence of professional learning communities on informal teacher leadership in a Chinese hierarchical school context. Educational Management Administration & Leadership 51(2): 324–344.

15.

Lemke

Sabelli

(2008) Complex systems and educational change: Towards a new research agenda. Educational Philosophy and Theory 40(1): 118–129.

16.

Lipscombe

Tindall-Ford

Lamanna

(2023) School middle leadership: A systematic review. Educational Management Administration & Leadership 51(2): 270–288.

17.

Margolis

(2012) Hybrid teacher leaders and the new professional development ecology. Professional Development in Education 38(2): 291–315.

18.

Margolis

(2020) The semiformality of teacher leadership on the edge of chaos. Harvard Educational Review 90(3): 397–418.

19.

McElrone

Choat

Gambetta

and Broderson CR (2013) Water uptake and transport in vascular plants. Nature Education Knowledge 4(5): 6.

20.

Meirink

Van Der Want

Louws

, Meijer P, Oolbekkink-Marchand H and Schaap H (2019) Beginning teachers’ opportunities for enacting informal teacher leadership: Perceptions of teachers and school management staff members. European Journal of Teacher Education 43(2): 243–257.

21.

Mital

Moore

Llewellyn

(2014) Analyzing K-12 education as a complex system. Procedia Computer Science 28: 370–379.

22.

Morrison

(2002) School Leadership and Complexity Theory. Taylor & Francis Group.

23.

Munsayac

(n.d.) Organizational communication theories under the cybernetic tradition. Available at: https://www.academia.edu/760701/Organizational_Communication_Theories_under_the_Cybernetic_Tradition (accessed 1 May 2024).

24.

New York City Department of Education (2020, September 16) Quality review 2020-2021: Big ideas by indicator and sub-indicator. https://infohub.nyced.org/docs/default-source/default-document-library/quality-review-big-ideas.pdf (accessed 8 November 2021).

25.

New York City Department of Education (2022) School quality. https://www.schools.nyc.gov/about-us/reports/school-quality (accessed January 2022).

26.

Nguyen

Harris

(2020) A review of the empirical research on teacher leadership (2003–2017): Evidence, patterns and implications. Journal of Educational Administration 58(1): 60–80.

27.

Nguyen

Yap

(2017) Instructional leadership structure in Singapore: A co-existence of hierarchy and heterarchy. Journal of Educational Administration 55(2): 147–167.

28.

Pan

HLW

Wiens

Moyal

(2023) A bibliometric analysis of the teacher leadership scholarship. Teaching and Teacher Education 121: 103936.

29.

Poekert

Alexandrou

Shannon

(2016) How teachers become leaders: An internationally validated theoretical model of teacher leadership development. Research in Post-Compulsory Education 21(4): 307–329.

30.

Resnick

Scherrer

(2012) Social networks in “nested learning organizations”—a commentary. American Journal of Education 119(1): 183–192.

31.

Sands

(2023) Complexity in action: Evaluating teacher leadership through the lens of complexity theory. SSRN.

32.

Sands

(2023) Leading from the classroom: Teacher leadership in New York City schools. PhD Thesis, Teachers College, Columbia University, USA.

33.

Sands

Margolis

Murphy

M J

(2022) Understanding teacher leadership in New York City through complexity theory. International Journal of Complexity in Education 3(1): 1–32.

34.

Santagata

Lee

Guarino

and Drake J (2023) Centering teacher voices in school-wide improvement: Possibilities and challenges of introducing change in complex systems. Professional Development in Education 49(6): 1167–1182.

35.

Shaked

(2023) Instructional leadership in school middle leaders. International Journal of Educational Management 37(6/7): 1288–1302.

36.

Siskin

(1993, August) Hermaphroditic roles: Teacher leadership in secondary schools. Paper presented at the International Study Association on Teacher Thinking, Gothenburg, Sweden.

37.

Smith

(2019) Leadership for “the all of it”: Formalizing teacher-leader networks. Teacher College Record 121(030308): 1–44.

38.

Snoek

Enthoven

Kessels

and Volman M (2017) Increasing the impact of a master’s programme on teacher leadership and school development by means of boundary crossing. International Journal of Leadership in Education 20(1): 26–56.

39.

Snyder

(2013) The simple, the complicated, and the complex: Educational reform through the lens of complexity theory, OECD Education Working Papers, No. 96, OECD Publishing, Paris.

40.

Supovitz

Comstock

(2023) The impact of a formal teacher leadership program on student performance. Education Policy Analysis Archives 31(74): 1–19.

41.

Taylor

Goeke

Klein

, Onore C

and Geist K (2011) Changing leadership: Teachers lead the way for schools that learn. Teaching and Teacher Education 27: 920–929.

42.

Thien

Uthai

Yeap

(2024) Does middle leaders’ learning-centered leadership matter in promoting teacher professional learning? A partial least squares analysis. Educational Management Administration & Leadership 52(4): 854–877.

43.

Warr

Mishra

(2023) Learning to see complexity: Teachers designing amidst indeterminacy. Professional Development in Education 49(6): 1036–1052.

44.

York

Duke

(2004) What do we know about teacher leadership? Findings from two decades of scholarship. Review of Educational Research 74(3): 255–316.

45.

Yost

Vogel

Liang

(2009) Embedded teacher leadership: Support for a site-based model of professional development. International Journal of Leadership in Education 12(4): 409–433.

46.

Zadok

Benoliel

Schechter

(2024) Organizational resilience and transformational leadership for managing complex school systems. Frontiers in Education 9(1333551): 1–13.

47.

Zhang

Wong

JLN

Wang

(2022) How do the leadership strategies of middle leaders affect teachers’ learning in schools? A case study from China. Professional Development in Education 48(3): 444–461.