How Districts Scale Instructional Improvement That Lasts

This practice brief, which extends the three-part PACE series Taking Reform to Scale, explores why many districts struggle to scale lasting instructional improvement and highlights the importance of district systems—structures and processes—in successfully scaling improvement initiatives. Drawing on data from districts’ work with California Education Partners (Ed Partners), the authors contrast effective district system engagement with common pitfalls such as delegating leadership to school principals, relying on teachers’ informal networks to spread improvements, and maintaining too many priorities to focus effectively on scaling up. District leaders play a pivotal role in engaging system components to support teachers and schools in districtwide improvement because only these leaders have the authority to repurpose existing system resources, create new structures as needed to fill gaps, and monitor systemwide progress to identify and adapt to challenges. A companion document, available at right, offers key, actionable implications of the main themes of the brief for district leaders.

Introduction

Why do so many school districts’ efforts to improve instruction fail? Coburn’s 2003 work examined districts’ implementation of literacy reforms, seeking to understand the features of implementation that led to durable improvements in student outcomes, which she described as reform at “scale.”1 She found that, in addition to spread (increasing the number of people doing a practice), truly scaling improvement requires that the new approaches have local ownership, be implemented at sufficient depth to lead to substantive changes in adult practices and beliefs, and be sustained over time. When we use the word scale, we are describing all four of these components.

The purpose of this brief is to lift up the role of district systems—structures and processes—in enabling successful scaling of instructional improvement across the entire district. In our work, we have seen that some districts already have the capacity to implement reform at scale2 whereas others are in earlier stages of building needed structures and processes.3 This brief relies on data from our ongoing research partnership with California Education Partners to illustrate how engaging a district’s system can support scaling instructional improvement to change student outcomes across a district. The brief contrasts such system engagement with other more common and less effective approaches to implementation. Although building ownership of improvement at the teacher and building level is critical, we find that district leaders play a pivotal role in enabling improvement to scale across a district because they are uniquely positioned to leverage the entire district system. Therefore, our findings have implications for the role leaders play when trying to scale improvement across their districts.

Background

Ed Partners currently has eight collaborations serving 51 teams from 45 districts and one County Office of Education (COE), along with 12 districts in that county. The goal of the collaborations is to help districts build their system capacity for instructional improvement in the service of better student outcomes. The collaborations provide participating districts with coaching and partnership opportunities with a cohort of other districts working on the same topic. Gallagher and Cottingham offered an overview of the Ed Partners model in a 2024 PACE report.4

Ed Partners currently offers collaborations on two topics, both of which give districts the opportunity to examine and improve their systems in a particular area, with the idea that districts can transfer some of what they learn to other priorities:

• The Pre-K to 3 Coherence Collaboration (P3CC) focuses on building shared learning expectations and a set of effective common instructional practices in mathematics that are vertically coherent from prekindergarten/transitional kindergarten (PK/TK) through third grade.
• 8th–9th On Track pairs schools within a district—typically, one middle school with an eighth grade and one high school with a ninth grade in the same feeder pattern—to align learning expectations and develop a small set of common effective practices in mathematics and English language arts (ELA), which enable more students to pass rigorous courses with a grade of C or better.

Across both collaborations, Ed Partners teaches a framework called the Fundamentals for System Coherence and Impact for thinking about a district’s capacity to improve instruction. The framework names four fundamentals—shared expectations for student learning, common effective practices, monitoring student progress, and systems for adult capacity building—that together enable districts to scale instructional improvements. Of these fundamentals, monitoring and adult capacity building are the most foundational because they are the central structures and processes that enable continuous improvement in instruction, while expectations and adult practices define the nature of the instructional improvement that is desired. Because earlier work by Policy Analysis for California Education (PACE) focused more heavily on the fundamentals,5 in this brief we focus on district actions to scale change, noting the fundamentals when relevant.

Many districts choose to work with Ed Partners recognizing that while they enter the collaborations with many assets, they need to build more widespread knowledge about how to engage district systems fully to support instructional improvement. At inception, Ed Partners requires each participating district to create an improvement team, which is carefully designed to facilitate the team’s testing and ultimately scaling new information and adult practices across the district’s system (as well as to ensure broad ownership of the changes). Teams typically have about 10 members representing key constituencies from the district:

• teachers who help decide which approaches to try and then test those approaches;
• principals of at least two participating schools who can set priorities and allocate resources at the school level, elevate needs or concerns to district leaders, and share knowledge and experiences with their peer leaders at other schools;
• at least one senior district leader—ideally, the superintendent or a cabinet member checking in regularly with the superintendent about the work’s progress—who is key to improving district systems and scaling the work; and
• other roles key to the work, such as teachers on special assignment (TOSAs) or counselors.

The structure of the team is critical because it creates multidirectional flows of information among senior decision makers and the educators who directly serve students. This helps spread the information necessary for improvement to scale by putting teacher leaders in direct communication with senior school and district leadership, who can make more informed decisions if they have access to clear information from teachers’ perspectives. Senior leaders’ decisions are necessary to engage a district’s system in an improvement effort, while the support of teachers and other educators is typically key for building support for new approaches.

Importantly, the team typically has representatives from only two schools in the district. This strategic choice helps maintain a manageable group size for the improvement team and gives the district practice with scaling new ideas and approaches to schools that are not directly working with Ed Partners a critical step in building districts’ capacity to scale other improvements on their own in the future. As we will illustrate in this brief, our data suggest that scaling improvements to schools where no staff work with Ed Partners requires the improvement team to engage system fundamentals in coherent ways.

Two goals of this brief are to illustrate what we mean by district system capacity for instructional improvement and to provide examples of approaches we have seen during the past 5 years that were most effective at scaling instructional improvement. Each district described in this brief has invested effort in its Ed Partners work and has some successes to show for that, but the work progressed relatively slowly in some districts and more quickly in others, and the results in some districts are more fragile than in others.

This brief starts by providing examples where the ways districts engaged in the work enabled progress initially but success was ultimately limited by their failure to make needed district-level changes. We then give examples of districts that engaged their systems more fully to support instructional improvement by modifying existing components and building new components that were needed to scale the improvement. These latter examples illustrate how senior leaders can strategically engage district systems to scale instructional improvement regardless of the attributes of their systems when they began the effort.

Attempting to Scale Without Leveraging District Systems

In our work with Ed Partners, we have seen several examples of districts that have made progress towards changing practice and improving student outcomes. We have also seen many districts face unnecessary challenges. The districts that made minimal progress while working with Ed Partners tend to have some common attributes. Although our data are based on Ed Partners initiatives, the findings have implications for any initiative designed to improve instruction and student learning throughout a district.

One critical misstep we have seen is a superintendent signing up their district for a collaboration but then largely disengaging after delegating leadership of that work to a member of central office staff. In cases where the superintendent is disengaged, their delegate is often unable to galvanize sufficient support for basic team requirements, such as attending meetings regularly or testing new approaches and reflecting on the results between meetings to move the work forward. Districts where senior leadership does not directly support the work often make minimal progress.

In other districts we have seen district leaders expressing three common beliefs that create very strong barriers to scaling improvement across the district, even if the leaders tacitly support the work:

• schools are the unit of instructional change,
• teachers are the primary mechanism for scaling improvement to other teachers, and
• districts can or should act on multiple high priorities simultaneously.

All three beliefs have some truth in them. Schools (and school leadership) play an important role in improving instruction. Teachers’ support for new practices is often crucial for building the goodwill of their colleagues for changing practices. And districts cannot ignore altogether the other domains of work that they need to do. But for meaningful and sustained improvement to occur, teachers and school leaders need to be supported by systems that align with the changes they are being asked to make. In the following sections, we illustrate how these beliefs about system support cut across district-specific contexts and limit the potential to scale improvement.

The scale of improvement is typically limited by the scope of its leaders’ influence; many districts delegate instructional leadership to principals, leading to inconsistencies across schools.

Instructional improvement clearly needs to engage teachers because they are the ones who provide instruction. District leaders who have chosen to work with Ed Partners generally realize this and understand the importance of principals’ support for instructional improvement in their own schools. In recognizing the importance of principals, though, some district leaders underestimate the significance of district-level involvement in supporting improved instruction to scale across a district.

Windy City’s (all district names are pseudonyms) work with Ed Partners focused on using specific instructional practices in mathematics in early grades (TK–3) to build students’ conceptual understanding. To understand student learning, the improvement team initiated the use of a test that provided a precise assessment of elementary school students’ numeracy skills; teachers could opt to administer the assessment. Over the duration of the collaboration, the instructional practices and numeracy test spread from just the improvement team to most of the teachers in the two principals’ schools. During the final year of the collaboration, the district decided to expand the work further by creating a voluntary districtwide team that met monthly to collaborate around math instructional practices. Primary-grade teachers at any school could join, and across the 14 elementary schools in the district, about 50 teachers participated. The spread of engagement with the work—in terms of use of new instructional practices and assessments by most TK–3 teachers in the two principals’ schools and an average of about three teachers per school in other district elementary schools—is apparent. But interviews with improvement team members in the district show a belief that the spread was fragile. Interviewees believed the improvements were unlikely to sustain and scale because the only district space where the work was prioritized was the newly created districtwide team, which was disconnected from the district’s central structures for curriculum and instruction.

One of the biggest challenges Windy City faced in scaling its math work districtwide is that the district office did not typically engage in instructional work and did not make leading the improvement a central part of any district leader’s role. Although the superintendent stated that he supported the math improvement work with Ed Partners, overseeing the initiative was one of many priorities for the district leader who was designated as the team lead. Most of her job focused on responding to the multiple issues and crises that arise regularly in large districts. She explained that even when she had time to devote to instruction, “considering the responsibilities, this [TK–3 math work] is not the end all, be all. … We have language arts, we have science, we’re trying to do STEM. … So there’s a lot of things [on my plate].”

The district decided to use time in regular principals’ meetings to propel the spread of this early math work to all elementary schools. But, as one of the principals engaged in the initiative described it, the math work rarely could garner meeting time because so many issues and other initiatives were also treated as priorities; meetings were used more for operational issues than as an opportunity to build instructional quality and coherence. One of the principals on the improvement team explained:

That has been a challenge, fighting to preserve time in our [district] agendas to talk about instruction. … I think there isn’t a culture for that. … We’re meeting two to three times a month, just as site administrators, and very few of those meetings involved instructional conversations or space for instructional conversations. And so step one has been through … sheer force of will, just pushing, pushing, pushing to make time and space for that. [The other principal on the improvement team and I] prepare every meeting to talk a little bit about math, and sometimes it really is like, “Hey, and don’t forget, we’re doing this math work.” Then that’s the extent of what we get on the agenda.

The regular principals’ meetings were correctly identified as a district structure that could have embedded the math improvements into the district system. But without time in the principals’ meeting dedicated to scaling the math instruction and assessment changes to other school leaders, the handful of teachers from elementary schools who participated in the voluntary districtwide meetings had no administrative support at their sites to implement new instructional practices. While their instruction may have improved, it was an individual act of improvement, not schoolwide or districtwide improvement. The district similarly failed to widely adopt any of the practices or the numeracy assessment tested by the improvement team.

Meanwhile, at the schools led by principals on the improvement team, the picture looked very different. Both principals on the improvement team reported that about 90 percent of their teachers used the focal math practices regularly and that grade-level teams used existing teacher collaborative time to analyze artifacts of student learning—and data from the numeracy assessment—to support teachers with refining their math instruction. Because the top leaders in the schools made the math improvement a priority—both by championing the work and allocating the critical resource of time to that work on a regular basis—math practices that were restricted to a few early adopters in most district schools became widespread and successful in those two schools. In effect, the two schools that had principals and teachers on the improvement team were able to change their school-level systems to support new practices in early grades math instruction; but without district support, they lacked the ability to scale those practices beyond their two schools.

Another example comes from Valley View, a relatively large district where the instructional changes introduced by Ed Partners were considered to be very successful, especially at the pair of schools that had teachers on the improvement team. But Valley View did not have much of a conception of the roles that district leaders typically needed to play to scale effective instructional practices (beyond such things as passing supportive board policy). One district leader explained, “I honestly believe that [instructional improvement] just has to live at the school site.” In the district’s final year working with Ed Partners, the work had spread from the two teachers on the original team to 80 percent of teachers in the one school whose principal was on the team—another great example of successful scaling at a site where the top leader was heavily engaged. But because the district did not have a clear vision of how it needed to engage in scaling instructional improvement across the district, the work had gained minimal traction in other schools.

The uneven quality and spread of the specific instructional practices at the center of the work illustrates a theme we have commonly seen. Improvements scale only as far as they are prioritized by top system leaders. Leaders need to champion the work by naming it as a priority and enacting that prioritization through the decisions they make and the resources they allocate to it. If the top leader prioritizing an improvement runs an entire district, then the improvement may scale to the whole district; if the top leader is a principal, it may scale schoolwide; and if it is a department chair or teacher, the maximum likely scale will be the department or the classroom.

‌While teachers’ buy-in to new ideas is critical, teachers’ informal networks are insufficient to scale improvement.

Over the past 5 years of studying Ed Partners collaborations (and other efforts at continuous improvement), we have seen that many districts approach the spread of improvement with an oversimplified view of Everett Rogers’s diffusion of innovation model—namely, that every new idea will have a few early adopters, and once it reaches a critical mass, it will naturally become widespread standard practice.6 In reality, Rogers’s theory was much more developed than that, emphasizing how social systems, communication channels, and the perceived characteristics of innovations shape adoption. The simplified version—namely that if new ideas have grassroots support, their spread will be almost automatic—nonetheless is prevalent. Some readers may immediately grasp the flaws in relying on early adopters as the main mechanism for spread, but we have seen this misunderstanding on numerous occasions.7

In one instance, Ed Partners presented information about scaling improvement to districts at a convening and gave teams time to plan how they were going to attempt scaling the work in their districts. Teachers from two of Sequoia Grove’s six elementary schools were members of the team. The district-level administrator who was also on the team recognized how important teachers can be as ambassadors of new instructional approaches and asked the teacher team members what they could do to spread the ideas to other schools. In the existing district structures, the teachers had no formal connections to their colleagues at other schools. As they brainstormed their informal connections, the teachers quickly realized that those connections were not sufficient for scaling substantive shifts in teacher practice and they worked with the district to create a pair of new, cross-school structures to support spreading to additional schools.

In Valley View, PACE observed a meeting of a voluntary afterschool professional development series open to teachers from all schools in the district, much like the monthly meeting in Windy City. A total of 32 teachers—several from the original site and an average of slightly more than two from each of the other dozen schools in the district—attended the monthly series and were trying the instructional changes at the core of the work. As an incentive for attending each meeting, the district gave every participant a raffle ticket, which was entered into a drawing for class sets of math manipulatives and other instruction-related prizes. At the close of the meeting we observed, the district administrator running the meeting asked teachers in attendance to “bring a friend” to the next event in the series to gain an additional raffle ticket. This approach, even when offered at the district level, fundamentally relies on the informal networks of individual early adopter teachers to spread the work. Our data suggest that this is a misunderstanding of what is required to reach all staff and achieve true systemwide change. Interested staff will engage with the work, but until learning about new practices is embedded in the district infrastructure in ways that both the early adopters and the more hesitant teachers are required to engage with, improvements will likely fail to scale districtwide.

Senior leaders must exercise their authority to deprioritize some things so that the organization can align around a few ongoing, critical goals.

“When everything is a priority, nothing is a priority” is a common saying, and its intuitive truth is easy to see. In districts and schools, it is very challenging for districts to consistently prioritize only the few most important things. The expectations placed on our public TK–12 education system have grown in recent years. We expect districts to teach all students many subjects, provide supports for social-emotional development and mental health, feed students, prepare students to enter a workforce where technology is accelerating the pace of change, and develop the “whole child” so that students will become happy and productive members of society. Every expectation is genuinely important, but districts do not have the resources—especially in terms of educator time—to improve in all these areas simultaneously.

Only senior leaders have the ability—often with substantial input from other educators, families, and the broader community—to deprioritize some areas in order to enable steady improvement over multiple years in others. Our data show that this is very difficult to do. For example, the relatively new superintendent in Foothills Unified had objective data indicating that the district needed to improve on multiple fronts. The superintendent simultaneously launched multiple major initiatives with professional development components for teachers. The approach was fragmented, and teachers were overwhelmed; as one educator explained:

At my site, there’s a lot of veteran teachers, and right now, there’s a lot of new initiatives.  … We have a new superintendent, and there’s a lot of new things coming our way, and we’ve had a lot of hesitancy from teachers to join new initiatives. … So far, our teachers have been pulled this year for 16 trainings. … They’ve been pulled all day. … That gives you an idea how overwhelmed they are.

Although this extent of pullout is an outlier, another district working with Ed Partners experienced similar amounts of all-day professional development in a year because it too was attempting multiple major initiatives simultaneously; in their interviews, teachers in other districts also reported initiative overload. Districts with too many “priorities” fail to build the collective ownership and deep understanding needed for high-quality implementation. The typical result is a diluted focus that prevents the district from giving sustained attention to any goal.

With these examples of common pitfalls and lessons learned, we turn next to what it looks like to engage a district system in instructional improvement.

A System for Instructional Improvement

Scaling instructional improvement beyond just a few teachers and their close colleagues requires engaging a district’s systems. Education researchers and leaders frequently use the term system reform, but our research with districts has shown that what is meant by system is often underspecified. The now-retired superintendent of an urban district, who led the district in all its work with Ed Partners through 2024–25, explained what she means when she talks about a district system:

So the system. There’s the operational side of the house …: maintenance, operation, nutrition, services, transportation, hiring, HR—all of those are systems and departments … [that need to] work efficiently and effectively. On the educational side [of the house], … it’s the systems of support to ensure that teachers are creating the right conditions  for kids. So, what does that look like? [It’s things like:] How do we go about adopting curriculum? How do we ensure that the professional development is targeted and  on point? … How do we give feedback so our principals are giving feedback to teachers? How we go about hiring? … How is that aligned with our instructional core? … How do those all work together? Because they can’t work in isolation of each other. Each one of those things that I identified has to have a coherent message, a coherent vision,  with everyone’s going towards the same … North Star. … And it’s not only the way that you think but the decisions that you make all aligned to a particular vision. So that’s  … a system. [emphasis added]

The task of system leaders is to determine how to set an instructional vision and align system components to support teachers to meet that vision consistently. We see leaders doing this by

• repurposing existing parts of their system and

• building new system components that are needed.

Districts can repurpose existing parts of a system to support scaling instructional improvement.

Ed Partners works with districts to identify components of their systems that leaders can align around efforts to scale instructional improvement. Districts often find they already have

• shared instructional resources (e.g., supplemental instructional tasks, common formative assessments, shared digital folders with instructional resources);
• meetings (principals’ meetings, staff meetings, grade-level/department meetings);
• professional learning opportunities (including coaching);
• instructional leaders (e.g., department chairs, TOSAs, administrators); and
• processes to support collaboration and build shared understanding of instruction (e.g., professional learning communities, nonevaluative instructional walkthroughs, co-teaching).

Allocating these resources behind an effort to improve instruction facilitates scale. For example, Copper Mine entered Ed Partners’ collaboration several years into a districtwide commitment to consistent implementation of a direct-instruction approach to teaching that adhered relatively closely to district-adopted curricular materials and district-developed pacing guides. Through working with Ed Partners, the district decided to make substantial changes to how it approached math instruction to better align with the state’s mathematics standards and instructional framework, and its preexisting system provided a strong foundation for its improvement efforts.

Copper Mine participated in Ed Partners’ P3CC collaborations, which include a “Red-Yellow-Green” exercise. During the first year of the collaboration, teachers on the improvement team tried replacing “red” lessons with one of the instructional routines taught by Ed Partners. They found that students were more engaged in math instruction and the routines supported differentiation, so they decided to scale the practice. They did an informal audit of their existing system and ultimately identified multiple structures they could use to spread changes in teaching mathematics throughout the district.

Shared instructional resources was the first system lever that Copper Mine activated to spread new approaches. Instruction in every teacher’s classroom prior to the work with Ed Partners was heavily shaped by the district’s pacing guide, which teachers were expected to use to ensure that they moved through the adopted curriculum at a roughly similar pace. Because the pacing guide included guidance to teach lessons that the team had identified as “red” and did not leave any time for inserting instructional routines with open-ended tasks, left alone it would likely have created a barrier to the spread of new practices.

During the summer after the district’s first year in the collaboration, the district invited a small team of teachers representing all grades TK–3 and all schools to revise the pacing guides. One teacher explained the detailed analysis process: “We have our standards—how does our curriculum line up? … Okay, if we take this out, what are we going to put in?” In the place of “red” lessons, teachers inserted open-ended tasks that were tightly aligned to standards, increasing the rigor of instruction. Because there was a shared expectation in the district that teachers would follow the pacing guides relatively closely, the revised guides essentially ensured that all students in the district would be exposed to more open-ended tasks during the second year of the collaboration.

Based on their own experiences trying the approaches in their classrooms the previous year, team members knew that teachers would need professional learning opportunities to implement the new approaches. Copper Mine identified meeting times during weekly staff meetings and regular professional development days to roll out these changes. Although district and school administrators reallocated how time was spent building teacher capacity for the new instructional approaches, they asked teachers on the team to lead the professional development because of their personal knowledge of the shifts and the fact that their positionality was most likely to build teacher support for the new approach. When teachers asked for additional support, the district reinstituted districtwide grade-level meetings, a format underutilized since the COVID-19 pandemic, to create a space for teachers in the same grade level across all schools in the district to work together.

Copper Mine’s approach purposely increased the number of instructional leaders in the district by activating the leadership capacity of teachers on the improvement team. Additionally, the district modified the role that their existing TOSAs played; TOSAs expanded from professional development facilitators to developers of the materials needed for districtwide training and supported teachers who needed more help with in-class implementation. Finally, the district redesigned the membership on its Ed Partners improvement team to include teachers from all schools on a “travel team” (the team that worked directly with Ed Partners) and added a larger “home team” composed of a broader set of educators from every school in addition to those on the “travel team.” Unlike the district teams in Windy City and Valley View, which engaged only interested volunteers, the structure of the home team was designed by engaged senior leaders to set expectations for doing these new practices across the district and to connect the travel team with all teachers in the district. As a result, Copper Mine’s home team built vertical and horizontal coherence for the Ed Partners work. All participating grade levels had representatives on the travel team (per Ed Partners guidance), TOSAs on the travel team supported improved instruction across the district, and all travel team members were also on the home team; additionally, all schools had teachers and site administrators on the home team, linking the two teams to every school.

Finally, Copper Mine used multiple processes to support collaboration and build shared understandings. In addition to the processes included in the meetings previously described, Copper Mine used existing walkthroughs to monitor implementation of the new approaches for the purposes of supporting teachers and informing needed professional learning. One educator explained that in Copper Mine, “our district has an expectation that our administrators are in classrooms, at minimum once a week—ideally, twice a week, all day, coaching teachers.… I would expect them to start to see grade-level standards being taught consistently.” District and school leaders included their Ed Partners program manager in one of the walkthroughs. When debriefing the walkthroughs during the second year of the collaboration, district and school leaders concurred with Ed Partners that new practices were widespread in the district, but implementation (in its first year for most teachers) was still uneven. This prompted the district to provide additional supports (such as the repurposed TOSAs) for teacher learning.

By repurposing components of its existing system, Copper Mine was quickly able to spread new practices across the district. Copper Mine’s approach made improvement efforts the default for all teachers in the district instead of for just a coalition of the willing. The district’s efforts to monitor the quality of implementation are an important—and often overlooked—part of scaling instructional changes. Administrators’ walkthroughs identified weaknesses in the depth of change, which are to be expected early in the process of making substantial changes in instruction. Armed with an understanding that teachers needed more support, district leaders could work on the depth of change so that the improvement could fully scale. As of this writing, the work to support teachers with using these practices consistently and well is ongoing. Over the course of the 2 years that the district has focused on TK–3 math instruction with Ed Partners, third-grade California Assessment of Student Performance and Progress (CAASPP) scores in mathematics have increased from 28 percent of students meeting or exceeding standards in 2022–23 to 40 percent in 2024–25.

Districts can build structures needed to scale improvement if they do not exist.

The example of Copper Mine shows how a district repurposed key components of its existing system to move forward quickly with shifting math instructional practices. However, districts vary in the extent to which they already have well-developed systems. Sometimes districts do not have a key structure or process necessary to scale a particular reform and must create one or more new ones.

When Center City started working with Ed Partners, it had built a set of vertically coherent systems that allowed all information and data to flow from classrooms through schools to the district office. But when the district sent a paired middle and high school to Ed Partners’ 8th/9th On Track collaboration, there were insufficient structures to facilitate spread of shared practices between the middle school and its high school pair. When the collaboration started, teachers on the improvement team in the middle school tested changes to grading practices, which reduced the number of students receiving Ds and Fs, and middle school teachers saw their course grades rapidly improve. Thanks to the district’s preexisting data system and its requirements for common formative assessments, the district was able to see that both course grades and interim assessment scores rose (which provided key evidence that students’ better grades were deserved because their skills were also rising). With help from its Ed Partners improvement team, the district created a new structure—a full-day release meeting for the math and ELA teachers in 8th and 9th grade at the paired schools at least twice a year—so the teachers could share what they were trying and how it worked. One administrator explained:

So one of our other next-step strategies that we have been doing regularly is we bring both schools, both departments together … to share what they’ve done, what they’re doing, and what they’re hoping for. And then we bring them back at the end of the semester, to share out their results. … What has happened is that [middle school] team that’s on fire has … demonstrated to the [partner high school] team, … [you] need to catch some of our fire. … It’s been really powerful. And now those teams are talking all the time. I think, actually, one of the English teachers from [middle school] was invited by the teacher from [high school] to come and show them how they do their [work].

As Center City has added additional pairs of schools to later cohorts, it has retained the vertical structure to support collaboration across the 8th- and 9th-grade math and ELA departments in paired schools and enable improvements to scale across each pair.

Discussion

Using examples drawn from six districts, we have illustrated key components of systems for instructional improvement and provided evidence that districts can adapt existing system components and build new ones to scale instructional improvement. Looking across these examples, we draw the following conclusions.

A critical role for top leaders is to prioritize very few things and allocate resources (including leaders’ time) to those things; if people perceive that there are too many priorities, nothing will be consistently prioritized.

For an instructional initiative to be sustainable, it must become part of the district’s standard operating procedure. A common pitfall is treating new improvement work as “one more thing”: an initiative piled on top of—rather than integrated into—the core functions of schools. Districts often create entirely new pathways and structures for the initiative instead of routing it through existing ones like professional learning communities or staff meetings. The experience in Windy City clearly illustrates this challenge. To spread new math practices, leaders created a districtwide team composed of “representatives, early adopters, interested folks” from every school. While this new structure generated energy, the work was not integrated into routine meetings, limiting its impact. In some cases, roles and responsibilities of key personnel—principals, coaches, and teacher leaders—are not formally redefined to include supporting specific improvements to instruction. When the work is not a core part of leaders’ job descriptions, its success depends on the benevolence and extra effort of willing individuals, which is not a viable strategy for long-term, systemic change.

Teachers and principals play an important role in scaling improvement across a district, but they cannot scale work beyond their spheres of influence—their grade level, department, or school.

Districts often select individuals to lead efforts without recognizing the importance of engaging district system components as well. This misstep is tempting because of the widespread recognition that teachers’ buy-in can help spread work to their colleagues. Districts may rely on teachers to spread new ideas without recognizing that they do not have the ability to set priorities and allocate resources in ways that move change past the early adopters; they may also select principals to pilot new ideas in schools without recognizing that principals do not have sufficient influence—much less authority—to spread the work to other schools in the district. Successful scale comes when district leaders strategically engage staff at all levels of the hierarchy in a way that coordinates grassroots supports and systemic action.

A critical step to sustaining work is embedding it in the structures and processes that make up district systems.

Too often, district leaders attempt to scale initiatives mainly through individuals—such as teachers or principals—rather than by engaging systems to support consistency in the adoption of new practices. Coburn noted that to scale improvements, it is critical that the work become institutionalized and part of the culture of the organization.8 Organizational culture is often described as “the way we do things here.” In Copper Mine, the district used the existing structure of the pacing guides to embed new math instructional approaches into all teachers’ classrooms on a regular basis. This made it possible to quickly begin to change “the way we teach math here.” This example provides helpful nuance for what it means for a district to lead instructional improvement. While top-down mandates alone typically face resistance, scaling instructional improvement requires steady support from top leaders who name an issue as a top priority, embed it in district structures and processes that engage every adult who needs to take on the focal practices, and allocate the resources necessary to support teachers to do it well.

To succeed in scaling instructional improvement across a district, leaders should start by analyzing their existing systems.

Identifying the relevant components of a system is itself an important element of a district’s capacity for improvement. Articulating the full set of existing system components can help a leader make strategic decisions about which levers to pull and where new structures or processes will need to be built to accomplish specific goals. Ed Partners works with districts to identify existing system components—such as meetings, classroom walkthroughs, and pacing plans—as likely starting places for embedding new approaches prior to attempting to create new structures or processes from scratch. Given the multiple demands on educators’ time and the constraints (especially for collaborative time), new structures and processes often struggle to gain traction unless they fulfill a vital need or build support through improved results.

Context matters in figuring out how to scale improvement, and because the path is uncertain, monitoring is critical for adaptation and success.

Ed Partners works with districts that range in size, assets, and needs—all of which affect efforts to scale improvement. As this brief has argued, structures and processes play a pivotal role. In our analysis of the data across the years, patterns emerged. As the number of units increases—with units being classrooms, grades, departments, or schools—the complexity of scaling also increases. Larger districts with more schools generally face more challenges when trying to scale improvements districtwide. The separation across disciplines in secondary schools, especially high schools, is typically greater than the separation across grade levels in elementary schools. The fact that elementary teachers teach the same subjects and often change grades over the years creates informal ties across grade levels; in high schools, the opposite conditions are present, making scaling across subjects more challenging. In general, the more that units are siloed, the harder it is to scale. Districts also have unique cultures and norms that affect efforts at systemic instructional improvement, such as how much teachers collaborate with one another and work with administrators on instruction. Finally, the extent of change desired and the resulting need for capacity building affect efforts to scale. For all these reasons, there is no clear playbook or timeline that can be expected when attempting to scale instructional improvement. The uncertainty requires continuous cycles of monitoring, reflection, and refinement to keep goals clear and relevant, provide sufficient supports, and adapt over time—all of which are made possible when district leaders to listen to individuals involved in the change at every level.

Conclusion

This brief illustrates what an instructional “system” is in a district as well as some of the key structures and processes that make up an instructional system. The pattern we see as we look across data from many districts is that scaling instructional improvement requires engaging the system; district leaders, whose roles allow them to allocate system resources and set goals, are key to engaging a system in support of scaling improvement. Leaders at every level—classroom, school, and district—all have important roles to play. We find that districts’ most senior leaders are often not sufficiently involved. Because senior leaders have the authority to reallocate system resources—such as teacher time and funding—as well as the ability to set a district’s vision and preserve attention on a steady, small set of the most important priorities, their direct involvement in supporting scaling of instructional improvement is essential.

• 1

  Coburn, C. E. (2003). Rethinking scale: Moving beyond numbers to deep and lasting change. Educational Researcher, 32(6), 3–12. doi.org/10.3102/0013189X032006003

• 2

  Cottingham, B. W., Gong, A., & Gallagher, H. A. (2019, October). A student-centered culture of improvement: The case of Garden Grove Unified School District [Report]. Policy Analysis for California Education. edpolicyinca.org/publications/student-centered-culture-improvement-case-garden-grove-unified-school-district

• 3

  Cottingham, B. W., & Gallagher, H. A. (2024, December). Scalable systems for adult capacity building: Translating Sierra House Elementary School’s pilot practices across Lake Tahoe Unified School District [Practice brief]. Policy Analysis for California Education. edpolicyinca.org/publications/scalable-systems-adult-capacity-building; Gallagher, H. A., & Cottingham, B. W. (2024a, December). Building momentum one step at a time: Grass Valley School District’s progress towards P–3 coherence in literacy [Practice brief]. Policy Analysis for California Education. edpolicyinca.org/publications/building-momentum-one-step-time

• 4

  Gallagher, H. A., & Cottingham, B. W. (2024b, December). Taking reform to scale: Learning from California Education Partners’ collaborations [Practice brief]. Policy Analysis for California Education. edpolicyinca.org/publications/taking-reform-scale

• 5

  Gallagher & Cottingham, 2024b.

• 6

  Rogers, E. M. (2003). Diffusion of innovations (5th ed.). Free Press.

• 7

  Gallagher, H. A., Cottingham, B. W., & O’Meara, K. (2022, July). Generating traction with continuous improvement: Lessons from two learning networks. Policy Analysis for California Education. edpolicyinca.org/publications/generating-traction-with-continuous-improvement

• 8

  Coburn, 2003.