Central Office Transformation for District-Wide Teaching and Learning Improvement

Click here to download the full report:
 Central Office Transformation for District-Wide Teaching and Learning Improvement

Research on learning underscores that people, such as the instructional leadership directors or ILDs, help others such as school principals improve the quality of their work not only when they assist the learners with those work practices but when they, too, receive assistance with their own work. Some learning theorists call this dynamic nested assistance relationships to capture the importance of assistance providers getting help with their own work assisting others (e.g., Tharp & Gallimore, 1991). Accordingly, we analyzed our data for evidence of support the ILDs received that seemed to matter to how they worked with principals on their instructional leadership. Four supports, summarized schematically in Figure 3, were promising in this regard.


More and less promising supports for ILDs’ work reflected the extent to which other central office administrators (1) engaged the ILDs in challenging conversations about their own work with individual school principals and how to improve the quality of that work, or (2) helped ILDs maximize the time they spent on support for principals’ instructional leadership. Here, we put together observations of central office administrators’ work with ILDs’ descriptions of these matters. To establish that the ILDs were indeed maximizing time with principals, we asked principals about the frequency of their interactions with their ILDs and reviewed a random sample of ILDs’ calendars, as part of interviews, three times over the course of the study period. While we were not able to associate particular supports for ILDs with specific assistance practices described in Chapter 2, the following four supports seemed essential to helping ILDs work with principals on their instructional leadership at even a basic level:

  • Professional development for ILDs that provided regular opportunities for challenging conversations about the quality of their work with school principals and how to improve it.
  • Taking issues and other demands off ILDs’ plates, thereby freeing up their time to work with principals on principals’ instructional leadership.
  • Leading through—rather than around—the ILDs, and otherwise supporting the leadership of ILDs, vis-à-vis principals’ instructional leadership.
  • Developing and using an accountability system in which ILDs did not act as the sole agents holding principals accountable for improvements in student performance.

Professional Development for Instructional Leadership Directors (ILDs)

Central office administrators in all three districts brought ILDs together in meetings ostensibly to help strengthen the ILDs’ work with individual school principals.15 In Atlanta these occasions happened as part of weekly meetings with the deputy superintendent, and in New York City and Oakland, ILDs came together twice each month in meetings at least partially dedicated to support for ILDs’ practice. These meetings, among other forms of professional development, gave ILDs a forum for examining their work with principals, and considering how to improve it.

However, the time other central office administrators actually dedicated to ILDs’ professional development was sometimes shortened or completely interrupted when other central office administrators shifted meeting agendas to address other, usually operational issues. Nor did all the professional development opportunities engage the ILDs in challenging conversations about the quality of their work with school principals and how to improve it. Not surprisingly, when their professional development time was shortened or not characterized by such challenging conversations, ILDs tended to report that the professional development they received did not help them improve the quality of their work with school principals.

New York City/Empowerment Schools Organization (ESO) stood out among our three sites for the amount of time actually dedicated to network leaders’ professional development and specifically to challenging conversations among them about how to improve the quality of their work with individual school principals. We observed almost 100 hours of twice-monthly network leader meetings that typically featured a significant amount of time on what some referred to as network leaders’ “inquiry.” Sometimes these inquiry sessions focused on ideas senior central office administrators brought in from the outside, as when staff dedicated over an hour to network leaders’ engagement with a school principal about how he was able to change his schools’ culture in ways supportive of gains in student achievement, or when ESO leaders hosted a retreat for half their network leaders with organizational change expert Peter Senge. More frequently, the staff provided network leaders with a prompt related to how they were handling certain situations, such as making accountability demands meaningful to principals or getting principals to trust their feedback.

For example, in the following exchange, a small group of network leaders responded to a prompt about one challenge they were facing in their work with school principals and how to address it. Network leader 1 introduced the problem of how to help principals understand what it means to exercise instructional leadership.

    Network Leader 1: How can I make that [focus on principals’ instructional leadership] actionable?

    Network Leader 3: [What I look at is] what are the different opportunities I can have with [principals] that will gain their trust, so I can have some influence? If that’s what your goal is. I think you almost have to go backward and have a set of experiences that aren’t so high stakes, so that when high stakes experiences arise, [the principals] trust you.

    Network Leader 2: But there’s no guarantee. But I hear underneath what you’re saying that you want some assurance [that the principal is going to engage in work with you on improving their instructional leadership]. The desire to be influential is partly rooted in belief that when we are confronted with situations we offer what we know is right, and when [principals] don’t take [our] advice, we are hurt.

    Network Leader 1: … I’m worried about [principal’s name] and his school … .

    Network Leader 3: What’s the work with this person? Because it seems to me this is someone who needs to feel he comes to decisions himself. How can you frame this so he feels he is coming to decisions on his own … . You have to take a tack … so you can meet [principals where they are] …

    Network Leader 1: Have you gained that respect from your principals in a year and three months?

    Network Leader 3: The goal is not so much about us gaining credibility, but about their development. It’s not really about us.

This exchange is typical of the conversations we observed in its focus on a specific challenge these ILDs were facing and their grappling with how they might take action on those challenges. New York had staff specifically dedicated to professional learning for the network leaders and other ESO staff. One of the staff reported that the job included “to continually protect the space of the network leader meetings for them to talk … [about their own learning].”

Some network leaders, especially those who had been in the original Autonomy Zone, reported that they did not see the need for the meetings and that they just wanted to work with their team members and their schools. However, the majority of network leaders we interviewed typically corroborated the value of these conversations, especially given the dynamic and largely uncharted nature of their work. For instance, one reported,

    I think [the network leader meetings] have contributed to how we operate and understand the role that I’m performing now. We continue to have clarity, although it changed from time to time. We were looking at being facilitators but we were also being looked at as being knowledge-able educators to influence decisions of principals.

Half the network leaders we interviewed reported that they facilitated such conversations for themselves by meeting with their colleagues either during or outside of network meetings. Accordingly, we viewed these ILDs as main agents of their own professional support. Similarly, in Atlanta at the time of our data collection, School Reform Team (SRT) executive directors convened in monthly meetings to share ideas and materials that they found useful in their work with principals. As one executive director reflected, “We meet once a month … . We all meet and talk about issues that are common to us and we share some ideas so that’s helpful … . I always need some learning supports [like those].” These and other executive directors emphasized that occasionally these meetings involved not simply the sharing of materials but extended conversations about how they were actually working with their principals.

By contrast, the biweekly meetings in Oakland that we observed infrequently focused directly on NExOs’ work with principals on their instructional leadership. On a few occasions, the meeting facilitator led these ILDs through a “consultancy protocol” that prompted one NExO to present the case of one school for feedback from ILD colleagues. However, those presentations tended to focus on the performance of the school with some discussion of the performance of the principal, and remarkably little discussion of how the NExO might support the principal. The short time for conversation allowed in that particular consultancy protocol curtailed critical conversations about the ILDs’ work with their schools, a sharp contrast to the more flexible, open-ended discussions among network leaders in New York City.

The NExOs received feedback on their work from senior central office staff mainly in the form of annual performance evaluations and through comments on how they completed various tools and rubrics. Our reviews of the tools and rubrics suggested they addressed the topic of NExOs’ work with and their knowledge of schools but did not penetrate deeply into matters of their practice. One senior central office staff member in Oakland described supporting NExOs this way:

    … they’ve been working on drafts [of their plans to support principals] and then I’ve been giving feedback on the drafts … . I’m basing my feedback on can they show the integration of the strategic practices, how does it connect to raising student achievement and meeting the student achievement targets, how does it leverage the small network model … and then how are they including mandatory elements in the plan … . And they are investing significant time and really building out their plans in much more detail, which is because they should be putting that kind of thought into it.

NExOs typically reported that this emphasis on documenting on paper how their plans for principals aligned with district priorities significantly took their time away from work in schools and did not contribute positively to the quality of that work. Perhaps as a result, the elementary school NExOs eventually met on their own to share their work and plan jointly together. Likewise, they convened themselves with an outside facilitator to receive professional support.

Taking Issues and Competing Demands off ILDs’ Plates

Though ILDs were supposed to focus their work on supporting principals’ instructional leadership, as mentioned earlier, various demands and distractions impinged on ILDs’ time and took them out of schools and away from work with principals. The central office–principal partnerships were assisted by intentional efforts of central office administrators, including ILDs themselves, to remove responsibilities or tasks that would make it harder for ILDs to maximize the time they spent helping principals exercise instructional leadership.

In Atlanta, the SRT executive directors frequently credited various central office staff, including their deputy, with helping them in this regard. As one recounted,

    … We had blackout days, right, and the blackout days were equivalent to one and a half days a week. And the blackout means that you don’t pull principals, you don’t pull school executive directors … because people are in schools working. And so … the school EDs asked for that time to be increased and it was increased to two and a half [blackout days per week], and basically our position was it’s a very poor commentary if this is our core business and we are only having blackout for less than half of the time [in the workweek]. And so [the deputy] was like, ‘You’re absolutely right. Two and a half days.’

Senior central office administrators were particularly instrumental in protecting ILDs’ time for work with principals when they reduced demands on ILDs that either threatened to consume too much time or that otherwise did not promise to strengthen ILDs’ support for principals’ instructional leadership. For example, one such central office administrator in Atlanta reported,

    I know I make a special effort when EDs call me or … when principals call or the ED has a principal call me … I try to make sure they get what they need as quickly as they can, because the bottom line is providing service to schools. That’s it. That’s it.

Another senior central office administrator staff person in Atlanta reported that the job of the executive director is to support schools and get schools what they need, but when these ILDs run up against barriers “and when they’ve exhausted … everything, they come to me and I take care of it.” Strikingly, every single network leader we interviewed in New York City repeatedly reported that any time they brought a challenging issue to the attention of senior staff people in the Empowerment Schools Organization, these staff either provided information that was needed to expedite the issue or handled the issue themselves. All these ILDs reported that these efforts on the part of senior staff helped them increase the time they spent working with school principals on their instructional leadership.

By contrast, some Oakland NExOs reported canceling meetings with individual principals, especially in the spring, due to personnel hearings or “fire drills,” urgent meetings back at the main central office building generally not related to their efforts to support principals’ instructional leadership. NExOs also reported having much of their time consumed by “homework assignments” from the senior central office staff, such as templates for recording all their schools’ assessment data and data on professional learning communities and family engagement, among others. One reported that, as a required part of the new accountability system, they had to give multiple presentations on their schools to executive staff during the fall of our data collection period because so many were on the list of low performing schools. When asked how, if at all, such “red school presentations” related to their own work with principals on instructional leadership, NExOs generally reported that the presentations kept them in compliance with the accountability system but did not contribute to their work with principals.

ILDs also took the initiative to protect their own time. As an Atlanta executive director reported, “I’ve learned to say no when other demands threaten my time with principals.” NExOs in Oakland likewise reported that they committed to spend at least 75 percent of their time working with school principals, and would also collectively resist attempts by district senior staff to load them up with other demands or take over too much of their twice-monthly meeting time for issues not related to improving principals’ instructional leadership. In all three districts, ILDs protected their own time, in part, by delegating certain matters not directly related to instructional leadership to their staff or other central office administrators who assisted them. Even in Oakland, where NExOs did not have team members like their counterparts in the other two districts, several reported, and observations confirmed, that they delegated certain issues to their administrators-on-specialassignment or to staff of other units who agreed to work with their networks in ways that freed up their time to focus on principals’ instructional leadership. However, some also reported not saying no often enough when other central office administrators placed demands on their time, for example, to review the district’s plan for getting off Program Improvement status. Some NExOs admitted they wanted to have influence over decisions within the central office and, in order to do so, they had to attend various planning meetings even though those meetings lessened the time they spent with principals.

Leading Through, Not Around, the ILDs

Another key support for ILDs included what we call “leading through the ILDs.” That is, based on a shared conception of the nature of the ILDs’ role vis-à-vis the principal, other central office administrators and board members did not circumvent the ILDs and work directly with principals, but through their actions reinforced the ILDs’ essential role in supporting principals’ instructional leadership development. ILDs generally pointed to such efforts as essential to their ability to work well with school principals, in part, by increasing their confidence that they were trusted and valued. When other central office administrators did not engage in such activities, ILDs generally reported spending time dealing with the resulting confusion and noted how the lack of reinforcement for their work undermined their relationships with some school principals.

One senior central office staff person in Atlanta described the importance of leading through, not around, the ILDs in this way,

    The cultural shift [involved in central office transformation is] for central office to become facilitators of the schools. Well, actually the … shift now is more of the central office supporting the EDs [executive directors] of schools, who in turn will provide support for the schools. Prior to that you had the EDs doing stuff for schools, you had central office doing stuff for schools, and in many instances they were both doing the same thing. You might have professional development doing professional development for principals, and the EDs doing professional development for the principals … . So you get kind of confused.

As a result, principals did not know whom to call when they needed professional support, and that this lack of clarity occasionally undermined the executive directors’ efforts to support principals. This respondent and others in Atlanta credited consensus that the executivedirectors’ role was to be the main point person for principals as fundamental to supporting the executive directors’ work and the central office’s ability to support principals in coordinated ways.

In Atlanta, executive directors frequently reported that senior central office staff trusted them to do their work, and that trust provided further evidence of “leading through the ILDs”. For example, when we asked what if any supports they received for their work with principals, one described in a typical comment,

    What I really enjoy is that the leadership trusted [me]. And my leadership that allowed me to do the work I needed to do without [their] hanging all over me and just expected the result. That’s all I really need.

Describing a key lesson they learned over time, two senior staff people noted that they cannot do everything and that effective executive leadership in a system that supports teaching and learning improvement helps point people, the executive directors, do their job well. In contrast, early in the implementation of the Atlanta transformation process, principals went directly to senior staff. Now, principals are sent back to an executive director, as one executive director noted:

    If principals go to [a senior staff person] … she] tells them things like ‘come to me’. I know it’s my job for them not to come to her because if they go to her I haven’t done what I am supposed to do. She will always send them back here generally … . I’ll tell you that’s the model from the top … None of us have time to get into the business of dealing with issues that other departments are existing to handle. So it’s a model from the top. This system is probably the most protocol and communication-driven system I’ve ever seen in my life. … I think before Dr. Hall got here there was a lot of monkeying around with the staff and board members … They tell me board members would come in, tell principals what to do, directing principals …

This ILD went on to describe how that kind of dynamic with board members has been significantly minimized in Atlanta through greater clarity about the executive directors’ role and consistent modeling by senior staff that the central office (and board) leads through the executive directors in supporting principals. This person reflected that there’s no memo describing the chain of command but “it is understood.”

ILDs obviously struggled when other central office administrators went around them or dealt directly with principals in these ways, further reinforcing how such challenges may throw up typical roadblocks for districts early in the implementation of a central office transformation effort. In one example, corroborated by multiple interviews and observations, senior central office staff in Oakland advised NExOs that several of their schools should be closed due to poor performance. The NExOs made a public presentation to the board with those recommendations and were met with resistance by community members, a typical response to decisions to close schools. Several weeks later the acting state administrator announced with little explanation that the schools would remain open. As NExOs saw it, such apparent “reversals” created confusion about their role and significantly undermined their work with principals. Several NExOs reported that they were often confused themselves about the nature of their responsibilities. In a typical comment, one reported, that when hired

    We were told we would be like the superintendent of our own little district and the word that … [was] used … . [Then later, senior staff said] ‘I know we told you all you were like area superintendents or assistant superintendents, but you are not. That is not how it turned out to be.’

This person went on to describe that when they were hired they were “running our own show” with principals, and that at the end of the year senior staff would look at their results and either keep them or fire them. However, over time, their job shifted so that at the time of the interview they described their job as one in which senior staff were “constantly bombarding us with homework assignments.”

By the end of our data collection, there was still a palpable ambivalence about the role of NExOs, including significant debates about whether or not Oakland should continue to have staff serve in this role, especially given the reportedly high cost of the model.

The System, Not Solely the ILDs, Holding Principals Accountable for Improving School Performance

ILDs also occasionally had their time consumed by participation in the systemwide accountability system, even in New York City where network leaders did not formally evaluate school principals. As noted above in the discussion of tools, ILDs sometimes participated in the accountability system in ways that focused on strengthening principals’ instructional leadership. In other instances, their participation seemed to interrupt or compromise the time they spent with their school principals on instructional matters.

For instance, network leaders in New York City all unanimously reported that one essential support for their work with principals on their instructional leadership was their charge to support, and not to evaluate, principals. They indicated that, in effect, “the data” held principals accountable for their performance. (Formal responsibility for annual performance evaluations of principals fell to the state-mandated community school superintendents, who, as part of central office transformation, no longer oversaw sub-district central offices but retained the authority to formally evaluate principals). As network leaders saw it, the separation of personnel evaluation from support meant that they could focus on supporting principals in ways that benefited their relationships with the school leaders and their focus on principal capacity building. The arrangement also signaled to principals that the network leaders were a different form of central office support from that available prior to central office transformation. As one network leader described in a typical comment, their work in some respects resembled an instructionally oriented supervisory role under the old regional central office structure called local instructional superintendent. Now,

    The major difference … you’re not the supervisor. So we probably perform a lot of the same functions, but that tension or that layer of ‘I’m coming in and I’m your supervisor’ and that power—whether or not that person had the personality—that power is always there … . [Now] folks can really feel comfortable saying certain things … asking certain questions that might be … on the edge or pushing the envelope … Whereas in the past, they got roiled in firepower and judgment.

In the two other districts, ILDs were as unanimous that their responsibility to conduct formal performance evaluations of principals and otherwise participate centrally in activities designed to hold schools accountable for student performance benefitted their work with their principals. These ILDs reported that having such positional authority over principals helped them exert more influence over principals’ practice. However, in Oakland, time for supporting principals’ instructional leadership was significantly compromised by time spent on activities related to principal evaluation, such as doing site plan reviews and other activities with schools that might be closing or that were performing poorly. One central office administrator in Oakland confided that often principals call him for advice rather than their NExO. When asked why, this person reflected,

    I’m not their supervisor, so I can truly offer pure support and sometimes it’s just a confidential ear even where people will just say “Hey, I’m having a really hard time.” I mean it’ll take all forms. “I’m having a really hard time, can you just listen. I have this PD session later today, can you help me think of a warm-up or here’s my agenda, what do you think?’

This person elaborated that they thought the “power dynamic” of NExOs evaluating principals

    … gets in the way for a lot of people … . I just know it does, having done a lot of leadership training and work and research. I mean the whole notion of evaluation and supervision—while I do believe it can be done—I mean a supervisor can find that balance, it’s very difficult. And Oakland, one of the biggest challenges we have in this district is trust and there’s a lot of people, and that’s across the organization— teachers, staff people, classified, custodians—you name it. People have felt burned, and so the lack of trust. So when a NExO turns to someone and says “Look, I’m here to support you. I’m not here to ding you,” they’re like, “Well, but look at my evaluation last time—you totally dinged me.” And the NExOs say that about their bosses, and their bosses say it about the board, and the teachers say it about the principals.

However, in Atlanta ILDs did not report or demonstrate that holding principals accountable for results interfered with their ability to support principals. Based largely on central office administrators’ reports, we attributed this difference in part to evidence that the system ultimately held principals responsible for their low performance through mechanisms that did not consume executive directors’ time. As one explained these systemic supports for principal accountability,

    When I came here [to interview for this job] … I asked … “So what happens when you’ve got principals who are not performing?’ You know, [you] work with them, develop them, try to build capacity, [but] they just don’t have it. They’re the typical mentality of a manager—not leadership, no instructional skills, all about keeping the building clean and the kids quiet. What do you do with those people?” I think my question was “Where do they go?” And she [the executive central office administrator] looked at me and she said “They go home [i.e., they are fired]” I remember thinking, “They go home? They go home!”

A district-wide accountability tool called the “Balanced Score Card” functioned as a key device for distributing responsibility for principal accountability across the central office in Atlanta. The superintendent and executive cabinet used a set of common indicators to differentiate performance targets for individual schools. These indicators—including student performance data on tests and common district assessments, attendance and graduation data—were used to evaluate the work in individual schools and the nature of central office support for schools, and to reestablish the next round of targets based on evidence of change over time.


We argued in this chapter that particular supports for the ILDs engaged them in challenging conversations about the nature and quality of their work with school principals and helped them increase the time they spent with school principals specifically focused on principals’ instructional leadership. In short, the central office supported ILDs’ work with principals through particular forms of professional development, as well as efforts by the ILDs and others to remove distractions, protect their time for instructionally-focused work, and limit the burden of holding principals’ accountable for school performance when this burden did not promise to build principals’ instructional leadership capacity.

< < Previous | Next > >


15. Interestingly, none of the professional supports for ILDs that we documented in any of our sites focused on how ILDs engage their principals in networks. One New York City network leader told us that professional development sessions the year prior to our data collection had addressed how to convene principals in networks but we did not observe similar conversations during the subsequent year. All the professional supports we captured were geared toward informing or otherwise supporting how ILDs worked with individual principals.