Now That You Know, What Will You Do?

The superintendency is among the most structurally difficult positions in American public life. It is not difficult primarily because of the work — though the work is genuinely hard — but because of the incentive architecture surrounding it. A superintendent serves at the pleasure of a board, operates within a political community, manages a workforce whose professional autonomy is protected by law and by culture, and is held accountable for outcomes that are shaped by forces almost entirely outside their control.

The average superintendent tenure in the United States is approximately three years — less in urban districts. Research on superintendent job loss finds that failure to raise student achievement is less likely to cause a superintendent's departure than poor interpersonal relationships with board members or community stakeholders. The political incentive structure does not reward instructional courage. It rewards relational management. And a leader who has absorbed that lesson — who has watched a colleague pushed out for a decision that was educationally sound but politically costly — has learned something accurate about the system they operate in, even if what they have learned is the wrong lesson for the students they serve.

Understanding this is not the same as excusing it. The structural difficulty of the position is real, and this paper is not interested in adding to the weight of a job that is already genuinely hard. But the research is also clear that the leaders who have produced durable instructional improvement — the ones whose names appear in the Wallace Foundation studies, the RAND principal pipeline research, the Bryk and Schneider longitudinal work — did not do it by waiting for political conditions to improve. They did it by building the conditions that make improvement possible, starting with the levers already within their authority, and accumulating the credibility that makes the next lever accessible.

This paper is about those levers. Not the ones that require a board supermajority, a legislative change, or a negotiated contract provision. The ones that belong to the superintendent right now — in the budget, in the calendar, in the placement decision, in the data meeting — and that the research identifies as having the largest effect on student outcomes.

The gap this paper is trying to close is not between knowing and caring. Most of the leaders this paper is written for care deeply about the students in their buildings. The gap is between caring and acting — between accepting the evidence and acting on it. That gap has a name in the research literature. Pfeffer and Sutton called it the Knowing-Doing Gap. Argyris and Schön called it the distance between espoused theory and theory-in-use. Heifetz called it the difference between technical problems and adaptive challenges. Whatever name it carries, it costs students every day it remains open.

American school districts have not lacked for examination. The components of the instructional system — curriculum adoption, teacher evaluation, professional development design, assessment selection, leadership preparation — have each been studied, debated, reformed, and re-reformed across decades of sustained policy attention. What has not become standard practice is the examination of the relationships between those components: whether the curriculum is actually being taught, whether the professional development is aligned to the curriculum, whether the leadership development is building the capacity the curriculum requires, and whether the data system is providing the feedback the whole chain needs to improve.

This distinction matters enormously. A district can have a high-quality curriculum adoption and a well-designed professional development calendar and a rigorous evaluation framework and produce no measurable improvement — if those elements are not connected to each other in a coherent system aimed at the same instructional outcomes. Fullan and Quinn named this phenomenon directly: districts that struggle are riddled with incoherence — mismatched strategies, competing cultures, and initiatives that cancel each other's effects before they reach a student.

The research on district coherence is specific about what a whole-system examination reveals and what it does not. It is not an audit designed to find fault with teachers, identify failing schools, or produce a ranked list of underperforming staff. It is a structured examination designed to answer a different set of questions: Do district goals connect to site goals? Do site goals connect to professional development? Does professional development connect to what is visible in classrooms? Does all of it connect to outcome data? When the chain breaks — and in most districts it breaks in multiple places — the examination identifies where, and the repair work can begin.

The examination also reveals something that most leaders do not initially expect: a superintendent's discretionary authority is considerably larger than most superintendents act as if it is. Sarason observed decades ago that leaders consistently perceive more barriers to change than actually exist — and that this perceptual inflation of constraint is itself one of the most powerful mechanisms keeping systems in place. The five levers in Section V of this paper are all within existing superintendent authority. None requires legislative action. None requires a board supermajority. None requires waiting.

What most leaders lack — and what political incentive structures consistently fail to supply — is an objective, ego-neutral picture of where the system is actually breaking. This is where AI enters not as a technology initiative, but as an analytical instrument. AI's ability to examine large volumes of evidence without institutional loyalty, without the need to protect relationships, and without the human tendency to explain away inconvenient findings — is precisely what a whole-system coherence examination requires.

Human examination of instructional systems is not unreliable. It is, however, shaped by the same political incentive structures this paper has already described. A central office administrator reviewing curriculum alignment has a stake in what that review finds. A principal observing classrooms has relationships with the teachers being observed. A consultant brought in by leadership has an interest in the client's continued confidence. None of these are failures of character. They are predictable features of how human beings operate inside institutions — and they are the reason that self-examination, however well-intentioned, consistently underestimates the depth of misalignment it is looking at.

What the examination produces is not a judgment. It is a map — a structured, evidence-based picture of where the instructional chain from district vision to classroom practice is intact and where it is broken.

Which essential standards are actually being taught across all classrooms, and which exist only in the adoption paperwork. Whether the professional development calendar is building the capacity the curriculum requires or consuming teacher time on initiatives disconnected from any instructional priority. Whether principals are observing instruction through a shared framework or through seventeen different personal lenses. Whether data is traveling up and down the system in ways that produce instructional decisions, or only in the direction that produces compliance reports.

The research on district improvement does not leave this question open. There is a substantial body of longitudinal evidence — from multi-year studies of hundreds of schools across multiple states — identifying the specific conditions that distinguish districts and schools that improve student outcomes over time from those that do not. The pattern that emerges from that evidence is not complicated. It is, however, consistently ignored.

The most comprehensive longitudinal study of school improvement in the United States was conducted by the Consortium on Chicago School Research. Beginning in 1990 and spanning fifteen years, researchers tracked the internal workings and external conditions of more than 400 elementary schools — identifying 100 that substantially improved student learning in reading and mathematics over a seven-year period and 100 that did not. The two groups of schools were drawn from the same city, served similar student populations, and operated under the same policy environment. What distinguished them was not funding, not demographics, and not the intensity of their reform efforts. It was how those efforts were organized.

Bryk and colleagues identified five essential supports for school improvement: a coherent instructional guidance system, the professional capacity of the faculty, strong parent-community-school ties, a student-centered learning climate, and leadership that drives change. The finding that matters most for the argument this paper is making is not any one of those supports in isolation. Schools strong in three to five of those essential supports were ten times more likely to improve student learning substantially compared to schools weak in three to five of those supports. Strength in one or two supports, without the others, produced no reliable improvement. The supports are interdependent.

Cobb and colleagues at Vanderbilt arrived at the same conclusion from a different entry point. Their eight-year research-practice partnership with four large urban districts produced a theory of action for advancing instruction at scale that centers on three things: the elements of a coherent instructional system, productive practices for school leaders in supporting teachers' growth, and the role of district leaders in developing school-level capacity. Teacher effectiveness, the variable most strongly associated with student outcomes, is not a fixed property of individual teachers. It is significantly shaped by the system surrounding them. Teachers in coherent systems improve. Teachers in incoherent systems plateau or leave.

The RAND Corporation's 2024 analysis of instructional system coherence adds an important dimension to this picture. More-coherent instructional systems supported teacher confidence, while incoherence evoked frustration and anxiety. This is not a peripheral finding. It is the mechanism through which incoherence produces the conditions that drain teachers — conditions that districts consistently misattribute to the difficulty of the work itself. When teachers receive conflicting guidance about what to teach, how to teach it, and what counts as success, the burden is not the work. The burden is the noise.

The pattern across all of this evidence converges on five specific practices that appear repeatedly as distinguishing markers of improving districts and schools: narrowing the instructional focus to a genuine guaranteed and viable curriculum; providing teachers with clear curricular guidance; aligning professional learning to that guidance; building principal and teacher-leader calibration; and using data to adjust instruction rather than to justify new programs.

What these practices have in common is that none is a program, a curriculum, a platform, or an external intervention. All are organizational decisions about how the system is structured, aligned, and aimed. They are available to every district leader right now — not as aspirational goals requiring years of preparation, but as the documented work that separates improving districts from those that don't improve. The districts that did this work did not do it by accident. They did it by deciding that coherence was a leadership responsibility — not an emergent property of good intentions, and not something a single exceptional teacher or principal could produce in isolation.

Every system that has survived for more than a generation has developed mechanisms for resisting examination. This is not a conspiracy. It is an organizational pattern so well-documented in the change literature that it has its own name: the institutional immune response. Fullan, Heifetz, and Argyris each describe it from different theoretical angles and arrive at the same conclusion: institutions instinctively protect themselves from the discomfort that honest self-examination produces.

The protection of professional identity.

Teaching is not a job most educators hold lightly. For many, it is a defining identity — and a whole-system examination, however carefully positioned, carries an implicit question about whether the work has been good enough. A teacher who has built an identity around what they do in their classroom experiences a whole-system examination as a potential threat to that identity, even when the examination is not evaluative. This is not a character flaw. It is a predictable response to a real vulnerability, and it requires a leader who can separate the examination of system coherence from the evaluation of individual performance — clearly, consistently, and with credibility.

Structural resistance through collective agreement.

Collective bargaining agreements define the boundaries of administrative discretion in most public school districts, and those boundaries are real. What is also real — and what many leaders underestimate — is that the examination of instructional coherence, the alignment of professional development to an instructional framework, and the strategic deployment of teacher leadership are all within administrative discretion in the vast majority of agreements. The structural resistance that collective agreements create is often less about what the agreement actually prohibits and more about what leaders assume it prohibits without testing that assumption.

The micropolitics of community pressure.

Community pressure functions as an immune response mechanism when it is organized and directed at instructional change. The most effective antidote is not better communication strategy but the prior establishment of a broader, more representative community coalition whose stake in the district's instructional direction was built before the opposition organized.

Institutional authority in service of institutional preservation.

The most consequential immune response mechanism is the one most rarely named directly: the use of institutional authority — at the board level, at the central office level, at the site level — to protect the institution from examination rather than to serve students. Heifetz names this pattern as the single most common failure mode in organizational leadership: the use of positional authority to manage down adaptive challenges rather than to mobilize people to face them. A leader who uses their authority to delay, deflect, or defund the examination of whether the system is serving students has inverted the purpose of their position.

The following five levers are drawn from the research on superintendent and principal effectiveness, district-level instructional leadership, and the implementation science of sustainable school improvement. Each is within existing administrative authority. Each is sequenced in the order most likely to build the credibility and system conditions that make the next lever possible. None of them adds to the burden teachers already carry. Each, executed well, addresses the conditions that created that burden — the misaligned expectations, the initiative overload, the absence of a guaranteed and viable curriculum that gives a teacher's effort somewhere coherent to land.

Every school district in the United States is operating under a federal mandate. The Every Student Succeeds Act requires local education agencies to meaningfully involve parents and family members in educational planning and school improvement. Not superficially. Not through a newsletter. Not through a back-to-school night that surfaces the same twelve parents who always attend. Meaningfully — which the research defines with enough specificity to make clear that most districts are not meeting the standard they are required to meet.

This matters to an instructional argument because family engagement is not a community relations function sitting adjacent to the work of teaching and learning. Research across 200 public elementary schools in Chicago found that schools with strong family and community engagement were ten times more likely to improve student learning outcomes — and that a sustained weakness in family engagement alone was sufficient to stall improvement even when every other school support was in place. The return on genuine family engagement is documented, replicated, and large enough that no superintendent can reasonably treat it as a secondary concern.

The gap between the federal mandate and operational reality is structural. Most district family engagement programs are built around informing families about decisions already made rather than involving them in the decisions themselves. Henderson and Mapp's comprehensive research review distinguishes sharply between programs that treat family engagement as communication — announcements, newsletters, attendance at events — and programs that treat it as partnership, defined as genuine shared decision-making about what a quality education looks like and how to build the conditions that produce it. The former keeps families at arm's length from the work. The latter makes them part of it.

Researchers call for a specific structural decision: creating a senior-level position dedicated to family and community engagement that reports directly to the superintendent, treating this work as a pillar of instructional improvement rather than an activity conducted after the important work is done. As Karen Mapp of the Harvard Graduate School of Education has stated plainly: it is real when it is on the budget sheet.

One of the most reliable ways to abandon work that would have succeeded is to evaluate it before it has had time to produce results. The research on district improvement is specific about timelines, and those timelines are longer than the political incentive structures described in Section I are comfortable with. A superintendent who begins this work in year one and expects to see measurable outcome gains in year one has not read the research. A board that evaluates an improvement initiative by its first-year test scores has not read the research either. The appropriate professional response is not to hide this from either of them — it is to name it clearly, early, and with the evidence in hand.

Year one: process change.

What year one looks like in improving districts is not outcome improvement. It is process change. Teachers can articulate the essential standards they are responsible for teaching. Professional development is visibly connected to those standards. Principals are conducting walkthroughs through a shared instructional framework. Data meetings are producing documented instructional decisions. These are the leading indicators — the structural shifts that the research identifies as necessary preconditions for the outcome changes that follow.

Year two: early signal.

Formative assessment data begins to reflect the curriculum students are actually receiving. Teacher confidence in the instructional framework increases. Principal feedback becomes more consistent and more connected to what teachers were professionally developed to do. The moratorium on new program adoption has protected the focus long enough for initial implementation to take hold.

Years three through five: outcome data.

What years three through five look like — in districts that hold the focus — is the outcome data the research promises. The Bryk longitudinal study identified its strongest improvement evidence in the fourth and fifth years of coherent, sustained organizational work. Cobb and colleagues' eight-year partnership produced its most significant student outcome findings after districts had sustained the coherent support structures long enough for them to become organizational habits rather than initiatives.

This timeline is not an excuse for slow leadership. It is a description of how improvement actually works — and a direct challenge to the strategic planning cycles, leadership transitions, and board election schedules that consistently interrupt coherent work before it can mature. The political case for this work is not "trust us, the scores will come." It is: "here are the leading indicators we committed to in year one, here is the evidence we have met them, and here is what the research says happens next if we hold the course."

This paper has made uncomfortable arguments. That is intentional. The research supporting those arguments is cited, verifiable, and not going away. But the people most positioned to act on them are also the people most accustomed to professional environments where difficult findings are softened before they reach the room. This section does not do that.

There is a moment every educator remembers — the moment they understood, with complete clarity, why this work matters. It is usually not a policy brief or a professional development session that produces it. It is a student. A specific child, in a specific classroom, on a specific day, whose potential was either honored or diminished by the decisions the adults around them had made. That moment is why people enter this profession. It is why teachers stay in difficult buildings, why coaches spend their careers trying to move practice one conversation at a time, and why the best superintendents carry the names of students in their heads when they sit in rooms full of adults arguing about adult priorities.

This paper has been written for the leaders who still carry those names.

Paper 1 asked a question that this profession has been avoiding for a long time: what is this system actually for? The evidence it presented was not ambiguous. The system sorts. It stratifies. It produces outcomes that are predictable not by student potential but by zip code, by economic circumstance, by the accident of which teacher a child happened to draw on the first day of school. That paper ended with a challenge — now that you know, what will you do?

This paper has answered that question as honestly and specifically as the research allows. Not with inspiration. Not with framework language dressed as action. With evidence from fifteen years of longitudinal research identifying what improving districts actually did. With five specific levers, each grounded in evidence, each within a superintendent's existing authority, each sequenced in the order most likely to produce visible results. With a clear-eyed account of what the institutional immune response will look like and how to move through it. With a realistic picture of what to expect in year one versus year three.

The argument of this paper is not that the work is easy. It is not. The argument is that the difficulty of the work does not change its necessity.

Every year a district operates without examined instructional coherence is a year students receive whatever instruction they happen to receive from whatever teacher they happen to have in whatever building happens to be in their neighborhood — rather than the guaranteed, viable, aligned instructional experience the district claims to be delivering. Every year a professional development calendar goes unaudited is a year teachers absorb the cost of misaligned initiatives in time, energy, and professional trust. Every year a principal is placed in a school on the basis of something other than instructional leadership capacity is a year the students in that building pay for the decision.

These are not hypothetical costs. They are the accumulated, documented, measurable costs of a system that has been treating adaptive challenges as if they were technical problems — and then expressing surprise when the solutions do not hold.

The leaders who changed things did not wait for the right moment. They understood that the right moment is not a feature of the system — it is something a leader builds, deliberately and sequentially, through the accumulation of honest decisions made in the direction of students. They absorbed the cost of being temporarily unpopular. They held the focus when the conversation tried to drift toward the comfort of the adults in the room. They did this not because they were exceptional people free of political pressure, financial anxiety, or professional risk. They did it because they had looked clearly at what the system was producing and had decided that the students in their buildings deserved a different answer.

That decision is available to every leader reading this paper. It does not require a perfect board, a cooperative union, a resolved political climate, or a budget surplus. It requires the willingness to look at the system honestly and respond to what the examination reveals — starting with the levers that are within reach, building the credibility that makes the next lever possible, and staying anchored to the students who cannot advocate for themselves in the rooms where these decisions get made.