The Turnaround Fallacy

Stop trying to fix failing schools. Close them and start fresh.

Video: Andy Smarick talks with Education Next about why the Obama administration needs to rethink its embrace of turnarounds and adopt a new strategy for the nation’s persistently failing schools.


For as long as there have been struggling schools in America’s cities, there have been efforts to turn them around. The lure of dramatic improvement runs through Morgan Freeman’s big-screen portrayal of bat-wielding principal Joe Clark, philanthropic initiatives like the Gates Foundation’s “small schools” project, and No Child Left Behind (NCLB)’s restructuring mandate. The Obama administration hopes to extend this thread even further, making school turnarounds a top priority.

But overall, school turnaround efforts have consistently fallen far short of hopes and expectations. Quite simply, turnarounds are not a scalable strategy for fixing America’s troubled urban school systems.

Fortunately, findings from two generations of school improvement efforts, lessons from similar work in other industries, and a budding practice among reform-minded superintendents are pointing to a promising alternative. When conscientiously applied strategies fail to drastically improve America’s lowest-performing schools, we need to close them.

Done right, not only will this strategy help the students assigned to these failing schools, it will also have a cascading effect on other policies and practices, ultimately helping to bring about healthy systems of urban public schools.

A Body at Rest Stays at Rest

Looking back on the history of school turnaround efforts, the first and most important lesson is the “Law of Incessant Inertia.” Once persistently low performing, the majority of schools will remain low performing despite being acted upon in innumerable ways.

Examples abound: In the first year of California’s Academic Performance Index, the state targeted its lowest-performing 20 percent of schools for intervention. After three years, only 11 percent of the elementary schools in this category (109 of 968) were able to make “exemplary progress.” Only 1 of the 394 middle and high schools in this category reached this mark. Just one-quarter of the schools were even able to accomplish a lesser goal: meeting schoolwide and subgroup growth targets each year.

In 2008, 52 Ohio schools were forced to restructure because of persistent failure. Even after several years of significant attention, fewer than one in three had been able to reach established academic goals, and less than half showed any student performance gains. The Columbus Dispatch concluded, “Few of them have improved significantly even after years of effort and millions in tax dollars.”

These state anecdotes align with national data on schools undergoing NCLB-mandated restructuring, the law’s most serious intervention, which follows five or more years of failing to meet minimum achievement targets. Of the schools required to restructure in 2004–05, only 19 percent were able to exit improvement status two years later.

A 2008 Center on Education Policy (CEP) study investigated the results of restructuring in five states. In California, Maryland, and Ohio, only 14, 12, and 9 percent of schools in restructuring, respectively, made adequate yearly progress (AYP) as defined by NCLB the following year. And we must consider carefully whether merely making AYP should constitute success at all: in California, for example, a school can meet its performance target if slightly more than one-third of its students reach proficiency in English language arts and math. Though the CEP study found that improvement rates in Michigan and Georgia were considerably higher, Michigan changed its accountability system during this period, and both states set their AYP bars especially low.

Though alarming, the poor record for school turnarounds in recent years should come as no surprise. A study published in 2005 by the Education Commission of the States (ECS) on state takeovers of schools and districts noted that the takeovers “have yet to produce dramatic consistent increases in student performance,” and that the impact on learning “falls short of expectations.”

Reflecting on the wide array of efforts to improve failing schools, one set of analysts concluded, “Turnaround efforts have for the most part resulted in only marginal improvements…. Promising practices have failed to work at scale when imported to troubled schools.”

Like Finding the Cure for Cancer

The second important lesson is the “Law of Ongoing Ignorance.” Despite years of experience and great expenditures of time, money, and energy, we still lack basic information about which tactics will make a struggling school excellent. A review published in January 2003 by the Thomas B. Fordham Foundation of more than 100 books, articles, and briefs on turnaround efforts concluded, “There is, at present, no strong evidence that any particular intervention type works most of the time or in most places.”

An EdSource study that sought to compare California’s low-performing schools that failed to make progress to its low-performing schools that did improve came to a confounding conclusion: clear differences avoided detection. Comparing the two groups, the authors noted, “These were schools in the same cities and districts, often serving children from the same backgrounds. Some of them also adopted the same curriculum programs, had teachers with similar backgrounds, and had similar opportunities for professional development.”

Maryland’s veteran state superintendent of schools, Nancy Grasmick, agrees: “Very little research exists on how to bring about real sea change in schools…. Clearly, there’s no infallible strategy or even sequence of them.” Responding to the growing number of failing Baltimore schools requiring state-approved improvement plans, she said, “No one has the answer. It’s like finding the cure for cancer.”

Researchers have openly lamented the lack of reliable information pointing to or explaining successful improvement efforts, describing the literature as “sparse” and “scarce.” Those attempting to help others fix broken schools have typically resorted to identifying activities in improved schools, such as bolstering leadership and collecting data.

However, this case-study style of analysis is deeply flawed. As the U.S. Department of Education’s Institute of Education Sciences (IES) has noted, studies “that look back at factors that may have contributed to [a] school’s success” are “particularly weak in determining causal validity for several reasons, including the fact that there is no way to be confident that the features common to successful turnaround schools are not also common to schools that fail.”

Researchers have noted that the Department of Education has signaled its own ignorance about what to do about the nation’s very worst schools. One study reported, “The NCLB law does not specify any additional actions for schools that remain in the implementation phase of restructuring for more than one year, and [the Department] has offered little guidance on what to do about persistently struggling schools.” Indeed, the IES publication, “Turning Around Chronically Low-Performing Schools” practice guide, purportedly a resource for states and districts, concedes, “All recommendations had to rely on low levels of evidence,” because it could not identify any rigorous studies finding that “specific turnaround practices produce significantly better academic outcomes.”

Still in Its Infancy?

The prevailing view is that we must keep looking for turnaround solutions. Observers have written, “Turnaround at scale is still in its infancy,” and “In education, turnarounds have been tried rarely” (see “The Big U-Turn,” features, Winter 2009). But, in fact, the number and scope of fix-it efforts have been extensive to say the least.

Long before NCLB required interventions in the lowest-performing schools, states had undertaken significant activity. In 1989 New Jersey took over Jersey City Public Schools; in 1995 it took over Newark Public Schools. In 1993 California took control of the Compton Unified School District. In 1995 Ohio took over the Cleveland Metropolitan School District. Between 1993 and 1997 states required the reconstitution of failing schools in Denver, Chicago, New York City, and Houston. In 2000 Alabama took over a number of schools across the state, and Maryland seized control of three schools in Baltimore.

Since NCLB, interventions in struggling schools have only grown in number and intensity. In the 2006–07 school year, more than 750 schools in “corrective action,” the NCLB phase preceding restructuring, implemented a new research-based curriculum, more than 700 used an outside expert to advise the school, nearly 400 restructured the internal organization of the school, and more than 200 extended the school day or year. Importantly, more than 300 replaced staff members or the principal, among the toughest traditional interventions possible.

Occasionally a program will report encouraging success rates. The University of Virginia School Turnaround Specialist Program asserts that about half of its targeted schools have either made AYP or reduced math and reading failure rates by at least 5 percent. Though this might be better than would otherwise be expected, the threshold for success is remarkably low. It is also unknown whether such progress can be sustained. This matter is particularly important, given that some point to charter management organizations Green Dot and Mastery as turnaround success stories even though each has a very short turnaround résumé, in both numbers of schools and years of experience.

Many schools that reach NCLB’s restructuring phase, rather than implementing one of the law’s stated interventions (close and reopen as a charter school, replace staff, turn the school over to the state, or contract with an outside entity), choose the “other” option, under which they have considerable flexibility to design an improvement strategy of their own (see “Easy Way Out,” forum, Winter 2007). Some call this a “loophole” for avoiding tough action.

Yet even under the maligned “other” option, states and districts have tried an astonishing array of improvement strategies, including different types of school-level needs assessments, surveys of school staff, conferences, professional development, turnaround specialists, school improvement committees, training sessions, principal mentors, teacher coaches, leadership facilitators, instructional trainers, subject-matter experts, audits, summer residential academies, student tutoring, research-based reform models, reconfigured grade spans, alternative governance models, new curricula, improved use of data, and turning over operation of some schools to outside organizations.

It’s simply impossible to make the case that turnaround efforts haven’t been tried or given a chance to work.

A Better Mousetrap?

Despite this evidence, some continue to advocate for improved turnaround efforts. Nancy Grasmick supports recognizing turnarounds as a unique discipline. Frederick Hess and Thomas Gift have argued for developing school restructuring leaders; Bryan Hassel and Emily Ayscue Hassel have recommended that states and districts “fuel the pipeline” of untraditional turnaround specialists. NewSchools Venture Fund, the Education Commission of the States, and the research firm Mass Insight have offered related turnaround strategies.

And the Obama administration too has bought into the notion that turnarounds are the key to improving urban districts. Education secretary Arne Duncan has said that if the nation could turn around 1,000 schools annually for five years, “We could really move the needle, lift the bottom and change the lives of tens of millions of underserved children.” In the administration’s 2009 stimulus legislation, $3 billion in new funds were appropriated for School Improvement Grants, which aid schools in NCLB improvement status. The administration requested an additional $1.5 billion for this program in the 2010 budget. This is all on top of the numerous streams of existing federal funds that can be—and have been—used to turn around failing schools.

The dissonance is deafening. The history of urban education tells us emphatically that turnarounds are not a reliable strategy for improving our very worst schools. So why does there remain a stubborn insistence on preserving fix-it efforts?

The most common, but also the most deeply flawed, justification is that there are high-performing schools in American cities. That is, some fix-it proponents point to unarguably successful urban schools and then infer that scalable turnaround strategies are within reach. In fact, it has become fashionable among turnaround advocates to repeat philosopher Immanuel Kant’s adage that “the actual proves the possible.”

But as a Thomas B. Fordham Foundation study noted, “Much is known about how effective schools work, but it is far less clear how to move an ineffective school from failure to success…. Being a high-performing school and becoming a high-performing school are very different challenges.”

In fact, America’s most-famous superior urban schools are virtually always new starts rather than schools that were previously underperforming. Probably the most convincing argument for the fundamental difference between start-ups and turnarounds comes from those actually running high-performing high-poverty urban schools (see sidebar). Groups like KIPP (Knowledge Is Power Program) and Achievement First open new schools; as a rule they don’t reform failing schools. KIPP’s lone foray into turnarounds closed after only two years, and the organization abandoned further turnaround initiatives. Said KIPP’s spokesman, “Our core competency is starting and running new schools.”

Start Schools from Scratch

Ask those who know how to run high-performing, high-poverty schools why they start fresh, and they’ll give strikingly similar answers—and make the case against turnarounds.

A study done for NewSchools Venture Fund found that the operators of school networks believed that “changing the culture of existing schools to facilitate learning was difficult to impossible.” One compared turnarounds to putting “old wine in new bottles.”

Tom Torkelson, CEO of the high-performing IDEA network agrees: “I don’t do turnarounds because a turnaround usually means operating within a school system that couldn’t stomach the radical steps we’d take to get the school back on track. We fix what’s wrong with schools by changing the practices of the adults, and I believe there are few examples where this is currently possible without meddling from teacher unions, the school board, or the central office.”

Chris Barbic, founder and CEO of the stellar YES Prep network, says that “starting new schools and having control over hiring, length of day, student recruitment, and more gives us a pure opportunity to prove that low-income kids can achieve at the same levels as their more affluent peers. If we fail, we have only ourselves to blame, and that motivates us to bring our A-game every single day.”

KIPP co-founder Mike Feinberg says simply, “The best way we can look a child in the eye and say with confidence what kind of school and environment we will provide is by starting that school and environment from scratch.”

A 2006 NewSchools Venture Fund study confirmed a widespread aversion to takeover-and-turnaround strategies among successful school operators. Only 4 of 36 organizations interviewed expressed interest in restructuring existing schools. Remarkably, rather than trusting successful school operators’ track records and informed opinion that start-ups are the way to go, Secretary Duncan urged them to get into the turnaround business during a speech at the 2009 National Charter Schools Conference.

The findings above deserve repeating: Fix-it efforts at the worst schools have consistently failed to generate significant improvement. Our knowledge base about improving failing schools is still staggeringly small. And exceptional urban schools are nearly always start-ups or consistently excellent schools, not drastically improved once-failing schools.

So when considering turnaround efforts we should stop repeating, “The actual proves the possible” and bear in mind a different Kant adage: “Ought implies can.”

If we are going to tell states and districts that they must fix all of their failing schools, or if we are to consider it a moral obligation to radically improve such schools, we should be certain that this endeavor is possible. But there is no reason to believe it is.

Turnarounds Elsewhere

Education leaders seem to believe that, outside of the world of schools, persistent failures are easily fixed. Far from it. The limited success of turnarounds is a common theme in other fields. Writing in Public Money & Management, researchers familiar with the true private-sector track record offered a word of caution: “There is a risk that politicians, government officials, and others, newly enamored of the language of failure and turnaround and inadequately informed of the empirical evidence and practical experience in the for-profit sector…will have unrealistic expectations of the transformative power of the turnaround process.”

Hess and Gift reviewed the success rates of Total Quality Management (TQM) and Business Process Reengineering (BPR), the two most common approaches to organizational reform in the private sector. The literature suggests that both have failed to generate the desired results two-thirds of the time or more. They concluded, “The hope that we can systematically turn around all troubled schools—or even a majority of them—is at odds with much of what we know from similar efforts in the private sector.”

Many have noted that flexibility and dynamism are part of the genetic code of private business, so we should expect these organizations to be more receptive to the massive changes required by a turnaround process than institutions set in what Hess calls the “political, regulatory, and contractual morass of K–12 schooling.” Accordingly, school turnarounds should be more difficult to achieve. Indeed, a consultant with the Bridgespan Group reported, “Turnarounds in the public education space are far harder than any turnaround I’ve ever seen in the for-profit space.”

Building a Healthy Education Industry

We shouldn’t be surprised then that turnarounds in urban education have largely failed. The surprise and shame is that urban public education, unlike nearly every other industry, profession, and field, has never developed a sensible solution to its continuous failures. After undergoing improvement efforts, a struggling private firm that continues to lose money will close, get taken over, or go bankrupt. Unfit elected officials are voted out of office. The worst lawyers can be disbarred, and the most negligent doctors can lose their licenses. Urban school districts, at long last, need an equivalent.

The beginning of the solution is establishing a clear process for closing schools. The simplest and best way to put this into operation is the charter model. Each school, in conjunction with the state or district, would develop a five-year contract with performance measures. Consistent failure to meet goals in key areas would result in closure. Alternatively, the state could decide that districts only have one option—not five—for schools reaching NCLB-mandated restructuring: closure.

This would have three benefits. First, children would no longer be subjected to schools with long track records of failure and high probabilities of continued failure.

Second, the fear of closure might generate improvement in some low-performing schools. Failure in public education has had fewer consequences (for adults) than in other fields, a fact that might contribute to the persistent struggles of some schools. We should have limited expectations in this regard, however. Even in the private sector, where the consequences for poor performance are significant, some low-performing entities never become successful.

Third, and by far the most important and least appreciated factor, closures make room for replacements, which have a transformative positive impact on the health of a field. When a firm folds due to poor performance, the slack is taken up by the expansion of successful existing firms—meaning that those excelling have the opportunity to do more—or by new firms. New entrants not only fill gaps, they have a tendency to better reflect current market conditions. They are also far likelier to introduce innovations: Google, Facebook, and Twitter were not products of long-standing firms. Certainly not all new starts will excel, not in education, not in any field. But when provided the right characteristics and environment, their potential is vast.

The churn caused by closures isn’t something to be feared; on the contrary, it’s a familiar prerequisite for industry health. Richard Foster and Sarah Kaplan’s brilliant 2001 book Creative Destruction catalogued the ubiquity of turnover in thriving industries, including the eventual loss of once-dominant players. Churn generates new ideas, ensures responsiveness, facilitates needed change, and empowers the best to do more.

These principles can be translated easily into urban public education via tools already at our fingertips thanks to chartering: start-ups, replications, and expansions. Chartering has enabled new school starts for nearly 20 years and school replications and expansions for a decade. Chartering has demonstrated clearly that the ingredients of healthy, orderly churn can be brought to bear on public education.

A small number of progressive leaders of major urban school systems are using school closure and replacement to transform their long-broken districts: Under Chancellor Joel Klein, New York City has closed nearly 100 traditional public schools and opened more than 300 new schools. In 2004, Chicago announced the Renaissance 2010 project, which is built around closing chronically failing schools and opening 100 new public schools by the end of the decade.

Numerous other big-city districts are in the process of closing troubled schools, including Detroit, Philadelphia, and Washington, D.C. In Baltimore, under schools CEO Andrés Alonso, reform’s guiding principles include “Closing schools that don’t work for our kids,” “Creating new options that have strong chances of success,” and “Expanding some programs that are already proving effective.”

Equally encouraging, there are indications that these ideas, which once would have been considered heretical, are being embraced by education’s cognoscenti. A group of leading reformers, the Coalition for Student Achievement, published a document in April 2009 that offered ideas for the best use of the federal government’s $100 billion in stimulus funding. They recommended that each state develop a mechanism to “close its lowest performing five percent of schools and replace them with higher-performing, new schools including public charter schools.”

A generation ago, few would have believed that such a fundamental overhaul of urban districts was on the horizon, much less that perennial underperformers New York City, Chicago, and Baltimore would be at the front of the pack with much of the education establishment and reform community in tow. But, consciously or not, these cities have begun internalizing the lessons of healthy industries and the chartering mechanism, which, if vigorously applied to urban schooling, have extraordinary potential. Best of all, these districts and outstanding charter leaders like KIPP Houston (with 15 schools already and dozens more planned) and Green Dot (which opened 5 new schools surrounding one of Los Angeles’s worst high schools) are showing that the formula boils down to four simple but eminently sensible steps: close failing schools, open new schools, replicate great schools, repeat.

Today’s fixation with fix-it efforts is misguided. Turnarounds have consistently shown themselves to be ineffective—truly an unscalable strategy for improving urban districts—and our relentless preoccupation with improving the worst schools actually inhibits the development of a healthy urban public-education industry.

Those hesitant about replacing turnarounds with closures should simply remember that a failed business doesn’t indict capitalism and an unseated incumbent doesn’t indict democracy. Though temporarily painful, both are essential mechanisms for maintaining long-term systemwide quality, responsiveness, and innovation. Closing America’s worst urban schools doesn’t indict public education nor does it suggest a lack of commitment to disadvantaged students. On the contrary, it reflects our insistence on finally taking the steps necessary to build city school systems that work for the boys and girls most in need.

Andy Smarick is a distinguished visiting fellow at the Thomas B. Fordham Institute and adjunct fellow at the American Enterprise Institute.

Last Updated


Notify Me When Education Next

Posts a Big Story

Business + Editorial Office

Program on Education Policy and Governance
Harvard Kennedy School
79 JFK Street, Cambridge, MA 02138
Phone (617) 496-5488
Fax (617) 496-4428

For subscription service to the printed journal
Phone (617) 496-5488

Copyright © 2024 President & Fellows of Harvard College