Spring 2010 Correspondence

Readers respond

School Budgets

In “The Phony Funding Crisis” (features, Winter 2010), James Guthrie and Arthur Peng examine the apparent phenomenon in which schools, while claiming continual underfunding and budget cuts, continue to open their doors and educate students. As logical as their argument sounds, we believe it comes from the 30,000 foot level and differs from what is experienced on the ground.

The authors seem to assume that, faced with an economic downturn, states will respond to their constitutional mandate and other pressures and automatically raise taxes. Our studies suggest that instead, many states have played a shell game with American Recovery and Reinvestment Act (ARRA) funds, backfilling cuts to education with those one-time dollars.

Over the past year, the American Association of School Administrators has monitored the impact of the economic downturn on public schools from the ground level. Our fall 2009 survey found that the financial crisis continues to threaten and impact the progress and stability schools have enjoyed in the past. Responding from 49 states and the District of Columbia, school district leaders observed they have yet to see concrete indicators of a rebound in the nation’s economy: When asked how ARRA dollars impacted their state and local revenues, 83 percent reported that ARRA dollars did not represent a funding increase. More than one-third (35 percent) of respondents were unable to save core teaching jobs as a result of ARRA monies.

America’s schools were not immune to the most recent economic downturn. The funding crisis is far from phony. Looking forward, the 2010–11 and 2011–12 school years pack a one-two punch, with school district leaders facing the end of ARRA dollars and answering tough questions about programs and personnel that have been (and will be) cut, while trying to figure out what, if any, economic recovery is in store at the state and local levels.

Robert S. McCord
University of Nevada at Las Vegas

Noelle M. Ellerson
American Association of School Administrators

Guthrie and Peng respond:

The civility of the McCord and Ellerson rejoinder is appreciated. Moreover, their self-reports of school-district financial shortfalls are no doubt heartfelt.

However, the 30,000-foot perspective for which we are impugned is the more accurate view. We report objectively derived federal government–calculated national averages over a century. The long-run overall pattern is crystal clear: America has supported its public schools plentifully for a century. Nothing McCord and Ellerson say refutes this.

By crying fiscal wolf year after year, in the face of contrary data, public school advocates deceive taxpayers and avoid the four-decade reality of stagnant student performance and woeful mismanagement in places such as Detroit.


Andy Smarick’s “The Turnaround Fallacy” (features, Winter 2010) suffers from three fallacies. First, Smarick erroneously believes that school turnarounds have been tried widely and haven’t worked. In fact, interventions in failing schools are typically lukewarm and reliant on coaching, new curriculum packages, or other rearranging of deck chairs. Real turnaround attempts, in which a district hires a highly capable leader with “the big yes” to do what’s needed to fix the school, almost never happen.

Second, he wrongly suggests that in healthy industries, leaders don’t try to fix failing units, which simply close and make way for new upstarts. In fact, large companies with failing units try many strategies. They typically start by enforcing faithful execution of practices that work in other areas. When that doesn’t work, they replace the leader and give the new manager a change mandate. Smarick is right that the threat of closure is essential; today, bad schools have more lives than cats. To rescue more schools from “the brink of doom,” policymakers must make the option of school “doom” real, but then vigorously try to fix failing schools in the meantime.

Third, he grossly overstates the potential of start-up schools. Don’t get us wrong. We strongly back a large-scale effort to start great new schools. But research from other sectors pegs the probability of start-up success at about 25 percent, comparable to the estimated 30 percent of major corporate-change efforts that succeed. Like Smarick, we’re fans of outliers like KIPP, but together these networks will add a few hundred schools, not the few thousand we need. Even if these networks are joined by other wildly successful upstarts, only a small fraction of students in failing schools will benefit.

That’s why the nation needs a portfolio of strategies to change the fortunes of kids trapped in failing schools. Clinging to just one approach will write off millions who desperately need something different. Let’s start great new schools and fix bad ones. Let’s expect both strategies to work some of the time, but not always. And when they don’t work, let’s try again, rapidly, so kids don’t continue to languish in schools that aren’t getting the job done.

Bryan C. Hassel
Emily Ayscue Hassel
Public Impact

Incremental improvement strategies are not the same as whole-school turnaround. Andy Smarick’s recent article applies the word “turnaround” to a wide range of efforts to help struggling schools, and while he makes an interesting case, I disagree with his premise that school-turnaround efforts “have consistently fallen far short of hopes and expectations.”

Using a very specific turnaround model, my organization, the Academy for Urban School Leadership (AUSL), has had success in transforming some of Chicago’s poorest-performing schools. The Chicago Public Schools first brought in AUSL in 2006 to turn around eight schools in which test scores, attendance, discipline issues, and graduation rates made it clear that the students were not getting the education they needed. Our top-to-bottom approach is like hitting a reset button for these schools.

We accomplish this without disruption to students. They return in the fall to their neighborhood school, which has been transformed with renovated facilities and a new principal who has handpicked a new team of teachers, a new curriculum, new conduct codes and disciplinary standards, and new expectations for student success. And we deliberately foster the direct involvement of parents and community members.

Our data—dramatically improved attendance, test scores, and attitudes toward learning—demonstrate success. We also have the overwhelming support of students, parents, and teachers who have participated in the process, some of whom had been among our loudest critics.

It’s not realistic to think that dozens of failing schools in Chicago (and thousands nationally) can be closed and effectively replaced with new-start schools. There aren’t sufficient budget and time resources. And our students do not have time to wait while we sort out new-start school options. At AUSL, we are transforming schools now because our students deserve nothing less.

Donald Feinstein
Executive Director
Academy for Urban School Leadership

Andy Smarick suggests that the energy we spend turning around failing schools would be better spent shutting them down and starting new ones. That’s part of the solution, but we should be skeptical that closure alone is the answer.

First, this argument assumes that “close and replace” always beats turnaround on results. Smarick cites literature on businesses in the private sector, where turnarounds work only one-third of the time. He’s right, but what’s not cited is the fact that less than 30 percent of new businesses last more than six years. Not the dramatically better results we’re looking for.

But maybe this churn means better student outcomes, which leads to my second point. The data don’t look good for that, either. A recent report by the Consortium on Chicago School Research studied the nation’s largest close-and-replace strategy. More often than not, the strategy meant worse outcomes for the children. Moreover, a recent study by CREDO (Center for Research on Education Outcomes) at Stanford tells us that only one-third of new charter schools are demonstrably better than their neighborhood comparisons. The rest are the same or worse.

Third, the word “turnaround” can be used to mean different things, and sometimes it’s code for weak interventions. I agree with Smarick when he talks about the failed strategies of the past, but a reasonable definition of turnaround should exclude those half measures. The new federal guidelines for school improvement adopt a more robust definition of turnaround, wherein the adults in a building, especially teachers and leadership, are subject to change, and outside organizations can manage schools under performance contracts.

Our research shows that high-poverty schools that outperform their peers share certain qualities: an intense focus on instructional practice, an integrated approach to student support services, and flexibility from bureaucratic operating conditions. The creative destruction and market competition inherent in closure are great for allocating scarce resources, but they’re not enough to ensure quality and equity for vulnerable children. We need to invest in turning around failing schools as well.

Justin C. Cohen
The School Turnaround Strategy Group
Mass Insight

Andy Smarick makes a compelling argument that we would be better off closing failing schools, but he doesn’t take into account the stark reality that often urban districts simply have too many “failing schools” to close them all. Closing a district’s most persistently underperforming schools must be an option, but districts cannot stop there. Even the most extreme school-closure program will leave the majority of students in their current schools, many of which are also inadequate. Instead, districts must develop a comprehensive approach to address all of their turnaround schools as well as low-performing schools that don’t quite qualify for turnaround attention.

At Education Resource Strategies (ERS) we believe that districts need to make decisions about failing schools as part of a long-range, districtwide strategy that incorporates all resources: people, time, and money. While we agree with Smarick that evidence is not clear on a single turnaround strategy that works, we do know that schools can accelerate improvement through strong, transformational leaders; collaborative teacher teams; and targeting expertise and resources to help students who have fallen behind. There is a lot that districts can do to increase the probability of success: 1) Implement a districtwide strategy for measuring school performance and determining appropriate action, including the possibility of school closure; 2) Recruit transformational school leaders who can establish high expectations for improved performance; 3) Implement strategies that give these leaders the flexibility to efficiently assign teaching staff and to assemble high-performing teams with appropriate expertise; 4) Ensure sufficient expert instructional support and collaborative time for teachers to adjust instruction based on data; 5) Fund targeted student support and take the time to accelerate student learning; and 6) Provide additional problem solving and support from central staff.

A successful turnaround strategy might be as ambitious as a “cure for cancer,” as Smarick claims. So just like medical researchers, we have to keep trying. Closing schools should unquestionably be part of a school district’s strategy, but only a districtwide transformation will result in improving education for all children that the district serves.

Karen Hawley Miles
President and Executive Director
Education Resource Strategies

D.C.’s Rhee

It’s hard to know what image the title of June Kronholz’s piece on Michelle Rhee—“D.C.’s Braveheart” (features, Winter 2010)—is meant to conjure up: a hopelessly romantic quest, a quixotic uprising against corrupt power, Rhee’s eventually being drawn and quartered. Or is it confrontation for confrontation’s sake?

None of the images it brings to mind are encouraging. Having just completed a study (Leading for Equity, Harvard Education Press, 2009) of the public schools in neighboring Montgomery County, Maryland (MCPS), which is encouraging, there is much to be said about Superintendent Jerry Weast’s nonconfrontational style. But MCPS and D.C. share only a common boundary. In no other respect are they comparable, and it is as hard to imagine Jerry’s approach working in D.C. as it is Rhee’s in MCPS. (For what it’s worth, Paul Vance was superintendent of both districts and left an indifferent legacy.) The underlying question, of course, is will Rhee’s in-your-face style work in D.C.?

To Rhee’s credit, she has gotten everyone’s attention. And she has effectively raised the issue of accountability. (Years ago, one of her predecessors, Vince Reed, told me he thought the whole system was hopeless.) I’m an MCPS alumnus and my sister is a D.C. alumna and we wish these very different districts well. I for one am skeptical about Rhee’s unvarnished approach: too much stick, too little carrot. But the jury is still out, and she still has time to build bridges to teachers. Without their support, all is lost; with their support, there is a fighting chance.

At minimum—in the one-size-fits-all No Child Left Behind era—her tenure reminds us of the real genius of the federal system, the opportunity to try many different approaches to a shared objective: increased academic achievement for all students.

Denis Doyle

Teacher Pensions

Golden Handcuffs” (research, Winter 2010) draws attention to the important incentives that are built into many teacher pensions, but shortcomings in the authors’ analysis lead them to spurious conclusions.

First, despite the authors’ claim, legally and otherwise, pension plans cannot and do not redistribute wealth.

Second, the authors incorrectly assert that there is no justification for the incentives embedded in teacher pensions. To the contrary, one of the most pressing issues facing schools is keeping top talent, especially in hard-to-staff areas. In light of this, retention incentives (which the authors refer to as “mobility penalties”) make perfect sense.

Economic research stretching back more than two decades has documented the strong retention effects embedded in traditional defined benefit (DB) plans, where benefits are based on an employee’s final pay. Authors Costrell and Podgursky describe the well-recognized pattern of benefit accruals in such plans and then jump to the conclusion that it serves no purpose. But, as economists have long known, it is precisely because of this pattern, which offers greater rewards for loyal employees who provide long service to an employer, that many employers offer DB pensions. Because turnover is costly (both in dollar terms and in terms of productivity), it makes sense for employers to build incentives into compensation that reward long service.

Indeed, until only recently, the vast majority of Fortune 500 companies relied heavily on DB pensions to retain top talent. Even as some private-sector employers have moved away from these plans in recent years, they have been careful to develop other compensation structures that mimic the incentives provided by DB pensions. Deferred compensation in the form of stock options or restricted stock awards is similar to pensions in that it encourages loyalty to an employer.

The problem is that school districts and government entities cannot offer stock options, restricted stock, or similar benefits. Thus, jettisoning DB pensions, as the authors recommend, can be expected to cause increased turnover and attrition of our most-effective teachers, hurting productivity and quality, in other words, exactly the wrong solution for our schools.

Beth Almeida
Executive Director
National Institute on Retirement Security

Costrell and Podgursky respond:

In Missouri, a teacher who retires at age 55 receives a lifetime pension worth 33 percent of cumulative earnings, but only 1 percent if she leaves at 35. Yet in both cases her employer contributed 12.5 percent of earnings each year. Redistribution or not? These incentives do encourage teachers, effective or not, to stay until early retirement (but no longer), while likely discouraging entry of mobile young teachers with 10 to 15 years to offer. Peculiar incentives or rational? The huge penalties for cross-state mobility may be rational for each state, but not for the U.S. K–12 system. Do we recommend jettisoning DB pensions? No, we recommend cash balance systems, which are DB but offer far greater portability and reward all years of service equally: benefits are tied to contributions, so no redistribution occurs.

Special Education Vouchers

Jay Greene and Stuart Buck (“The Case for Special Education Vouchers,” features, Winter 2010) are correct that some children with disabilities have unique needs that require private schooling. That’s why the federal Individuals with Disabilities Education Act (IDEA) allows children with disabilities to attend private schools at public expense when their districts cannot provide a free, appropriate public education (FAPE).

But only a small percentage of children with disabilities have such placements, and not, as Greene and Buck contend, because the law’s processes for securing private placements are inadequate, but because the vast majority of children with disabilities can, and do, receive FAPE in the public schools. That’s not to gloss over the shortcomings in our special education system, or the difficulties some parents face in obtaining services for their children. But there’s no evidence that children with disabilities need additional education options more than any other youngsters in underperforming schools, or that vouchers address the underlying problems in special education. Rather, voucher proponents have seized on this population because they are more sympathetic beneficiaries than poor and minority youngsters. Using children with disabilities to increase public support for vouchers may be smart politics, but it doesn’t mean that special education vouchers are good policy.

Policymakers must take steps to expand education options for children with disabilities and make it easier for their parents to access needed services. But special education vouchers are not the best way to do this; they create other, adverse consequences, such as further segregating or perpetuating double standards for children with disabilities and creating perverse incentives for parents and educators.

Other approaches—expanding high-cost/low-incidence pools, improving IDEA’s dispute-resolution and placement processes, enhancing charter schools’ capacity to serve children with disabilities, and authorizing more charters with specific missions to serve disabled youngsters—have more promise to expand high-quality, accountable options for youngsters with disabilities.

Sara Mead
Senior Research Fellow
New America Foundation

Greene and Buck respond:

Sara Mead makes several assertions that are contrary to the findings of our article, but she presents no evidence to substantiate those assertions or to contradict the evidence we presented.

For example, she says that there are so few private placements of special education students “not…because the law’s processes for securing private placements are inadequate, but because the vast majority of children with disabilities can, and do, receive FAPE in the public schools.” How does she know this? She doesn’t say. Nor does she do anything to refute the evidence we presented that shows the inadequacy of the current private placement system.

It is insufficient simply to contradict claims. One needs to address evidence and Mead fails to do so.

Federal Lunch Program

[an unabridged version of the following letter appears at the end]

In “Fraud in the Lunchroom?” (check the facts, Winter 2010), David Bass presents evidence of substantial error in students’ eligibility for free or reduced-price school meals through the National School Lunch Program (NSLP), citing a recent Mathematica study that found most errors result from misreporting of household income. The title of Mr. Bass’s article implies that these errors may be intentional.

Our research suggests that fraud is not a major factor in explaining errors. Households that fail to respond to a district’s request for income verification are not necessarily engaging in fraud. We examined a randomly selected set of households that did not respond, finding that most were eligible for free or reduced-price meals. In the Access, Participation, Eligibility, and Certification (APEC) study, we found that, in more than 40 percent of household misreporting errors, parents overreported, rather than underreported, their income. If fraud were rampant, we would have expected much less of this type of error. Instead, we believe that most errors are unintentional: parents do not understand which household members should be included, forget about a minor income source, report net instead of gross income, or incorrectly enter the frequency of income receipt.

Even if fraud is minimal, the resulting costs to taxpayers are substantial. How might policymakers respond? We caution against requiring income documentation from all applicants. As Mr. Bass notes, our research showed that a test of this approach not only failed to reduce benefit receipt for ineligible households, but also reduced benefit receipt for eligible households.

A simple approach that could reduce error by one-third would eliminate the distinction between free and reduced-price benefits, since much program error results from misclassification. We could also build on current federal initiatives such as direct certification to improve NSLP certification accuracy. Under this policy, now required in all districts, households receiving benefits from other federal programs with more rigorous income-verification requirements are automatically eligible for NSLP. The U.S. Department of Agriculture is also considering using existing surveys to estimate the proportion of eligible children in selected schools, and then developing schoolwide reimbursement rates. This would eliminate the need for districts to certify households through the current process.

Philip Gleason
Senior Fellow
Michael Ponza
Senior Fellow
Mathematica Policy Research

In “Fraud in the Lunchroom?” (check the facts, Winter 2010) David Bass presents evidence of substantial error in students’ eligibility for free or reduced-price school meals. He notes high rates of nonresponse when districts try to verify students’ eligibility for benefits. He also cites Mathematica’s recent Access, Participation, Eligibility, and Certification (APEC) study, which found that most error results from households’ misreporting of eligibility information on their applications (Ponza et al. 2007). Bass raises the question of whether this error is the result of fraud—in other words, ineligible households lie about their income to receive benefits.

Our research suggests that fraud is not a major factor in explaining this error. We examined a randomly selected set of households that did not respond to a request for income verification (Burghardt et al. 2004). Over half were eligible for the benefits they were receiving. Furthermore, 80 percent were eligible for either free or reduced-price meals. This suggests that they failed to respond to the verification request because they forgot, were too busy, could not locate the documentation requested, or some similar reason other than fraud.

Bass notes that the APEC study found 15 percent of participants received more benefits than they were eligible for. However, many of these participants were eligible for some level of benefits; the errors involved misclassifications between free and reduced-price status. We believe the type of error most indicative of fraud would involve ineligible students receiving free meals. We found that only about 6 percent of students receiving free meals were ineligible.

Most certification errors result from households’ reporting less income or a larger household than they should have, resulting in higher benefits than they are eligible for. However, we also found substantial error in the other direction: more than 40 percent of cases involved the parent reporting more income or fewer persons in the household than they should have. If there were rampant fraud, we would have expected fewer reporting errors of this type.

The presence of substantial over- and underreporting of income suggests that many applicants have difficulty understanding how to complete the application and that most errors are unintentional. Although districts provide detailed instructions on how to complete meal applications, unintentional errors can occur for many reasons: applicants do not understand which household members to include, forget about a minor income source, report net instead of gross income, or incorrectly enter the frequency of income receipt (e.g., record “monthly” when it should be “weekly”).

Even if the level of fraud is minor, evidence suggests that the level of reporting error is substantial. Given the costs of this error to taxpayers and the widespread use of program eligibility as a poverty indicator, how might policymakers respond? We caution against requiring income documentation from all applicants. As Bass notes, our research showed that a test of this approach not only failed to reduce the rate at which ineligible households received benefits, but it also reduced access to the program for eligible households (Gleason et al. 2008).

A simple approach that could reduce error by as much as one-third would eliminate the distinction between free and reduced-price benefits altogether. Because a lot of program error results from misclassification, removing this distinction would both reduce the error rate and cut administrative costs. To be budget-neutral, the federal reimbursement for this new category combining free and reduced-price benefits could be set somewhere between the current free and reduced-price rates.

Current federal initiatives such as direct certification aim to improve the accuracy of National School Lunch Program (NSLP) certification. Under this policy, now required in all districts to take advantage of more intensive income-verification efforts in other federal programs, households receiving benefits from these programs are automatically eligible for the NSLP. In addition, the Food and Nutrition Service of the U.S. Department of Agriculture is considering using extant surveys to provide small-area estimates of the proportion of children eligible for free and reduced-price benefits. These data would be used to develop the percentages by which USDA sets school- or districtwide meal reimbursement rates for some schools, eliminating the need to certify households through the current process.

The research upon which some of these ideas are based includes:

Burghardt, J., Silva, T., and Hulsey, L. “Case Study of National School Lunch Program Verification Outcomes in Large Metropolitan School Districts.” Special Nutrition Program Report Series, No. CN-04-AV3. U.S. Department of Agriculture, Food and Nutrition Service, Office of Analysis, Nutrition and Evaluation. Alexandria, VA: 2004.

Ponza, M., Gleason, P., Moore, Q., and Hulsey, L. NSLP/SBP Access, Participation, Eligibility, and Certification Study?Erroneous Payments in the NSLP and SBP, Vol. I: Study Findings, U.S. Department of Agriculture, Food and Nutrition Service, Office of Research, Nutrition and Analysis. Alexandria, VA: 2007.

Gleason, P., Burghardt, J., Strasberg, P., and Hulsey L. “Tightening Income Documentation in a Means-Tested Program: Who Stays Away?” Evaluation Review 32, no. 3 (2008), 273–97.

Philip Gleason
Senior Fellow
Michael Ponza
Senior Fellow
Mathematica Policy Research

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
Email Education_Next@hks.harvard.edu

For subscription service to the printed journal
Phone (617) 496-5488
Email subscriptions@educationnext.org

Copyright © 2024 President & Fellows of Harvard College