Executive Summary
This report highlights patterns of charter school expansion across several large and mid-size U.S. cities since 2000. In this report, the focus is the loss of enrollments and revenues to charter schools in host districts and the response of districts as seen through patterns of overhead expenditures. I begin by identifying those cities and local public school districts that have experienced the largest shifts of students from district-operated to charter schools, and select from among those cities illustrative examples of the effects of charter school expansion on host district finances and enrollments.
Effects of charter expansion
District schools are surviving but under increased stress
In some urban districts, charter schools are serving 20 percent or more of the city or districtwide student population. These host districts have experienced the following effects in common:
While total enrollment in district schools (the noncharter, traditional public schools) has dropped, districts have largely been able to achieve and maintain reasonable minimum school sizes, with only modest increases in the shares of children served in inefficiently small schools.
While resources (total available revenues to district schools) have declined, districts have reduced overhead expenditures enough to avoid consuming disproportionate shares of operating spending and increasing pupil/teacher ratios.
Despite expenditure cutting measures, districts simultaneously facing rapid student population decline and/or operating in states with particularly inequitable, under-resourced school finance systems have faced substantial annual deficits.
Charter expansion is not driven by well-known, high-profile operators
Most charter expansion in these cities has occurred among independently operated charter schools.
High profile, frequently researched nonprofit charter school operators including the Knowledge is Power Program (KIPP) have relatively small shares of the charter school market in all cities except Newark.
In many of these cities, some of the leading charter operators (those with the most market share) have been the subject of federal and state investigations and judicial orders regarding conflicts of interest (self-dealing) and financial malfeasance. These operators include Imagine Schools, Inc., White Hat Management, National Heritage Academies, and Concept Schools.
The varied and often opaque financial practices across charter school management companies, while fitting with a competitive portfolio conception, leads to increased disparities across students, irregularities in the accumulation of additional public (publicly obligated) debt, and inequities and irregularities in the ownership and distribution of what were once commonly considered public assets—from buildings and vehicles right down to desks, chairs, and computers.
Charter schools are expanding in predominantly low-income, predominantly minority urban settings
Few are paying attention to the breaches of legal rights of students, parents, taxpayers, and employees under the increasingly opaque private governance and management structures associated with charter expansion.
Expansion of charter schooling is exacerbating inequities across schools and children because children are being increasingly segregated by economic status, race, language, and disabilities and further, because charter schools are raising and spending vastly different amounts, without regard for differences in student needs. Often, the charter schools serving the least needy populations also have the greatest resource advantages.
With the expansion of charter schooling, public districts are being left with legacy debts associated with capital plants and employee retirement systems in district schools while also accumulating higher risk and more costly debt in the form of charter school revenue bonds to support new capital development.
In many cases, the districts under investigation herein are large enough to be cut in half or thirds while still being financially viable, at least in terms of achieving economies of scale. In effect, charter expansion has already halved the size of many urban districts. Similar charter expansion in smaller districts, however, may lead the districts to enroll fewer than 2,000 pupils in district schools and suffer elevated costs. Given the literature on costs, productivity, and economies of scale, it makes little sense in population-dense areas to promote policies that cause district enrollments to fall below efficient-scale thresholds (around 2,000 pupils) or that introduce additional independent operators running below efficient-scale thresholds. It makes even less sense to introduce chartering to rural areas where schools and districts already operate below efficient scale.
Beyond issues of economies of scale, charter expansion can create inefficiencies and redundancies within district boundaries, from the organization and delivery of educational programs to student transportation, increasing the likelihood of budgetary stress on the system as a whole, and the host government in particular. In addition to increasing per pupil transportation expense, ill-planned (or unplanned) geographic dispersion may put more vehicles on already congested urban streets, contributing to traffic and air quality concerns, and significantly reduces the likelihood that children use active transportation (walking or biking) to school (Baker 2014b; Davison, Werder, and Lawson 2008; Evenson et al. 2012; Merom et al. 2006; Rosenberg et al. 2006; Wilson, Wilson, and Krizek 2007).
Policy recommendations
I conclude with policy recommendations for moving toward more equitable systems of excellent schooling. First, state policymakers must rethink charter laws that deregulate both the operators and regulators (authorizers) of charter schools, applying the following two key principles:
Authorizers must work in collaboration with districts to ensure that the mix of providers in any context provides the best possible array of opportunities
Authorizers and providers must be sufficiently publicly accountable and transparent
Current systems involving multiple, competing government and nongovernment authorizers are unlikely to ever achieve these goals, especially when the objective of both school operators (management companies) and those who authorize and oversee them is to maximize revenue by maximizing market share.
Looking beyond waiting lists and proficiency rates
When cities and school districts are debating whether to expand charter schools in the jurisdiction, decision-makers must look beyond facile claims of miraculous proficiency rates in cherry-picked charter schools (serving cherry-picked or culled populations) and reports of long waiting lists. Policymakers should consider a much longer checklist, to include but not be limited to the following preliminary set of items:
How is the whole system, not just a subset of the system, meeting performance measures, such as assessment scores and gains, and is the performance both adequate and equitable?
What are the cost and equity implications of sorting students into different schools based on needs, and how are resources, programs, and services going to be reallocated to ensure equitable access to adequate educational opportunities for all children?
Will management structures and service delivery be efficient or lead to inefficient duplication?
Are seemingly mundane operations management issues such as logically, spatially distributing enrollments; optimally using facilities space; and optimizing transportation services/networks getting the proper attention?
How will new systems affect quality of life factors such as transportation time, school/neighborhood walkability, and numbers of schooling disruptions faced by children and families?
Can we evaluate entrants to the market (based on their prior behavior and practices) and regulate both their practices and those of providers already operating in the space to ensure that the rights of students, taxpayers, and employees are protected equitably?
If we consider a specific geographic space, like a major urban center, operating under the reality of finite available resources (local, state, and federal revenues), the goal is to provide the best possible system for all children citywide, given the resources available. That is, resources should be used most efficiently and equitably to achieve the best possible system of schools for all children. Chartering, school choice, or market competition are not policy objectives in-and-of-themselves. They are policy alternatives—courses of policy action—toward achieving these broader goals and must be evaluated in this light. This checklist will help reveal whether charter expansion or any policy alternative increases inequity, introduces inefficiencies and redundancies, compromises financial stability, or introduces other objectionable distortions to the system, so that those costs can be weighed against expected benefits.
Moving toward efficient and effective unified education systems
If the broad, long-term policy objective is to move toward the provision of a “system of great schools” in each of America’s communities, then those systems must be responsibly, centrally managed to achieve an equitable distribution of excellent (or at the very least adequate) educational opportunities for all children, while protecting the interests and legal rights of children, parents, taxpayers, and employees. Achieving this lofty goal requires determining which functions of the system must be centrally and publicly regulated and governed. Systemwide public responsibilities include but are not limited to:
The equitable management of enrollments and schooling access
The equitable distribution of financial and other resources across the system, including allocation of resources to centralized functions that serve all schools
The centralized management and equitable use/allocation, maintenance, and operations of the public’s capital stock of schools and related land and facilities
The centralized management of systemwide debt obligations and long-term liabilities including employee retirement and health benefits
Numerous analyses have found chartering to lead to an imbalanced distribution of students by race, income, language proficiency, and disability status. So too does magnet schooling, or concentration of any specialized services across buildings within districts. The point is not that all such variations must necessarily be erased, or even could be, but that these variations must be acknowledged, and managed for the good of the system as a whole. To the extent that student needs continue to vary across school settings, resources must be targeted to accommodate those needs. This is a central function, and includes budget allocations, space allocations, and personnel allocations that draw on a substantial body of research on costs associated with providing equal educational opportunities (Duncombe and Yinger 2008).
Capital stock—publicly owned land and buildings—should not be sold off to private entities for lease to charter operators, but rather, centrally managed both to ensure flexibility (options to change course) and to protect the public’s assets (taxpayer interests). Increasingly, districts such as those discussed herein, have sold land and buildings to charter operators and related business entities, and now lack sufficient space to serve all children should the charter sector, or any significant portion of it, fail. Districts and state policymakers should not put themselves in a position where the costs of repurchasing land and buildings to serve all eligible children far exceed fiscal capacity and debt limits.
Finally, pension and health care costs are systemwide concerns that cannot be ignored by shifting students, and thus teachers and public dollars, across sectors.
Introduction
This report highlights patterns of charter school expansion across several large and mid-size U.S. cities over the past decade. The public’s interests lie in providing the highest quality educational opportunities for all children at an expense the public is willing and able to support. If we consider a specific geographic space, such as a major urban center, operating under the reality of finite available resources (local, state, and federal revenues), the goal is to provide the best possible system for all children citywide (in that space and under the policy umbrellas governing that space), given the resources available. That is, resources should be used most efficiently and equitably to achieve best possible system of schools for all children. Chartering, school choice, or market competition are not policy objectives in-and-of-themselves. They are merely policy alternatives—courses of policy action—toward achieving these broader goals and must be evaluated in this light. To the extent that charter expansion or any policy alternative increases inequity, introduces inefficiencies and redundancies, compromises financial stability, or introduces other objectionable distortions to the system, those costs must be weighed against expected benefits.
In this report, the focus is on the host district, the loss of enrollments to charter schools, the loss of revenues to charter schools, and the response of districts as seen through patterns of overhead expenditures. I begin by identifying those cities and local public school districts (hosts) that have experienced the largest shifts of students from district-operated to charter schools. I also explore the average size of schools as the charter sector grows. Next, I evaluate equity concerns related to the sorting of students by their population characteristics, and variations in classroom resources across schools and children. Finally, I discuss frequently overlooked concerns such as substantive changes to the rights of children, parents, and employees under privately governed and managed systems, and emerging concerns over accumulating high risk debt incurred by charter operators through municipal bond markets.
An opportunity for scalable innovation
Since its origins in the early 1990s, the charter school sector has grown to over 6,500 schools serving more than 2.25 million children in 2013.1 In some states, the share of children now attending charter schools exceeds 10 percent (for example, Arizona and Colorado), and in select major cities that share exceeds one-third (for example the District of Columbia, Detroit, and New Orleans)(Weber and Baker 2015a). Modern day charter schools were conceived by union leader Albert Shanker in the 1980s as a way to provide opportunities for creative, independent educators to collaborate and test new ideas with lessened policy constraints (Shanker 1988). To the extent these innovations were successful they could inform practices in traditional district schools, Shanker posited. Over the next few decades, states adopted statutes providing opportunities for individuals and organizations, including traditional districts, to create these newly chartered schools. In some cases, statutes allowed for the creation of charters governed and financed by existing districts, and in other cases, for the creation of charters independent of district governance, while operating within the boundaries of and in competition with local public districts.
While charter schooling was conceived as a way to spur innovation—try new things, evaluate them, and inform the larger system—studies of the structure and practices of charter schooling find the sector as a whole not to be particularly “innovative” (Preston et al. 2012). Analyses by charter advocates at the American Enterprise Institute find that the dominant form of specialized charter school is the “no excuses” model, a model that combines traditional curriculum and direct instruction with strict disciplinary policies and school uniforms, in some cases providing extended school days and years (McShane and Hatfield 2015). Further, charter schools raising substantial additional revenue through private giving tend to use that funding to a) provide smaller classes, and b) pay teachers higher salaries for working longer days and years (Baker, Libby, and Wiley 2012). For those spending less, total costs are held down, when necessary, through employing relatively inexperienced, low-wage staff and maintaining high staff turnover rates (Epple, Romano, and Zimmer 2015; Toma and Zimmer 2012). In other words, the most common innovations are not especially innovative or informative for systemic reform.
Emergence of private managers
The early charter movement coincided with the emergence of private management firms interested in public schooling. Two private for-profit companies tried their hand at providing school management services for public districts in the 1990s: Edison Schools, Inc., and Education Alternatives, Inc. (EAI) (Richards 1996). Education Alternatives, Inc., a publicly traded for-profit company, failed financially while holding an operating contract for nine (then 11) schools within Baltimore City Public Schools and soon after signing a contract with Hartford Connecticut Public Schools. The company failed prior to taking full responsibility for schools in Hartford. Edison Schools expanded cautiously in the wake of EAI’s failure, operating a school in Wichita, Kansas, in 1995 and 25 schools nationally by the end of 1996 (Steinberg 1997). Edison also faced financial troubles as a publicly traded stock, eventually buying back its company stock in 2003 and reverting to privately held status (Las Vegas Sun 2003).
As charter schools expanded, including online and hybrid schooling options, Edison Schools and other upstart for-profit companies shifted their growth strategy to the charter sector, where they could control employment contracts, increasing financial flexibility and profit potential. Coinciding with these developments, many now high-profile nonprofit charter management firms got their start as founders of single charter schools, including the Knowledge is Power Program (KIPP), which actually had two schools, a middle school in Houston and another in New York City; Uncommon Schools, founded from North Star Academy in Newark, New Jersey; and Achievement First, founded from Amistad Academy in New Haven, Connecticut. Now the charter school landscape consists of a mix of schools operated by major nonprofit charter management organizations, schools operated by for-profit managers, schools operated by other education management organizations described by Miron and colleagues as nonprofit in formal status but engaging in contractual arrangements more similar to for-profit organizations, and schools that remain independently operated, i.e., “mom-and-pop” (Miron and Gulosino 2013).
From portfolios to parasites?
As early as the mid-1990s, authors including Paul Hill, James Guthrie, and Lawrence Pierce (1997) advocated that entire school districts should be reorganized into collections of privately managed contract schools (Hill, Pierce, and Guthrie 2009). This contract school proposal emerged despite the abject failure of Education Alternatives, Inc., in Baltimore and Hartford. This proposal provided a framework for renewed attempts at large-scale private management including the contracting of management for several Philadelphia public schools in the early 2000s. Philadelphia’s experiment in private contracting yielded mixed results, at best (Mac Iver and Mac Iver 2006).2 Notably, Hill and colleagues’ contract school model depended on a centralized authority to manage the contracts and maintain accountability, a precursor to what is now commonly referred to as a “portfolio” model. In the portfolio model, a centralized authority oversees a system of publicly financed schools, both traditional district-operated and independent, charter-operated, wherein either type of school might be privately managed (Hill 2006).3 The goal as phrased by former New York City schools’ chancellor Joel Klein is to replace school systems with systems of great schools (Patrino 2015).
A very different reality of charter school governance, however, has emerged under state charter school laws—one that presents at least equal likelihood that charters established within districts operate primarily in competition, not cooperation with their host, to serve a finite set of students and draw from a finite pool of resources. One might characterize this as a parasitic rather than portfolio model—one in which the condition of the host is of little concern to any single charter operator. Such a model emerges because under most state charter laws, locally elected officials—boards of education—have limited control over charter school expansion within their boundaries, or over the resources that must be dedicated to charter schools. Thus, there is no single, centralized authority managing the portfolio—the distributions of enrollments and/or resources—or protecting against irreparable damage to any one part of the system (be it the parasites or the host).
Figure 1 displays a system in which a set of district operated schools (DOS), district-charter schools (DCS), and independent charter schools (ICS) serve a geographic space previously governed and operated entirely by local elected officials. In many states, independent charter schools may be only authorized by a government or government-appointed entity—a single authorizer. Nonetheless, these schools are not required to be responsive to local elected governance (unless required under state charter law) and have little or no incentive to be concerned with the financial condition of their host. In other states, additional entities may authorize charter schools to compete for students and resources in the same geographic space. This approach further disperses authority for schools serving any geographic area. Among other issues, dispersed authorization provides the opportunity for potential charter operators, including those with previously failing track records, to “shop” for authorizers who will more readily permit their expansion and more likely turn a blind eye to academic failure and/or financial mismanagement (Journal-Gazette 2015).
Figure 1
Proponents of the dispersed governance model in Figure 1 assert that competition both for governance/accountability and for management/operations of schools provides greater opportunity for rapid expansion and innovation. However, some of the more dispersed multiple authorizer governance models have been plagued by weak accountability, financial malfeasance, and persistently low-performing charter operators, coupled with rapid, unfettered, underregulated growth (Center for Education Reform 2011; Rowland 2015; Dixon 2014).
Nonetheless, charter advocacy organizations including Bellwether Education Partners (BEP) continue to argue for more rapid growth and increased market share for charter schooling. BEP provides a facile extrapolation (along unconstrained linear trajectories) to claim that at current rates, the charter sector will grow to serve 20 percent to 40 percent of all U.S. students by 2035. For charter expansion advocates such as BEP, however, even this rate is too slow, constrained by having too few authorizers, caps on new charters in some states, and unwillingness of districts as authorizers to approve new charter applications (Mead, LiBetti Mitchel, and Rotherham 2015). For charter advocacy organizations, tight centralized regulation and slow or limited growth of charters is a non-starter, with the optimal balance somewhere between approximately 40 percent (as in Washington, D.C.) and nearly 100 percent (as in New Orleans) of children served by independent charter schools (Pearson, McKoy, and Kingsland 2015).
Fiscal stress, inefficiency, and charter expansion
Increased attention is being paid to the fiscal and enrollment effects of charter schooling on host districts. These concerns come at a time when municipal fiscal stress and the potential for large-scale municipal and school district bankruptcies are in the media spotlight (Governing 2015). Many high profile cases of municipal fiscal stress are in cities where the charter sector is thriving, for example Chester Upland, Pa., Detroit, and Philadelphia (Layton 2015; Graham 2015; Pierog 2015). Some charter advocates have gone so far as to assert that school district bankruptcy presents a “huge opportunity” to absolve the taxpaying public of existing debts and financial obligations and start fresh under new management, reallocating those funds to classrooms (Persson 2015). Of course, this strategy ignores the complexities of municipal bankruptcy proceedings, and the contractual, social, and moral obligations for the stewardship of publicly owned capital (and other) assets and responsibility to current and retired employees.
Advocates for charter expansion typically assert that charter expansion causes no financial harm to host districts. The logic goes, if charter schools serve typical students drawn from the host district’s population, and receive the same or less in public subsidy per pupil to educate those children, then the per pupil amount of resources left behind for children served in district schools either remains the same or increases. Thus, charter expansion causes no harm (and in fact yields benefits) to children remaining in district schools. The premise that charter schools are uniformly undersubsidized is grossly oversimplified and inaccurate in many charter operating contexts (Baker 2014c). In addition, numerous studies find that charter schools serve fewer students with costly special needs, leaving proportionately more of these children in district schools. Perhaps most important, the assumption that revenue reductions and enrollment shifts cause districts no measurable harm for host ignores the structure of operating costs and dynamics of cost and expenditure reduction.
Moody’s Investors Service opined in 2013 that “charter schools pose greatest credit challenge to school districts in economically weak urban areas.” Specifically, Moody’s identified the following four areas posing potential concerns for host urban districts with growing independent charter sectors:
Weak demographics and district financial stress, which detract from the ability to deliver competitive services and can prompt students to move to charter schools
Weak capacity to adjust operations in response to charter growth, which reduces management’s ability to redirect spending and institute program changes to better compete with charter schools
State policy frameworks that support charter school growth through relatively liberal approval processes for new charters, generous funding of charters, and few limits on charter growth
Lack of integration with a healthier local government that can insulate a school system from credit stress (D’Arcy and Richman 2013)
Moody’s reiterated these concerns in a follow-up report (Moody’s Investors Service 2015).
District officials in Nashville, Tennessee, recently contracted consultants to evaluate the impact of charter expansion on their district. The consultants’ report noted that charters:
Will continue to cause the transfer of state and local per student funds without reducing operational costs
Will continue to increase direct and indirect costs
Will continue to negatively impact deferred maintenance at leased buildings
May have an offsetting impact on capital costs, but only where available space can be re-allocated efficiently (MGT of America 2014)
Recently published academic analyses raise similar concerns. Bifulco and Reback (2014) evaluate the fiscal impact of charter expansion on two midsize upstate New York cities, Albany and Buffalo. They find that charter schools have had negative fiscal impacts on these districts, and argue that there are two reasons for these impacts. First, districts are generally unable to adjust their expenditures on a student-by-student basis, because costs range from fixed costs (districtwide and school overhead costs that are not reduced by the transfer of individual pupils), to step costs (including classroom level costs, also not reduced by the transfer of individual pupils) to variable costs, which are most easily reduced on a student-by-student basis, but constitute a relatively small share of school district budgets. These concerns echo those of consultants to Nashville Public Schools. Further, Arsen and Ni (2012b) find that higher levels of charter school enrollments in Michigan school districts are strongly associated with declining fund balances, and that revenues declined more rapidly than costs in districts losing students to charter schools.4
Second, Bifulco and Reback (2014) point out that “operating two systems of public schools under separate governance arrangements can create excess costs,” or inefficient expenditures. Baker, Libby, and Wiley (2012) have raised similar concerns about additional, often exorbitant, overhead expenses created by introducing school systems within school systems (independent charter schools within districts). That is, while inducing fiscal stress on host districts, charter expansion may also be increasing total overhead costs. Two studies of Michigan charter schools, which operate fiscally independently of local public districts, have found them to have particularly high administrative expenses and low direct instructional expenses. Arsen and Ni (2012a, 1, 13) find that “controlling for factors that could affect resource allocation patterns between school types, we find that charter schools on average spend $774 more per pupil per year on administration and $1,141 less on instruction than traditional public schools.” Further, they find “charter schools managed by EMOs [Education Management Organizations] spend significantly more on administration than self-managed charters (about $312 per pupil). This higher spending occurs in administrative functions traditionally performed at both the district central office and school building levels.”
Izraeli and Murphy (2012, 265) find that district schools in Michigan tended to spend more on instruction per student than did charter schools, and the gap increased by another 5 percent to nearly 35 percent percent over the period studied (1995–96 to 2005–06). Further they find the spending gap for instructional spending to be greater than that for general spending. The overall funding gap between district and charter schools was approximately $230. The spending gap for basic programs was $562 and for total instruction $910. The authors note “much like a profit-maximizing firm, charter schools generate a surplus of revenue over expenditure.”
Baker and Miron (2015) show that in New Jersey, charter school administrative expenses are “nearly $1,000 per pupil higher than those of other regular school district types, and the share of budgets allocated to administration is nearly double.”5 Further, they show that local public school districts maintain responsibility for providing some services to charter school students, and thus, the administrative overhead associated with those responsibilities. That is, on a per pupil basis, district administrative expenses are being overstated and charter school administrative expenses understated. Additionally, these publicly reported administrative expenses do not include, for example, expenses (including executive salaries) from regional or national management organizations above and beyond management fees, further potentially understating total administrative expenses of the charter schools.
In addition, the uneven reshuffling of children and resources across schools within geographic boundaries can exacerbate inequities. Numerous studies find charter schools to increase segregation across schools of children by their income status, language proficiency, disability, and race (Stein 2015; Ladd, Clotfelter, and Holbein 2015; Kotok et al. 2015; Logan and Burdick-Will 2015; Moored 2015). Under some models, such as the New Orleans charter system, stratification exists by design (Adamson, Cook-Harvey, and Darling-Hammond 2015). Baker, Libby, and Wiley (2015) find that through the sorting of children and resources, charter expansion induces inequities within districts—inequities among charter schools and between charter and district schools. But, to the extent that charter share remains small, these inequities remain limited. Profit status of charter operators also may lead to very different available school-level resources available to children, as for-profit schools seek to achieve profit margins, while nonprofits seek to enhance revenues through tax exempt private giving (Weber and Baker 2015a). Finally, the concentration of needy children in some schools can dramatically increase the costs of improving outcomes for those students. That is, uneven sorting of children by needs can create additional inefficiencies (Baker 2011).
Finally, it is conceivable that the dissolution of large centralized school districts and introduction of multiple school operators into a single geographic space could compromise efficiency associated with economies of scale, which operate at both the school and district level. Numerous studies of education costs have found that the costs of providing comparable services rise as district enrollments drop below 2,000 students and rise sharply at enrollments below 300 students. Further, a comprehensive review of literature on economies of scale in education by Andrews, Duncombe, and Yinger (2002, 245) find “there is some evidence that moderately sized elementary schools (300–500 students) and high schools (600–900 students) may optimally balance economies of size with the potential negative effects of large schools.” To the extent that charter expansion creates independent “districts” operating with fewer than 2,000 pupils and/or increases shares of children attending schools with smaller enrollments than those noted above, inefficiencies may be introduced.
Charter school market share growth in U.S. cities
In this section, I identify cities and school districts that have experienced dramatic growth of their charter school sectors over the past few decades. Specifically, I focus on cities and school districts where total numbers of enrolled students exceed 20,000 and where charter school shares of enrolled students exceeded 20 percent by the year 2013. From these districts, I hand select illustrative cases across regions and states, and across different financing models and enrollment constraints. While revealing some commonalities, the selected examples illustrate the need to explore this issue on a case-by-case, city-by-city, and state-by-state basis.
The goal is to understand how the growth in charter schools’ share of district enrollment, across various urban contexts, affects enrollments and in turn finances of traditional public school districts.
For each city context, in the “Enrollment and charter operators in U.S. cities” section, I summarize:
Trends in charter school and host district (or citywide) enrollments
Distributions of enrollments by specific management company operators
Changes to average charter and district school enrollments
Resulting school enrollment size distributions
Of particular interest in this discussion is whether the shifts in enrollments from district to charter schools have resulted in increased shares of children enrolled in what would be considered “inefficiently small” schools, whether upstart charters or declining district schools. Of secondary interest, discussed further below, are the types of operators that have penetrated different markets.
In the “Fiscal effects of charter market growth” section, I summarize a series of fiscal indicators over time for each urban context, including:
Host district per pupil revenues by source (local, state, federal) and where possible (and relevant), district transfers to charter schools
Host district per pupil expenditures and expenditure shares on overhead functions, including:
Building and general administration
Plant operations and maintenance
Transportation
Of particular interest here is whether the reduction of enrollments caused by students transferring from district to charter schools is resulting in a manageable decline in total revenues, given declining enrollments of host districts. Potential indicators of fiscal stress include increased per pupil expense and greater budget shares allocated to administrative overhead, increased per pupil expense and greater budget shares allocated to maintaining an underutilized aging capital stock, and/or increased transportation expense resulting from remaining student populations that are more dispersed.
A number of background issues are important here. First, there are different types of charter schools. I have already addressed above the different authorizer alternatives. Expanding on the previous description, charter schools may either be fiscally dependent on local public school districts or fiscally independent.
Fiscally dependent charter schools rely on local public school districts to pass along the revenues associated with the children to be educated by the charter school. Under these models, host districts often retain responsibility for providing some services, including such things as transportation, food service, special education services, enrollment management systems, and in some cases facilities. As such, districts may retain some share of revenues to cover the costs of these services. That share may be defined in state charter law, or may be a matter of local board policy. In some cases, districts are also fiscal dependents of municipalities. That is, the district budget is part of the city budget. In these cases, the district operating within the city and charter schools operating within the city are each dependents of the city (charters not dependent on the district). As noted previously, districts may or may not have leverage over the rate of growth of these schools, depending on how they are authorized.
Fiscally independent charter schools often receive funding directly from their host states, based on the pupils they serve. These schools essentially operate as independent local education agencies, located geographically within an existing traditional district LEA (local education agency) and drawing students from within and outside of the host district geographic boundaries. The subsidy sent to the charter school by the state is then not sent to the district that would have otherwise served the students (sending district), i.e., it is deducted from the state aid. Similar arrangements exist between school districts in states with interdistrict choice programs. Geographic hosts to fiscally independent charter schools may or may not be the primary senders of students to those schools. This is certainly the case for cyber charter schools. Geographic hosts also likely have little leverage over the growth of fiscally independent charter schools, even if those schools draw primarily from the geographic host district.
Fiscal dependence matters in many ways, but most notably in terms of policy implications. Fiscally dependent charter schools operate under the policy umbrella of the district responsible for serving children in the same urban space. Thus, it makes sense that they be considered a fully integrated part of this local system, where district leadership—or portfolio managers at the district level—take responsibility for management and distribution of enrollments, annual operating resources, and capital assets. Fiscally independent charters present a different set of issues, competing as independent entities in the same space as host districts. But those districts are merely their spatial/geographic host, not their financial or management host. Thus, state policymakers rather than local districts must regulate rational, equitable distribution of schools, children, and resources, with consideration for the equity and adequacy of the system for all children.
These complexities introduced by chartering merely add to existing layers of governance complexity. In some states local public school districts are largely aligned with other geographic governing units, such as cities or municipalities or counties. That is, one school district serving elementary through secondary grades, in one city or county. In other cases smaller cities or towns might operate only K through sixth-grade or K through eighth-grade schools, and send their high school students to regional districts. These arrangements often exist in densely populated metropolitan areas as well as remote rural areas. Phoenix, Arizona, for example is home to a complex overlay of elementary school districts and high school districts, along with numerous fiscally independent charter operators. It can be incredibly difficult to determine charter school impact on any one geographic host under these circumstances.
There are also those cases where school district boundaries simply have no relation to municipal boundaries for a variety of historical reasons (Fischel 2010; Holme and Finnigan 2013). Such is the case in Kansas City, Missouri, a district with significant charter school market share growth, in a state that restricted, until recently, charter schools to the boundaries of Kansas City and St. Louis school districts. But, Kansas City Public Schools is only a small portion of the city—a predominantly black, low-income portion carved out of the central city, with boundary changes leading to further racial isolation as recently as 2006 (Green and Baker 2006).
In this section, we also provide an overview of the mix and distribution of charter management companies that have entered various large urban markets. We introduce these managers for a variety of reasons. First, leading (by market share) charter operators in many cities studied include those that have been the subject of federal and state investigations and judicial orders regarding conflicts of interest (self-dealing) and financial malfeasance, including Imagine Schools, Inc., White Hat Management, National Heritage Academies, and Concept Schools (Baker and Miron 2015). The rate of growth of suspect providers may be a result of lax authorization and oversight and may in fact be coupled with more rapid expansion of chartering in general. Second, these variations in financial and educational practices post not only overall quality and accountability concerns, but also equity concerns, where access to different providers across the urban space is uneven.
Figure 2 provides a nationwide map of the distribution of charter schools as of 2013. The population dense Northeast Corridor from Washington, D.C., to Boston certainly has its share of charter schools. So do large metropolitan areas in California and Texas. But charter schools are notably widespread and dispersed throughout other states including North Carolina, Arizona, Michigan, Ohio, Wisconsin, and Florida.
Figure 2
Figure 3 uses the city or municipal boundaries as the unit of analysis for identifying locations with the greatest penetration of charter schooling. The notable outlier is New Orleans, where, following Hurricane Katrina’s destruction of the city’s school system and displacement (much of it permanent) of the city’s student population, the state pursued a strategy of replacing the remaining district-operated schools with a centrally governed Recovery School District consisting entirely of charter schools. In the past few years (since these data) the system has converged on over 90 percent charter enrollment. New Orleans is not shown in the graph but the data are available in the underlying table. Although not shown in the figure, Highland Park, Michigan, a much smaller city carved out of Detroit, also shifted control of its schools to a single charter operator (Leonia Group) in 2012 (Banchero and Dolan 2012).
Figure 3
Charter market penetration in other cities has grown more steadily and linearly ending at between 25 percent and 41 percent by 2013. Major cities where school district and city boundaries are largely contiguous include Cleveland, Columbus, and Toledo in Ohio; Detroit; Philadelphia and Harrisburg, Pennsylvania, and Washington, D.C. Other cities are embedded within Florida counties (Pembroke Pines and Homestead). Charter market share for their host counties, however, remains much lower.
Figure 4 applies the host district geographic boundaries as the unit of analysis for portraying market share in the locations with the greatest penetration of charter schooling. There is some overlap between figures 3 and 4. Again, New Orleans, the notable outlier, is not shown in the graph but the data are available in the underlying table. Other Rust Belt cities in Pennsylvania, Ohio, and Michigan, where school district and city boundaries are contiguous, remain on the list. Deletions to the list (including Florida cities) bump Indianapolis up into the top 10. The notable addition to the list is the Kansas City school district, where the bounded school district within the city has very high charter market share, even though the city as a whole does not.
Figure 4
Enrollment and charter operators in U.S. cities
Here, I take a closer look at the enrollment effects and provider distributions across seven cities, including New Orleans. Charter advocates frequently cite research studies on the successes of high profile charter school operators including the Knowledge is Power Program and Harlem Children’s Zone. Far less attention has been paid to the vast majority of management organizations that actually run charter schools across the country, especially in more saturated markets in less regulated states such as those I discuss here. Because there are similarities across Ohio cities with large charter market shares, rather than discuss them each herein, I substitute one city, Newark, New Jersey, from just outside the top 10 in market share, a city in a state with more tightly regulated (single, state authorizer) charter sector.
New Orleans and the Katrina effect
As noted above, the conversion of the New Orleans public school system to a more than 90 percent charter system was conscious and deliberate. As such, it was governed by a central authority, involved a centrally managed transportation and student assignment plan, and centralized accountability over charter operators. It also involved, as shown in Figure 5, complete displacement and selective repopulation, racially, economically, and geographically, of the student and family population of the city. It is a model of so many moving parts and unique circumstances that its applicability to other contexts is extremely limited. Immediately after Hurricane Katrina in 2005, the district and charter populations, which had been around 66,000 students combined (and 80,000 in 2000), dropped to under 10,000. Since that time, total enrollment has come back to about half of its original (albeit already declining) levels of the early 2000s. Traditional district schools have now been eliminated.
Figure 5
Figure 6 shows the distribution of charter operators in New Orleans, based on data from 2011–12, and classifications from Miron and Gulosino (2013) adapted by Weber and Baker (2015a). At the time of these data, district schools (within city boundaries) still served about 29 percent of children. While about 5 percent of enrolled children are in KIPP schools (higher than most other cities) and some children are in schools operated by UNO Charter School Network and Edison Learning, Inc., most charter schools in New Orleans are operated by “other” operators, not major national or regional providers. Many are local or regional. Algiers Charter School Association is a locally based but apparently large chain. Again, all Recovery School District schools operate under a centralized enrollment/assignment system and accountability system.6
Figure 6
Washington, D.C.
Washington, D.C., presents a more typical example of charter school enrollment growth within a context of relatively constant total school enrollment (Figure 7). Total charter and district enrollment (excluding private enrollment) in the District of Columbia has remained constant at around 70,000 students since 2000. As charter enrollment has increased, district enrollment has dropped under 50,000, but appears to have stabilized from 2009 to 2013.
Figure 7
The mix of operators in Washington, D.C., is also diverse (Figure 8). As noted above, the district continues to serve the lions’ share (62 percent) of enrollments. Some areas of the district, notably the more affluent northwest quadrant, are home to far fewer charter schools. “Other charter” operators remain the largest group of charter operators. KIPP schools serve 3 percent of total enrollments. An array of other regional and national providers each serves relatively small shares. Again, these schools operate under a single authorizer and a central application and assignment system, permitting the district to maintain, and financially plan for, a balanced system (My School DC n.d.).
Figure 8
Figure 9 shows the average enrollments per school for all schools and for charter schools from 2000 to 2013. In a city with relatively constant citywide population, school average enrollments have declined, but leveled off as new schools—charter schools—have been introduced and expanded. By 2013 average school sizes for both charter and district schools were similar at about just below 300 pupils for charter schools and above 300 for all schools. But this is a relatively small school enrollment total, even for elementary schools, and near the lower bounds for efficient operation (Andrews and Yinger 2002).
Figure 9
Detroit, Michigan
Figure 10 displays total and district enrollment trends from 2000 to 2013 in Detroit, Michigan. Detroit presents a unique case of rapid and long-term school depopulation. Total enrollment is declining and district enrollment has a similar rate of decline, with charter enrollments holding relatively constant. As such, the charter share is increasing. Whereas traditional district schools once served over 170,000 pupils, they now serve fewer than 50,000.
Figure 10
Figure 11 shows the distribution of providers in Detroit. These providers are authorized by multiple entities, including universities, and are fiscally independent of the district. Thus they can grow to the extent their authorizers will allow (or shop for new authorizers) and largely set their own rules regarding enrollments, lotteries for oversubscription, and backfilling (whether they will take on new students to fill seats if/when other students exit). Detroit includes a diverse array of nonprofit and for-profit operators, including one with a suspect track record, National Heritage Academies (4 percent of enrollments) (Singer 2014). The largest portion of charter enrollments are in the group “other charter” operators. With 5 percent of enrollments, Leona Group holds the largest share among single operators, and is also the operator of Highland Park Schools, a separate smaller district surrounded by Detroit. Notably absent are any high profile providers on which rigorous peer reviewed studies have been conducted.
Figure 11
Average per school enrollments in all Detroit schools and Detroit charter schools have converged on the 350 to 400 pupils per school range, slightly larger than in Washington, D.C., and seemingly stabilizing in recent years (Figure 12).
Figure 12
Columbus, Ohio (and other cities)
Columbus, Ohio, and other major Ohio cities present more modest cases of charter sector growth, but as with Detroit, operate in a context in which the city has little leverage over charter growth within its boundaries. Ohio “community schools” are authorized by a number of independent authorities and funded directly by the state, with the state funding removed from district allotments. As will be seen in the next section, however, this funding is reported as a pass-through from sending districts.
Columbus presents a somewhat typical example of charter expansion and enrollment growth in Ohio (Figure 13). Total students enrolled in district and charter schools within city boundaries remained somewhat constant around 70,000 until an uptick to nearly 78,000 in recent years. District enrollment slid from about 70,000 at the outset of charter growth in 2001 to around 50,000 by 2013 with some apparent leveling off in recent years.
Figure 13
Figures 14, 15, and 16 show the mix of for-profit and nonprofit operators of charter schools in Columbus, Dayton, and Cleveland. KIPP has a minor presence in Columbus. But across these three cities, operators with a presence include White Hat, Imagine Schools, and Concept Schools. According to news reports, White Hat has engaged in questionable practices (Akron Beacon Journal 2015). Specifically it was chastised by the Ohio Supreme Court for engaging in questionable, self-enriching contracts with charter school boards. Imagine Schools recently lost a federal court challenge in Kansas City, Missouri, over its questionable management (“double dealing”) and financial practices (Robertson 2015), and Concept Schools are also no stranger to legal claims of financial impropriety (Saunders 2015). Imagine Schools was also under fire in St. Louis (Crouch 2011), Columbus (Candisky and Siegel 2014), and in Florida (Matus and Solochek 2012) for suspect real estate dealings and charging exorbitant lease costs to charter school boards, at taxpayer expense.
Figure 14
Figure 15
Figure 16
In Dayton, White Hat controls 7 percent (as of 2011–12) of charter enrollments while Imagine and Concept control a percent and 2 percent respectively. Edison Learning also holds a 3 percent share. In Cleveland, the distribution is similar with White Hat at 4 percent and Concept and Imagine at 3 percent and 2 percent respectively. The Cleveland market includes an additional manager, Constellation, serving 3 percent of enrollments.
Figure 17 takes us back to Columbus, and shows the average school level enrollments over time. Enrollments for charter schools and for all schools have leveled off around 400 pupils.
Figure 17
Kansas City, Missouri
Kansas City, Missouri, presents a peculiar case in many ways. Figure 18 shows that total enrollments in the school district (which covers only part of the city) were already in decline from 2000 to 2007, at which time the remaining predominantly white northeast corner of the district (which was within the city boundaries of Independence) was permitted to annex itself. Thus, the sharp dip in total enrollment that follows. Also during this period, the state of Missouri permitted charter schools to be established only within the boundaries of Kansas City Public Schools and the St. Louis Public School District.
Figure 18
Since 2000, total district and charter enrollments combined are down from just over 38,000 to just over 26,000, with district enrollment down to about 17,000. Charter enrollment has grown over that time, but charter market share has grown largely as a function of total enrollment decline (including annexation). Enrollments seem to have stabilized in recent years.
As Figure 19 shows, average enrollments per school, which dipped below 300 in the late 2000s, appear to have rebounded, to over 350 students.
Figure 19
Newark, New Jersey
Newark, New Jersey, presents another more modest and more regulated case of charter expansion. New Jersey has only one authorizer, the state itself, and charter schools receive pass-through funding from host districts. Unlike Ohio, Michigan, and Pennsylvania, New Jersey does not permit for-profit management companies from operating charter schools. Total student enrollments in Newark have remained relatively constant (Figure 20). District enrollments have declined from near 43,000 in 2004 to around 33,000 by 2013, with the biggest dips in