The Washington PostDemocracy Dies in Darkness

The link between charter school expansion and increasing segregation

Placeholder while article actions load

One thing that proponents of the broad expansion of charter schools never talk about is the evidence of how charters are leading to increasing segregation by race, ethnicity and income. Here George Washington University Research Professor Iris C. Rotberg explores this connection. She is Research Professor of Education Policy at The George Washington University. She focuses on school reform, the education of low-income children, school choice, international education, and international comparisons of student achievement. She is the editor of Balancing Change and Tradition in Global Education Reform, which describes education reforms in 16 countries. This first appeared in Kappan magazine, a publication of  PDK International, a professional education association.

By Iris C. Rotberg

In remarks at the 2013 annual meeting of the American Educational Research Association, Secretary of Education Arne Duncan emphasized the importance of “compelling educational research” and expressed concern “that today educators and policy makers still have a large unmet need for relevant research. . . . Sadly, school leaders and educators too often have to guess when they make education policy.”

The fact is we don’t have to guess about the consequences of one of the Obama administration’s most visible policies: the national expansion of charter schools. We need only turn to a large body of relevant research showing that charter schools, on average, don’t have an academic advantage over traditional public schools, but they do have a significant risk of leading to increased segregation.

In spite of this, the policy on charter schools remains a centerpiece of the administration’s initiatives (as it was, in a different form, in the Bush administration), despite abundant evidence that the policy is inconsistent with the longstanding goal of promoting school integration.

Although there has been considerable public attention to test-based accountability and to comparing student achievement in charter and traditional public schools, there has been less attention to the link between charter schools and increased segregation. A policy that exacerbates existing levels of segregation should be a major concern, particularly in the current environment: large inequalities in income and wealth, a widening gap in student achievement between affluent and low-income students and implementation of state voucher and tax plans, which further contribute to student stratification.

This article considers how a policy promoting the expansion of charter schools risks increasing segregation based on race, ethnicity, and income. It also considers the potential for increasing the segregation of special education and language-minority students and for contributing to religious and cultural stratification not typically found in U.S. public education.

Federal policy and research evidence

The Obama administration has promoted expanding the number of charter schools, both through its public advocacy and through the Race to the Top (RttT) competition. RttT gave states a strong incentive to reduce or eliminate caps that had previously limited charter school expansion. Nationally, the proportion of charter schools to public schools has tripled since 2000 and, in the last several years, some states have accelerated that trend in response to RttT.

The conclusions summarized in the sections that follow are based on a wide array of research in the United States and in other countries. The research review of school choice programs in the United States is focused on charter schools to reflect the focus of the Obama administration. The research review in other countries includes a broader set of programs because the 10 countries reviewed use a variety of school choice initiatives — academies, vouchers, or subsidies — in structuring their education systems. For purposes of analyzing segregation effects, however, these various initiatives operate in very similar ways.

The studies reviewed used a mix of methodologies. Some compared the demographic characteristics of students in school choice programs with those in the traditional public schools they would have attended. Others compared the characteristics of students in school choice programs with those in the surrounding communities. Case studies were also conducted to increase understanding of the reasons for the choices families and schools make. Regardless of the specific methodology used, however, the preponderance of research evidence leads to the following conclusions.

#1. There is a strong link between school choice programs and an increase in student segregation by race, ethnicity, and income.

Studies in a number of different states and school districts in the United States show that charter schools often lead to increased school segregation, a finding that is consistent with research in a number of other countries, including Australia. In many cases — not all — school choice programs exacerbate current school segregation and, in more heterogeneous settings, lead to the stratification of students who were previously in integrated environments.

The primary exceptions to increased student stratification are in communities that are already so highly segregated by race, ethnicity, and income that further increases are virtually impossible, or they occur in school choice programs that are targeted to increase diversity — not a goal of most charter schools or school choice programs generally.

#2. The risk of segregation is a direct reflection of the design of the school choice program.

Certain design features magnify the risk of segregation. For example, a growing number of charter schools target specific racial or ethnic groups and therefore lead directly to increased segregation. In addition, several other designs are particularly vulnerable to increased segregation. Segregation effects are especially pronounced in charter schools run by education management organizations, as well as in large, unregulated choice programs. Partial government vouchers or subsidies to which families must add the remaining tuition costs virtually guarantee increased segregation because many families can’t afford the costs. Increased segregation is also a predictable outcome for programs that select students based on their achievement levels because of the high correlation between socioeconomic status (SES) and achievement, compounded by the fact that low-SES students are often less likely to be referred to selective programs even when their achievement levels are high.

Some school choice programs do have a positive effect on integration. A small proportion of charter schools are designed specifically to increase diversity. In addition, magnet school programs, which were originally started to increase integration, have often succeeded in doing so. However, like other school choice programs, magnet schools tend to segregate when diversity is no longer a specific goal.

The uphill battle faced by magnet schools demonstrates how strongly the odds are against programs that do not focus on diversity. Families often choose schools based on their perceptions of the extent to which other families in the school community are “similar” to them. Even when the choice is based on other considerations, such as the characteristics of the educational model, high-SES families have far greater leverage in gaining access to the most competitive schools, both in finding the information needed to choose among schools and in having the resources (for example, to support transportation costs) to enable their children to attend the schools.

Charter schools, even under a lottery system, also choose — sometimes explicitly and sometimes indirectly — and increase the probability of segregation. They limit the services they provide, thereby excluding certain students, or offer programs that appeal only to a limited group of families. Some charter schools also exclude students from consideration because their parents can’t meet the demanding parent involvement requirements, or they expel students who haven’t met the school’s academic or behavioral requirements. Charter schools also choose where to locate which, in turn, influences enrollment options given the transportation difficulties for low-income students.

In some communities, charter schools have a higher concentration of minority students than traditional public schools. In others, charter schools serve as a vehicle for “white flight.” School segregation increases in both cases — in the charter schools students attend and in the traditional public schools they would have attended. This outcome can be offset only if the choice program has a specific goal to increase diversity.

However, the federal role in encouraging charter school diversity has been minimal. Although legislation in some states includes provisions on diversity, without oversight, the legislative language has had little effect. Advising charter schools to be diverse will not make it happen.

#3. Even beyond race, ethnicity, and income, school choice programs result in increased segregation for special education and language-minority students, as well as in increased segregation of students based on religion and culture.

Special education and language-minority students are under-represented in charter schools, unless the schools are specifically targeted to these population groups. Even when the students are selected in a lottery, they are discouraged from attending charter schools when the schools do not provide the services they require.

Perhaps less visible, but clearly growing, are charter schools that target specific religious and cultural groups. Some of these schools were formerly private religious schools, or schools that are likely to attract specific religious groups (for example, by offering extensive language instruction in Hebrew, Arabic, or Greek), or schools designed to appeal to families with particular social or political values. These “niche” schools, therefore, often result in the segregation of students by religion or by social values — a type of stratification many countries now struggle with that has not traditionally been prevalent in U.S. public education. As charter schools proliferate, so do these schools — a trend that will almost inevitably lead to a public school system that is increasingly fragmented.


The research evidence shows the risk of policies that have led to a largely uncontrolled expansion of charter schools. Yet, the evidence has had little influence on public policy. Despite hundreds of studies on school choice, the general perception is that we have little research information or that the information we do have is ambiguous.

Researchers bear some responsibility. Research reports often conclude by saying, “We need more research.” The conclusion apparently stems from a belief that inconsistent results are the same as ambiguous results and, therefore, of little use in policy formulation. Yet, the inconsistency is a reflection of the reality that charter schools vary depending on the purpose and design of the programs and the settings in which they’re implemented. We won’t discover the single “right” answer about the effects of charter schools no matter how many studies we conduct.

Moreover, when researchers simply conclude that we need more research, they miss the opportunity to communicate to policy makers the implications of the large body of research on school choice that already exists. The variance in findings is not a negative; it is an essential basis for policy formulation. If we fail to communicate clearly to policy makers the implications of the evidence we have, we should not be surprised when research is not used. The fact is we know why certain initiatives lead to increases in segregation along different dimensions: race, ethnicity, income, religion, and social values, as well as for students who need special education services or English-language instruction. We also know why certain initiatives — although relatively few — achieve diversity and how that is accomplished.

Federal policy, however, applies to school districts nationwide, regardless of their characteristics or the design of their charter school initiatives. It does not distinguish among initiatives based on their probability of increasing student integration. A policy that encourages states to expand charter schools applies across the board both to programs that are designed to facilitate integration and to the far larger number that are likely to increase segregation.

I am not under the illusion that by modifying federal policy on charter schools we would solve the basic problem of segregation. But we could at least eliminate one factor exacerbating it: the federal pressure on states and school districts to proliferate charter schools, even in situations that might lend themselves to increased segregation. Instead of serving as a cheerleader for charter schools, the federal government might instead support diversity in schools and, at the same time, publicize the risks of increased student stratification.

Even apart from the negative effect of increased segregation, justifying federal advocacy of charter school expansion is difficult when there’s no evidence that charter schools, on average, are academically superior to traditional public schools or even that they can be more innovative given the Common Core State Standards and the testing associated with them

The finding in Brown v. Board of Education that “separate education facilities are inherently unequal” has been demonstrated repeatedly in the United States and throughout the world in the 60 years since that decision. It has also been demonstrated in the results of the 2009 Program for International Student Assessment. It is ironic that the political rhetoric surrounding this assessment focuses almost exclusively on test-score differences among countries, which account for only about 11% of the variance, while little attention is paid to the far more important finding that the remaining variance is accounted for by differences within countries. On average, almost 60% of the differences in reading test scores within member countries in the Organization for Economic Cooperation and Development are explained by the SES of students and schools. In the United States, the SES of students and schools explains almost 80% of the variance in performance. That finding is certainly not a strong recommendation for policies that further increase the segregation of schools.

It is also ironic that as other countries become increasingly concerned about the social implications of their school choice programs, the United States is promoting the expansion of these programs. Until now, the link between charter schools and segregation has been partially masked by the fact that a large proportion of charter schools are in urban areas that are already highly segregated. However, as charter schools expand into areas with more diverse student bodies, their segregating effects will become even more extensive and visible. That expansion is already underway; although the largest increases in charter schools to date have occurred in cities, significant increases are also occurring in towns, suburbs, and rural areas that are more diverse.

It is not that government has an agenda to increase segregation. Proponents of charter schools believe they’re giving low-income and minority students opportunities they otherwise would not have had. That belief is true in some cases; all charter schools do not result in segregation. But far too many do, and the trend is unfavorable. It takes a lot of care through targeted funding and oversight to mitigate the pressures that lead to yet more segregation. But whatever motivations drive the choices families and schools make, it is important that government does not exacerbate the problem of segregation by ignoring the unintended consequences of its policies. The risk is an increasingly divided public education system.

The Internet version of this article ( includes extensive citations and references.