Civil Rights Project’s Response to Re-analysis of Charter School Study

By Guest Bloggers Erica Frankenberg, Genevieve Siegel-Hawley, and Gary Orfield

The Civil Rights Project was founded, in part, to bring rigorous social science inquiry to bear on our most pressing civil rights issues. On-going trends involving public school segregation have been a primary focus of the CRP’s research, and the expanding policy emphasis on school choice prompted analysis of the much smaller – but politically potent – charter sector. In 2003, and again this year, we have released reports examining charter school segregation.  Both times we have been subject to attacks, often for doing or saying things we never claimed to undertake in the reports.  On April 27, 2010, Education Next posted a reanalysis and commentary, by Gary Ritter and several colleagues, of our February 2010 charter school report. Below, we respond to the team’s claims, seeking to accurately explain what we did, why we did it, and the actual nature of our conclusions.  Before proceeding, however, it is important to revisit what we know about the continued significance of school segregation.

Decades of social science research describing the harms of racially isolated schools and the benefits of diverse schools has informed the development of federal policy–and even many states’ charter school legislation.[1] In 1954, the Supreme Court declared that separate was inherently unequal, and as recently as 2007, a majority of Justices affirmed that there are compelling reasons to voluntarily pursue integrated schools and to prevent racially isolated schools.  Though extraordinary high-poverty, high-minority schools exist, social science repeatedly shows them to be the exception to the rule. The reason: racially and socioeconomically isolated schools tend to have unequal educational resources, higher drop out rates, and prepare students poorly for life after high school.  Furthermore, schools of white segregation – as well as those that concentrate students of color – do not provide the educational opportunities for students to learn to challenge stereotypes and live and work in a diverse society.  These are vital skills for 21st century citizens and workers, and are among the reasons that it is critical to consider not just whether families have access to schools, but also the demographic composition of students in those schools.

Responding to Methodological Criticism

The major empirical distinction between our analysis and the Ritter team’s reanalysis is the level to which we aggregated school-level student composition patterns.  In other words, trends in the racial makeup of charter schools were examined at different levels of geography. As we describe in detail in our report, we aggregated school-level charter enrollment data to the national, state, and metropolitan level.  The use of the latter geographic unit, metropolitan areas, stemmed from a deliberate methodological decision.  As 1998 guidance from the U.S. Department of Education noted, “…charter schools often draw students from outside their home district’s attendance boundaries and are sometimes treated as a school district or a separate entity.”[2] We also cited a study from Arizona that found that charter schools within one traditional public school district pulled students from 21 distinct districts.[3] Indeed, charter advocates originally promoted not being attached to particular school districts as one of the strengths of the movement.

Ironically, the Ritter team reanalysis used district boundary lines in an effort to justify the charters’ segregated enrollments–even though charter schools are often not constrained by those boundaries.  They described their decision to compare the racial enrollment of charter schools to central city schools as “the best available unit of comparison,” arguing that the geographic concentration of charter schools in urban areas merits a comparison of schools only located within urban districts.

We disagree. Given that charter schools can and do enroll students across traditional boundary lines, our analysis took into account the demographic composition of students in the entire metro area, as opposed to a single school district.  Using a metropolitan area as point of comparison allowed us to consider segregation within a smaller geographical area – compared to our state-level analysis – where students can conceivably choose to attend either traditional public or charter schools.[4] The urban concentration of charter schools is irrelevant if charter schools are drawing students from across boundary lines. And, similar to the results from our analyses at the national or state level, in many of the metro areas containing at least twenty charter schools, minority segregation was higher in charter schools than in the metro’s regular public schools.

While the national, state, and metro area analysis comprised the bulk of our report, we did, in fact, examine the segregation of students in charter and traditional public schools by geography—comparing students in these school sectors within cities, suburbs, and rural areas.  Significantly, we found that the geographical skew of charter schools mitigates very little of the differences in minority segregation.  Fifty-two percent of city charter school students were in 90-100% minority schools, compared to only 34% of traditional public school students—a difference of eighteen percentage points, very similar to the overall difference of twenty percentage points between the two sectors of schools (Table 22 on p. 63 of our report).

Even with a more circumscribed approach, the authors of the reanalysis still found that charters are more segregated than traditional public schools—which themselves are extremely segregated.[5] This central finding, together with our study, only reinforces our ultimate conclusion: it is critical to consider what kinds of choices we are offering families in urban, suburban and rural areas across the country, and in charter or traditional public schools alike. Currently, the data show that we are in the process of subsidizing an expansion of a substantially separate – by race, class, disability and possibly language[6] – sector of schools, with little to no evidence suggesting that it provides a systematically better option for families–or that access to these schools of choice is fairly available to all.

Contextualizing our Report

Our study does not stand in isolation from the growing research consensus concerning segregation in charter schools. Numerous studies, including six separate analyses by the U.S. Department of Education (each of which relied on state-level data), have concluded that charter schools are more segregated than traditional public schools.[7] Ritter and his colleagues described their findings as being similar to a recent report conducted by RAND. Our reading of the same report led us to include the following statement in our literature review:

The [RAND] study determined that in five of the seven locales, the movement of black students to charter schools meant these students attended more segregated schools (Zimmer, et al., 2009; see also Bifulco & Ladd, 2007). In the remaining two locations, Chicago and Milwaukee, black students attended slightly less segregated charter schools than they would have if they remained in public schools, though both traditional school systems contained very low percentages of white students (Zimmer, et al., 2009). The study also found more mixed enrollment patterns for white and Latino students (p. 11).

Using a district-level comparison, then, the RAND study, like the Ritter et al. reanalysis, found evidence of higher racial segregation among charters.

It is extremely important to emphasize that our report focused on a number of other civil rights dimensions in charter schools, none of which have been addressed in these recent postings.  Although the racial composition of schools is critically important to student outcomes, we also examined the extent to which charter schools are serving low-income and English Language Learner students.  California charter schools provided a stunning example of the gaps in the data that exist for ELL students, reporting that just seven ELLs enrolled in state charter programs. We also found that, depending on the data source, between 20 and 25 percent of charter schools show no evidence of offering the National School Lunch Program, thus calling into question whether they are enrolling low-income students.  Schools not offering the subsidized lunch program also tended to overlap with schools having a higher concentration of white students, highly suggestive of the existence of a set of charter schools serving disproportionate numbers of non-poor, white students.


Though charter schools, in many instances, have not yet successfully utilized choice as a means to create diverse schools, we have longstanding evidence that it can be done.  Across the country, some of the most diverse, sought-after schools are magnet programs that combine choice with civil rights provisions to insure access and integration.  Inter-district magnet schools in Connecticut provide a current example outside the scope of traditional school districts as to the way charters might draw students across district boundary lines to create high-quality, integrated schooling options.  Today, existing school segregation typically occurs across district boundary lines, so schools that enhance inter-district choice are extremely important.  Charter schools have the opportunity to contribute to integration across those lines, and our report offered concrete suggestions for incorporating civil rights provisions into charters as the sector expands.

Having set forth our disagreements, we conclude by noting that we find common agreement with Ritter and his colleagues on the need for more complete and accurate data on charter schools. The data is essential if we are to fully understand the role charter schools play in the racial, economic, and linguistic segregation of students. Based on a wealth of existing evidence, however, we are unable to share in the team’s optimism that more complete data might show narrower differences in segregation between charter and traditional public schools.  We urge the federal and state governments to improve publicly available data about charter schools and to monitor the civil rights of all students who attend or wish to attend charters, in addition to further examining the effects charter schools have on surrounding public schools.

[1] A number of states have legislation promoting or mandating the establishment of racially diverse charter schools. See Frankenberg, E. & Siegel-Hawley, G. (2009). Equity Overlooked: Charter Schools and Civil Rights Policy. Los Angeles: UCLA Civil Rights Project. Available at:

[2] U.S. Department of Education. The Charter School Roadmap: Finance and Funding (Archived information). Washington, D.C.: U.S. Department of Education. Available at:

[3] Gifford, M., Ogle, M., & Solomon, L. (1999). Who is Choosing Charter Schools? A snapshot of geography and ethnicity of charter school students. Center for Market-based Education.

[4] In many states, considerable inter-district choice options also exist.

[5] Frankenberg, E., Lee, C., & Orfield, G. (January 2003). A multiracial society with segregated schools: Are we losing the dream? Cambridge, MA: The Civil Rights Project at Harvard University; Frankenberg E. (2009).  The demographic context of urban schools and districts. Equity and Excellence in Education 42(3): 255-271.

[6] Frankenberg, E., Siegel-Hawley, G., & Wang, J.  (February 2010).  Choice without equity: Charter school segregation and the need for civil rights standards. Los Angeles: Civil Rights Project/ Proyecto Derechos Civiles; Mathis, W., Miron, G., Tornquist, E. & Urschel, J. (2010). Schools without Diversity: Education Management Organizations, Charter Schools and the Demographic Stratification of the American School System. EPIC/EPRU. Available at:  Finnegan, K., Adelman, N., Anderson, L., Cotton, L., Donnelly, M. B., & Price, T. (2004).Evaluation of Charter Schools Program: 2004 Final Report. U.S. Department of Education. Policy and Programs Study Service. Available at: Nelson, B., Berman, P., Ericson, J., Kamprath, N., Perry, R., Silverman, D. & Solomon, (2000). The state of charter schools 2000: Fourth-year report. Washington, D.C.: U.S. Department of Education, Office of Educational Research and Improvement.

[7] For a summary of these studies, please refer to pp. 9-13 in our report. Available at:

Last Updated


Notify Me When Education Next

Posts a Big Story

Business + Editorial Office

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

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

Copyright © 2024 President & Fellows of Harvard College