Charter Schools and Race : A Lost Opportunity for Integrated Education

Segregation patterns in the nation's charter schools are studied. After reviewing state charter legislation that directly addresses issues of racial and ethnic balance of student enrollment, we briefly examine the racial composition and segregation of the charter school population nationally. School-level analyses, aggregated by state constitute the primary method of studying segregation in charter schools. First, we look at racial composition and segregation of charter schools by state. Then, we consider the differences in segregation between non-charter public schools (or simply "public schools" for convenience) and charter schools, as well as segregation within the charter school sector. We conclude with a discussion of the article's findings and recommendations to promote further racial

inequality on many dimensions, this isolation can have serious consequences for students.This article details a disappointing set of findings regarding its central question-charter schools are largely more segregated than public schools.Segregation is worse for African American than for Latino students, but is very high for both.In some states, white student isolation in charter schools is as high as that of African Americans.The problems reported here may not be due either to the intent or the desires and values of charter school leaders.They may reflect flaws in state policies, in enforcement, in methods of approving schools for charters, or the location where charter schools are set up.
The justification for segregated schools as places of opportunity is basically a "separate but equal" justification, an argument that there is something about the schools that can and does overcome the normal pattern of educational inequality that afflicts many of these schools.Charter school advocates continually assert such advantages and often point to the strong demand for the schools by minority parents in minority communities, including schools that are designed specifically to serve a minority population.It is certainly true that minority parents are actively seeking alternatives to segregated, concentrated poverty, and low-achieving public schools.(Note 6) White parents have also shown strong interest in educational alternatives as evidenced by the strong demand for magnet schools.
Unfortunately, despite claims by charter advocates, there is no systematic research or data that show that charter schools perform better than public schools.Since charter schools embody wildly different educational approaches and since charter and public schools obtain their enrollment in very different ways, evaluation and comparisons between the two require very careful analysis.At a minimum, it is certainly safe to say that there is little convincing evidence for the superiority of charter schools over public schools in the same areas.In fact, some of the studies suggest that charter schools are, on average, even weaker.(Note 7) Authorization of charter schools is different in each state that has approved them.Charters permit and even welcome an enormous variety of innovative educational approaches, though they support very traditional approaches as well.Some of the charter founders are idealistic education leaders with a great new idea, strong imagination and inexhaustible energy, while some are committed community activists who have longed to run their own schools, or to serve only one group in a community, and many are managed by corporations that hope to profit from their operation.For many charter school founders, there is an implicit assumption that less government control and oversight will produce positive educational benefits.
One of the problems in evaluating the academic effectiveness of charter schools is that their effect is normally examined by comparing them to regular public schools, but their student body and parent groups are not the same, which makes the comparison of academic achievement inaccurate.Even if one were able to control for income, parent education, and other relevant, easily measurable family resources, there are several kinds of selection bias that make such comparisons virtually impossible.First, the families who are informed enough to choose a school and make the effort to get their child to a more distant school every day are not the same as the families who do not.(Note 8) Second, charter schools commonly lack the expertise and programs to serve students who are English Language Learners or severely disadvantaged children such as those in Special Education.As these students tend to score lower on standardized tests, if students from lower achieving groups do not enroll, the school's average scores will tend to rise.(Note 9) Third, many charters seek applications from students they believe would succeed, or who would respond to their approach, while not recruiting others.Some schools have screening procedures that public schools are prohibited from using because the public schools are required to serve all students.These biases mean that even if there were higher test scores or lower dropout rates for charter schools it might well be because of selective recruitment-students from families with more resources and/or fewer students with special needs-than because of the school's superior educational approach.
Curiously, in an era in which tests and accountability have been the hallmark of education policy, there has been little serious accountability for charter schools.Theoretically charter schools must meet the terms of their charter or they will be terminated.In most states, however, there are few resources for oversight of schools and revocations of charters for educational failure, as opposed to financial problems, are rare.(Note 10) Often their impact on racial segregation is ignored by the policymakers, despite the growing body of research evidence that has documented a trend of segregation in charter schools.If there is no real evidence linking superior performance to educational program rather than admissions selectivity, looking at general characteristics of the student body that are usually linked to educational inequality, such as levels of segregation, certainly deserves attention.On this front, there is little positive to say about these schools.
One might well think that charter schools would have a better chance to be integrated than public schools.Like magnet schools a generation earlier, charter schools offer distinctive curricula and the opportunity to create and manage schools with freedom from many normal constraints in large districts.Unlike magnet schools, charter schools have the added advantages of even greater freedom to innovate and for the most part, are not tied to geographically fixed attendance boundaries in residentially segregated communities as are neighborhood public schools but can draw from wherever interested students can be found (in some places where school districts grant charters, they are limited to the school district boundaries.) The high level of racial segregation in charter schools is not a surprise when viewed in light of segregation in many aspects of American life.Those who think that charter schools are inherently likely to be free of racial inequality need to reflect on the racial consequences of other markets operating in areas of housing, employment, health care, etc. , where the markets have worked more to perpetuate and spread racial inequality than to cure it.One could accurately say that the normal outcome of markets when applied to a racially stratified society is a perpetuation of racial stratification.This is why early educational choice programs were often found to produce white flight from integrated schools and to contribute to segregation in many school desegregation trials.(Note 11) Those experiences were apparently unknown or overlooked by designers and supporters of many charter school policies.
In looking at the data presented here it is worth considering the experience of magnet schools.There have been a handful of highly selective schools in American public school systems, such as Boston Latin, San Francisco's Lowell High, and New York City's Stuyvestant High, which have produced remarkable students for generations.Overall, however, choice of schools and specialized curriculum for schools (except for vocational schools) were very rare in the U. S. until desegregation policies produced the magnet school movement in the mid-1970s.Magnet schools, like charter schools, grew rapidly in response to federal grant programs.The magnet school programs funded by the Emergency School Aid Act, however, had desegregation policies while the federal charter school law did not.The charter school law was a movement backward to the unregulated choice policies common 40 years ago across the South and in many big cities.Those did not work to produce integration and charter school policies do not either.
Racial segregation in charter schools needs to be considered as both a critical problem and a lost opportunity.Experience shows, that segregation is not inevitable and that it is possible to produce quite different outcomes with appropriate civil rights policies.As we approach the 50th anniversary of Brown v. Board of Education, this issue should be addressed and resolved.If we are to be serious about the impact of charters on minority opportunity in American schools, we need to look with considerable suspicion on unfounded claims of sweeping benefits, insist that accountability be extended fully to this sector, and not reach conclusions on the basis of assumptions rather than evidence.This article should broaden the discussion of the future development of charter schools.Certainly any publicly funded schools should not be run in ways that either intensify racial isolation or undermine integrated schools in integrated neighborhoods.Charter schools offer opportunities, like good magnet schools, to create successful and voluntary diversity.Clearly there are some very ambitious and

Introduction
In the school year 2000-01, 1,855 charter schools were operating in 34 states that had passed legislation authorizing the creation of charter schools.(Note 12) Charter schools educate fewer than one percent of all public school students yet can have a substantial local impact on surrounding districts in terms of student enrollment.Most of these charter schools are concentrated in a few states, and in most states are located in urban areas.Charter schools in the sixteen states covered in this article (see Table 6 for list of states and their enrollments) make up more than 95% of the population of charter school students.(Note 13) Among different states there is great variation in the percentage of minority students attending charter schools.(Note 14) One reason for this variation could be that charter school reform has been supported by a diverse array of politicians and educators.Nonetheless, as publicly-funded schools of choice, it is important to examine whether these schools offer white and minority students interracial exposure when segregation across the country is increasing for black and Latino students, and white students are more racially isolated than students of any other racial/ethnic group (Frankenberg, Lee & Orfield, 2003).
In the past, most educational choice options (such as magnet schools) arose from desegregation plans (American Institutes for Research, 1993).In l973 the U. S. Supreme Court extended desegregation requirements to northern and western cities.However, just a year later, the Court rejected the lower Detroit court's proposition that integrating minority students in heavily minority and rapidly changing districts required including the suburbs to produce long lasting desegregation.Big cities looking at demographic facts and seeing the conflict over mandatory reassignments of students in cities such as Boston looked for a way to accomplish desegregation through voluntary choice.The problem was that very few whites had ever voluntarily chosen to attend black schools or to transfer for integration purposes.The idea of the magnet schools movement was to create specialized schools that could offer unique opportunities that would create a demand for voluntary transfers from both white and minority students and result in a student population that would meet desegregation standards (American Institutes for Research, 1993).By establishing special programs and curricular offerings in inner-city areas, school systems used magnet schools and programs to attract white students to predominantly minority schools.This movement became central to the desegregation strategies of cities such as Milwaukee, Cincinnati, and Buffalo.Furthermore, a title was written into federal law offering funds for such schools when they served desegregation purposes.The idea led to the creation of many highly popular and often well-integrated schools in districts that had few such opportunities and was strongly supported by school superintendents and boards.Because of the explicit emphasis on racial/ethnic balance, magnet schools are often among the most integrated schools in their district (Blank, Levine and Steele, 1996).By 2001 there were a reported 1,736 magnet schools in the county and there had been federal support for them for a quarter century.They enrolled 3. 0% of American students, compared to the 1. 2% in charter schools (Hoffman, 2003).Thus, magnet schools were a well-established model long before the charter school movement began.
Recently educational choice options have proliferated, through the growth of charter schools, vouchers, inter-and intra-district choice, magnet schools, and private schools.Building on the increasing belief of the importance of parents to have choice in their child's education, the No Child Left Behind Act (NCLB) of 2001 (Public Law 107-110) further expanded school choice by allowing students in failing schools to transfer.(Note 15) The rationale is that the achievement of poor and minority students will improve if they have access to schools that have demonstrated higher levels of student performance.It also assumes that parents will be able to make decisions about what education is best for their children, which will force schools to compete-and ultimately improve-to keep and/or attract students.
Charter schools, a form of school choice that is almost a decade old, represent a further attempt to institute school choice within the public education sphere.A charter is merely a political, legal, administrative and financial arrangement of relative autonomy, created in a somewhat different form in each state that has authorized them.The belief is that by introducing such choice options into the public schools, students and their parents could choose the school that was most appropriate, which would create incentives for all schools to improve in order to compete for students (Apple, 2001).Literature on school choice is mixed as to whether these assumptions are correct and actually result in improved education for all (for a brief discussion see Kim and Sunderman, 2003).
Since its inception, the charter school movement has been politically charged for both its proponents and opponents.Even within the charter school reform there is a diffuse group of supporters who favor charter schools for widely varying reasons; two of the major driving forces behind the charter school reform have been the excellence movement including high standards for all students and market-driven reforms aimed at making schools more efficient.Charter supporters say that such schools give important new options for parents, allow for educational innovation, and are not constrained by typical school district boundaries and student assignment practices that produce segregated patterns of schooling in many neighborhood school systems (Finn, Manno & Vanourek, 2000;RPP International, 2000).
Politicians have also supported the rapid growth of charter schools: NCLB also provided monetary assistance to increase the number of charter schools in states with charter legislation.(Note 16) In fact, while many public schools and districts across the nation are facing substantial budget cuts, President Bush has proposed $700 million in spending for charter schools.
Critics of school choice, however, argue that competition among schools will only improve student achievement if all schools are able to compete and students are equally free to choose.Otherwise, those students who are left behind by those who choose or are chosen in more competitive environments will have even less resources with which to compete (Arsen, Plank & Sykes, 1999).Those opposed also suggest that choice systems can compromise the public good by educating students in isolation from others for their private good, often further stratifying students along racial and socioeconomic lines (Cobb & Glass, 1999;Horn & Miron, 2000;Hochschild & Scovronick 2003).Additionally, school choice policies can allow schools to exclude students with special needs if it does not fit within their mission (Howe and Welner, 2002).The theory of market solutions rests on assumptions about choice in charters-that there is full information for everyone, that there are not economic or other barriers to attendance, an that the school will welcome students from all backgrounds.A great deal of experience with choice plans and magnet schools before the charter experiment show that knowledge and access were often very unequal, that families with the most resources and information often got access to the most highly regarded schools, that students from other races often felt unwelcome unless there were special efforts to recruit and support them in the new school, and that free transportation was essential to assure choice and access for lower income families (Fuller and Elmore, 1996).Choice plans that operate without these kinds of supports permit choice for only those who know what the choices are, how to access them, and do not need support to get to school.
Given these claims and counterclaims about charter schools, there remain important questions that should be addressed, particularly in the area of racial/ethnic segregation of students, which has been largely ignored in the overall debate about charter schools.Are charter schools offering students better opportunities for interracial exposure than the increasingly segregated public schools?Past research has shown that minority students attending integrated schools are more likely to attend and succeed in college, as well as to live and work in interracial settings (Wells & Crain, 1994;Eaton, 2001;Braddock II, 1980).Additionally, recent research by The Civil Rights Project has documented a number of important educational and civic benefits for students of all races in desegregated high schools (Kurlaender & Yun, 2001).Segregated minority schools, where minority students experience little interracial exposure, are highly correlated with schools of concentrated poverty.Eighty-six percent of the students in all public schools that have greater than 90 percent black and Latino students of their total enrollment are also in schools where at least half of the student body is poor.(Note 17) These schools are more likely to have lower average test scores, less qualified and experienced teachers, and fewer advanced courses (Young & Smith, 1997).Moreover, research on charter schools has shown that charter schools with higher proportions of minority students tend to have fewer resources and less academic curricula than charter schools serving mainly white students (Fuller, Gawlik, Gonzales, Park & Gibbings, 2003).As publicly-funded schools, it is essential that charter schools provide equal educational opportunity for all students.
Charter school proponents claim that charter schools provide options for low socio-economic students (Finn, Manno & Vanourek, 2000).Preliminary analyses question whether charter schools are, in fact, achieving this goal of educating low-income students.At the national level, in 1997-8, 39% of charter school students versus 37% of public school students received free and/or reduced lunch.Miron and Nelson (2002) report that, based on data from half of Michigan charter schools, when examining the student poverty composition of charter schools in comparison to their surrounding districts, charter schools serve a slightly lower percentage of low-income students; there are similar and even stronger trends in California, Massachusetts, and Colorado (SRI International, 1997;Wood, 1999;Clayton Foundation, 1999).At the district level, Ascher and colleagues (1999) found that only 35% of charter schools were socio-economically diverse (between 20% and 80% of students on free/reduced lunch) as compared to 72% of public schools in surrounding districts.However, it is difficult to determine the level of student poverty in charter schools because many of these schools do not participate in the federal free/reduced lunch program, which is the most common measurement of the socio-economic status of students (Wells, Holme, Lopez, & Cooper, 2000).(Note 18) More analysis is needed to accurately ascertain the levels of student poverty in all charter schools-particularly in the many segregated charter schools that exist across this country.

Research Questions
Recently, issues of accountability and equity for charter schools have come under greater scrutiny (Cobb & Glass, 1999;Wells, 2002).However, as the 2001 RAND book, Rhetoric versus Reality: What We Know and Need to Know about Vouchers and Charter Schools, concludes, given the different conditions under which charter schools operate, we really do not know much about the issue of racial segregation in charter schools (Gill, Timpane, Ross, & Brewer, 2001).Because of their complexities-they are both public but also independent from the public school system, they can choose their students but also need to attract students, and they are governed by state charter legislation yet are influenced by their local context and mission-it is difficult to know how to even evaluate charter schools.
There has been some research to suggest that black students have a relatively high level of access to charter schools (Wells, et al. , 2000).Recent state evaluations have found that even though the aggregate racial composition of charter schools is similar to host districts, there are great differences at the school level in enrollment compositions (Cobb & Glass, 1999;Miron, Nelson and Risley, 2002).
In this article, we address one key aspect of the multi-faceted charter school phenomenon with the following questions: What is the racial/ethnic composition of charter schools?What is the average exposure of charter school students to students of other races in their schools?How are charter school students distributed among the charter schools?Are students more racially isolated in charter schools than in public schools?
There is strong evidence that many Americans believe in the importance of integrated education.Sixty percent of blacks in 1998 and 34% of whites believed that it is "absolutely essential" for schools to "have a diverse student body with kids from different ethnic and racial backgrounds" (Farkas and Johnson, 1998).Further, a national poll in 1999 reveals that 68% of all respondents believe that integration had "improved the quality of education" for blacks and 50% believe that it had made education better for whites.By 1999, almost three-fifths of Americans believed that we needed to do more to integrate schools (Gallup, 1999).Certainly there is also substantial support for choice policies; in 1993, 65% of the public were in favor of allowing students and parents to choose what schools they attended, regardless of where they lived (Elam, Rose & Gallup, 1993).However, as subsequent discussion will illuminate, despite many parents' preferences for integrated schools and choice policies, many state charter laws are not explicitly supportive of racial diversity in charter schools.

Data and Methods
We compare the racial composition of charter schools with that of all non-charter public schools by examining who is enrolled in charter schools and the extent to which they are segregated.Although in 2000-01 charter schools enrolled fewer than one percent of all public school students in the country, many of these schools are concentrated in certain areas and states, and can have a substantial local impact on surrounding public school district enrollment and racial diversity.We focus on the sixteen states that had total statewide charter enrollments of at least 5,000 students in 2000-01.Charter students in these sixteen states account for 95.4% of the entire U. S. charter school population.The data analyzed for this article are from the National Center for Education Statistics 2000-01 Common Core of Data (CCD).The CCD is a comprehensive, yearly national dataset of all operational public schools and includes school information on student characteristics such as enrollment and racial counts (Note 19) that are comparable across states and between charter schools (Note 20) and non-charter public schools.
In examining issues of charter school segregation, we use several measures to evaluate different school-level dimensions of segregation.By aggregating the school-level data to the state level we are able to compare charter and public schools within a particular state as well as charter school segregation across states.The exposure index provides an average picture of the interracial exposure of students: the index can be interpreted as the percentage of students of a particular racial group in the school of the average student of another group (Massey & Denton, 1988;Orfield, Bachmeier, James & Eitle, 1997;Reardon & Yun, 2002).For example, Michigan's charter school white-black and white-white exposure rates (Note 21) of 16% and 78%, respectively, (Table 8), mean that, on average, Michigan's white charter school students attend a school where 16% of students are black and 78% of the students are white.If students were evenly distributed (e.g., no black-white segregation), all Michigan charter school students would, on average, attend schools that are 54% black and 40% white, respectively, a racial composition equal to the proportion of white and black students in Michigan's total charter school enrollment (Table 6).These exposure indices demonstrate that white charter school students in Michigan, on average, attend schools that disproportionately enroll high levels of white students and low levels of black students.
Examining the exposure index gives us an average picture of interracial exposure in charter schools.However, this measure, which is essentially a weighted average of the racial composition of schools of students from each race, can mask the variation and distribution of students in schools.For example, if black exposure to white students in charter schools is 50%, that could describe two schools that are both 50% white, or could be one school that is 90% white and one school that is 10% white.These two examples would have very different implications in terms of the interracial experience of students in charter schools.To explore the distribution of students in charter school, we examine the concentration of students of all races in predominantly minority schools (greater than 50% of the student body is non-white), intensely segregated minority schools (90-100% minority), and intensely segregated white schools (90-100% white).Together, these measures portray both the actual level of interracial exposure i schools as well as the percentage of students attending racially imbalanced and isolated schools.
It is important to note that using schools as our unit of analysis, this article analyzes the racial composition and exposure at the state level.Previous studies at the district-and school-level have shown that when examined in terms of their local contexts (comparing the racial enrollments of charter schools to that of the surrounding public school district or the closest public school), charter schools are less racially diverse than local public schools and districts (Wells, et. al, 2000;Ascher, Jacobwitz, & McBride, 1999;Cobb & Glass, 1999).We recognize that the context of where schools are situated locally and how districts choose to interpret state charter legislation are important considerations that likely influence the outcomes we examine.However, we do not specifically address that in this article.One reason is because our data do not allow us to examine these questions at a more local level.However, it is potentially misleading to look at charter schools at the district level, (Note 22) because in many states charters are not necessarily part of a school district or confined to drawing students only from surrounding districts.(Note 23) One characteristic common across all charter schools is the statewide nature of charter school legislation.This orientation influences the context in which all charter schools throughout the state must operate.In addition, who can attend charter schools, how many can be established, and by what means they enroll students are just some of the stipulations in charter school legislation that differ widely among states.Demographic contexts of the entire state population also vary across the country and these variations can affect the racial composition of the students in charter schools.Furthermore, although charter schools can enroll students across district and county lines throughout metropolitan areas, charter schools do not enroll students across state lines.A comparison between charter schools and public schools at the state level gives us important comparisons of the racial composition and segregation in the small but growing sector of charter schools within legislatively defined geographic boundaries Our purpose in this article is not to discount the variation that occurs at the district-and school-level, but simply to focus on state-level observations of differences in racial composition between public schools and charter schools and how students are distributed among charter schools.

Findings
In the sixteen states with charter school enrollments greater than 5,000, we find that charter schools in most of these states enroll disproportionately high percentages of minority students, particularly African American students.Over half (56%) of all charter schools in these states are located in central cities.Specifically, we find the following trends for charter school students by race: Seventy percent of all black charter school students attend intensely segregated minority schools compared with 34% of black public school students.In almost every state studied, the average black charter school student attends school with a higher percentage of black students and a lower percentage of white students.White students in every state studied attend schools with a much higher white percentage than their overall share of the charter school population.In many states, however, white charter school students are exposed to substantial percentages of non-white students.Furthermore, there are pockets of white segregation where white charter school students are as isolated as black charter school students.The pattern for Latino segregation is mixed; on the whole, Latino charter school students are less segregated than their black counterparts.
In sum, although many of the charter laws require compliance with desegregation orders or mandate specific racial/ethnic balance in charter schools, there is little evidence of serious effort at the state level to ensure racial balance.

Charter Legislation
In a reform with such a diverse array of schools and ideologies, one of the few consistencies for the charter schools in a state are the state charter school legislation and guidelines under which all schools are supposed to operate.This legislation and regulations vary significantly among states.More than half of all states with charter school laws have policies that require charter schools to comply with desegregation standards or reflect student racial/ethnic populations in the state (see Table 1).In most cases, the state or local education agency (usually a school district or the state department of education but can vary in some states), and not the state itself, authorizes the charter schools and reviews and regulates the schools.
Although the charter school reform is primarily governed by policies set by each state, there are federal regulations and programs that may also affect the composition of the student body of charter schools.In 1994, a new federal grant program was implemented to support charter schools as part of the Improving America's Schools Act.(Note 24) Charter schools can receive funding through federal programs such as, the Eisenhower Professional Development Program, the Safe and Drug Free Schools Act, and the Perkins Occupational Education Act.However, federal funding can only be used if charter schools comply with federal civil rights statutes such as Title VI.NCLB provides funding to schools with high levels of student poverty (formerly known as Title I of the Elementary and Secondary Education Act) -but accepting NCLB money means that these schools must comply with federal civil rights provisions.(Note 25) Likewise, althoug states individually pass their own charter legislation, if charter schools receive money from the federal Public Charter Schools Program, they are required to use a lottery to admit students in the event that there are more applicants than available slots for the school.
Courts have held that in school districts under federal court desegregation orders, charter schools will not be allowed to impede compliance with a court's or administrative entity's desegregation plan.(Note 26) However, even in this instance, the federal guidelines as to the responsibilities of charter schools in such situations are unclear, at best.(Note 27) Moreover, there is not a general framework to support such efforts.Charter schools are often given little support in implementing these guidelines, and in general, there is not a framework to support those who value racially diverse schools (Wells, 2002).(Note 28) Of all states with charter school legislation, nineteen states have specific racial/ethnic balance enrollment guidelines for their charter schools (Table 1).(Note 29) Without these rules, charters have little incentive to maintain racial/ethnic balance in their schools.Two of the four states with the largest enrollment of charter school students (Arizona and Texas) have no racial/ethnic guidelines.There are also some states that include equity provisions-such as providing free transportation to all students or requiring information to be widely available-that are important in ensuring that students from all backgrounds are truly able to choose to enroll in charter schools.Nine states with racial balance policies are included in the state-level analysis of this article (those states with charter enrollment greater than 5,000).Interestingly, six of the nine states in our analysis that have specific racial/ethnic guidelines are southern states.(Note 30) Among the seven southern states in our group of sixteen states with at least 5,000 charter students, only Georgia has no racial balance provision.
Moreover, the language of the racial/ethnic balance provisions varies from state to state.In some states, general guidelines regarding non-discrimination on the basis of race is used; fewer than ten states require compliance with desegregation orders.We find that despite the specific racial/ethnic balance guidelines in charter legislation, many states still have racially imbalanced enrollments.Because many state regulations call for district proportionality and this analysis is primarily state-level, more research is needed at the district level to determine the impact of the guidelines.Perhaps even state charter laws with racial/ethnic balance language are still too weak; without other equity provisions built in to this market-based reform, charter schools are unlikely to overcome the persistent segregation of our larger society.

Rhode Island 557
Charter school must have a program to encourage the enrollment of a diverse student population, and the makeup of the school must be reflective of the population of the district.

South Carolina 483
Racial composition of charter school enrollment may differ by no more than twenty percent from school district or targeted student population, but local school district may find charter school not operating in racially discriminatory manner without regard to twenty percent requirement.

National Trends (Note 31)
While they can mask considerable variation among the states' implementation of the charter school reform, national statistics provide a helpful background in which to consider charter school students and their distribution among schools.In the 34 states with charter schools in 2000-01, less than half (43%) of all charter schools students were white.Another one-third (33%) were black and one-fifth (19%) were Latino.Asian and Native American students make up a very small percentage of the charter school enrollment.The national non-charter public school population has a much higher percentage of white students (a difference of sixteen percentage points) and a lower percentage of black students than charter schools (Table 2).(Note 32) The percentage of black students in charter schools is almost twice the total black public school enrollment.The share of Latino students in charter schools versus public schools is comparable.The fraction of Asian students in charter schools is slightly less than their proportion of the total public school population, while that of Native Americans is slightly more.Almost ninety percent of black charter school students are in predominantly minority schools where minority students are more than 50% of the student body (see Table 3).Seventy percent of all black charter school students, over 100,000 students, are in 90-100% minority charter schools.This number is striking when compared to the 34 percent of black public school students who attend 90-100 % minority schools.Although the public school figure (34%) is the highest it has been in three decades, the charter school distribution suggests even more segregation.(Note 33) These numbers indicate that black students are not only disproportionately over-enrolled in charter schools, but that they are enrolled at a much higher rate than other black public school students in intensely segregated minority schools.
White charter school students are also more likely to be in predominantly minority and intensely segregated minority schools than white public school students.The percentage of white students in such schools, however, is much lower than students of any other race, in both charter and public schools.(Note 34) Higher percentages of Latino and Asian charter school students attend intensely segregated minority schools than their public school peers, but their rates of attendance in predominantly minority schools are similar.Eighty-three percent of white charter school students are in majority white schools (Table 4).About one-fifth (22%) of all white charter school students nationwide are in schools that have a student body that is more than 90% white, a rather high percentage due to the fact that the majority of students in charter schools are minority students.
Not surprisingly, given their high concentration in minority schools, black charter school students are the least commonly found in predominantly and intensely segregated white schools.Ten percent of black charter school students attend majority white schools and only about one percent is in 90-100% white charter schools.These rates are substantially lower than those of students of other racial groups except Latinos.Interestingly, Latino students are the most segregated from whites in public schools, but Latino charter students-while still highly segregated from white students-are less segregated than black charter students.Just over one-fifth (22%) of all Latino charter school students are in majority white schools, twice the percentage of black students in such schools.While Native American public school students are exposed to a higher share of white students than students of any other minority group, in charter schools, Asian students are more commonly enrolled in white schools than other minority students.Nationally, the average white charter school student attends a school that is 72% white.White exposure to black and Latino students is fairly even: the percentage of black and Latino students in the average white charter student's school is 12 and 11 percent, respectively (see Table 5).White exposure to other racial minorities is low, in part due to the small percentages of Asian and Native American students attending charter schools.
On average, black and Hispanic students are disproportionately exposed to higher percentages of students of their own race in charter schools.For example, the average black charter school student attends a school that is 73% black and only 14 percent white.The percentages of Latino and white students in the charter school of the average Latino are 52% and 26%, respectively.Perhaps due to their low enrollment in charter schools, Asians and Native Americans are exposed to more whites than are either black or Latino students.Note: Totals may not add to 100 due to rounding.
In sum, at the national level, blacks are over-enrolled and whites are under-enrolled in charter schools relative to public school enrollment.Black charter school students are overwhelmingly found in intensely segregated minority schools, and are more segregated from white students than black public school students.However, for white charter school students, the story is quite different.Because whites make up a relatively small percentage of the charter school population, they are exposed to more blacks and Latinos and to fewer white students in charter schools than in the public schools at the national level.(Note 35) For Latino students, at the national level, public and charter school segregation rates are similar.
Because aggregation of racial composition and segregation at the national level can obscure more localized variation, it is also important to see how these trends vary by state between charter and public schools.The over-enrollment of black students in charter schools indicates segregation between charter schools and public schools; therefore, it is important to also examine the distribution of students within the charter sector.This paper, then, looks at each of these issues in turn.

State-Level Trends: Racial Composition
In 2000-01 there were sixteen states with at least 5,000 students in charter schools, but the number of students enrolled in these schools and the racial composition of the schools varied widely across states.California, the most populous state, has the largest charter school population with over 100,000 students in charter schools during 2000-01.On the other end of the spectrum, there are 18 states whose charter school enrollment totals less than 5,000 and are not included in our state-level analysis.Of the sixteen states with substantial charter school enrollment, nine have guidelines specifying racial balance in the state charter school legislation (see Table 1 above for racial/ethnic balance guidelines in all states with charter legislation).
Among all public school students, only six states have a majority non-white student body (Frankenberg, Lee & Orfield, 2003).For charter school students, the picture is very different: only six of the states with a substantial charter population have a majority of the charter school enrollment that is white (see Table 6).In fact, six states have charter enrollments that are more than 50% black.Eight states have at least 15% of the charter school enrollment composed of Latino students.Asian students account for a very small percentage of students enrolled in charter schools; only in California and Minnesota are Asian enrollments greater than 5% of the total charter population.The Native American population is also small in all states except Minnesota and Arizona.(Note 36) In the sixteen states included in this study, charter schools were predominantly located in cities.Table 7 displays the location of charter school students in each of the sixteen states, ranked by percentage of white charter students in each state.(Note 37) States with higher percentages of charter school students in cities were less likely to have large white enrollments, similar to trends in large central city public school districts.(Note 38) Overall more than half of the charter school students in these sixteen states attended schools that were located in central cities (56%) while a third (34%) were in schools located in suburban areas.Missouri (100%), Ohio (98%), Illinois (94%), and Texas (87%) had the highest proportion of their charter school students in cities and were four of the five states with the lowest percentage of white students of their total charter enrollment.Charter schools in these four states educated almost one-sixth of all charter school students.Only three states, Florida (52%), Georgia (63%), and Colorado (46%) had greater percentages of their charter school students enrolled in schools in suburbs than in cities. (Note 39) Generally, as can be seen from Figure 1, states with the lowest proportion of white students in their charter schools also had the highest proportions of their charter school students in central city schools while states with the highest proportion of white charter school students were those that have higher proportion of charter school students enrolled in suburban areas.Note:States with zero percent of charter schools in a given location may have less than three bars.For example, 100% of Missouri's charter schools are in urban areas, so there is no suburban or rural bar for Missouri.
As mentioned above, the demographics of the states' populations and public school enrollments vary widely.Thus, we examine how the state's charter school racial composition compares to the state's public school enrollment by race.In almost every instance, the white percentage of charter school students is smaller than in public schools.In ten of these states, the white percentage in public schools is at least twenty percentage points higher than the white share of total enrollment in charter schools.Half of these states are Midwestern states.Missouri shows the starkest contrast between public and charter white enrollment: the white percentage of the public school enrollment is more than eight times greater than the white charter school proportion.(Note 40) Four states have a greater proportion of white students in charter schools than in public schools (see Figure 2).
The reverse trend holds for black enrollment: in every state except Georgia, charter schools have a higher black enrollment share than public schools (see Figure 3).For example, in New Jersey, Ohio, and Missouri, although black students are less than 20% of total public school enrollment, black students make up more than 70% of charter students in these states, despite specific racial guidelines in the state charter legislation in all three states.Interestingly, Georgia has the highest black percentage of total public school enrollment and is the only state in which charter schools disproportionately enroll a lower proportion of black students.
In most states, the differences between Latino public and charter school enrollment are far smaller than for white and black students (see Figure 4).The largest difference is in California, where the Latino portion of charter school enrollment (34%) is ten percentage points lower than the Latino portion of public school enrollment (44%).The states with the largest under-enrollment of Latino students in charter schools (California, Arizona, and Colorado) are all in the West.

Exposure to Students of Other Races
We have already documented that charter schools, when compared to public schools at the state level, disproportionately enroll higher percentages of black students and lower percentages of white students relative to non-charter public schools, which suggests that segregation between charter and public schools exists both nationally and state by state.It is critical to more closely examine these distributions, to see whether students are enrolled evenly across charter schools or whether they are isolated in schools with students of their own race.One commonly used measure of segregation is the exposure index, which describes racial composition of the school attended by the typical student of a given race.

White Student Exposure
Within charter school sector As seen in Table 8, white students in every state attend schools with a much higher white percentage than their overall share of the charter school population.For example, although Missouri's white charter students are exposed to a lower percentage of white students (23% on average), this is more than twice the white share of charter enrollment (9%).Even in states where they are only a small percentage of charter school enrollment, whites are generally concentrated in schools with other white students and substantially isolated from students of other races.For example, in Illinois, Texas, and Ohio, where less than one in four charter school students is white, the average white charter student attends a school where more than 50% of the student body is white (Table 8).The isolation of white students in Illinois is particularly marked.Despite white students comprising less than 9% of the overall charter enrollment, the typical white student is in a school which is 54% white, a percentage that is six times higher than the white share of the state's charter school enrollment.
As a result of these relatively high levels of white isolation in charter schools, white students, in general, are exposed to lower percentages of students from other racial groups than would be expected by enrollment share alone.Except in four states with the highest black enrollment share (i.e. , Illinois, Missouri, New Jersey, and Ohio), the average white student attends a charter school where fewer than one in five students is black.Even in states where over half the charter school population is black, white students, on average, attend schools with more white students than black students.In Michigan, where over half (54%) of charter students are black, the average white student is exposed to five times as many white students as black students (white students, on average, attend charter schools that have 16% black students and 78% white students).
In most state, white exposure to Latino students in charter schools is lower than white exposure to black students, which might be due to the lower enrollment of Latino students in charter schools in some states.The four exceptions are in the West (Texas, California, Arizona, and Colorado).In ten states the average white charter school student attends a school with less than 10% Latino students.The high isolation of white charter school students in Illinois, however, prevents substantial white exposure to Latinos despite a relatively high charter school enrollment of Latinos.Illinois has the fourth highest percentage of Latino students in charter schools (23%); yet the average white student in Illinois attends a school that is only 9% Latino.
Charter vs. public schools Regardless of the type of school (i.e. , charter or public), the average white student attends a school with a higher proportion of white students than the state's aggregate percentage of white students, which suggests some sort of segregation mechanism at work.(Note 41) However, as noted above, white charter students in ten states are less isolated than public school students.This could be due to a lower percentage of white students enrolled in charter schools than public schools in these states, which would make it more difficult to create schools that were predominantly white.However, in states where the white share of total enrollment is similar in both public and charter schools (Note 42) (California, Florida, North Carolina, Arizona, and Georgia), the average white charter student is equally as isolated or more isolated in schools with other white students than the average white public school student.This provides support to the contention that it is not that charter schools are inherently doing a better job of integrating students, but rather that low white enrollments are responsible for the lower levels of white racial isolation in charter schools in most states.(Note 43) In terms of white students' exposure to minorities in charter versus public schools, white students in most states, on average, are more exposed to black students in charter schools than in public schools.In fact, the average white charter school student in all states except Colorado, Florida, North Carolina, and Georgia has greater exposure to black students (see Table 8) than does the average white public school student.This could be due to the disproportionately high enrollment of black students in charter schools.
There are few differences between public and charter schools for white exposure to Latino students.In four states-mainly in the West-white exposure to Latinos is lower in charter schools, than in public schools.However, these differences tend to be small.

Minority Student Isolation
Within the charter school sector Black charter students are heavily isolated in overwhelmingly black schools.This could be due partially due to their disproportionately high enrollment in charter schools relative to non-charter public schools.However, black isolation indices are well above proportional representation (e. g. , black share of total enrollment), which suggest something in the structure of charter school enrollment that acts to segregate black students, such as the large percentage of charter schools located in central cities.The exposure and isolation indices for black and Latino students in charter and public schools are presented in Table 9.Except in two states (Arizona and Colorado), black charter school students, on average, attend majority black charter schools.In almost half of the states, the average black charter student attends a school that is at least three-quarters black.Illinois provides an interesting example.In Illinois, 68% of the charter school enrollment is black and the typical black charter student's school is 77% black.Despite the fact that whites comprise fewer than 9 percent of Illinois's charter school population, however, the average white charter student's school is 54% white and only 34% black (Table 8).
Latino charter school enrollment patterns are mixed.In only eight of the sixteen states analyzed, Latinos comprise a higher percentage of the total charter school enrollment than the state's public school enrollment.In some states (Texas, California, Arizona, and Pennsylvania) there are relatively high levels of Latino isolation for the average Latino charter school student.Latino students in Texas experience the highest isolation of all Latino charter students with the typical Latino charter student attending schools where two-thirds of the student body is Latino.In most states, however, Latinos are less racially isolated than either black or white charter school students.
Charter vs. public schools Not surprisingly, given the higher proportion of black students enrolled in charter schools when compared to public schools, the average black charter school student is more isolated than his or her public school counterpart.Georgia is the only state (Table 9) in which black students are less isolated, on average, in charter schools than in public schools.This could be due to the fact that of all states, Georgia has the lowest percentage of charter school students in central cities (see Table 6).Whereas in eight states, black public school students attend schools where black students compose more than half of the student body, the typical black charter school student attends a majority black school in fourteen states.
The pattern of segregation for Latino charter school students is more varied: in six states (Florida, Massachusetts, Michigan, Minnesota, Missouri, and Pennsylvania), Latino students are more isolated in charter schools than in public schools.For example, the schools of the average Latino charter school student in Minnesota have three times as many Latino students as those of their Latino public school counterparts.However, there are also eight states in which Latino charter students are less isolated than Latino public school students.Overall, there is no clear pattern for Latino charter school student isolation.

Minority Student Exposure to White Students
Within the charter school sector Given the relatively high percentage of black students in charter schools and the levels of black isolation, we would expect to see low black exposure to whites in charter schools.It is surprising that even in states where there are more white students than black students in charter schools (e. g. , California, Wisconsin, North Carolina, Massachusetts, and Florida), the average black charter school student still attends schools with three to four times more black students than whites (see Table 9 and Table 10).For example, in Minnesota, where black students comprise fewer than one-quarter (23%) of all charter students (Table 6), the average black charter school student attends a school that is 70% black (Table 9) and only 17% white (Table 10).The average white student attends a school with a very different racial composition, one that is 83% white and only 7% black (Table 8).
In every state, Latino charter school students experience similar, or greater, exposure to white students when compared to the black charter school students.There are five states in which the average Latino charter school student attends school where there are at least 40% white students.While this could be due to the fact that in most states there is a much larger black share of charter school students than Latino students, we see that in states such as Texas, where the Latino and black student composition is similar (37% and 41%, respectively), the average Latino student attends a school with a greater percentage of white students (14% versus 9%, respectively).
Charter vs. public schools When comparing charter schools to public schools for minority students, in every state except Georgia (which as mentioned earlier, is the only state in which black charter student isolation is lower than black public student isolation), black exposure to whites is higher in public schools than in charter schools (see Table 10).Black students in the public schools of these fifteen states, on average, attend schools with substantially higher proportions of white students than the average black charter school student.In Illinois, both black charter and public school students have the lowest exposure to whites than in any of the sixteen states.
Whereas the state's black public school student exposure to white students is just under 20%, black charter school student exposure to white students is only 4%.
In eleven states, the average Latino student has lower exposure to white students in charter schools than public schools, and in some states they are substantially less exposed to white students than in public schools.For example, in Texas, a Latino public school student attends a school, which is, on average, 23% white; the typical Latino charter school student in this state has just over 14% white students in his or her school.In summary, the exposure and isolation indices suggest that, due to the disproportionately high enrollment of blacks and under-enrollment of white students in charter schools when compared with public school enrollment, the average white charter student attends a school with more minority students than the average white public school student.Conversely, because of the small proportion of whites, the average black-and to a certain extent, the average Latino-student is generally more isolated in charter schools than in public schools.Although white isolation among public school students is the highest, among charter school students, black isolation is as high as white isolation.Even in states in which white enrollment is higher than black enrollment in charter schools, blacks still attend schools with three to four times the number of white students.Latino charter school student segregation from white students is lower than that of black charter students, and is not uniformly more segregated in comparison to public school students in these states.While Latino charter student exposure to whites is higher than blacks, in most states it is still lower than that of Latino public school students.

Racial Segregation
Because the exposure index only shows what the average student experiences, we now turn to other segregation measures that examine how students are distributed across schools.To gain a clearer picture of the distribution of charter school students, we examine the percentage of students of each race that attend predominantly minority schools, intensely segregated minority schools, and intensely segregated white schools.

Predominantly Minority Charter Schools
Within the charter school sector The proportion of white, black, and Latino students attending charter and public schools where more than 50% of the student body is minority is presented in Table 11.The white share of enrollment in both charter and public schools-as well as the difference in white enrollment between the two-are in columns 1 to 3. For example, 9% of Illinois's charter school enrollment is white and 60% of its public school enrollment is white, a difference of 51 percentage points.Columns 4 to 6 show the percent of white, black, and Latino students who are enrolled in charter schools that are predominantly minority.In Illinois, 32% of white charter school students, 98% of black charter students, and almost all Latino charter school students are enrolled in 50-100% minority schools.As columns 7 to 9 show, 8% of whites, 82% of blacks, and 74% of Latinos attend 50-100% minority public schools in Illinois.Regardless of race, a higher percentage of charter school students attend predominantly minority schools when compared to public school students, which is not surprising given the much smaller percentage of white students in charter schools than in public schools in Illinois.
As discussed above, charter schools in twelve of the sixteen states enroll, in aggregate, a lower percentage of white students than public schools.In some states, these differences are stark.As column 4 shows, low percentages of white students in many states attend predominantly minority charter schools, regardless of the white share of enrollment.For example, charter school enrollment in Pennsylvania is 30% white, yet only 13% of white charter school students attend predominantly minority schools (Table 11).In fact, ten of the sixteen states have fewer than one-fifth of white charter school students attending predominantly minority schools.However, there are variations.In New Jersey, white charter school students are exposed to large proportions of students from other racial groups: 61% of white charter students in the state attend predominantly minority charter schools.This could be due to the small percentage of whites in charter schools (12%).Yet, in Illinois, a state with smaller proportion of white students in charter schools (9%), only 32% of white students attend predominantly minority charter schools, which seems to indicate that the charter school segregation of whites and blacks in Illinois is more extreme than in New Jersey.
For black students there is less variation in the percentage attending predominantly minority schools: in virtually every state there is a majority-and often an overwhelming majority-of black charter school students that attend schools with at least 50% minority students, regardless of the white proportion of the state's charter school enrollment (see column 5).In fact, in half of the sixteen states, over 90% of black students attend predominantly minority schools.This may be due to the low white charter enrollment in some of these states.However, even in some states where at least half of the charter school population is white (e. g. , Arizona, Georgia, Massachusetts, North Carolina, and Minnesota), at least three out of every five black students attend predominantly minority charter schools.In the case of Colorado, where 74% of the charter school enrollment is white, almost 60% of black students attend predominantly minority schools.Minnesota and North Carolina, states with racial guidelines in their charter legislation and where over 50% of the charter school enrollment is comprised of whites, have an overwhelming percentage of black students attend predominantly minority charter schools (91% and 83%, respectively).
As shown in column 6, Latino-white charter school segregation is less severe than black-white student segregation but is still high.For example, 60% of Latino charter students in North Carolina attend predominantly minority schools, whereas 83% of black charter school students attend such schools.By contrast, only 11% of white charter school students attend predominantly minority schools.Except in two states (Georgia and Colorado), at least half of Latino charter school students are in predominantly minority schools.In most states, however, a lower share of Latino charter students are in predominantly minority schools than are black charter school students.
Charter vs. public schools Comparing the enrollment rates of predominantly minority charter schools (columns 4 to 6) to that of predominantly minority public schools (columns 7 to 9) illustrates that in a majority of states, regardless of race, students are more likely to attend predominantly minority charter schools than predominantly minority public schools.This is especially true for black students.A higher proportion of blacks attend predominantly minority charter schools than public schools in all except two states (Georgia and Colorado).For Latino students, this is true in all except five states (California, Florida, Arizona, Georgia, and Colorado) (see Table 11).One possible explanation could be the relatively higher enrollment of white students in charter schools in these states by comparison to other states.In most states with lower white charter school enrollment than white public school enrollment, a higher percentage of white charter school students than white public school students are enrolled in predominantly minority schools.
Charter-Public Difference

Intensely Segregated Minority Schools
Within the charter school sector Examining the distribution of students in intensely segregated minority schools, it becomes even more apparent how isolated minority students are in charter schools.(Note 44) The percentage of white, black, and Latino students that are attending charter and public schools where more than 90% of the student body is minority is shown in Table 12.Columns 1 to 3 show the white share of enrollment in both charter and public schools.The percentage of white, black, and Latino students who are enrolled in intensely segregated minority charter schools are in columns 4 to 6, and the share of students by race enrolled in intensely segregated minority public schools are in columns 7 to 9. For example, Massachusetts, a state where white students comprise 54% of total enrollment in its charter schools, has 2% of white charter students, 56% of black charter students, and 40% of Latino charter students attending intensely segregated minority charter schools.In Massachusetts' publi schools, which have a greater percentage of white students enrolled compared to charter schools, a lower percentage of all students are in intensely segregated minority schools.Less than one-half of one percent of white public school students, twenty-three percent of black students and 18% of Latino students are attend these intensely segregated minority schools.
As column 4 shows, low percentages of white charter students are in intensely segregated minority charter schools.Except in three states (Illinois, Missouri, and New Jersey), fewer than 10% of white students in charter schools attend 90-100% segregated minority schools.Even in states where the white share of charter enrollment is very low, such as Illinois (9%) and Missouri (9%), only 25% and 21% of white students, respectively, attend these intensely segregated minority charter schools (see Table 12).However, if students were evenly distributed in Illinois charter schools, for example, every school would be 9% white and thus all white charter students (as well as all minority charter school students) would be attending the intensely segregated minority schools.
In every state except Arizona, Georgia, and Colorado, at least half of black charter school students attend 90-100% minority schools (see column 5 in Table 12).A striking example is Pennsylvania, where 80% of black charter school students attend intensely segregated minority schools.
Latino charter school students experience higher segregation than that of whites and lower segregation than blacks (column 6).Five states have more than half of Latino charter school students in intensely segregated minority schools.However, except for Illinois, the attendance of Latino students at 90-100% minority schools, while still high, is less severe than that of blacks.In Minnesota, the first state to enact a charter law (which includes racial/ethnic balance guidelines), and a state with very high white charter school enrollment, demonstrates high levels of charter segregation for minority students with roughly two out of every three black and two out of every five Latino charter school students attending intensely segregated schools.
Part of this segregation may be due to the higher percentage of minority students enrolled in charter schools, which results in more predominantly minority schools.But the racial disparities among these schools suggest that there is another factor aside from the racial composition in the state's charter schools that is driving these numbers.For example, as we have seen earlier, even if they are a small proportion of students in charter schools, whites are not as likely as black and Latino students to attend heavily minority schools.This indicates that the over-enrollment of minority students in charter schools is more likely to result in highly segregated schools for minorities than for whites.These trends of disproportionately high enrollment of minority students in intensely segregated schools could also be due to the fact that some of the charter schools are located in segregated central city neighborhoods.It is worth remembering, however, that charter schools as schools of choice are not limited to neighborhoods or even public school districts, but can draw students from a larger geographical area.
Charter vs. public schools Charter school students across all racial groups in most of the sixteen states are more likely to attend intensely segregated minority schools than are public school students (see columns 4 to 9).In both sectors, however, attendance at such schools differs substantially by race.In every state, a higher percentage of black students in charter schools than in public schools are enrolled in intensely segregated schools (see Table 12).In California, Arizona, and Texas, the three states with the largest charter school enrollment, black charter school students are attending intensely segregated charter schools at rates almost two times higher than black public school students.In Illinois, Missouri, New Jersey, Pennsylvania, Minnesota, and North Carolina, the share of black students attending intensely segregated charter schools is more than thirty percentage points greater than those in intensely segregated minority public schools.Of these states, North Carolina, Missouri, New Jersey, and Minnesota have racial guidelines in their charter legislation.
In all states, regardless of the type of school (i.e. , charter or public), fewer than 25% of white students attend 90-100% minority schools.It is worth noting that a higher percentage of white charter school students than white public school students are in intensely segregated minority schools in twelve of sixteen states.In three states, fewer than one percent of white charter school students are in 90-100% minority schools.

Intensely Segregated White Schools
Within the charter school sector Table 13 displays the percentage of students by race that attends intensely segregated white charter schools.(Note 45) Despite relatively low white charter school enrollment rates, there are only 2 states (Illinois and Missouri) without any students attending intensely segregated white charter schools (see columns 4 to 6), and in some states, white isolation is particularly stark.For example, despite the fact that about 60% of Michigan's charter school students are minority, 40% of white students attend intensely segregated white charter schools (see Table 13).In fact, in 10 states at least 15% of white charter school students attend intensely segregated white schools; in six states, over one-quarter of all white charter school students are in intensely segregated white schools.
In every state except Massachusetts, blacks are the least likely of all students to enroll in intensely segregated white charter schools.(Note 46) In no state are there greater than 4% of black students in intensely segregated white charter schools and, further, regardless of the white share of total enrollment, fewer than 10% of black students-public or charter-are enrolled in intensely segregated white schools in all states (column 5).
When compared to black students, higher percentages of Latino students are in intensely segregated white charter schools but still fewer than 10% of Latino charter school students in every state attend such schools (column 6).
Charter vs. public schools In most states, a lower percentage of white charter school students attend intensely segregated white schools than white public school students, which would be expected given the lower percentage of white students in charter schools.There are five states in which a higher proportion of white charter school students by comparison to public school students attend90-100% white schools (i.e. California, Florida, North Carolina, Arizona, and Colorado).Interestingly, in North Carolina, a state with racial/ethnic balance guidelines in their charter legislation, and where white share of the charter school enrollment (53%) is smaller than the public school enrollment (61%), there is a higher percentage of white charter school students in intensely segregated white schools than white public school students, indicating that, on average, white students in charter schools are more isolated than in the public schools in North Carolina.
For minority students, in states where there is a higher percentage of black students in 90-100% white charter schools than in the public schools (i.e. , Arizona, California, Georgia, and Colorado), there is still only a very small presence of black students in intensely segregated white schools.For Latino students, there are six states in which more than 10% of Latino public school students are enrolled in 90-100% white schools, but Latino charter school students are generally less likely to attend intensely segregated white schools than Latino public school students.However, in four states (i.e. California, Arizona, Florida, and Colorado), a higher share of Latino charter school students are in intensely segregated white schools than are Latino public school students.
In general, minority students in charter schools are less likely to be in heavily white schools than minority students in public schools.This could be due to the larger enrollment share of minority students in charter schools.In most states, white charter school students are less likely to be in intensely segregated white schools than public school students, but on average, they enroll in intensely segregated white charter schools at rates much higher than black and Latino charter school students.In earlier sections of this article, analysis using the exposure index showed that the average white charter school student was less segregated from minorities than the average white public school student.Conversely, black charter school students are more isolated than their public school counterparts, and the record was mixed for Latino charter school students.In this section, we examine the distribution of races within predominantly minority, intensely segregated minority, and intensely segregated white schools.
Students of all races are more likely to enroll in predominantly and intensely segregated minority charter schools than their public school counterparts.However, the percentages of white students in these minority charter schools were still much lower than those for black and Latino students.We speculate that the over-enrollment of minority and under-enrollment of white students in charter schools might result in more students attending predominantly minority and intensely segregated minority schools.This over-enrollment of minority students should make it possible to expose whites (as well as black and Latino students) to greater percentages of minority students, and we have seen that, in fact, white charter school students in many states are less isolated than their public school counterparts.However, given the high white isolation of public school students, white charter school students are still heavily enrolled in intensely segregated white schools.Even in states with a predominantly minority population in their charter school population, few white charter school students attend heavily minority charter schools.Thus the possibility of substantial interracial exposure of white students to minority students is largely unrealized.In most of these sixteen states, black and Latino charter school students are attending segregated minority schools at an even higher rate than those in the increasingly resegregating public schools.

Conclusion
The driving idea behind the charter school movement has been allowing schools greater autonomy in exchange for greater accountability.After a decade of rapid expansion and huge increase in public support for charter schools, often on the basis of arguments that they improve equity in school systems, it is time to hold these schools accountable for their accomplishments.
Our study shows that charter schools face high levels of segregation.Certainly there is tremendous variation among schools: some are highly diverse while others have high levels of isolation, particularly for black students.Although these schools have the potential to transcend high residential segregation created by neighborhood assignment and school district boundary lines, in many cases they are even more segregated than regular public schools.This might be due to the fact that many charter schools are located in segregated neighborhoods; establishing charter schools on boundaries between white, black and Latino neighborhoods could increase the likelihood of drawing a diverse student body.
Our state data suggest that black students are enrolled in charter schools-as well as intensely segregated minority charter schools-at a rate nearly twice their share of the public school population.Despite higher minority enrollments in charter schools, however, we still see in a number of states that whites are racially isolated.We find that regardless of white share of the entire charter school enrollment, black students in charter schools experience high levels of racial isolation and are exposed to very low percentages of white students.There is little evidence from this analysis that the existence of charter schools helps to foster more integrative environments, especially for minority students.At a time when the public schools are more segregated for minority students than thirty years ago, any reform that is publicly funded and intensifying the increasing public school segregation deserves very careful evaluation.
We continue to learn about the benefits of racial and ethnic diversity in schools for students of all races and at the same time, according to public opinion polls, public support for racial diversity is increasing.(Note 47) Further in a recent case concerning affirmative action in higher education, Grutter v. Bollinger, the Supreme Court recognized the importance of diversity as a compelling state interest.This article shows that instead of creating schools of diversity, many charter schools are places of racial isolation, particularly for minority students.Based on lessons learned in other school choice programs, such as magnet schools, the following conditions may help to address issues of racial isolation by creating a system that allows students to choose to attend charter schools on an equitable basis: Full information: The theory of choice as an equitable system has always depended on full information to all families.Information about charter schools and application procedures are often linked to social networks.Information must be made available to all potential students and parents, and in a language that all can understand.This might be aided by centralizing means of charter information dispersal in state departments of education and/or charter offices regardless of which agencies and organizations are allowed to grant charters. 1.
The provision of free transportation for all students, even across school district boundaries, is essential to ensuring that all interested students can choose to attend charter schools.Students of poorer families will see their opportunities to choose constrained where charter schools are not required to provide transportation.

2.
Providing for and welcoming all groups, including students from all racial/ethnic groups, English Language Learners, and special education students.In many ways, both implicitly and explicitly, charter schools can make their environment unwelcoming for a diverse array of students.Simply put, any publicly funded school should be a place where all students could be effectively educated.

3.
No screening of children for charter schools, both academic and otherwise.
Although most states require that charter schools enroll students on a first-come, first-serve basis, legislation in some states allows schools to employ both academic and non-academic criteria in student enrollment.Admissions procedures that might unfairly prohibit any child from enrolling (such as pre-admissions interviews or a requirement of parental involvement in the school) should be eliminated.Some states, such as Michigan, have tried to address this by specifying that admissions processes be made public.

4.
No Child Left Behind provides an opportunity for all students in low-performing schools to attend better schools, including moving to charter schools.We believe that this transfer opportunity should include a majority-to-minority transfer to all charter and magnet schools where room is available, and that the transfer will increase racial integration in the sending and receiving schools.As such, transportation should be provided for students across a metropolitan area.
To ensure that choice policies and charter schools promote racial equity and integrated schools, a number of political scientists and policymakers have underscored the need for government regulation of education markets (Cobb & Glass, 1999;Moe, 2002;Taebel et al, 1997).For example, Hill and Guin (2002) assert that "choice programs must be carefully designed to prevent segregation, and any program that produces levels of segregation as great as those now prevailing in the public education system should be scrapped or redesigned" (p.49).Our findings suggest that many state charter laws need to be redesigned to include stronger enforcement mechanisms to ensure racial integration.State education agencies should develop policies to ensure that the four conditions above exist wherever charter schools are authorized.They should provide support and encouragement for schools to create a diverse student body and to recruit students of all races.Indeed, charter laws should incorporate lessons learned from regulated choice plans, such as controlled open enrollment and magnet schools, that have produced stable, integrated schools in many districts including Minneapolis, Minnesota and Cambridge, Massachusetts (Willie, 2000).On the other hand, permissive charter school laws and unregulated choice policies have increased racial isolation for black students and facilitated white flight from integrated schools in Arizona (Cobb & Glass, 1999).Given the increasing ethnic separation in Arizona charter schools, Cobb and Glass (1999) argue that charter schools "should be required to actively pursue ethnic representation" (p.31).
If charter schools are to be an educational reform that provides an alternative means to broaden access to high quality education, issues of racial/ethnic segregation and practices that create the disturbing patterns of racial isolation in charter schools in many of our states, as detailed in this article, must be closely examined.In addition to monitoring student achievement and financial management, charter granters must hold charter schools to racial/ethnic balance guidelines in those states and districts with such legislation or court orders.
Ultimately, the extent of public oversight over school choice will determine, to a large extent, whether charter schools support or undermine racial integration in public education.
Unless otherwise indicated, all the authors' tabulations are from the 2000-01 NCES Common Core of Data.
For example, of the 1855 schools, only 291 schools reported free and reduced lunch data.Of these, 63% of the schools had student bodies with than 10% black and Latino students.While it is interesting to note that segregated white charter school are more likely to offer the free and reduced lunch program than other charter schools, these data are not reliable enough to draw any conclusions as to the correlation of racial minority and poverty concentration. 124.
Note: The CCD racial categories, as derived from information submitted by each state, are: non-Hispanic White, non-Hispanic Black, Hispanic, Asian, and Native American.Thus, our analysis is limited to these categories and cannot include, for example, biracial students. 125.
NCES defines a charter school as, "a school that provides free elementary and/or secondary education to eligible students under a specific charter granted by the state legislature or other appropriate authority."http://nces.ed.gov/ccd/data/txt/psu00lay.txt126.
The term isolation is used to denote the exposure of one race to itself, for example, white to white.This is another measure of segregation, which shows how concentrated students are with other students of their own race.We use the terms white-white exposure and white isolation interchangeably throughout the report to refer to the exposure of white students to other white students in their school.

127.
There are certain shortcomings to comparing individual charter schools to district averages since these averages are, in general, more diverse than 128.
There is some evidence that supports the idea that charter schools are attracting students from a broader geographic area than other public schools.
In Pennsylvania, Miron, Nelson, and Risley (2002) found that charter school students traveled an average of 5.6 miles from their home to charter school whereas other public school students traveled 2.4 miles.In theory, local districts are responsible for transportation arrangements, yet Miron and colleagues note that that some districts are still working out these details.Miron and Horn (2002) found similar patterns of longer distances to charter schools than traditional public schools in Connecticut as well. 184.
Essentially, the 2000 U.S. Department of Education guidelines only tell prospective charter school founders to determine whether their proposed school is in a district with a school desegregation plan, and, if so, to consult with Department of Education officials.(See Parker, 2001).Note: Some states are now starting to address this in revising their charter laws.

188.
Statutes governing charter schools are found using Westlaw and Lexis-Nexis.The specific citations of statutes are available upon request from the authors.

189.
The states comprising our definition of the South, as traditionally used in documenting school segregation trends, are the former slave states that practiced legally mandated segregation: Alabama, Arkansas, Florida, Georgia, Louisiana, Mississippi, North Carolina, South Carolina, Tennessee, Texas, and Virginia.Our definition of other regions is as follows: Border: Delaware, Kentucky, Maryland, Missouri, Oklahoma, and West Virginia; Northeast: Connecticut, Maine, Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania, Rhode Island, and Vermont; Midwest: Illinois, Indiana, Iowa, Kansas, Michigan, Minnesota, Nebraska, North Dakota, Ohio, South Dakota, and Wisconsin; West: Arizona, California, Colorado, Montana, Nevada, New Mexico, Oregon, Utah, Washington, and Wyoming.Note: Alaska and Hawaii are excluded because of their unique ethnic compositions and isolation from the regions studied here. 190.
This section analyzes data from the 34 states with operational charter schools.Thus, public school trends in these 34 states may be slightly different 191.
than national trends based on all 50 states.Throughout this report, in all data presented in tables comparing public and charter schools, we have removed charter schools from the public school data.Therefore, we can compare charter schools with non-charter public schools.

218.
For the remainder of the report, we use the term "predominantly minority" to designate schools where at least 50% of the student body is minority.Likewise, we use the term "intensely segregated minority" to designate schools where at least 90% of the student body is minority.

219.
Of course, at least in intensely segregated minority schools, by definition there will be a small percentage of white students.

220.
To compare to charter student exposure in Table 6 of this report, see Table 4, page 27 in Frankenberg, Lee, & Orfield (2003).

221.
Due to the small numbers of Asian and Native American students in charter schools in most states (although there are exceptions such as Minnesota and Arizona), the state-level analysis of racial/ethnic segregation will not include these students.

222.
The Common Core of Data has eight categories for locale: large city, mid-size city, urban fringe of large city, urban fringe of mid-size city, large town, small town, rural outside metropolitan statistical area (MSA), and rural inside MSA.We defined the three categories of urban, suburban, and rural based on NAEP's definitions.As defined by NAEP, central cities include all central cities in Standard Metropolitan Statistical Areas (SMSAs) as determined by the Office of Management and Budget.Urban Fringe/Large Town denotes large towns that are located within SMSA's that are urban but not defined as central city.Rural/Small Town areas include all areas that are classified as rural by the Census.For the purposes of this report, we will use central cities, suburban for urban fringe or large town areas, and rural for small town and rural areas.

223.
For data on the racial composition of the largest public school districts, see Frankenberg, Lee, and Orfield (2003), pp.53-57.

224.
The states where charter schools over-enroll white students are either in the South or West (see Figure 2).One reason suggested for this trend is that in states with large and/or diverse public school systems, charter schools might provide a means for white students to avoid racially diverse schools (Wells, et al. 2000).The South and the West are also the two regions of the country with the highest percentages of minority public school students, which are almost 50% (Frankenberg, Lee, and Orfield, 2003).

225.
It should be noted, however, that Missouri only authorizes charter schools in St. Louis and Kansas City.As these urban areas are heavily minority, it is not surprising that Missouri charter schools enroll such a high percentage of African-American students.

226.
There are exceptions to this trend for white public school students in four states (Colorado, Arizona, Florida, and California) in which white isolation is actually lower than the white percentage of the state's total enrollment.

227.
E.g., the difference in the white enrollment share is less than ten percentage points.

229.
Racial isolation also has a high correlation with student poverty; of all public schools nationwide, 86% of schools in 2000-01 that had 90-100% minority 230.
students were schools in which at least half the student body was poor or near poor (Frankenberg, Lee & Orfield, 2003).
Intensely segregated white schools tend to be schools with a lower percentage of poor or near poor students; nationally, less than 15% of schools that are 90-100% white are likely to be schools of concentrated poverty (Frankenberg, Lee & Orfield, 2003).

44.
Latino students in Massachusetts are enrolled in intensely segregated white schools at a lower percentage than blacks (2% for Latino students to 3% for black students).

Table 1 Racial/Ethnic Guidelines in State Charter School Legislation, 2003(Enrollment 2000-01)
Charter must specify means by which a school's student body will reflect racial and ethnic balance of the general population living in the school district granting the charter.Colorado 20,155A charter school shall be subject to any court-ordered desegregation plan in effect for the school district in which it operates.Charter must specify procedures to promote a diverse student body and state board will give preference to granting charters in districts that have 75% or more minority students.Commissioner must assess whether charter has a segregative effect on other districts sending pupils to the charter.
Arkansas 708Charters in districts under court-ordered desegregation plans must use a weighted lottery in student selection as well as issues relative to funding.Charter must have a plan to enroll cross-section of school-aged population including racial and academic factors.Commissioner of Education must assess whether charter will have segregative effect on district of residence of the charter school, and after the charter is operating,

Table 13 Percentage of Charter and Public School Students in Intensely Segregated White Schools, by Race/Ethnicity and by State, 2000-01 (Ranked by Percent White of Charter School Students)
v. Council of Emporia, 407 U.S. 451, 460-462 (1972)(a new school district could not be created if its effect would be to impede progress of dismantling an existing dual system).Also, for more recent cases that specifically pertain to charter schools, see Berry v. School District of the City of Benton Harbor, 56 F.Supp.2d 866, 872 (W.D. Mich.1999) (when considering charter school application to operate within a dual school system, court will consider interference with remedial order and effect on court's ongoing ability to eliminate vestiges of discrimination); Beaufort County Bd. ofEduc.v.Lighthouse Charter School et.al., 516 S.E.2d 655, 659 (S.C. 1999) (upholding a school board finding that a prospective charter school failed to adhere to same reporting requirements under OCR Title VI desegregation plan as other public schools in the district); Davis v. East Baton Rouge Parish SchoolBoard, et al, C.A. No. 56-1662 (M.D. La. 1999) (stating that charter schools in district remain subject to court's orders relating to desegregation of district).