Full opinion text
AMENDED MEMORANDUM AND ORDER GERTNER, District Judge. TABLE OF CONTENTS I.INTRODUCTION. CO CO CO II.PROCEDURAL HISTORY. lO CO CO A. The Comfort Litigation. CO CO 1. Parties. CO CO 2. Preliminary Injunction.'. CO CO 3. Motions to Dismiss. CO CO B. The Bollen Litigation. CO CO III. TRIAL. 00 CO CO A. Plaintiffs’ Case. 00 CO CO B. Defendants’ Case. 05 CO CO 1. The Administrators: . CO CO 2. The Parents and Students:. ^ CO 3. Defendants’ Experts:. ^ CO C. Plaintiffs’Rebuttal. ^ cO IV. FINDINGS OF FACT. A. The Racial Imbalance Act. B. Racial Imbalance in Lynn’s Public Schools. 1. 1977: The First Warning. 2. 1979: Washington, the First Magnet School. 3. 1980s: Profound Changes in Lynn. 4. 1986: A Series of Failed Voluntary Plans. 5. 1987-1988: Greater Imbalance; More Accusations. 6. 1988-1990: Drafting the Current Plan. C. The Current Plan. D. Continuous Monitoring. E. A Current Snapshot of the Lynn School District. 1. Residential Segregation and Geographical Separation. 2. ‘White Flight” and Its Decline after Implementation of the Lynn Plan. 3. Racial Balance or Imbalance. 4. The Special Problem of Poverty. lO 5. School Construction and Renovation.,. to 6. “Magnet" Schools. to F. The Lynn Schools at Present. to 1. Observations by Participants.. 1C 2. Expert Testimony. lO a. Dr. Oifield: Desegregation Expert. ic b. Drs. Dovidio and Kitten: Social and Developmental Psychologist, Respectively . 00 Ol Ci (1) Intergroup Contact Theory. OO Cn CJi (2) “Critical Mass”. 00 Oi <1 (3) Impact of Resegregation . 00 OJ 00 c. Nancy McArdle: Limitations Imposed by the Demographics in Lynn. 00 Ol 00 d. Plaintiffs ’ Rebuttal. 00 io CO Y. LEGAL ANAÍYSIS . A. Jurisdictional Issues . 1. Amendments to Prior Decisions (Comfort Plaintiffs) 2. Partial Motion to Dismiss (Bollen Plaintiffs) . a. Claims for Injunctive and Declaratory Relief... b. Nominal Damages. B. Equal Protection. 1. Strict or Intermediate Scrutiny?. 2. Facial Challenge to the Racial Imbalance Act. 3. The Strict Scrutiny Standard . a. Compelling State Interest. b. Narrow Tailoring. (1) Are the means necessary; are there adequate race-neutral alternatives?. CO -Q M (2) Is the policy proportional to the compelling interest. CO to (3) What Is the Impact on Third Parties?. CO CO (4) Miscellaneous Concerns; Deference to School Boards’ “Narrow Tailoring. CO CO 4. The Goals of the Plan . CO CR a. Curricular Goals: “Promoting Racial and Ethnic Diversity,” “Increasing Educational Opportunities for All Students and Improving the Quality of Education,” “Ensuring Safety” 1C 1> CO (1) Are These Curricular Goals Compelling State Interests? ... lO l> CO (2) Is the Plan Narrowly Tailored to These Compelling Interests? . vO CO (a) Are the Plan’s Means Necessary to Achieve its Ends? ÍO CO (b) Proportionality of the Means. CO (c) Minimal Burden on Third Parties; the Issue of Stigma c* 00 (3) Plaintiffs’Arguments Do Not Apply in Lynn. 00 CO (a) A White/Nonwhite Distinction Is Appropriate. o CO (b) Additional Resources Would Not Have Been Adequate to Accomplish the Curricular Goals; the Significance of “Critical Mass”. © 00 CO b. Remedying the Effects of De Facto Segregation; “Reducing Minority Isolation”.. 00 CO (1) Is this Remedial Interest Compelling?. 00 CO (2) Is the Lynn Plan Narrowly Tailored to this Compelling Interest?. © 00 CO (3) Race-Neutral Alternatives are not Feasible. t> 00 CO e. Interest (5): “Providing an Education to All Students that Satisfies Federal and State Constitutional Requirements” CO 00 © (1) The Command and Promise of Brown v. Board of Education . CO (2) State Constitutional Requirements. © C. Other Federal Claims. © 1.. Title VI. © 2. 4% U.S.C. § 1981. © 3. 42 U.S.C. §§ 1985, 1986. © D. Article 111 of the Massachusetts Declaration of Rights. © 1. Applicable Principles of Constitutional Interpretation. © 2. The Lynn Plan and the Purpose of Article 111. © 3. SJC Interpretation of Similar Language. to 4. State Constitutional Harmony and Federal Constitutional Doubt © VI. CONCLUSION 400 1. INTRODUCTION The issues raised in this litigation are critically important, not just for the parties, but for the nation. This case and others like it around the country require courts to grapple with whether and how public school officials may implement race-conscious programs in order to fulfill the Constitution’s promise of the Equal Protection of the laws, a promise articulated with special force fifty years ago in Brown v. Board of Education, 347 U.S. 488, 74 S.Ct. 686, 98 L.Ed. 873 (1954). Plaintiffs, parents of elementary school children in Lynn, Massachusetts, challenge their city’s school assignment plan (the “Lynn Plan”) because it takes race into account in permitting children to transfer from their neighborhood schools to other schools within the district. Since the implementation of the Lynn Plan entitles the Lynn Public Schools to certain additional aid from the Commonwealth of Massachusetts under the state’s Racial Imbalance Act (“RIA”), the plaintiffs also challenge the state law on its face and as applied in this case. Both the Lynn Plan and the state laws under which it was enacted, they say, violate the Constitutions of the United States and Massachusetts, as well as various federal civil rights statutes. They petition the Court to invalidate the Lynn Plan and to enjoin the defendants from employing racial classifications in student assignments and in the distribution of state aid. For the reasons set forth below, I DISMISS plaintiffs’ claims and enter JUDGMENT in favor of the defendants. To the plaintiffs, the issues could not be simpler: Taking race into account in school transfers violates the Equal Protection Clause. They contend that it sends the wrong message to the children of Lynn, namely, that their race matters when, in fact, the Constitution requires color blindness. Plaintiffs concede that reducing racial isolation and educating students to be citizens in a multiracial nation are important goals. They also concede that Lynn has accomplished those goals and has vastly improved its schools since the Plan’s implementation. But they claim that neither these goals nor this record are sufficiently compelling under the Constitution to justify Lynn’s race-conscious school transfer policy. To the defendants, the Lynn Plan not only complies with the Equal Protection Clause but is critical to maintaining that compliance. The Plan seeks to encourage learning and good citizenship in a racially diverse environment. The message it conveys to the students is that our society is heterogeneous, that racial harmony matters — a message that cannot be conveyed meaningfully in segregated schools. In order to teach that the “content of [one’s] character” does not depend on color, a child must interact with children of other races, an interaction that necessarily challenges nascent stereotypes. Without meaningful social contact, talk of tolerance and cooperation is nothing but an abstraction. If the Lynn school transfer plan were eliminated, the elementary schools of Lynn would become more and more racially segregated, with a host of pernicious consequences. While there has been a rising tide of litigation challenging the government’s use of racial preferences, both in and outside of the educational context, and while courts increasingly treat such programs with suspicion, it is not established — as the plaintiffs, not to mention many in the media, contend' — -that any government use of race for the purpose of promoting diversity is unconstitutional. The answer to the question “Can schools constitutionally use race in furtherance of education in a multiracial society?” is and has always been “It depends.” It depends upon the setting — for example, K-12 education, in contrast to higher education or employment, raises very different issues requiring distinct legal analysis. And it depends upon the nature of the plan — its purposes, its flexibility, the level of coercion involved, its history, its administration. On both axes — the setting and the nature of the plan — the Lynn Plan is different from race-conscious plans that have been successfully challenged. In constitutional parlance, the Lynn Plan passes muster even under the most stringent “strict scrutiny” test: It serves “compelling” state interests and is “narrowly tailored” to achieve them. On the setting: Unlike cases involving law schools, undergraduate degree programs, or even elite “magnet” high schools, Lynn’s school transfer policies are not about admissions or rejections in a competitive environment where merit supposedly determines “winners” and “losers” in a zero-sum game. The fact that one child may transfer to a particular elementary school while another may not does not affect the quality of the educational experience for either. The parties agree that, under the Plan, all Lynn schools are not only comparable, but by 2003, equally successful. Indeed, K-12 education involves a setting in which diversity has a different resonance than in any other. The goal of elementary education is, as the Supreme Court noted nearly fifty years ago, to foster good citizenship — not, for example, to train skilled professionals or to engage in a commercial enterprise. And in the twenty-first century, good citizenship necessarily entails the ability to function in an ever more heterogeneous democracy. On the Plan: The Plan that Lynn officials created, and that the state authorities approved, is minimally intrusive. It uses public funds for two critical public goals— to enable parents to choose integrated schools over segregated ones and to minimize racial imbalance across the school system. Since a racially diverse learning environment is essential for citizens-to-be, the Plan is a critical part of a comprehensive, districtwide plan to improve the quality of education for all Lynn’s children. Unlike many of the school desegregation efforts that have roiled courts and legislatures in the years since Brown, the Lynn Plan does not entail coercive assignments or forced busing; nor does it prefer one race over another. It allows every child to attend his or her neighborhood school. It also allows — and indeed encourages— elective transfers to schools outside the neighborhood, not to offer “choice” for its own sake, but to promote as much integration as possible while maintaining a neighborhood school system. Space permitting, transfers are approved unless they thwart the goal of reducing segregation in the city. The history and track record of the Plan also set it apart. Before the Plan, Lynn’s neighborhood schools were troubled, overcrowded, and racially segregated. City officials were accused of exacerbating these problems by allowing white students to transfer out of minority schools, but not vice versa. And even when official discrimination ended, racial divisions persisted, with documented results: Schools that were largely populated by minority students received less funding, had high teacher turnover, and had lower test scores. Throughout the system, even in the largely white schools, racial tensions and divisions persisted. White enrollment was declining at an alarming rate — the phenomenon known as “white flight.” After experimenting with race-neutral alternatives without success, Lynn officials sought the advice of experts, including social psychologists, educators, and even demographers. They concluded that the only way to improve the schools was to implement a comprehensive program that employed not only the race-conscious transfer policy at issue here, but a host of other innovations and resource infusions. Their goal was not simply to create a diverse learning environment, but also to support it and to ensure its success through curricular changes and materials, teacher and staff training, as well as improved facilities. They created a flexible plan, based on sophisticated data collection and analysis, that changes with the conditions in Lynn. Anyone denied the placement of his or her choice can appeal, as many of the plaintiffs in this case have done. By 2002, when the trial of this case began, it was clear that the Lynn Plan played an important part in creating a thriving, diverse, and integrated urban school system, successful on all fronts and by all measures — where race relations are positive and racial and ethnic tensions are absent; where students from diverse backgrounds maintain friendships and are well represented in student government and extracurricular activities; where student attendance rates are uniformly high and test results reflect substantial gains, particularly in the schools located in Lynn’s urban center; and where there are extraordinarily low levels of student conflict, crime, and violence. Nothing about the plaintiffs’ challenge or the government’s constitutional and statutory obligations obliges the Court to dismantle this Plan. II. PROCEDURAL HISTORY Before me presently are two civil actions, Comfort v. Lynn School Committee, Civ. No. 99-11811, and Bollen v. Lynn School Committee, Civ. No. 01-10365. A. The Comfort Litigation 1. Parties Samantha J. Comfort, Rhonda Campbell, Karen Agnew, Andrew and Cattibell DiGaetano, and Jean and William O’Neil, all parents of school children enrolled in the Lynn district, brought the Comfort action in 1999. The Comfort plaintiffs sued the Lynn School Committee, its individual members, the Superintendent of Lynn Schools, the City of Lynn and its Mayor. In December 1999 the Commonwealth of Massachusetts intervened as a party defendant for limited purposes. The Comfort plaintiffs challenged the RIA and the Lynn Plan under the Equal Protection Clause of the United States Constitution, Article 111 of the Massachusetts Declaration of Rights, and several federal civil rights statutes, including 42 U.S.C. §§ 1981, 1983, and 2000d. The RIA obliges the state Board of Education to address de facto segregation in Massachusetts’ public schools, either by funding voluntary efforts of individual school districts to integrate or, if necessary, by compelling them to adopt plans to improve racial balance in school populations. Comfort v. Lynn School Comm., 100 F.Supp.2d 57, 62 (D.Mass.2000). The Lynn Plan, as I describe more fully below, entails a neighborhood school system that permits transfers to out-of-neighborhood schools, unless such transfers would result in increased “racial isolation (too low a minority percentage) or racial imbalance (too high a minority percentage)” in a particular school. Id. at 61. 2. Preliminary Injunction The Comfort plaintiffs sought a preliminary injunction to enjoin the district’s use of race in the Plan. I denied the motion, finding that the plaintiffs showed neither a likelihood of success on the merits nor irreparable harm. Id. at 59-60. The First Circuit in Wessmann v. Gittens, 160 F.3d 790 (1st Cir.1998), held that diversity “might be sufficiently compelling, in specific circumstances, to justify race-conscious actions.” Id. at 796. The intensely fact-bound question of what educational circumstances would permit race-conscious actions was hardly amenable to resolution on the truncated record of a preliminary injunction. Comfort, 100 F.Supp.2d at 66. Nor could the Comfort plaintiffs demonstrate irreparable harm from the continuation of the Plan since, as described below, most were content with the schools their children were in and did not intend to seek further transfers. Id. at 63-64. The plaintiffs did not appeal the denial of their Motion for a Preliminary Injunction. 3. Motions to Dismiss The defendants moved to dismiss a number of the Comfort plaintiffs’ claims on the grounds that (1) the Constitution does not allow a state to be sued for damages, and (2) the plaintiffs lacked standing to sue because their children had been placed in the schools of their choice and no longer wished to transfer. With certain exceptions, I granted these motions. I dismissed the federal statutory civil rights claims for damages against the Commonwealth on Eleventh Amendment grounds, since there was no question that the state may not be sued for damages. Comfort ex rel. Neumyer v. Lynn School Comm., 131 F.Supp.2d 253, 254 (D.Mass.2001). I also held that the Eleventh Amendment barred declaratory relief against the Commonwealth. Id. at 256. See infra Section V.A.I. In a subsequent opinion, Comfort v. Lynn School Comm., 150 F.Supp.2d 285 (D.Mass.2001), I held that the plaintiffs lacked standing to seek injunctive relief against any of the defendants. Id. at 288. The plaintiffs had children who were contentedly enrolled in schools of their choosing and could not “demonstrate either actual present harm or a likely danger of direct injury in the future.” Id. at 295-98. I held that the Comfort plaintiffs could pursue limited declaratory relief, as well as nominal damages. Id. at 298-99, 302. Specifically, I found that “prospective” declaratory relief was unavailable to the plaintiffs, id. at 302, but that they were nonetheless entitled to sue for a “retrospective” declaration that the Lynn Plan violated their legal rights. I now conclude that this finding was erroneous: a declaration, prospective or retrospective, is available only to plaintiffs that have standing because of a present case or controversy. See also infra Section V.A.1. In any event, the Comfort plaintiffs were free to litigate the substance of their claims for nominal damages. Comfort, 150 F.Supp.2d at 298-99. Of the original Comfort plaintiffs, only Samantha Comfort remains in this case. B. The Bollen Litigation Rather than amend the existing action, plaintiffs’ counsel added new parent plaintiffs in a separately filed lawsuit. The Bollen plaintiffs, Todd and Laurie Bollen, Janeen Goodwin, Gina Leone, LeAnne Manuel, Michael and Meta Stinson, and Karen Tsaltas sued the same defendants but added claims under 42 U.S.C. §§ 1985 and 1986 and listed members of the Massachusetts State Board of Education as defendants in their official capacities. Defendants move to dismiss a number of the Bollen plaintiffs’ claims on standing grounds. That motion [document # 174] is GRANTED IN PART AND DENIED IN PART. See infra Section V.A.2.a. III. TRIAL The parties filed multiple summary judgment motions, all of which I denied. In June 2002, the parties presented evidence in an eleven-day bench trial. A. Plaintiffs’ Case Consistent with their view that the case was a simple one, plaintiffs offered one live witness, Meta Stinson, a Lynn parent and plaintiff; the deposition testimony of state Board of Education member Abigail Thernstrom; and several exhibits. Ms. Stinson, who has three daughters currently attending Lynn schools, sought to have her daughter, Angelica Jackson, transferred from Breed Middle School to Pickering Middle School. That transfer was initially denied because Stinson had listed her daughter as “white” when she first registered her for school. Although Angelica’s father is white, Stinson herself is of mixed racial and ethnic background— French Canadian, Irish, Jewish, Barbadian, and African American. Upon learning of the denial, Stinson went to the Parent Information Center (“PIC”) and added an additional racial designation of “black,” but to no avail. Stinson conceded that although she was advised of her right to appeal, and indeed, even though that appeal would have been successful, she elected not to do so. In the interim, the district announced that at the Sewell-Anderson school fifth graders could remain for the sixth grade, and Stinson was satisfied with that placement. Significantly, Stinson testified that Angelica maintains a number of interracial friendships with her classmates, has a positive attitude about race, and gets along well with students of other races, even though Stinson has not personally discussed matters of race with her daughter. Stinson conceded that it is an important component of Lynn’s educational mission to ensure that its students overcome racial stereotypes and acquire a better understanding of students with different racial backgrounds. The plaintiffs also introduced the deposition testimony of Abigail Thernstrom, a member of the Massachusetts Board of Education. Thernstrom, though not offered as an expert, took issue with the views of certain school desegregation experts — including Dr. Gary Orfield, a nationally renowned expert who testified in the defendants’ case. She challenged Orfield’s view that America’s schools are growing increasingly racially segregated principally because she did not believe that “[t]he dividing line in America” is “between whites and nonwhites.” Deposition of Abigail Thernstrom (“Thernstrom Deposition”), Oct. 9, 2001, at 22. Although a member of the Board of Education and charged with enforcing the RIA Thernstrom opposed the Act in principle. She conceded that she could not speak with authority on the law since she lacked day-to-day familiarity with its terms. Moreover, she acknowledged that she had no personal knowledge or experience of specific conditions in Lynn. Finally, plaintiffs offered exhibits that pertained to the operation of the Lynn Plan and RIA and the extent to which racial categories are used in making transfer decisions. B. Defendants ’Case Defendants countered with ten witnesses, five of whom were district administrators and educators with twenty to thirty years’ experience in the Lynn system variously as students, teachers, and administrators. Significantly, these witnesses were in a position to describe the troubled state of Lynn schools before the Lynn Plan was implemented, and to attest to its impact afterward. They described pre-plan efforts to improve the schools, all to no avail. And they lauded the current state of the school system under the Lynn Plan and the substantial improvements it has effected. Defendants also offered the testimony of a Lynn parent and a student, who described the actual workings of the system in the schools with which they were familiar. Finally, defendants offered the testimony of a number of nationally known experts. The experts based their testimony on their personal observations of Lynn schools, a survey of Lynn students, interviews with teachers and administrators and other Lynn-specific data. The experts affirmed the importance of Lynn’s race-conscious transfer policy to the district’s educational objectives. 1. The Administrators: Janet Birchenough has been Director of Equity and Program Support for the Lynn public schools since January 1992. Birche-nough managed and supervised Lynn’s Parent Information Center, the school district’s central registration office, which oversees the day-to-day implementation of the Lynn Plan. In addition, the PIC gathers the data that Lynn officials use to certify that the Plan remains precisely tailored to accomplish the Plan’s goals — data about the school system, its schools, its students, simulations about demographic patterns in Lynn, and the likely composition of the schools if the Plan were to be eliminated. Moreover, Birchenough testified from the perspective of nearly forty years’ experience in the Lynn schools — as an elementary school teacher, as an art teacher working in eleven of the district’s eighteen elementary schools and ultimately as an administrator. I found her testimony credible and relevant, and her knowledge of the system — of individual schools and the history of the Plan — encyclopedic. Patricia Barton, currently the principal of the Fecteau-Leary Middle School, also has 30 years of professional experience in the Lynn system. She testified credibly about her experience as a student attending racially isolated schools, as a teacher in a segregated system, and as an administrator who presided over that system’s transformation under the Lynn Plan. Nicholas Kostan, the current Superintendent of the Lynn School Department, also testified with considerable authority about the changes he has seen. He has been a teacher and a professional administrator in the Lynn school system for 31 years, first as a principal of the Breed Middle School. He serves as liaison between the Lynn School Committee and the Lynn School Department and at the same time oversees the PIC and works with teachers, parents, students, and the community. Patricia Mallett is currently a teacher in the media center at the Brickett school, having taught in Lynn schools for twenty-six years. During the transition to magnet school programming, she was the system’s Magnet School Facilitator. Mallett testified from that unique perspective on the race-neutral alternatives that were tried and considered, as well as the success of the current Plan. 2. The Parents and Students: Karen Homer is an African American parent of two current Lynn public school students. She is active in and serves as the Acting Secretary for the Lynn branch of the NAACP. She moved from Boston to avoid living in a segregated neighborhood and to provide her children with an integrated education. Horner’s children, Kyle, age 17, and Erroll, age 14, have attended neighborhood and out-of-neighborhood schools in Lynn, and she spoke movingly about the quality of her children’s experiences. Nicole Oak is a white student who had just graduated from the Lynn school system in June 2002 and described her educational experience there. 3. Defendants’Experts: Dr. Gary Orfield is a nationally recognized political scientist and one of the leading national experts in the field of education and equal educational opportunity, which encompasses issues such as racial segregation, racial isolation, desegregation, housing and racial change, and their effects on students in primary, secondary, and higher education. He is a professor of education and social policy at Harvard University, where he is co-director of the Harvard Project in School Desegregation. The Project in school desegregation is an interdisciplinary research center that commissions research across the nation on issues of civil rights policy, racial change, and racial inequality. Significantly, Dr. Orfield based his testimony not only on his own research, experience, and studies from other jurisdictions but also on an extensive amount of data and information about Lynn and its public schools in particular. Orfield put the full resources of the Harvard Civil Rights Project (“Harvard Project”) (a separate entity from the Project in school desegregation with a broader research focus) into his study of Lynn and gathered information firsthand during a site visit as well. Nancy McArdle testified as an expert on the demographic and housing trends in Lynn. She was a researcher with the Joint Center for Housing Studies at Harvard for fourteen years. McArdle has researched, consulted, and published in the field of housing policy, with specific attention to demographic, population, and immigration trends. Like Dr. Orfield, McArdle relied on a wide variety of data sources specific to Lynn. Dr. Melanie Killen, a developmental and education psychologist, testified on how racial segregation and racial diversity impact the social and moral development of children and adolescents. She is a Professor of Human Development in the College of Education at the University of Maryland, and she presently serves as Associate Director for the Center for Children, Relationships and Culture, a research department with faculty from developmental psychology and education psychology. Dr. Killen has focused specifically on how children and adolescents evaluate exclusion based on race and ethnicity, as well as on intergroup relationships and conflict resolution across lines of ethnicity. She, too, based her expert testimony on a series of direct observations and interviews on site in the Lynn Public Schools, as well as her review of the Harvard data. Finally, Dr. John Francis Dovidio testified as a nationally renowned social psychologist with a particular interest in the subdiscipline of intergroup relations and the development of racist attitudes. Dr. Dovidio is presently the Charles A. Dana Professor of Psychology and Dean of Faculty and Provost at Colgate University. His research has focused on social psychology and race relations among elementary school children, secondary school students, adolescents, and college students. He testified about how and when stereotypes are formed and what strategies can be deployed to prevent their formation. He offered his expert opinion based on personal visits and the Harvard data. C. Plaintiffs’Rebuttal Plaintiffs offered one witness in rebuttal, Dr. Christine Rossell, a professor of political science at Boston University. Dr. Ros-sell’s field of interest is the comparative efficacy of different approaches to school integration. While Dr. Rossell had been a paid consultant to Lynn in the development of its original school desegregation plan in 1987 (when she supported an earlier, and more intrusive iteration of the Plan), at the time of her testimony, she had no current knowledge of the Plan. She based her conclusions on her research involving other school systems and what she had remembered of that early draft of the Lynn Plan from fifteen years earlier. IV. FINDINGS OF FACT A. The Racial Imbalance Act In 1965, Massachusetts became the first state in the nation to enact a law addressing racial imbalance in the public schools. The Racial Imbalance Act (“RIA”), Mass. St.1965, c. 641, §§ 1 et seq. (codified at Mass. Gen. Laws c. 71 §§ 37C, 37D, and c. 15 §§ II, 1J, IK), had its origin in a legislative finding that racial imbalance in Massachusetts public schools was so dramatic as to reach a crisis level, with damaging effects on the students. The Massachusetts Board of Education’s “Kieman Report” not only concluded that “[r]acial imbalance represents a serious conflict with the American creed of equal opportunity,” School Comm. v. Bd. of Education, 366 Mass. 315, 318 n. 5, 319 N.E.2d 427 (1974) (quoting the Kiernan Report) (internal quotation marks omitted), it also underscored the extraordinary impact of racial imbalance throughout the school system. Racial imbalance was found to encourage prejudice among students of all races, inadequately prepare students for life in multiracial communities, and produce inferior educational facilities for African Americans. Id. As a result, the legislature found that racial imbalance in state public schools posed an “emergency” situation, see School Comm. v. Bd. of Education, 352 Mass. 693, 698, 227 N.E.2d 729 (1967), and enacted the RIA to correct it. Section 1 of the RIA provides: It is hereby declared the policy of the commonwealth to encourage all school committees to adopt as educational objectives the promotion of racial balance and the correction of existing racial imbalance in the public schools. Mass. St.1965, c. 641, § 1 (codified with amendments at Mass. Gen. Laws e. 71 § 37C). Toward this end, the RIA encourages — but does not require — all schools to devise “plans” to promote racial balance. The only instance in which the law imposes a duty upon a school committee to prepare such a plan is when the school committee has proven unable to satisfy requests from nonwhite students attending racially isolated (predominantly minority) schools to transfer to racially imbalanced (predominantly white) schools. Mass. Gen. Laws. c. 71, § 37D; see also School Comm., 366 Mass, at 322-23, 319 N.E.2d 427. The state Board of Education is to assist in the development of any plan, voluntary or mandatory, and must approve it prior to implementation. Mass. Gen. Laws c. 76, § 12A. The plan must set forth in detail the district’s proposed changes in school attendance zones; any possible alteration to or expansion of existing facilities, as well as any planned addition of new school buddings; and any “other measures” that the district anticipates implementing to ensure space for students seeking “desegregative transfers,” i.e., transfers that will reduce racial imbalance. Mass. Gen. Laws c. 71, § 37D. Under the RIA, the Commissioner of Education has the power to withhold school construction funds and other state aid if a school district does not act within a reasonable time to reduce an identified racial imbalance in its system. Mass. Gen. Laws c. 15, § II, ¶2; see also School Comm., 366 Mass, at 323, 319 N.E.2d 427. A 1974 amendment provided affirmative financial rewards to school districts that undertake voluntary plans. Mass. St.1974, c. 636, § 1. Such incentives included 100% state reimbursement of certain student transportation costs, substantial funding of the costs of establishing magnet schools, and payments of $500 to the district for each student transfer that reduces racial imbalance or isolation. Mass. Gen. Laws c. 15, § II, ¶¶ 3, 4. The city of Lynn submits yearly “entitlement reports” certifying the number of desegregative student transfers that occurred in its school system in the prior school year; the state responds with aid monies based on the numbers Lynn provides. From 1974 until 1984, the Act authorized state reimbursement for up to 75% of the cost of approved school renovation or construction directed at reducing racial imbalance in school populations. By 1984, the rate of reimbursement was increased to 90%, Mass. St.1984, c. 394, § 5, where it remained until 2001, when the legislature eliminated the reimbursement, Mass. St.2000, c. 159, § 36. In Lynn’s case, state assistance provided under the RIA supplied 90% of the funding for school renovations and construction that the city outlined in its 1990 voluntary plan. B. Racial Imbalance in Lynn’s Public Schools In the nearly ten years preceding the adoption of the Lynn Plan, Lynn faced two serious problems — substantial overcrowding of its neighborhood schools and significant levels of racial imbalance. Significantly, the state attributed the latter not only to growing residential segregation within the community, but also to the school district’s own policies and practices. Indeed, correspondence during this period suggests that state officials felt Lynn’s actions and inactions made the city vulnerable to state and/or federal lawsuits alleging de jure segregation. The emergence of increasingly racially identifiable schools in Lynn created a crisis. A number of witnesses told of conditions not unlike those targeted in the Kier-nan Report. Janet Birchenough observed racial polarization and severe resource inequalities between the predominantly minority and identifiably white schools where she taught. These inequalities were apparent on all fronts, including building facilities, learning materials, and teacher commitment. The predominantly white schools were well-maintained and well-managed; parents were deeply involved in their children’s education. The minority schools, in contrast, limped along with crowded classes and outdated materials. There were discipline problems: parents and administrators alike were apathetic; students were angry, felt abandoned, and often lashed out. Attendance rates were low; achievement at all levels suffered. Class sizes were larger in identifiably minority schools than in identifiably white schools. Wherever possible, teachers with seniority — who had priority in choosing new openings- — transferred out of minority schools into the identifiably white, wealthier schools, where the school climate was more conducive to learning and teaching. There was no professional training for teachers to prepare them for teaching a diverse student population, and little or no curricular support. Birchenough observed that the racially charged and intolerant atmosphere in the schools led to cross-racial conflict, the students’ frequent resort to racial slurs, as well as tendencies to self-segregate by race at recess and in cafeterias and classrooms. This was so in all of Lynn’s schools, in the better-off, predominantly white schools that harbored small minority populations as well as in the more racially mixed schools. Birchenough’s observations were confirmed by other participants, including Principal Barton, Superintendent Rostan, and Ms. Mallett. 1. 1977: The First Warning In 1977, the Board of Education first confronted Lynn about the racial imbalance in its schools. Specifically, the Washington Community School (“Washington”), one of Lynn’s elementary schools, was found to have a 57.2% minority student population, when the overall district was only 9% minority. The Board of Education attributed the concentration of minority students at Washington in part to residential patterns, but more significantly to the district’s school assignments, which the state roundly condemned. Administrators regularly allowed minority transfers into the already racially isolated Washington Community School and assigned all bilingual classes to that school’s annex. Over the next decade the Board of Education regularly warned Lynn that its action and inaction exposed it to charges of de jure segregation. In contrast to other cities across the nation, however, by implementing its voluntary Plan, Lynn officials ultimately headed off court involvement and dramatically changed its schools for the better. 2. 1979: Washington, the First Magnet School In 1979 a magnet school program was established at the Washington school. The state offered supplemental funding to the Washington magnet school program, conditioned on the district’s adoption of a voluntary transfer policy that would attract white students. Lynn accepted this invitation and adopted a voluntary transfer policy. In April 1980 the district announced a more comprehensive plan. Neighborhood schools remained the centerpiece of the plan with these exceptions: White students from schools that were 70% or more white had the right to transfer to the Washington school, whereas white students already enrolled in that school would not be permitted to transfer out of it. Nonwhite students in the Washington district and elsewhere had the right to transfer to any school that was more than 70% white, but white students were not permitted to make such transfers. Again, there was official manipulation of the policy. Officials were accused of bending the rules for white parents. Administrators regularly approved the requests of white parents to transfer their children out of predominantly minority neighborhood schools. For example, in the 1987-1988 school year 107 out-of-neighborhood white students were attending the 93% white Aborn high school as a result of transfers. Of these 107 students, more than half of them resided in the attendance zone that fed into the Ingalls high school, which was located in a minority neighborhood. Lynn officials conceded that these transfers were in blatant violation of the school assignment policy in place at the time. 3. 1980s: Profound Changes in Lynn Between 1980 and 2000 racial and ethnic minorities moved into Lynn in considerable numbers, transforming a city that was 93% white to 63% white. Defense expert Nancy McArdle, citing census data, noted that the child population of Lynn went from 90% white to more than half minority during this period. More and more, Lynn’s residents began to self-sort into neighborhoods by race. The 1980 Census found that the northern and western areas of Lynn were 90% white, while a belt of moderately integrated neighborhoods surrounded a minority enclave in south central Lynn. By the end of the decade, these trends intensified. The schools, like Lynn’s neighborhoods, likewise grew more and more racially polarized. While the magnet program at Washington — the only such program in Lynn in the early 1980s — saw minor changes in minority concentration, dropping from well above 50% to 44% during this period, the minority share in other minority-identifiable elementary schools in Lynn dramatically increased. The school system was troubled, with high absentee rates, racial tension and conflict, and chronically low test scores. Racial polarization in Lynn’s neighborhoods and schools continued into the mid-1980s. By 1984, notwithstanding the city’s demographics, then roughly 83% white and 17% minority, four of Lynn’s seventeen elementary schools had minority populations of between 35% and 50%. The remaining schools were overwhelmingly white and remained so until 1987, even as the minority student population in Lynn climbed to 26% over the intervening three years. 4. 1986: A Series of Failed Voluntary Plans The state Department of Education urged Lynn to adopt a long-term, comprehensive plan to defeat racial isolation and imbalance. Lynn formulated such a plan in April 1986 but elected not to seek state approval (as the RIA requires). As a result, grant monies offered by the state for Lynn’s “desegregation coordinator” were suspended mid-year, due to Lynn’s stalled progress. The district ultimately drafted a second voluntary plan, which the state approved in September 1986 but Lynn never implemented. 5. 1987-1988: Greater Imbalance; More Accusations By 1987 four elementary schools had become minority dominated, four other elementary schools were at least 95% white, and three more were hovering at around 90% white student enrollment. In April 1987, the Lynn School Committee developed a third voluntary plan, which the Board of Education approved. Lynn would launch and develop five schools, designated “magnet” schools, in the first year of implementation, with five more to follow in the next year. With the state’s approval would come additional funds to support the necessary construction and renovation of facilities to improve these schools. The plan also approved a redrawing of elementary and junior high school attendance zones to assist with the desegregative effort. However, certain aspects of the plan, slated for implementation in September of 1987, were delayed. The district did not implement its proposed class-size máxi-mums, and the school year began with the elementary and middle school attendance zones unchanged. Significantly, school officials continued to approve improper segregative transfers. By January 1988, state officials made their accusations more pointed than ever before. In a letter to Lynn’s Superintendent they noted: It appears that the condition of minority identifiable schools in Lynn is directly [attributable to past actions and inac-tions by Lynn School officials. The most significant finding is that the School Committee failed to enforce its own controlled transfer policy and has admitted to that fact. Again, the message was clear: Lynn’s “actions and inactions,” its failure to enforce its own plan in an evenhanded way, and its own admissions of these facts made the district vulnerable to charges of de jure segregation. The state withdrew its approval of Lynn’s Plan. 6. 1988-1990: Drafting the Current Plan In its February 1988 Plan, Lynn officials admitted that official actions had exacerbated racial isolation and imbalance. They redoubled their efforts, this time soliciting feedback from the community, which — no doubt because of Lynn’s history — eyed Lynn’s proposed actions warily. A letter from Christine Rossell — who consulted for Lynn during this period but now serves as the plaintiffs’ expert in this case — was more sanguine, observing that Lynn stood “an excellent chance of desegregating its six minority schools and minority isolated schools.” A key point of contention was student assignment. The state Department of Education was concerned that Lynn’s use of voluntary transfers and magnet programs would not be sufficient to reverse the decades-long trends toward racial imbalance, a trend exacerbated by official action. Lynn, however, remained committed to a voluntary transfer plan and refused to assent to “restricted choice” or “controlled choice” regimes that the Department was recommending. Finally, in September 1989, Lynn devised yet another amended plan, which the state accepted. This new Plan, the terms of which I will set forth in detail below, guaranteed that every student could attend his or her neighborhood school. In addition, a student could transfer from the neighborhood school to another school as long as the transfer improved the racial balance in either the neighborhood or the destination school. The Plan gave assurances that there would be space for such transfers in new and renovated schools. In February 1990 the Plan was amended to allow for “neutral transfers,” that is, transfers that would neither improve nor adversely affect racial balance in the schools that were party to the transfer. The state approved this amendment. A 1999 amendment to the Plan added more flexibility to the transfer system, instituting an appeals process for transfer denials and certain exemptions for bi- and multiracial students and cases of extreme hardship. This Plan is the one that the plaintiffs challenge. C. The Current Plan It is important to note at the outset that under the Lynn Plan every student in Lynn is entitled to attend the school in his or her neighborhood. Students have options beyond their neighborhood schools if their proposed transfers are “desegrega-tive” — ie., when they contribute to the districtwide integration effort. Additional “neutral transfers” are allowable as well, provided that they are approved by the school district’s PIC and the sending and receiving principals. Whether or not a transfer is desegrega-tive or neutral turns on the interplay of three factors: the racial composition of the sending and receiving schools and the student’s race. Schools are classified as either “racially balanced,” “racially isolated,” or “racially imbalanced.” An elementary school is considered “racially balanced” if its minority population falls within -1-/15% of the overall percentage of minority students in the Lynn district. For the middle and high schools racial balance is +/10% of the district population. When the proportion of minority students in a school exceeds the range of racial balance, that school is classified as “racially imbalanced.” Likewise, a school in which the number of white students surpasses the outer bounds of racial balance is “racially isolated.” For example; Lynn’s student population for the 2001-2002 school year was 42% white and 58% nonwhite. By the Lynn Plan’s definitions, then, an elementary school that enrolled between 48% and 78% minority students would qualify as racially balanced. Middle and high schools required a tighter fit of between 48% and 68% minority students. Elementary schools with more than 73% minority students in the 2001-2002 school year — and middle or high schools with more than 68% minority students — were designated as racially imbalanced. Conversely, an elementary school that was fewer than 43% minority or a middle or high school that was fewer than 48% minority would be considered “racially isolated.” A proposed transfer is desegregative— always allowed, space permitting — when it would improve the racial balance of the sending or the receiving school. For example, a minority student may always transfer out of a racially imbalanced school (“School A”) or into a racially isolated school (“School B”). Conversely, a white student may always transfer out of the racially isolated School B and into the racially imbalanced School A. A transfer is segregative, and never allowed, when it would exacerbate an already existing condition of racial imbalance in the sending or receiving school. A requested transfer that is neither desegregative nor segre-gative is “neutral” and conditionally allowable, as I explained above. At a district level, white and nonwhite students are equally subject to the Plan; sometimes a white student may be denied a transfer, sometimes a nonwhite student. But at the individual level, there are times when, all else being equal, a student’s ability to transfer turns on his or her race. To continue the above example, with School A as a racially imbalanced school and School B as a racially isolated school: A student who seeks a transfer from School A to School B may do so (again, space permitting) if he or she is a minority student. A white student requesting the same transfer, however, would not be eligible, unless she or he qualified for certain exceptions. Such a student, denied a transfer because it would exacerbate racial imbalance in the system, may appeal the decision to the PIC’s Director, who may occasionally refer the matter to or consult with the district Superintendent. The appeals process is not an empty formality. The district will override transfer denials on appeal when the denial would result in siblings attending different schools or when parents can make a showing of medical, safety, or other extreme hardship. Moreover, the PIC goes out of its way to make the appeals process accessible to everyone. It assists parents in preparing documentation required to establish hardship. It then arranges a meeting with the parents to discuss the student’s options. If conditions at the destination school preclude the transfer, the PIC will present the student with alternative destination schools. Finally, the unique considerations posed by bi- and multiracial students have led Lynn to amend its Plan to permit parents to appeal the district’s race designations directly to the Superintendent. Significantly, a number of the plaintiffs in this case have successfully invoked the appeals process. The Plan’s drafters also recognized that integration involves more than race-conscious school assignment policies, more than simply the mixing of students of different racial backgrounds. Thus, the Plan included substantial curricular innovations designed to ensure positive racial interaction; training and development of staff to address the challenges of teaching children of diverse backgrounds; programs that would create opportunities for positive interaction among students, school personnel and parents from different racial and ethnic groups, which are not normally found within regular school programming; integrated leadership opportunities and training to give students the skills necessary to deal effectively with racial tension and conflict, etc. In addition, the Plan’s drafters acknowledged that the improvements it sought could not be sustained in the long term unless all the schools were made attractive to all Lynn parents, whatever their race. Thus, the Plan included an ambitious construction program, largely funded by the state, to ensure sufficient space for out-of-neighborhood transfers. It involved the development and standardization of curriculum so that there would be equal instructional opportunities across Lynn; development of indicators of performance and achievement for individual schools, programs and students; development of measures designed to improve school attendance; and creation of business/college partnerships with the schools to improve the quality of instruction. D. Continuous Monitoring The district’s PIC oversees the ongoing implementation of the Lynn Plan. In effect, it is the vehicle through which the Plan administrators certify that the Plan is narrowly tailored to meet its goals. Chief among the PIC’s responsibilities is the processing of all admissions and re-admissions to the district and all transfers within the school system. It assesses students for special needs and maintains an ongoing database of each student’s transfer history. That database, which features thirty-five data fields per student, enables Birchenough to monitor class size and enrollment by school, as well as the racial composition of individual schools and the district more generally. The PIC prepares monthly reports tracking movements of students into and out of the district, as well as between the schools, to ascertain how many of these transfers qualify as desegregative. It provides regular reports to the state Department of Education on the progress and status of the district’s magnet school programs, as well as the number of desegre-gative transfers. Significantly, the data that the PIC gathers enable it to continuously monitor the need for the Plan. The PIC tracks the patterns of Lynn residential segregation. It attempts to predict the choices of white or minority parents if there were no restrictions on transfers. Based on that data, Birchenough (and the defendants’ experts) predicted that in a pure choice model, white parents would seek transfers to predominately white schools, and minority parents to minority schools. E. A Current Snapshot of the Lynn School District 1. Residential Segregation and Geographical Separation According to McArdle’s expert testimony, recent censuses and the data kept by the Lynn PIC reveal that Lynn remains geographically segregated by race. In the 1990s white populations continued to move to the northern — and particularly northeastern — areas of Lynn, while the concentration of racial minorities in south central Lynn expanded to consume all of southern Lynn and more of central Lynn, with a transition zone of racially mixed neighborhoods between these two enclaves. In addition, the elementary schools located in the predominantly white sections of Lynn and the elementary schools in the predominantly minority sections are separated by significant distances. The travel time between the two areas, whether by private car or public transportation, is not insubstantial. 2. “White Flight” and Its Decline after Implementation of the Lynn Plan In 1979, before the Lynn Plan’s implementation, when the city’s schools were the most polarized, statistics show that the overall number of students in the Lynn public schools began to decline — a direct result of the decline in white enrollment, what has been described as “white flight.” Significantly, after the Plan’s implementation, this trend began to reverse. The number of students now enrolled in Lynn schools is similar to enrollment in the 1970s, just before the most significant increase in the number of minority students enrolled in the school system. While the percentage of Lynn students attending private and parochial schools rose as high as 17.4% in 1987, it has since declined dramatically to a low of 10% in 2001. Of the 15,444 students attending Lynn public schools in the 2001-2002 school year, 42% were white and 58% were nonwhite (15% African American, 29% Hispanic, and 14% Asian). 3. Racial Balance or Imbalance There are presently twenty-five schools in the Lynn system — eighteen elementary schools, four middle schools, and three high schools. Under the Lynn Plan, “racially balanced” elementary schools in 2001-2002 had minority populations of between 43% and 73%; nine schools qualified. Six schools were racially isolated. Four elementary schools were “racially imbalanced.” Under the Plan’s stricter requirements for the middle and high schools, namely, a racial makeup falling within +/10% of the districtwide breakdowns, which in 2001-2002 translated into 48% and 68% minority enrollment to qualify as racially balanced, only one middle school of the four was racially balanced, while all three high schools met the test. The Breed middle school (49%) was racially balanced, while the Marshall (72%) and Fecteau-Leary (81%) middle schools were racially imbalanced, and Pickering (33%) was racially isolated. Lynn’s high schools were all racially balanced with the following percentages: Lynn Classical (51%), English (51%) and Lynn Vocational Technical (63%). 4. The Special Problem of Poverty Many of the city’s resident families live at or near the poverty level, a situation that greatly complicates any school’s educational mission. In the 2001-2002 school year, 65% of all Lynn students received free or reduced-cost lunch (eligibility for which is based on family income), including 40.7% of white students, 72% of the African American students, and 79.8% of the Asian American students. Fourteen of Lynn’s eighteen elementary schools enroll a sufficient number of impoverished students to qualify for federal Title I assistance. Significantly, the four that do not receive assistance are the four most identi-fíably white schools. 5. School Construction and Renovation Since the Plan was implemented, Lynn has been able to renovate and expand six of its elementary schools, for which the state provided 90% of the funding. Lynn is currently planning another phase of construction that would include renovations to existing facilities and the construction of two new middle schools and two or three elementary schools. Elimination of Lynn’s restrictions on segregative transfers would render this project ineligible for 90% funding under the RIA. 6. “Magnet” Schools Of the ten magnet programs envisioned in early iterations of the Plan, only seven were ultimately developed. The district added two more programs in September 1999. However, the term “magnet school” in this setting needs to be clarified. It does not connote a competitive admissions process or the provision of a more elite education. See infra note 72. Lynn’s “magnets” differ from its other schools only insofar as they have adopted certain “educational themes.” The magnet themes currently include “Brickett by the Sea,” “Reading and Writing Literary and Whole Language,” “Technology and Language/Citizenship Skills for the Future,” “Life Science,” “Rainbow Connection Society,” “User Friendly Society,” “Pickering After School Support,” and “Healthy Schools Make Healthy Communities.” While the schools offer and provide varying academic programs, which are designed to draw students from other neighborhoods elsewhere in Lynn in order to further the integration effort, the parties stipulate that “the education provided ... in each of the elementary, middle, and high schools in Lynn is comparable in quality, resources, and curriculum.” F. The Lynn Schools at Present By all accounts, and by all measures, since the implementation of the Plan the Lynn schools have become a success story. That success was recounted in the moving testimony of the participants, in expert testimony, and in the data. The defendants have identified certain compelling interests to justify the Plan: First, to prevent racial isolation; second, to promote racial and ethnic diversity; third, to increase educational opportunities for all students and to improve the quality of education; fourth, to provide a sufficient education by state constitutional standards; fifth, to ensure the safety of Lynn’s public school students; and sixth, to implement the clear command of the Supreme Court’s decision in Brown v. Board of Education. 1. Observations by Participants Since the Plan has been in place, Lynn school superintendent Nicholas Kostan has observed a “steady progression” of improvement in the schools. Racial tensions have subsided, attendance and test scores have increased, and suspensions of students have declined. Indeed, according to Kostan, Lynn’s school attendance rates are remarkable for an urban school district. He attributed this attendance record to the Lynn Plan, which he believes has cultivated feelings of comfort and safety among Lynn students that make them want to go to school. He has also noticed that students no longer self-segregate by race in their social interactions, which he believes is also a result of the Lynn Plan. Patricia Mallett, a teacher at the Brick-ett elementary school, testified that since the Lynn Plan’s implementation, she has witnessed a stunning improvement in student race relations at Brickett. Prior to the Plan, she observed self-segregation and racial tension in Brickett, a predominantly white school. Now, students of different races seek out one another and form friendships, a result she attributes to the Plan’s fostering of cross-racial interaction in the crucial formative years. 2. Expert Testimony The expert testimony in this case corroborated the moving observations of the teachers, students, and administrators who testified on behalf of the district. McArdle and Drs. Orf