Opinion En Banc
LIPEZ, Circuit Judge.This appeal requires us to review certain features of a voluntary plan designed to achieve the educational benefits of racial diversity in the public schools of Lynn, Massachusetts (“Lynn Plan” or “Plan”). *6The Plan addresses resource allocation, curricula, and other aspects of the classroom experience. Relevant to this appeal, it also controls school assignments and transfers. Under the Plan, each student is entitled to attend his or her neighborhood school. Students who do not wish to attend their neighborhood school may apply to transfer to another school. Approval of a transfer depends, in large part, on the requesting student’s race and the racial makeup of the transferor and transferee schools.
Parents whose children were denied transfers on race-conscious grounds challenged the transfer provisions of the Lynn Plan, claiming, inter alia> that the provisions violate the Fourteenth Amendment Equal Protection Clause. The district court rejected the parents’ challenges and upheld the Plan. A panel of this court reversed, finding that the Plan was not narrowly tailored to the defendants’ compelling interest in achieving the benefits of educational diversity. We granted review en banc and now affirm.
Our review of the equal protection challenge is informed by the Supreme Court’s recent decisions regarding affirmative action in higher education, Grutter v. Bollinger, 539 U.S. 306, 123 S.Ct. 2325, 156 L.Ed.2d 304 (2003), and Gratz v. Bollinger, 539 U.S. 244, 123 S.Ct. 2411, 156 L.Ed.2d 257 (2003). We conclude, based on those cases, that Lynn has a compelling interest in securing the educational benefits of racial diversity. Applying the analytic framework set forth in Grutter and Gratz to the context of a K-12, non-competitive transfer plan, we hold that the Lynn Plan is narrowly tailored to meet this compelling interest.
The plaintiffs assert a number of other claims as well. We do not reach the merits of their facial challenge to the Massachusetts Racial Imbalance Act, which fails for lack of standing. We treat the plaintiffs’ federal statutory claims as foreclosed by our equal protection ruling and reject their challenge to the Plan under Article 111 of the Massachusetts Declaration of Rights. Finally, we conclude that the district court properly denied the plaintiffs’ motion for recusal.
I. Background
This case comes to us with a rich factual background, described in detail in a series of district court rulings. See Comfort v. Lynn Sch. Comm., 283 F.Supp.2d 328 (D.Mass.2003) (Comfort IV); Comfort v. Lynn Sch. Comm., 150 F.Supp.2d 285 (D.Mass.2001) (Comfort III); Comfort v. Lynn Sch. Comm., 131 F.Supp.2d 253 (D.Mass.2001) (Comfort II); Comfort v. Lynn Sch. Comm., 100 F.Supp.2d 57 (D.Mass.2000) (Comfort I). We set forth only those facts necessary to put this case into context, drawing upon the largely unchallenged findings of the district court.
A. Lynn Public Schools
Lynn is the ninth-largest city in Massachusetts, with a population of approximately 89,000. At all relevant times, its school system has been neighborhood-centered, entitling students to attend their local schools as a matter of right. By the mid-1970s, several of Lynn’s schools were experiencing significant racial imbalance. In 1977, for example, the Washington Community Elementary School had a nonwhite student population of 57%, more than six times the nonwhite percentage in the school system as a whole. Predominantly nonwhite schools suffered disproportionately from resource shortages, overcrowding, discipline problems, and teacher apathy. The school system was plagued by high absentee rates, racial tension, and low test scores.
*7In an effort to combat these problems, Lynn established its first magnet school in 1979. At the same time, it inaugurated a voluntary transfer program aimed at attracting white students to that school (which apparently was located in a predominantly nonwhite area of the city). The magnet school was only modestly successful in alleviating racial imbalance.
In the meantime, Lynn was undergoing a demographic shift. Between 1980 and 2000, the city went from being 93% white to 63% white, with the school-age population becoming more than half nonwhite by 2000. Residential segregation by race increased during this period as whites clustered in the northern and western areas of Lynn and nonwhites concentrated in its south-central region.
Because of the neighborhood school system, these residential patterns heightened the racial imbalance of Lynn’s schools. By 1987, seven of eighteen elementary schools had white enrollments of 90% or more, while four others had predominantly nonwhite student bodies. Lynn responded by developing a plan to launch ten magnet schools,1 but city leaders did not believe that the magnet program, on its own, would effectively combat the growing racial imbalance. In September 1989, the Lynn School Committee (“Lynn”) adopted the Plan that is the subject of this litigation.2
B. The Lynn Plan
The defendants describe the Lynn Plan as a voluntary plan for school improvement and the elimination of minority isolation. The Plan begins with the premise that every child is entitled to attend his or her neighborhood school. Race is taken into account only when a student seeks to transfer to a school other than his or her neighborhood school.
Lynn operates eighteen elementary schools (six of which are magnets), four middle schools (three of which have magnet programs), and three high schools.3 In the 2001-02 school year, 15,444 students were enrolled in the Lynn public schools. Out of this group, approximately 42% of students were white, 15% African-American, 29% Hispanic, and 14% Asian (for a total “minority” or nonwhite population of roughly 58%).
For purposes of the Lynn Plan, schools are placed in one of three categories. A “racially balanced” school is one in which the percentage of nonwhite students falls within a set range of the overall proportion of minorities in Lynn’s student population. The range is +/15% for elementary schools and +/10% for other schools. For example, an elementary school with between 43% and 73% nonwhite students during the 2001-02 school year was considered racially balanced, as was a middle or high school that had a nonwhite enrollment *8of 48% to 68%. In the 2001-02 school year, nine of Lynn’s elementary schools, one of its middle schools, and all three of its high schools were racially balanced.
If a school’s nonwhite population falls below the racially balanced range (i.e., if the percentage of nonwhite students in 2001-02 fell below 43% for an elementary school or 48% for a middle or high school), it is “racially isolated.” Conversely, a school whose nonwhite population rises above the racially balanced range (i.e., over 73% for an elementary school or 68% for a middle or high school) is considered “racially imbalanced.” In 2001-02, five of Lynn’s elementary schools and one of its middle schools were classified as racially isolated, while four elementary schools and two middle schools were racially imbalanced.
The transfer policy is straightforward. Space permitting, a student whose neighborhood school is racially balanced may transfer to another racially balanced school without regard to race. Because all three of Lynn’s high schools are currently racially balanced, for example, students may transfer freely among them. Students are also permitted to make “de-segregative” transfers. That is, a white student may transfer out of a racially isolated school and into a racially imbalanced school (i.e., to a school with a lower percentage of white students), and a nonwhite student may transfer out of a racially imbalanced school and into a racially isolated school (i.e., to a school with a lower percentage of nonwhite students). By contrast, absent certain exceptions, students may not make “segregative” transfers. A segregative transfer is one that would exacerbate racial imbalance in the sending or receiving school (i.e., a white student may not transfer to a racially isolated school, and a nonwhite student may not transfer to a racially imbalanced school).4
A student whose transfer request is denied is entitled to appeal. Roughly half of all appeals are granted. Common grounds for successful appeals are medical and safety concerns, daycare issues, and other types of hardship. Appeals will also be granted when the denial would result in siblings attending different schools. The Plan is implemented by the Parent Information Center (“PIC”), which processes all admissions and transfers, works with parents on appeals, and monitors enrollment and racial composition of individual schools and the district in general.
As the plaintiffs point out, the Lynn Plan can result in unequal treatment based on race. Consider, for example, two children, one white and one African-American, who are initially assigned to the same neighborhood elementary school for the 2001-02 school year. The school is racially isolated (i.e., less than 43% minority). Both children request a transfer to a nearby school that is racially imbalanced (i.e., greater than 73% minority). Under the Plan, the white student will be permitted to transfer, and the African-American student will not.
Although the race-conscious transfer policy is the focus of this appeal, the Lynn Plan also includes numerous other provisions aimed at improving the quality of all schools in the district. It calls for curricular programs and teacher training designed to foster cross-racial understanding and reduce racial tension. It also implemented a standardized curriculum throughout Lynn’s schools; developed performance indicators for schools, programs *9and students; took measures to improve student attendance; and created business/college partnerships with schools to improve the quality of instruction. Finally, the Plan envisions a construction program to upgrade school facilities and alleviate overcrowding.
All parties agree that Lynn’s public schools have improved markedly since the Plan’s inception, although they dispute which aspects of the Plan are responsible for this improvement. In any event, students’ standardized test scores have increased, absentee levels have decreased, and racial tensions have diminished under the Plan.
C. The Racial Imbalance Act
The Racial Imbalance Act (“RIA”), Mass. Gen. Laws ch. 15, §§ 1I, 1J, 1K; id. ch. 71, §§ 37C, 37D, directs the Massachusetts Board of Education to remedy de facto segregation in the public schools throughout the state. See Sch. Comm. of Boston v. Bd. of Educ., 352 Mass. 693, 227 N.E.2d 729, 732 (1967). The legislature enacted the RIA in response to findings that dramatic levels of racial imbalance in the public schools threatened to harm students’ educational opportunities. See id. at 733-34. The RIA has two main effects: it authorizes the Board to fund voluntary efforts to improve racial balance, Mass. Gen. Laws ch. 15, § 1I, and it allows the Board to require that school districts adopt integration plans in certain circumstances, id. ch. 71, § 37D.
The Lynn school system has received significant state aid under the voluntary provisions of the RIA. These funds have helped pay for new construction and school renovations.5 Lynn also receives a state stipend of $500 for each desegregative student transfer, and the state defrays certain costs associated with cross-neighborhood transportation and the creation of magnet schools.
D. Procedural History
In 1999, the parents of children who had been denied transfers under the Lynn Plan (“Comfort plaintiffs”) brought a civil action against the Lynn School Committee, its individual members, and several governmental officials. They claimed that the Lynn Plan and the RIA violate the Fourteenth Amendment Equal Protection Clause, several federal civil rights statutes, and Article 111 of the Massachusetts Declaration of Rights. The Commonwealth intervened as a party defendant for the limited purpose of defending the RIA. See 28 U.S.C. § 2403(b). The district court denied a motion to preliminarily enjoin the race-conscious aspects of the Lynn Plan, see Comfort I, 100 F.Supp.2d at 59-60, and dismissed several of the plaintiffs’ claims, see Comfort III, 150 F.Supp.2d at 289, 296-97, 302; Comfort II, 131 F.Supp.2d at 254, 256. Only one of the original Comfort plaintiffs remains in the case. Other parents (“Bollen plaintiffs”) then filed a second action that included the Comfort plaintiffs’ claims, as well as other statutory claims. The Bollen plaintiffs also added as defendants the members of the Board of Education in their official capacities. The district court consolidated the two cases.
Following an eleven-day bench trial, the district court issued a lengthy opinion dismissing a number of the Bollen plaintiffs’ claims on standing grounds. Comfort IV, 283 F.Supp.2d at 361-63. It rejected the facial attack on the RIA, id. at 366-68, *10and determined that the Plan’s transfer provisions are narrowly tailored to several compelling state interests, and thus constitutional. Id. at 375-92. The court also rejected the plaintiffs’ remaining federal statutory claims, treating the statutory provisions as co-extensive with the Equal Protection Clause. Id. at 392-93. Finally, the court held that the transfer provisions of the Lynn Plan did not violate Article 111 of the Massachusetts Declaration of Rights. Id. at 393-400.
A panel of this court reversed, holding that the Plan could not the survive strict scrutiny review required by the Equal Protection Clause. Relying on the Supreme Court’s decision in (Lrutter upholding a race-conscious admissions policy at the University of Michigan Law School, the panel recognized a compelling interest in “obtaining the educational benefits of a racially diverse student body.” Comfort v. Lynn Sch. Comm., No. 03-2415, slip op. at 30 (1st Cir. Oct. 20, 2004), withdrawn by 2004 WL 2348505 (1st Cir. Nov.24, 2004). It concluded, however, that the Plan is not narrowly tailored to that interest because it uses race “mechanically” and “forgoes individualized consideration of transfer applications.” Id. at 40. The panel also cited other narrow tailoring flaws, including the Plan’s breadth and indefinite duration. We granted on banc rehearing and now affirm.6
II. Standing
“[T]he general rule is that a court should first confirm the existence of rudiments such as jurisdiction and standing before tackling the merits of a controverted case.” Berner v. Delahanty, 129 F.3d 20, 23 (1st Cir.1997). This is because “standing is a necessary concomitant to the court’s power to adjudicate a case.” R.I. Ass’n of Realtors v. Whitehouse, 199 F.3d 26, 30 (1st Cir.1999). Thus, we begin by considering the plaintiffs’ standing.
To establish standing in federal court, a party must demonstrate three things:
First, the plaintiff must have suffered an “injury in fact”—an invasion of a legally protected interest which is (a) concrete and particularized, and (b) actual or imminent, not conjectural or hypothetical. Second, there must be a causal connection between the injury and the conduct complained of—the injury has to be fairly traceable to the challenged action of the defendant, and not the result of the independent action of some third party not before the court. Third, it must be likely, as opposed to merely speculative, that the injury will be redressed by a favorable decision.
Lujan v. Defenders of Wildlife, 504 U.S. 555, 560-61, 112 S.Ct. 2130, 119 L.Ed.2d 351 (1992) (citations and internal quotation marks omitted). The plaintiffs must have standing to obtain both forms of relief they seek: an injunction against the race-conscious aspects of the Plan and a declaration that the RIA is facially unconstitutional. See Friends of the Earth, Inc. v. Laidlaw Envtl. Servs. (TOC), Inc., 528 U.S. 167, 185, 120 S.Ct. 693, 145 L.Ed.2d 610 (2000).
With respect to the Lynn Plan, one Bollen plaintiff (Gina Leone) clearly meets the threshold standing requirement. Leone sues on behalf of her minor son, Troy Lamothe, whose transfer request was denied on the ground that it would be segregative. The fact that Troy was allowed to attend the school of his choice pending the outcome of this litigation does not defeat standing. See Nat'l Amusements, Inc. v. Town of Dedham, 43 F.3d *11731, 735 n. 3 (1st Cir.1995) (accepting plaintiffs standing despite agreement not to enforce disputed ordinance pending outcome of litigation). So long as one plaintiff has standing to seek a particular form of global relief, the court need not address the standing of other plaintiffs seeking the same relief. See Watt v. Energy Action Educ. Found., 454 U.S. 151, 160, 102 S.Ct. 205, 70 L.Ed.2d 309 (1981); Houlton Citizens’ Coalition v. Town of Houlton, 175 F.3d 178, 183 (1st Cir.1999).
Standing to seek injunctive or declaratory relief against the RIA is a different matter. The parties stipulated, and the district court confirmed, that the mandatory portion of the RIA (i.e., the provision under which the Board of Education can require a district to adopt a plan to reduce racial imbalance) has not been applied to Lynn. Comfort IV, 283 F.Supp.2d at 367. The plaintiffs nevertheless launch a facial attack against the RIA’s mandatory provisions, Mass. Gen. Laws ch. 71, § 37D, contending that they offend the Fourteenth Amendment Equal Protection Clause and Article 111 of the Massachusetts Constitution by giving white children a right to transfer out of isolated schools and nonwhite children a right to transfer out of imbalanced schools, but not vice versa (i.e., white children cannot transfer out of imbalanced schools and nonwhite children cannot transfer out of isolated schools). Even if that is true in theory, the plaintiffs cannot overcome the fact that only a person who was denied a transfer on the basis of the mandatory provisions of the RIA has standing to challenge them.
The mandatory provisions of § 37D apply only to school districts that refuse to create voluntary plans to combat identified racial imbalance. See Sch. Comm. of Springfield v. Bd. of Educ., 366 Mass. 315, 319 N.E.2d 427, 429 (1974). Lynn never refused to take action; rather, it drafted and implemented a voluntary plan. The plaintiffs therefore have not suffered a cognizable injury from § 37D of the RIA.7 Accordingly, they lack standing to seek a declaration as to its validity.
The plaintiffs also lack standing to seek injunctive or declaratory relief against the RIA provisions that offer incentives to districts that voluntarily adopt plans to combat racial imbalance. See Mass. Gen. Laws ch. 15, §§ 1I, 1J, 1K; id. ch. 71, § 37C. Redressability, one of the prerequisites for standing, see N.H. Right to Life PAC v. Gardner, 99 F.3d 8, 13 (1st Cir.1996), requires a substantial likelihood that the relief sought will in fact remedy the alleged injury, Bonas v. Town of N. Smithfield, 265 F.3d 69, 73 n. 4 (1st Cir. 2001). Plaintiffs describe their injury as both a racially-based denial of transfers and the stigma of their inability to “compete” on equal terms for transfers. Even assuming arguendo that this asserted injury is somehow traceable to the RIA, the plaintiffs cannot show that an injunction against the RIA’s incentive provisions will lead to racially unrestricted transfers within the Lynn public schools or eliminate the perceived stigmatic harm.
This point is apparent from the record. Under the terms of the RIA, Lynn has received state funding for construction and busing based on its voluntary efforts to combat racial imbalance. Even if we directed the Board to distribute aid without regard to racial balancing efforts, the plaintiffs have not demonstrated that redress would likely follow. There is no reason to believe that Lynn would cancel *12its transfer program merely because state funding was no longer contingent on it.
In an effort to sustain their claim, the plaintiffs and amicus Pacific Legal Foundation also present a second theory for prospective relief. They assert that the incentive provisions of the RIA are effectively mandatory because they coerce school districts to adopt race-conscious plans by tying state aid to the adoption of those plans. There is a flaw in this argument. While the RIA’s incentive provisions reward schools that address racial imbalance, they do not dictate a procedure or methodology by which schools must do so. See Mass. Gen. Laws eh. 15, § 1I; id. ch. 71, § 37C. Given the absence of a requirement that schools adopt race-conscious plans to comply with the RIA, the plaintiffs have not shown that the incentive provisions of the RIA are causally responsible for the voluntary adoption of race-conscious transfer policies.
Even if the plaintiffs did have standing to mount a facial challenge to the voluntary provisions of the RIA, we agree with the district court that such a challenge would fail. Comfort IV, 283 F.Supp.2d at 367-68. “A facial challenge to a legislative Act is ... the most difficult challenge to mount successfully, since the challenger must establish that no set of circumstances exists under which the Act would be valid.” United States v. Salerno, 481 U.S. 739, 745, 107 S.Ct. 2095, 95 L.Ed.2d 697 (1987). Such circumstances exist here. As the district court recognized, “[a] school district may theoretically adopt a plan that improves racial imbalance without explicitly introducing race-based criteria at all.” Comfort IV, 283 F.Supp.2d at 367-68 & n. 77 (noting that “if a small district with racially identifiable school attendance zones built a single central elementary, middle, and high school to absorb its entire enrolled student populations, this strategy would qualify as a racial balancing plan under the RIA, and it would not trigger any equal protection scrutiny”); see also Boston’s Children First v. Boston Sch. Comm., 260 F.Supp.2d 318, 327 (D.Mass.2003) (finding that the RIA’s goals may be satisfied by race-neutral methods), aff'd sub nom. Anderson v. City of Boston, 375 F.3d 71 (1st Cir.2004).
III. Federal Equal Protection Claims
The main issue on appeal is the constitutionality of the Lynn Plan’s race-conscious transfer restrictions. The plaintiffs contend that by mechanically taking race into account, the Plan violates the Equal Protection Clause of the Fourteenth Amendment and various federal civil rights statutes. The resolution of the federal statutory claims depends on the fate of the constitutional challenge. See infra Part III.D. Consequently, we focus on the equal protection issue.
A. Standard of Review
We review the court’s findings of fact for clear error and its legal conclusions, including its application of the law to the facts, de novo. See Wessmann v. Gittens, 160 F.3d 790, 795 (1st Cir.1998).
The Supreme Court has reviewed racial classifications under the strict scrutiny standard, which requires that the policy be narrowly tailored to a compelling state interest. Grutter, 539 U.S. at 326, 123 S.Ct. 2325; Adarand Constructors, Inc. v. Pena, 515 U.S. 200, 227, 115 S.Ct. 2097, 132 L.Ed.2d 158 (1995). The defendants urge us to apply a more relaxed standard here. They emphasize that although the Plan is race-conscious, it is unlike affirmative action because it affects whites and nonwhites equally.
*13This argument is foreclosed by the Supreme Court’s recent decision in Johnson v. California, — U.S. -, 125 S.Ct. 1141, 160 L.Ed.2d 949 (2005). There, the Court considered an unwritten policy of the California Department of Corrections whereby inmates are segregated by race for up to sixty days after entering a new correctional facility. Rejecting the State’s argument that its policy should be subjected to relaxed scrutiny because it “neither benefits nor burdens one group or individual more than any other group or individual,” id. at 1147, the Court explained that all racial classifications
raise special fears that they are motivated by an invidious purpose. Thus, we have admonished time and again that, “[ajbsent searching judicial inquiry into the justification for such race-based measures, there is simply no way of determining ... what classifications are in fact motivated by illegitimate notions of racial inferiority or simple racial politics.” Richmond v. J.A. Croson Co., 488 U.S. 469, 493, 109 S.Ct. 706, 102 L.Ed.2d 854 (1989) (plurality opinion). We therefore apply strict scrutiny to all racial classifications to smoke out illegitimate uses of race by assuring that [government] is pursuing a goal important enough to warrant use of a highly suspect tool.
Id. at 1146 (internal quotation marks omitted). This rule applies in the present context just as firmly. The Plan must be reviewed under strict scrutiny.
This standard is not “strict in theory, but fatal in fact.” Id. at 1151; see also Grutter, 539 U.S. at 326-27, 123 S.Ct. 2325 (“Although all governmental uses of race are subject to strict scrutiny, not all are invalidated by it.”). Strict scrutiny “is designed to provide a framework for carefully examining the importance and the sincerity of the reasons advanced by the governmental decisionmaker for the use of race in that particular context.” Grutter, 539 U.S. at 327, 123 S.Ct. 2325. We therefore bear in mind the court’s admonition that “[cjontext matters when reviewing race-based governmental action under the Equal Protection Clause.” Id.
B. Compelling State Interest
Until recently, there was some question as to whether diversity could constitute a compelling interest in the educational context. See Wessmann, 160 F.3d at 795-96. The Supreme Court has now answered that question in the affirmative, holding in Grutter that a law school’s interest in obtaining the educational benefits that flow from a diverse student body was compelling enough to justify the narrowly tailored use of race in admissions. 539 U.S. at 343, 123 S.Ct. 2325.
Grutter involved a challenge to the University of Michigan Law School’s admissions policy, which took into account racial and ethnic background as one of several “soft variables” used in assessing applicants. Id. at 315, 123 S.Ct. 2325. The Law School justified this strategy as furthering its goal of assembling a class that was both “exceptionally ... qualified and broadly diverse.” Id. at 329, 123 S.Ct. 2325. It also sought to enroll a “critical mass” of minority students, thereby enhancing its quest for broad diversity. Id. at 330, 123 S.Ct. 2325.
The Grutter Court stressed that the Law School’s plan did not pursue a critical mass of minority students for its own sake, but rather for the sake of obtaining the educational benefits that flow from having a racially diverse student body. Id. at 329-30, 123 S.Ct. 2325 (acknowledging that racial balancing for its own sake is unconstitutional). These educational benefits include promoting cross-racial understanding, breaking down stereotypes, fostering *14livelier and better informed class discussions, and preparing students to succeed in an increasingly diverse society. Id. at 330, 123 S.Ct. 2325. The Court largely deferred to the Law School’s educational judgment not only in determining that diversity would produce these benefits, but also in determining that these benefits were critical to the school’s educational mission. Id. at 328-33, 123 S.Ct. 2325. The Court warned, however, that “scrutiny of the interest asserted by the Law School is no less strict for taking into account complex educational judgments in an area that lies primarily within the expertise of the university.” Id. at 328, 123 S.Ct. 2325. Nevertheless, the Court concluded that the pursuit of these benefits constituted a compelling state interest. Id. In so ruling, it recognized “the overriding importance of [education in] preparing students for work and citizenship.” Id. at 331, 123 S.Ct. 2325.
Against this background, we consider the interest that Lynn’s race-conscious Plan seeks to advance. This is not a case where the racial classification is aimed at remedying past segregation. See Comfort IV, 283 F.Supp.2d at 390 n. 101. Rather, the parties stipulated that Lynn’s interests
include fostering integrated public schools and what Lynn believes are [their] positive effects; reducing minority isolation and avoiding segregation and what Lynn believes are their negative effects; promoting a positive racial climate at schools and a safe and healthy school environment; fostering a cohesive and tolerant community in Lynn; promoting diversity; ensuring equal education and life opportunities and increasing the quality of education for all students.
The district court grouped these interests into two categories: (i) reaping the educational benefits that flow from having a racially diverse student body in each of Lynn’s public schools, and (ii) avoiding the negative educational consequences that accompany racial isolation.
Although there are some differences between these interests, we conclude that they are essentially two sides of the same coin. The negative consequences of racial isolation that Lynn seeks to avoid and the benefits of diversity that it hopes to achieve are rooted in the same central idea: that all students are better off in racially diverse schools. We therefore restate the interests at stake here as obtaining the educational benefits of a racially diverse student body.
Lynn maintains that ensuring a racially diverse student body in its schools has produced, and will continue to produce, many of the same benefits cited by the Grutter Court: disarming racial stereotypes, increasing racial tolerance, and preparing students to live and work in an increasingly multi-racial society. The defendants’ expert evidence also suggests that racially isolated students often feel psychological burdens that can lead to poor attendance and academic woes, and that these effects can be combated by racial integration. Consistent with these assertions, Lynn’s schools have indeed experienced many positive developments— including higher attendance rates, declining suspension rates, a safer environment, and improved standardized test scores— since the Plan’s inception.
In Lynn’s view, these developments can be explained by the intergroup contact theory. This theory holds that “under certain conditions, interaction between students of different races promotes empathy, understanding, positive racial attitudes!,] and the disarming of stereotypes.” Comfort IV, 283 F.Supp.2d at 356. Under the intergroup contact theory, there are four basic conditions for success: (1) equal sta*15tus among racial groups, (2) the presence of teachers and staff trained to facilitate interactions between members of different groups, (3) common goals and cooperative activities, and (4) opportunities for personalized contact with a sufficient number of children from different racial groups to disrupt stereotypes. Id. at 356-57.
Lynn’s experts explained that meaningful intergroup contact (the fourth condition of intergroup contact theory) requires that a school have a “critical mass” of students of each group, i.e., white and nonwhite. Id. at 357. Lynn’s experts also testified, and the district court found, that the benefits of intergroup contact continue to accrue as a school becomes increasingly diverse. Id. Citing this theory and crediting the defense experts who explained its application in Lynn, the district court agreed that there was a causal link between improvements in the school system and increased racial diversity. Id. at 353-54.
While acknowledging improvements in the Lynn schools since the Plan’s inception, the plaintiffs disagree that these changes can be attributed to the race-conscious aspect of the Lynn Plan. More significantly, they also contend that regardless of whether there are educational benefits to racial diversity, Lynn does not have a compelling interest in achieving those benefits. We disagree.
Lynn’s transfer policy expressly aims at attaining racial diversity in the city’s schools. Where a community does not seek racial diversity for its own sake, but rather to advance a compelling interest in the educational benefits that diversity provides, there is no absolute bar to pursuing racial diversity. See Grutter, 539 U.S. at 330, 123 S.Ct. 2325. The district court found that this was Lynn’s purpose, Comfort IV, 283 F.Supp.2d at 375-76, and the record supports that finding. We see no reason to second-guess it. Cf. Grutter, 539 U.S. at 328, 123 S.Ct. 2325 (stating that, typically, a school’s “educational judgment that ... diversity is essential to its educational mission is one to which we defer”).
The plaintiffs assert that, unlike Grutter, this case does not implicate a compelling interest that would justify the pursuit of racial diversity. The admissions plan at issue in Grutter strove for diversity along many axes, including race, in an effort to create a student body with diverse viewpoints, thereby enriching classroom discussion and academic experiences. See 539 U.S. at 329, 123 S.Ct. 2325 (“As part of its goal of assembling a class that is ... broadly diverse, the Law School seeks to enroll a critical mass of minority students.” (internal quotation marks omitted)). The plaintiffs contend that Grutter’s recognition of a compelling interest in “the educational benefits that flow from student body diversity,” 539 U.S. at 330, 123 S.Ct. 2325, is thus limited to the benefits that flow from viewpoint diversity in the higher education context and does not extend to the benefits that flow from racial diversity in the K-12 context.
Again, we disagree. Lynn’s asserted interests bear a strong familial resemblance to those that the Grutter Court found compelling. There is no reason to believe that these interests are advanced by viewpoint diversity but not racial diversity, or that they are substantially stronger in the context of higher education than in the context of elementary and secondary education. See McFarland v. Jefferson Cty. Pub. Schs., 330 F.Supp.2d 834, 852-53 (W.D.Ky.2004) (reasoning that the benefits recognized in Grutter also “accrue to students in racially integrated public schools”); cf. Plyler v. Doe, 457 U.S. 202, 221, 102 S.Ct. 2382, 72 L.Ed.2d 786 (1982) (emphasizing the importance of K-12 education “in maintaining the fabric of our society”). In fact, there is significant evi*16dence in the record that the benefits of a racially diverse school are more compelling at younger ages. See, e.g., Comfort IV, 283 F.Supp.2d at 356 (summarizing expert’s testimony that “[i]t is more difficult to teach racial tolerance to college-age students; the time to do it is when the students are still young, before they are locked into racialized thinking”).
The plaintiffs correctly point out that the benefits attributed to the Lynn Plan are not identical to those described in Grutter. But Grutter teaches that the compelling state interest in diversity should be judged in relation to the educational benefits that it seeks to produce. 539 U.S. at 330, 123 S.Ct. 2325. The Lynn Plan uses race in pursuit of many of the same benefits that were cited approvingly by the Grutter Court, including breaking down racial barriers, promoting cross-racial understanding, and preparing students for a world in which “race unfortunately still matters.” Id. at 333, 123 S.Ct. 2325.8 There are, of course, some variances between the benefits sought. For example, the law school plan at issue in Grutter focused on the advantages of viewpoint diversity in the classroom, while Lynn emphasizes the positive impact of racial diversity on student safety and attendance. But it is natural that safety and attendance issues will loom larger in elementary and secondary schools than in graduate schools. Conversely, lively classroom discussion is a more central form of learning in law schools (which prefer the Socratic method) than in a K-12 setting. These differences do not negate a compelling interest in racial diversity in a K-12 setting. Instead, they are the logical result of context.
We are persuaded by the extensive expert testimony in the record, rooted in observations specific to Lynn, that there are significant educational benefits to be derived from a racially diverse student body in the K-12 context. Lynn has a compelling interest in obtaining those benefits. See Brewer v. W. Irondequoit Cent. Sch. Dist., 212 F.3d 738, 752 (2d Cir.2000); McFarland, 330 F.Supp.2d at 855.
C. Narrow Tailoring
Recognizing that public schools have a compelling interest in obtaining the educational benefits of racial diversity does not give schools a blank check to adopt race-conscious policies. Rather, the government’s use of race must be narrowly tailored to achieve its compelling interest. See Grutter, 539 U.S. at 333, 123 S.Ct. 2325. “The purpose of the narrow tailoring requirement is to ensure that ‘the means chosen ‘fit’ ... th[e] compelling goal so closely that there is little or no possibility that the motive for the classification was illegitimate racial prejudice or stereotype.’ ” Id. (quoting Richmond, 488 U.S. at 493, 109 S.Ct. 706).
Narrow tailoring generally requires the proponent to show that a plan or practice is (i) necessary to the declared purpose, (ii) proportional to the declared purpose, and (iii) not more burdensome than necessary on third parties. See United States v. Paradise, 480 U.S. 149, 171, 107 S.Ct. 1053, 94 L.Ed.2d 203 (1987) (plurality opinion). It is a context-specific inquiry that must be “calibrated to fit the distinct issues raised” in a given case, taking “relevant differences into account.” Grutter, 539 U.S. at 334, 123 S.Ct. 2325 (internal quotation marks omitted).
*17Although the Supreme Court has not yet considered a constitutional challenge to a voluntary race-based transfer policy for elementary and secondary schools, its recent opinions in Grutter and Gratz provide some guidance for our narrow tailoring inquiry into the use of race to obtain the educational benefits of diversity. Thus we consider these cases further.
1. Gratz and Grutter
Gratz involved a challenge to the University of Michigan’s undergraduate admissions program. The University automatically assigned twenty points — one-fifth of the 100 points necessary to guarantee admission — to an applicant from an underrepresented racial or ethnic minority group. Gratz, 539 U.S. at 255, 123 S.Ct. 2411. This twenty-point bonus effectively made race/ethnicity determinative for minimally qualified minority applicants. Id. at 272, 123 S.Ct. 2411. Grutter involved a challenge to the University of Michigan Law School’s admissions policy. The Law School took race into account as one of several variables in an individual’s application. Grutter, 539 U.S. at 340, 123 S.Ct. 2325. It assigned no mechanical score based on an applicant’s race; instead, it considered race only as one of several possible ways in which an applicant could enrich the diversity of the student body. Id. at 315-16, 123 S.Ct. 2325.
The Supreme Court struck down the undergraduate admissions plan in Gratz while upholding the law school admissions policy in Grutter. In arriving at these decisions, the Court followed a four-part narrow tailoring inquiry. First, a race-conscious program cannot institutionalize a quota system or otherwise insulate one category of applicants from competition with another solely because of race. Id.; Gratz, 539 U.S. at 334, 123 S.Ct. 2325. Second, the government must consider whether there are any workable, race-neutral alternatives. Grutter, 539 U.S. at 339, 123 S.Ct. 2325. Third, the plan must not “unduly harm members of any racial group.” Id. at 341, 123 S.Ct. 2325. Fourth, the use of racial distinctions must be limited in time. Id. at 342, 123 S.Ct. 2325.
Much of this inquiry is relevant here despite significant differences between the competitive admissions plans at issue in Gratz and Grutter and the Lynn Plan, which is non-competitive and governs only student transfers, not initial assignments. The requirement that the court consider race-neutral alternatives addresses whether the Plan is necessary; if there were a race-neutral way to achieve the benefits of diversity and reduced racial isolation, the use of race would be unnecessary and therefore not narrowly tailored. The requirements that a race-conscious policy not unduly harm members of any racial group and that it be limited in time minimize the scope of the Plan, ensuring that its use of race is no broader than necessary. The weight of these considerations may vary somewhat from the Grutter setting to ours, but they remain applicable and we will return to them shortly.
The first Grutter criterion relating to competition, however, is less useful to our narrow tailoring inquiry. The University of Michigan admissions policies were designed to “assemble a student body that is diverse in ways broader than race.” Grutter, 539 U.S. at 340, 123 S.Ct. 2325. Individualized assessments, in which race was only one consideration among many, were the most narrowly tailored way to achieve such diversity. The mechanical use of race, by contrast, would preclude an admissions committee from considering students’ “background, experiences, and characteristics to assess [their] individual ‘potential contribution to diversity.’ ” *18Gratz, 539 U.S. at 274, 123 S.Ct. 2411 (quoting Regents of Univ. of Cal. v. Bakke, 438 U.S. 265, 317, 98 S.Ct. 2733, 57 L.Ed.2d 750 (1978) (opinion of Powell, J.)).
Unlike the Gratz and Grutter policies, the Lynn Plan is designed to achieve racial diversity rather than viewpoint diversity.9 The only relevant criterion, then, is a student’s race; individualized consideration beyond that is irrelevant to the compelling interest. Cf. Brewer, 212 F.3d at 752 (“If reducing racial isolation is — standing alone — a constitutionally permissible goal, ... then there is no more effective means of achieving that goal than to base decisions on race.”)
The concerns motivating the individualized consideration requirement in a competitive, race-preferential admissions context that focuses on diversity along a number of axes (e.g., the Gratz and Grutter policies) are simply not present in a non-competitive K-12 transfer policy aimed at racial diversity. Because transfers under the Lynn Plan are not tied to merit, the Plan’s use of race does not risk imposing stigmatic harm by fueling the stereotype that “certain groups are unable to achieve success without special protection.” Bakke, 438 U.S. at 298, 98 S.Ct. 2733 (opinion of Powell, J.) (raising the possibility of stigmatic harm in the affirmative action context). There is also little chance that the decisive use of race in a plan concerned strictly with racial diversity creates the unwarranted presumption that race is a proxy for viewpoint. See Gratz, 539 U.S. at 271, 123 S.Ct. 2411 (recognizing this as a risk when members of a group are favored based on a presumption that “persons think in a manner associated with their race”). Indeed, the Plan strives for exactly the opposite result — that is, to preempt racial stereotypes through intergroup contact.
The plaintiffs emphasize that the Supreme Court has also criticized the mechanical use of race on the ground that it may breed cross-racial tension. As the Court recently explained in considering a prison policy of segregating prison inmates by race,
racial classifications threaten to ... incite racial hostility. Indeed, by insisting that inmates be housed only with other inmates of the same race, it is possible that prison officials will breed further hostility among prisoners and reinforce racial and ethnic divisions. By perpetuating the notion that race matters most, racial segregation of inmates may exacerbate the very patterns of [violence that it is] said to counteract.
*19Johnson, 125 S.Ct. at 1147 (internal quotation marks, citations, and emphasis omitted). These concerns, however, are not applicable to the Lynn Plan, which takes race into account to foster intergroup contact rather than to segregate. As the Johnson Court acknowledged, “racial integration ... tends to diffuse racial tensions and thus diminish interracial violence.” Id. (citing the opinion of former corrections officials and a study finding that “the rate of violence between inmates segregated by race ... surpassed the rate among those racially integrated”). The Lynn Plan validates this conclusion: by reducing racial isolation and increasing intergroup contact, it has ameliorated racial and ethnic tension and bred interracial tolerance. Comfort, 283 F.Supp.2d at 376. We therefore see no reason to impose a blanket prohibition on the use of race as a decisive factor in a student transfer plan to further a compelling interest in obtaining the educational benefits of racial diversity. If a non-competitive, voluntary student transfer plan is otherwise narrowly tailored, individualized consideration of each student is unnecessary.10
2. The Lynn Plan
The district court determined, and we agree, that the Plan’s use of transfer limits to achieve racial diversity has produced benefits central to Lynn’s educational mission.11 Under the general narrow tailoring framework, however, we must also consider whether the Plan’s use of race is no broader than necessary and whether race-neutral alternatives are available. See Grutter, 539 U.S. at 339-42, 123 S.Ct. 2325; see also Wygant v. Jackson Bd. of Educ., 476 U.S. 267, 280 n. 6, 106 S.Ct. 1842, 90 L.Ed.2d 260 (1986) (noting that the term “narrowly tailored” requires “consideration of whether lawful alternative and less restrictive means could have been used” or that “the classification at issue must ‘fit’ with greater precision than any alternative means”).
a. Breadth
The defendants maintain that the Plan’s use of race is minimally invasive. First, it governs only voluntary transfers, rather than initial student assignments. Instead of forcing children to attend schools far from their homes, as might be the result of a controlled choice plan,12 the Lynn Plan *20preserves the traditional neighborhood school model. Second, the Plan allows students to transfer freely between racially balanced schools and provides an appeals process for students whose transfer requests are denied on racial grounds.13
The Plan is also less burdensome on third parties here than in other contexts because of the nature of the “benefit” at issue, namely the grant of a transfer request. Every child in Lynn is guaranteed a seat in a district where, as the parties have stipulated, every school provides a comparable education. The denial of a transfer under the Plan is therefore markedly different from the denial of a spot at a unique or selective educational institution. See, e.g., Gratz, 539 U.S. at 251, 123 S.Ct. 2411 (University of Michigan); Wessmann, 160 F.3d at 793 (Boston Latin School); cf. Wygant, 476 U.S. at 282-83, 106 S.Ct. 1842 (“Though hiring goals may burden some innocent individuals, they simply do not impose the same kind of injury that layoffs impose. Denial of a future employment opportunity is not as intrusive as loss of an existing job.”). This is not to say that the denial imposes no harm at all; the transfer request itself indicates that despite the availability of a comparable education at any school in Lynn, students (or their parents) do not view the schools as fungible. But in construing the narrow tailoring requirement that a race-conscious plan not unduly harm members of any racial group, we view the diminished nature of any harm here as significant.
i. Calibration
Despite the minimally invasive nature of the Plan, the plaintiffs contend that it imposes undue harm because of its calibration. Emphasizing the defense experts’ testimony that the educational benefits of diversity are predicated on the presence of a critical mass of white and nonwhite students, a figure that social science literature approximates at 20%, the plaintiffs assert that the Plan’s numerical guidelines are substantially more restrictive than necessary. In their view, a plan narrowly tailored to the defendants’ compelling interest in the benefits of educational diversity would prohibit only those transfers that would upset critical mass. They point out that because the Plan is calibrated around district demographics rather than around critical mass, it prohibits some transfers that do not bring a school population below 20% white. For example, because nonwhites made up 58% of Lynn’s student population at the time of trial, an elementary school with a 40% nonwhite enrollment qualified as racially isolated, and therefore subject to transfer limits, even though it contained a critical mass of white and nonwhite students.14
In response, the defendants rely on expert testimony that while critical mass is *21the point at which educational benefits begin to accrue, those benefits increase as a school nears an even balance between white and nonwhite students. Relying on this evidence, the district court found that “gains occur along a continuum: as the racial composition of school populations creeps closer to balanced, racial stereotyping and tension is [sic] reduced and racial harmony and understanding increase.” Comfort IV, 283 F.Supp.2d at 357. It thus concluded that the Plan was narrowly tailored, despite its orientation around district demographics rather than critical mass. See id. at 384 (The Plan “effectively generates integration in Lynn’s schools in such quantity as to catalyze intergroup contact while still respecting the neighborhood school principle and Lynn’s ever-changing demographics.”).
We agree with the district court’s reasoning. The Plan does not seek racial balancing for its own sake, nor does it use rigid quotas to ensure a pre-determined level of diversity at each of Lynn’s schools. See Grutter, 539 U.S. at 335-36, 123 S.Ct. 2325 (“The ... goal of attaining a critical mass of [nonwhite] students does not transform [a] program into a quota.”). Rather, the transfer policy conditioned on district demographics (+/10 — 16%) reflects the defendants’ efforts to obtain the benefits of diversity in a stable learning environment.15 The Plan thus provides a sufficiently close “fit” to the defendants’ compelling interest to ensure that “ ‘the motive for the classification was [not] illegitimate racial prejudice or stereotype.’ ” Id. at 333, 123 S.Ct. 2325 (quoting Croson, 488 U.S. at 493, 109 S.Ct. 706 (plurality op.)).
The plaintiffs launch a second attack at the Plan’s calibration on the grounds that it is inconsistent with the defendants’ statements that the benefits they seek maximize as a school moves closer to 50% white/nonwhite. They point out that as of December 2004, Lynn’s student population was more than 61.9% minority. A middle school that is 50% minority (the proportion that the defendants have described as ideal) would now fall outside of the +/10% range for racial balance and would instead be considered racially isolated, resulting in transfer limitations.
This argument misses the mark. The Lynn Plan’s goal is to improve the racial balance not of any particular school, but across the school system as a whole. The optimal balance for each school might well be 50%, but Lynn’s 61.9% minority population means that for every school closer to that ideal, another will be further away from it. Evaluating schools by reference to the racial composition of the city’s population is a sensible way for Lynn to strive for the best racial balance attainable across its entire school system, while acknowledging that practical constraints make it impossible for Lynn to have an equal population of minority and non-minority students in every individual school.
ii. White/nonwhite distinction
In addition to challenging the Plan’s numerical ranges, the plaintiffs also argue that the Plan is not narrowly tailored to advance a compelling interest in racial diversity because it paints with too broad a brush by distinguishing only between white and nonwhite students, thereby blurring the many subgroups within each category. However, this white/nonwhite *22distinction reflects the reality of Lynn’s experience. As the district court found, before the Plan, “racial divisions and ethnic conflict between students occurred predominantly along a white/nonwhite axis. The growing gap in understanding between these groups burdened the schools in ways that more precise shades of racial and ethnic difference did not.” Comfort IV, 283 F.Supp.2d at 379. By increasing diversity along the white/nonwhite axis, the Plan reduced racial tensions and produced positive educational benefits. Narrow tailoring does not require that Lynn ensure diversity among every racial and ethnic subgroup as well. See Grutter, 539 U.S. at 316, 123 S.Ct. 2325 (noting that the Law School sought to enroll a critical mass of “minority” students, a category that included African-Americans, Hispanics, and Native Americans).
iii. Duration
A narrowly tailored plan must be limited not only in scope, but also in time. See id. at 342, 123 S.Ct. 2325. The Court held in Grutter that this durational requirement can be met by “periodic reviews to determine whether racial preferences are still necessary to achieve student body diversity.” Id. The Lynn Plan includes such review. The PIC continuously monitors the schools’ demographies, gathering data on racial composition and transfers. Under the Plan, transfer limits are suspended among schools that are racially balanced. Comfort IV, 283 F.Supp.2d at 377. This feature is not merely theoretical. Students may now transfer freely among all three Lynn high schools. Lynn has also periodically reevaluated the calibration of its Plan with an eye toward maximizing the availability of transfers while maintaining diverse schools. Id. at 348 n. 38 (noting that the Plan’s original 10% range was expanded to 15% for elementary schools to “permit more choice” and that Lynn considered a 20% range in 1994 but determined that it would compromise student body diversity). We expect that Lynn will continue to do so, presuming, as did the Grutter Court, that school officials will demonstrate a good faith commitment to monitoring the continued need for racial restrictions. See 539 U.S. at 343, 123 S.Ct. 2325.
b. Consideration of race-neutral alternatives
Because narrow tailoring dictates that the government use race only when necessary to achieve a compelling interest, it requires “serious, good-faith consideration of workable race-neutral alternatives that will achieve the diversity [the government actor] seeks.” Grutter, 539 U.S. at 339, 123 S.Ct. 2325; see also Wygant, 476 U.S. at 280 n. 6, 106 S.Ct. 1842. Here, the defendants have met their burden. The record reflects that they seriously considered, and plausibly rejected, a number of race-neutral alternatives. These included (i) a no-transfer policy, see Comfort IV, 283 F.Supp.2d at 387-88 (crediting evidence from a demographics expert that instituting such a policy would throw several elementary schools into racial imbalance); (ii) a policy of unrestricted transfers, see id. at 388 (crediting evidence that instituting such a policy would result in 500 to 800 segregative transfers per year); (iii) a redrawing of district lines, see id. at 387-88 (noting that this would be impractical); (iv) forced busing, see id. at 387-88 (concluding that the problems that accompany forced busing justified Lynn’s rejection of a controlled choice scheme); (v) a lottery system, see id. at 389 (finding that demographic and scheduling factors made this impracticable); and (vi) a plan conditioning transfers on socioeconomic status, rather than race, see id. at 389 n. 100 (noting that because of residential patterns, this system *23would exacerbate existing racial imbalance).
The plaintiffs argue that there are several other alternatives that the defendants failed to consider. They point specifically to a Department of Education study reviewing successful race-neutral programs based on socioeconomic status or a lottery, see U.S. Dep’t of Educ., Achieving Diversity: Race-Neutral Alternatives in American Education (Feb.2004), available at http://www.ed.gov/abouVoffiees/list/ocr/ra-ceneutrahhtml, and to the race-neutral student assignment plan adopted in Boston, see Anderson, 375 F.3d at 76-77. As noted, Lynn has already considered, and rejected, the possibility of basing student assignments on socioeconomic status or a lottery. While the record does not reflect whether Lynn has considered the Boston plan in depth, we note that the Boston plan is specific to the residential patterns in Boston, which differ from those in Lynn. Lynn must keep abreast of possible alternatives as they develop, see Grutter, 539 U.S. at 342-43, 123 S.Ct. 2325, but it need not prove the impracticability of every conceivable model for racial integration. It is sufficient that it demonstrate a good faith effort to consider feasible race-neutral alternatives, as it has done here. We therefore hold that the Lynn Plan is narrowly tailored to the defendants’ compelling interest in obtaining the benefits of racial diversity.
D. Related Federal Claims
The plaintiffs also advance several statutory equal protection claims, contending that the Lynn Plan violates, inter alia, 42 U.S.C. §§ 1981 and 1983,16 and Title VI of the Civil Rights Act of 1964, 42 U.S.C. § 2000d.17 Our resolution of the constitutional equal protection challenge controls those claims. Title VI “ ‘proscribe[s] only those racial classifications that would violate the Equal Protection Clause.’ ” Alexander v. Sandoval 532 U.S. 275, 280-281, 121 S.Ct. 1511, 149 L.Ed.2d 517 (2001) (quoting Bakke, 438 U.S. at 287, 98 S.Ct. 2733 (Powell, J.)). Courts have also treated the bar on racial discrimination imposed by § 1981 and § 1983 as coextensive with the protections of the Equal Protection Clause. See Mescall v. Burrus, 603 F.2d 1266, 1271 (7th Cir.1979) (“The relationships of §§ 1981 and 1983 to the Fourteenth Amendment are so close ... that we believe the use of each section must be guided by the principles announced by the Supreme Court for application of the Fourteenth Amendment to discrimination cases.”); see also Anderson, 375 F.3d at 77 n. 7 (concluding that plaintiffs’ claims under Title VI, § 1981 and § 1983 “turn on the resolution of the equal protection claim”). The district court was therefore correct in holding that the plaintiffs are not entitled to federal statutory relief.
IV. Article 111
Article 111 of the Massachusetts Declaration of Rights provides that “[n]o student shall be assigned to or denied admittance to a public school on the basis of race, color, national origin[,] or creed.” Plaintiffs contend that when a child is prevented from making a segregative transfer under the Lynn Plan, he is illegally “de*24nied admittance to a public school on the basis of race.” Mass. Const, amend, art. 111. The district court rejected this position as inconsistent with the meaning of Article 111, which was designed to limit forced busing and to protect neighborhood schools. Comfort IV, 283 F.Supp.2d at 393. The court also emphasized that although “no court has ever expressly interpreted Article 111, the SJC [ie., the Massachusetts Supreme Judicial Court] has consistently construed similarly worded statutes narrowly, holding that they do not categorically ban suspect classifications but rather merely subject them to strict scrutiny.” Id. Finally, the district court suggested that if Article 111 were read broadly to prohibit any race-conscious student assignment plan, as the plaintiffs urge, it would likely run afoul of the federal Constitution and other sections of the Massachusetts Declaration of Rights. Id. We review the district court’s interpretation of the Article de novo.18 See Blockel v. J.C. Penney Co., Inc., 337 F.3d 17, 29 (2003).
Although the SJC has not yet considered Article 111, it has identified the considerations that guide judicial interpretation of the state’s constitution:
In determining the meaning of a constitutional provision, we look to the language and structure of the provision, so that it is construed so as to accomplish a reasonable result and to achieve its dominating purpose. We do so bearing in mind the Constitution was written to be understood by the voters to whom it was submitted for approval. It is to be interpreted in the sense most obvious to the common intelligence. Its phrases are to be read and construed according to the familiar and approved usage of the language. The words of a constitutional provision are to be given their natural and obvious sense according to common and approved usage at the time of its adoption.
Moreover, the Constitution is to be interpreted in the light of the conditions under which it and its several parts were framed, the ends which it was designed to accomplish, the benefits which it was expected to confer, and the evils which it was hoped to remedy.
McDuffy v. Sec’y of the Executive Office of Educ., 415 Mass. 545, 615 N.E.2d 516, 523 (1993) (citations and internal quotation marks omitted).
To determine whether Article 111 prohibits a race-conscious plan for voluntary student transfers, we begin, as the SJC has instructed, with the language of the provision. Id. at 524. The plaintiffs contend that the plain language of the Article unequivocally invalidates the Lynn Plan. They reason that when a student is precluded from transferring because of the transfer’s impact on racial balance, that student is “denied admittance to a public school on the basis of race” within the meaning of Article 111.
It is not readily apparent, however, that a student denied a transfer is “denied admittance” to a school within the meaning of the provision. The Article’s pairing of the terms “assigned” and “denied admittance” suggests that it contemplated initial student placements, not subsequent transfers. Under the Lynn Plan, students are assigned to their neighborhood schools *25without regard to race; conversely, no student is denied that placement based on race. Only after a student has been assigned to a school and wishes to transfer does race enter the calculus.
The “specific circumstances of the adoption” of Article 111, see McDuffy, 615 N.E.2d at 528, convince us that the provision does not preclude such a plan. Article 111 was adopted as an amendment to the Massachusetts constitution in 1978. The House and Senate bills that became Article 111 were introduced by legislators on behalf of Massachusetts Citizens Against Forced Busing. See Comfort IV, 283 F.Supp.2d at 395. When the bill was presented to voters for ratification, the constitutionally required description that accompanied it19 explained:
A “YES VOTE” would guarantee the right of parents or guardians of school-age children to educate those children free from any arbitrary assignment by school authorities to schools outside the school district. Any public assignment to a school outside the school district, based on achieving any established racial quota-system or ethnic balance^] would require the permission of a parent or guardian.
Id. (citing Massachusetts Information for Voters (1978)); see also Bates v. Dir. of Office of Campaign and Political Finance, 436 Mass. 144, 763 N.E.2d 6, 22-23 (2002) (using Information for Voters statement as an aid to statutory construction).
We would be hard-pressed to extrapolate from these circumstances an intent that Article 111 bar an entirely voluntary student transfer program narrowly tailored to a compelling interest in the educational benefits of diversity. As the district court noted, “ ‘the benefit [Article 111] was expected to confer’ was preservation of neighborhood schools, and the ‘evil which it was hoped to remedy’ was the politically divisive resort to forced busing.” Comfort IV, 283 F.Supp.2d at 395 (quoting McDuffy, 615 N.E.2d at 523). The Lynn Plan advances those goals. It protects neighborhood school assignments and does not entail any forced non-neighborhood assignments. Race-conscious transfers occur only if initiated by a student’s parent or guardian, a result entirely consistent with the explanation of the Article put before the voters.
The plaintiffs do not dispute that Article 111 was designed to prevent forced busing. Rather, they argue that any inquiry into the Article’s legislative history is improper because the provision is clear on its face. As we have already explained, the application of the Article to the context of voluntary transfers is not entirely clear from the statutory language. Moreover, the SJC has expressly provided that the state constitution’s words “must be given a construction adapted to carry into effect its purpose.” Cohen v. Att’y Gen., 357 Mass. 564, 259 N.E.2d 539, 543 (1970) (quoting Tax Comm’r v. Putnam, 227 Mass. 522, 116 N.E. 904, 906 (1917)). The Article’s legislative history helps us determine its purpose, and thereby points us to a construction that furthers that purpose. Looking to legislative history is therefore entirely consistent with the principles of constitutional interpretation set forth by the SJC.
Our conclusion that the Lynn Plan does not violate Article 111 would be the same *26even if the Article did apply to a voluntary transfer system. As the district court recognized, the SJC has treated provisions similar to Article 111 not as banning the use of race outright, but rather as subjecting the use of race to strict scrutiny. Comfort IV, 283 F.Supp.2d at 396; see also McDuffy, 615 N.E.2d at 545 (noting that the court’s decision on a matter of first impression was consistent with earlier decisions).
In Attorney General v. Massachusetts Interscholastic Athletic Association, Inc., 378 Mass. 342, 393 N.E.2d 284 (1979), the SJC considered a challenge, brought under the Massachusetts Equal Rights Amendment (“ERA”) and a state statute, to a rule prohibiting boys from joining girls’ sports teams. The ERA provides that “[ejquality under the law shall not be denied or abridged because of sex, race, col- or, creed or national origin.” Mass. Const, pt. 1, art. 1, as amended by art. 106. The relevant statute provides: “[n]o person shall be excluded from or discriminated against in admission to a public school ... or in obtaining the advantages, privileges and courses of study of such public school on account of race, color, sex, religion, national origin or sexual orientation.” Mass. Gen. L. ch. 76, § 5. Although both of these provisions speak in absolute terms, the SJC did not treat them as automatically invalidating the gender-based prohibition at issue. Rather, it subjected the “no boys on girls’ teams” rule to something akin to strict scrutiny. 393 N.E.2d at 291-93. We have already determined that the Lynn Plan survives the strict scrutiny required by the federal Equal Protection Clause. Therefore, it would also survive review under Article 111, if the Article did apply in this context. Either way, the plaintiffs’ claim fails.
Y. Recusal
Finally, the plaintiffs assert that the district court judge should have recused herself from this ease. Their argument is as follows: (1) prior to her appointment to the federal bench, Judge Gertner was a member of the Lawyers’ Committee for Civil Rights (“LCCR”), a nonprofit organization; (2) LCCR unsuccessfully moved to intervene in this litigation on the side of the defendants, and therefore (3) the law required that Judge Gertner recuse herself. Judge Gertner denied the recusal motion in an order dated March 21, 2002. We review that ruling for abuse of discretion. Camacho v. Autoridad de Telefonos, 868 F.2d 482, 490 (1st Cir.1989).
The controlling statute is 28 U.S.C. § 455, which sets forth the standards for recusal. That statute provides that a judge “shall” recuse herself “in any proceeding in which [her] impartiality might reasonably be questioned.” Id. § 455(a). A party who suggests that recu-sal is appropriate must support the motion with facts that “provide what an objective, knowledgeable member of the public would find to be a reasonable basis for doubting the judge’s impartiality.” In re Boston’s Children First, 244 F.3d 164, 167 (1st Cir. 2001) (quoting In re United States, 666 F.2d 690, 694 (1st Cir.1981)). The plaintiffs have failed to make this showing.
Every judge comes to the bench with a lifetime of background experiences, a roster of associations, and a myriad of views. This past history, in and of itself, is seldom sufficient to require recusal. Brody v. President & Fellows of Harvard Coll., 664 F.2d 10, 11 (1st Cir.1981) (per curiam). Unless there is a direct link that establishes a reasonable basis for doubting impartiality, the judge should not step aside. In re United States, 158 F.3d 26, 31 (1st Cir.1998); cf. United States v. Giorgi, 840 F.2d 1022, 1035 (1st Cir.1988) *27(explaining that unless a party can establish a reasonable factual basis to doubt a judge’s impartiality “by some kind of probative evidence,” then the “judge must hear a case as assigned”) (internal quotation marks omitted).
These principles govern our decision here. LCCR is not a party to this case. Even if it were, Judge Gertner’s relationship with that organization ended when she took the bench on February 14, 1994. There is no allegation that she has maintained ties with the LCCR. Given the eight-year interval between the end of the judge’s connection with LCCR and the recusal motion, her prior association with that organization cannot be the basis for a reasonable attack on her impartiality. Thus, recusal was not obligatory. See, e.g., Sierra Chib v. Simkins Indus., Inc., 847 F.2d 1109, 1117 (4th Cir.1988) (holding that an association with a nonprofit organization that ended a decade before the proceedings commenced does not form a reasonable basis for questioning the trial judge’s impartiality, even though the nonprofit organization was a party to the litigation); cf. Veneklase v. City of Fargo, 236 F.3d 899, 901 (8th Cir.2000) (finding recusal unnecessary where judge’s former law firm had represented parties tangentially involved in civil rights suit); United States v. Story, 716 F.2d 1088, 1090 (6th Cir.1983) (finding recusal unnecessary where judge had represented victim in will contest matter thirteen years earlier).
The plaintiffs’ argument that our decision in Boston’s Children First dictates recusal here ignores an important difference between the two cases. In Boston’s Children First, we ordered recusal not because of the judge’s past association with civil rights organizations, but because the judge had publicly commented on a pending case. See 244 F.3d at 169-70. The judge’s passing comment in this case that LCCR was eligible to apply for ami-cus status does not constitute a basis on which to require recusal.
The plaintiffs’ claim of bias is equally unpersuasive. The only evidence of bias they cite involves their view of the judge’s policy preferences. That perception, standing alone, does not warrant our interference with the district judge’s recusal decision. If judges were subject to disqualification on such a basis, our judicial system would be paralyzed. See Camacho, 868 F.2d at 491. We therefore conclude that the court did not abuse its discretion in denying the motion for recusal.
Affirmed.
.Ordinarily, the label "magnet school” describes an elite public school with a competitive admissions policy. Lynn's magnet schools, however, do not use a competitive admissions policy. Rather, Lynn uses the term for schools that feature an educational theme beyond the standard curriculum, designed partially to attract cross-neighborhood transfers. Examples of such themes include “Life Science” and "Reading and Writing Literary and Whole Language.” Despite this specialization the parties have stipulated that "the education provided ... in each of the elementary, middle, and high schools in Lynn is comparable in quality, resources!,] and curriculum.”
. The Plan was amended in 1990 and again in 1999.
. Lynn also operates six alternative schools, offering such things as special needs education and vocational training. These schools are not subject to the transfer provisions of the Lynn Plan.
. Any student who qualifies as '‘multi-racial’' is not subject to the race-conscious transfer limits.
. Before 2001, the RIA provided reimbursement for school construction and renovations undertaken for the purpose of reducing racial imbalance. The law has been amended to change this funding structure, but that amendment applies prospectively and does not affect Lynn's state funding.
. We express our appreciation to the many amici curiae for their valuable assistance.
. Nor have the plaintiffs shown that they are under any imminent threat of being subjected to these mandates.
. Notably, one of the studies that the Supreme Court cited as demonstrating that diversity produces educational benefits was authored by the defendants’ expert in this case, Dr. Gary Orfield.
. As we have already discussed, see supra Part III.B., the Lynn Plan's focus on racial diversity rather than viewpoint diversity is the result of contextual differences between higher education, where the emphasis is on the exchange of ideas, and primary education, where the emphasis is on fostering interracial cooperation. The district court explained this point in distinguishing Grutter, which was then pending before the Supreme Court:
In contrast [to Grutter ], the “critical mass” sought by the Lynn Plan is different, because Lynn's goal is not viewpoint diversity. As I have said, at the elementary, middle, and high school level, the goal of teaching socialization is at least as important as the subject matter of instruction. The value of a diverse classroom setting at these ages does not inhere in the range of perspectives and experience that students can offer in discussions; rather, diversity is valuable because it enables students to learn racial tolerance by building cross-racial relationships. In this context a meaningful presence of racial minorities — and of whites at minority-dominated schools — is crucial not only to reducing feelings of tokenism, but also to disarming stereotypes that students in the classroom majority might harbor about students of other races.
Comfort IV, 283 F.Supp.2d at 381 n. 90.
. We note that this conclusion in no way rests on the administrative difficulties that would be inherent in individually considering each of the thousands of transfer requests that Lynn receives each year. Administrative difficulty "does not render constitutional an otherwise problematic system.” Gratz, 539 U.S. at 275, 123 S.Ct. 2411.
. Plaintiffs argue that improvements in Lynn schools cannot be attributable to racial diversity. They emphasize (i) that levels of diversity vary at schools in the district; (ii) that at least one school slipped below critical mass during the 2000-01 academic year; and (iii) that defense experts testified that all schools they visited — including schools that had slipped below critical mass — demonstrated the benefits that intergroup contact theory attributes to racial diversity. Plaintiffs reason that the experts' uniformly positive testimony is inconsistent with a theory that increased diversity produces increased benefits, and that the benefits must be attributable instead to race-neutral factors present in equal measure throughout the district. We do not find this argument persuasive. The defendants point out that although some of Lynn's schools occasionally fall below critical mass, those deviations are small and temporary. Students do not automatically forfeit the lessons learned from integration when they attend a school with relatively short-term or marginal deviation from critical mass.
.Controlled choice plans are an alternative to neighborhood school assignments. See, e.g., Anderson, 375 F.3d at 74-77 (describing controlled choice plans used by the Boston Public Schools). "Under such programs, parents can choose among a select number of schools, but their choices and their likelihood *20of getting their choice are controlled to help ensure a particular racial balance.” Wendy Parker, The Legal Cost of the "Split Double Header" of Gratz and Grutter, 31 Hastings Const. L.Q. 587, 603 n. 76 (2003).
. Appeals are granted to unite siblings or when parents can show a medical, safety, or other hardship, including one based on daycare arrangements. The district court found that the Parent Information Center (Lynn’s central registration office) "goes out of its way to make the appeals process accessible to everyone.” Comfort IV, 283 F.Supp.2d at 349. Additionally, a student whose appeal is denied will be presented with alternative placement options.
. Under the Plan, an elementary school is racially isolated if its nonwhite enrollment falls more than 15% below the percentage of Lynn’s total student population that is nonwhite. If Lynn's student population was 58% nonwhite, as it was during the 2001-02 academic year, a school whose student body was less than 43% nonwhite (i.e., more than 15% below 58%) was racially isolated.
. This conclusion is bolstered by the testimony of Dr. Orfield, a nationally recognized expert on school desegregation, who concluded that the Plan "used race no more than was necessary to allow Lyrm to meet its educational goal of preparing students to live in a multiracial society.” Comfort IV, 283 F.Supp.2d at 355.
. 42 U.S.C. § 1981 provides that "[a]ll persons within the jurisdiction of the United States shall have the same right ... to the full and equal benefit of all laws.” 42 U.S.C. § 1983 provides a cause of action based on the deprivation of constitutional rights “under color of any statute, ordinance, regulation, custom, or usage, of any State.”
. Title VI forbids racial discrimination “under any program or activity receiving Federal financial assistance.” 42 U.S.C. § 2000d.
. The defendants assert that we should not consider the substance of the Article 111 claim at all, maintaining that the Commonwealth was a necessary party to adjudicate this claim but could not be joined because of the Eleventh Amendment, see Fed.R.Civ.P. 19(b). Alternatively, they urge us to certify a question regarding the proper interpretation of Article 111 to the SJC pursuant to Mass. R. Sup. Jud. Ct. 1:03. We reject both contentions, finding it appropriate to reach the merits as the district court did.
. The Massachusetts Constitution requires the Secretary of State to publish and distribute information regarding ballot questions, including the full text of every measure, "a fair, concise summary of the measure as such summary will appear on the ballot [and] other information and arguments for and against the measure.” Mass. Const. art. 48, Gen. Prov., pt. IV, as amended by arts. 74 and 108.