Parents Involved in Community Schools v. Seattle School District, No. 1 ( 2005 )


Menu:
  •                                                 Volume 1 of 2
    
                       FOR PUBLICATION
      UNITED STATES COURT OF APPEALS
           FOR THE NINTH CIRCUIT
    
    PARENTS INVOLVED IN COMMUNITY            
    SCHOOLS, a Washington nonprofit
    corporation,
                        Plaintiff-counter-
                    defendant-Appellant,
                       v.
    SEATTLE SCHOOL DISTRICT, NO. 1, a
    political subdivision of the State
    of Washington; JOSEPH OLCHEFSKE,
    in his official capacity as
    superintendent; BARBARA SCHAAD-               No. 01-35450
    LAMPHERE, in her official capacity
    as President of the Board of                   D.C. No.
                                                 CV-00-01205-BJR
    Directors of Seattle Public
                                                    OPINION
    Schools; DONALD NEILSON, in his
    official capacity as Vice President
    of the Board of Directors of
    Seattle Public Schools; STEVEN
    BROWN; JAN KUMASAKA; MICHAEL
    PRESTON; NANCY WALDMAN, in
    their official capacities as
    members of the board of
    Directors,
                     Defendants-counter-
                     claimants-Appellees.
                                             
           Appeal from the United States District Court
              for the Western District of Washington
         Barbara Jacobs Rothstein, District Judge, Presiding
    
    
                                 14653
    14654    PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
                 Argued and Submitted En Banc
            June 21, 2005—San Francisco, California
    
                     Filed October 20, 2005
    
     Before: Mary M. Schroeder, Chief Judge, Harry Pregerson,
    Alex Kozinski, Andrew J. Kleinfeld, Michael Daly Hawkins,
              William A. Fletcher, Raymond C. Fisher,
             Richard C. Tallman, Johnnie B. Rawlinson,
      Consuelo M. Callahan and Carlos T. Bea, Circuit Judges.
    
                    Opinion by Judge Fisher;
                 Concurrence by Judge Kozinski;
                     Dissent by Judge Bea
              PARENTS INVOLVED v. SEATTLE SCHOOL DIST.       14659
                             COUNSEL
    
    Harry J.F. Korrell (argued) and Daniel B. Ritter, Davis
    Wright Tremaine LLP, Seattle, Washington, for the plaintiff-
    counter-defendant/appellant.
    
    Michael Madden (argued) and Carol Sue Janes, Bennett Bige-
    low & Leedom, P.S., Seattle, Washington, and Mark S.
    Green, Office of the General Counsel, Seattle School District
    No. 1, Seattle, Washington, for the defendants-counter-
    claimants/appellees.
    
    Sharon L. Browne, Pacific Legal Foundation, Sacramento,
    California, and Russell C. Brooks, Pacific Legal Foundation,
    Bellevue, Washington, for amici curiae Pacific Legal Founda-
    tion, American Civil Rights Institute, American Civil Rights
    Union and Center for Equal Opportunity in support of
    plaintiff-counter-defendant/appellant.
    
    Paul J. Lawrence, Preston Gates & Ellis LLP, Seattle, Wash-
    ington, for amicus curiae American Civil Liberties Union in
    support of defendants-counter-claimants/appellees.
    
    
                             OPINION
    
    FISHER, Circuit Judge, with whom Chief Judge Schroeder
    and Judges Pregerson, Hawkins, W. Fletcher and Rawlinson
    join concurring; Judge Kozinski, concurring in the result:
    
       This appeal requires us to consider whether the use of an
    integration tiebreaker in the open choice, noncompetitive,
    public high school assignment plan crafted by Seattle School
    District Number 1 (the “District”) violates the federal Consti-
    tution’s Equal Protection Clause. Our review is guided by the
    principles articulated in the Supreme Court’s recent decisions
    regarding affirmative action in higher education, Grutter v.
    14660        PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    Bollinger, 
    539 U.S. 306
     (2003), and Gratz v. Bollinger, 
    539 U.S. 244
     (2003), and the Court’s directive that “[c]ontext mat-
    ters when reviewing race-based governmental action under
    the Equal Protection Clause.” Grutter, 539 U.S. at 327. We
    conclude that the District has a compelling interest in securing
    the educational and social benefits of racial (and ethnic)
    diversity, and in ameliorating racial isolation or concentration
    in its high schools by ensuring that its assignments do not
    simply replicate Seattle’s segregated housing patterns.1 We
    also conclude that the District’s Plan is narrowly tailored to
    meet the District’s compelling interests.
    
                               I. Background2
    
      A. Seattle Public Schools: A Historical Perspective
    
       Seattle’s historical struggle with the problem of racial isola-
    tion in its public school system provides the context for the
    District’s implementation of the current challenged assign-
    ment plan. Seattle is a diverse community. Approximately 70
    percent of its residents are white, and 30 percent are non-
    white. Seattle public school enrollment breaks down nearly
    inversely, with approximately 40 percent white and 60 per-
    cent nonwhite students. A majority of the District’s white stu-
    dents live in neighborhoods north of downtown, the
    historically more affluent part of the city. A majority of the
    city’s nonwhite students, including approximately 84 percent
    of all African-American students, 74 percent of all Asian-
    American students, 65 percent of all Latino students and 51
       1
         The terms “racial diversity,” “racial concentration” and “racial isola-
    tion” have been used by the District to encompass racial and ethnic diver-
    sity, concentration and isolation. For the purposes of this opinion, we
    adopt this shorthand.
       2
         We draw the following restatement of facts largely from the district
    court opinion, see Parents Involved in Cmty. Schs. v. Seattle Sch. Dist. No.
    1, 
    137 F. Supp. 2d 1224
     (W.D. Wash. 2001) (“Parents I”), and the Wash-
    ington Supreme Court Opinion, see Parents Involved in Cmty. Schs. v.
    Seattle Sch. Dist. No. 1, 
    72 P.3d 151
     (Wash. 2003) (“Parents IV”).
                PARENTS INVOLVED v. SEATTLE SCHOOL DIST.                14661
    percent of all Native-American students, live south of down-
    town.
    
       The District operates 10 four-year public high schools.
    Four are located north of downtown — Ballard, Ingraham,
    Nathan Hale and Roosevelt; five are located south of down-
    town — Chief Sealth, Cleveland, Franklin, Garfield and Rai-
    nier Beach; one is located west of downtown — West Seattle.
    For over 40 years, the District has made efforts to attain and
    maintain desegregated schools and avoid the racial isolation
    or concentration that would ensue if school assignments repli-
    cated Seattle’s segregated housing patterns. Since the 1960s,
    while courts around the country ordered intransigent school
    districts to desegregate, Seattle’s School Board voluntarily
    explored measures designed to end de facto segregation in the
    schools and provide all of the District’s students with access
    to diverse and equal educational opportunities.
    
      In the late 1950s and early 1960s, school assignments were
    made strictly on the basis of neighborhood.3 In 1962, Garfield
    High School reported 64 percent minority enrollment and it
    accommodated 75 percent of all African-American students.
    Meanwhile, the eight high schools serving other major areas
    of the city remained more than 95 percent white.
    
       The District responded to this imbalance, and racial ten-
    sions in the de facto segregated schools, in various ways. In
    the early 1960s, the District first experimented with small-
    scale exchange programs in which handfuls of students
    switched high schools for five-week periods. In 1963, expand-
    ing on this concept, the District implemented a “Voluntary
      3
        The history that follows comes principally from two documents in the
    district court record. One is a report entitled, “The History of Desegrega-
    tion in Seattle Public Schools, 1954-1981,” which was prepared by the
    District’s desegregation planners. The other is the “Findings and Conclu-
    sions” adopted by the Board in support of the current assignment plan.
    (They are cited as History of Desegregation and Findings and Conclu-
    sions, respectively.)
    14662      PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    Racial Transfer” program through which a student could
    transfer to any school with available space if the transfer
    would improve the racial balance at the receiving school. In
    the 1970s, the District increased its efforts again, this time
    adopting a desegregation plan in the middle schools that
    requested volunteers to transfer between minority- and
    majority-dominated neighborhood schools and called for
    mandatory transfers when the number of volunteers was
    insufficient, though this portion of the plan was never imple-
    mented. The District also took steps to desegregate Garfield
    High School by changing its educational program, improving
    its facilities and eliminating “special transfers” that had previ-
    ously allowed white students to leave Garfield. Finally, for
    the 1977-78 school year, the District instituted a magnet-
    school program. According to the District’s history:
          While it appeared evident that the addition of magnet
          programs would not in itself desegregate the Seattle
          schools, there was supportive evidence that volun-
          tary strategies, magnet and non-magnet, could be
          significant components of a more comprehensive
          desegregation plan.
    History of Desegregation at 32.
       By the 1977-78 school year, segregation had increased:
    Franklin was 78 percent minority, Rainier Beach 58 percent,
    Cleveland 76 percent and Garfield 65 percent. Other high
    schools ranged from 9 percent to 23 percent minority enroll-
    ment.
       In the spring of 1977, the Seattle branch of the National
    Association for the Advancement of Colored People
    (“NAACP”) filed a complaint with the United States Depart-
    ment of Education’s Office of Civil Rights, alleging that Seat-
    tle’s School Board had acted to further racial segregation in
    the city’s schools. Several other organizations, principally the
    American Civil Liberties Union (“ACLU”), formally threat-
    ened to file additional actions if the District failed to adopt a
    mandatory desegregation plan. When the District agreed to
                PARENTS INVOLVED v. SEATTLE SCHOOL DIST.              14663
    develop such a plan, the Office of Civil Rights concomitantly
    agreed to delay its investigation, and the ACLU agreed to
    delay filing a lawsuit.
       During the summer of 1977, the District and community
    representatives reviewed five model plans. Ultimately, the
    District incorporated elements of each model into its final
    desegregation plan, adopted in December 1977 and known as
    the “Seattle Plan.” The Seattle Plan divided the district into
    zones, within which majority-dominated elementary schools
    were paired with minority-dominated elementary schools to
    achieve desegregation. Mandatory high school assignments
    were linked to elementary school assignments, although vari-
    ous voluntary transfer options were available. With the Seattle
    Plan,
          Seattle became the first major city to adopt a com-
          prehensive desegregation program voluntarily with-
          out a court order. By doing so the District
          maintained local control over its desegregation plan
          and was able to adopt and implement a plan which
          in the eyes of the District best met the needs of Seat-
          tle students and the Seattle School District.
    History of Desegregation at 36-37. Opponents of the Seattle
    Plan immediately passed a state initiative to block its imple-
    mentation, but the Supreme Court ultimately declared the ini-
    tiative unconstitutional. Washington v. Seattle Sch. Dist. No.
    1, 
    458 U.S. 457
    , 470 (1982).
       The Seattle Plan furthered the District’s school desegrega-
    tion goals, but its operation was unsatisfactory in other ways.4
    In 1988, a decade after its implementation, the District aban-
    doned the Seattle Plan and adopted a new plan that it referred
    to as “controlled choice.” Under the controlled choice plan,
    schools were grouped into clusters that met state and district
      4
        For example, the Seattle Plan was confusing, required mandatory bus-
    ing of nonwhite students in disproportionate numbers, made facilities and
    enrollment planning difficult and contributed to “white flight” from the
    city schools. Findings and Conclusions at 30.
    14664      PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    desegregation guidelines, and families were permitted to rank
    schools within the relevant cluster, increasing the predictabil-
    ity of assignments. Because of Seattle’s housing patterns, the
    District’s planners explained that “it was impossible to fash-
    ion clusters in a geographically contiguous manner”; some
    cluster schools were near students’ homes, but others were in
    “racially and culturally different neighborhoods.” Findings
    and Conclusions at 30-31. Although roughly 70 percent of
    students received their first choices, the controlled choice plan
    still resulted in mandatory busing for 16 percent of the Dis-
    trict’s students.
       In 1994, the Board directed District staff to devise a new
    plan for all grade levels to simplify assignments, reduce costs
    and increase community satisfaction, among other things. The
    guiding factors were to be choice, diversity and predictability.
    Staff developed four basic options, including the then-existing
    controlled choice plan, a regional choice plan, a neighborhood
    assignment plan with a provision for voluntary, integration-
    positive transfers and an open choice plan.
       Board members testified that they considered all the
    options as they related to the District’s educational goals —
    with special emphasis, at the secondary school level, on the
    goals of choice and racial diversity. Neighborhood and
    regional plans were viewed as unduly limiting student choice,
    on which the District placed high value because student
    choice was seen to increase parental involvement in the
    schools and promote improvements in quality through a mar-
    ketplace model. The District sought to maintain its commit-
    ment to racially integrated education by establishing diversity
    goals while moving away from the rigid desegregation guide-
    lines and mandatory assignments prevalent in the 1970s and
    1980s.
       The Board adopted the current open choice plan (the
    “Plan”) for the 1998-99 school year. Under the Plan, students
    entering the ninth grade may select any high school in the
    District. They are assigned, where possible, to the school they
    list as their first choice. If too many students choose the same
                 PARENTS INVOLVED v. SEATTLE SCHOOL DIST.                14665
    school as their first choice, resulting in “oversubscription,”
    the District assigns students to each oversubscribed school
    based on a series of tiebreakers. If a student is not admitted
    to his or her first choice school as a result of the tiebreakers,
    the District tries to assign the student to his or her second
    choice school, and so on. Students not assigned to one of their
    chosen schools are assigned to the closest school with space
    available; students who list more choices are less likely to
    receive one of these “mandatory” assignments. The most
    recent version of the Plan, which the School Board reviews
    annually, is for the 2001-02 school year and is the subject of
    this litigation.
       B. The Plan
       The District has sought to make each of its 10 high schools
    unique, with programs that respond to the continually chang-
    ing needs of students and their parents. Indeed, the District
    implemented the Plan as part of a comprehensive effort to
    improve and equalize the attractiveness of all the high
    schools, including adoption of a weighted funding formula, a
    facilities plan and a new teacher contract that would make
    teacher transfers easier. Nevertheless, the high schools vary
    widely in desirability. Three of the northern schools — Bal-
    lard, Nathan Hale and Roosevelt — and two of the southern
    schools — Garfield and Franklin — are highly desirable and
    oversubscribed, meaning that more students wish to attend
    those schools than capacity allows.5 The magnitude of the
      5
       The current popularity of Ballard High School is illustrative of the con-
    stantly changing dynamic of Seattle’s public high schools. In the fall of
    1999, Ballard moved to a new facility under the leadership of a new prin-
    cipal. Prior to the move, Ballard was not oversubscribed; now it is one of
    the most popular high schools in Seattle.
       Similarly, the popularity and demographics of Nathan Hale High School
    changed significantly when it acquired a new principal who instituted a
    number of academic innovations, including joining the “Coalition of
    Essential Schools” and instituting a “Ninth Grade Academy” and “Tenth
    Grade Integrated Studies Program.” Prior to 1998, Nathan Hale, a north
    area high school, was not oversubscribed, and the student body was pre-
    dominantly nonwhite. Starting in 1998, the high school began to have a
    waitlist, and more white students, who had previously passed on Nathan
    Hale, wanted to go there. As a result, the number of nonwhite students
    declined dramatically between 1995 and 2000.
    14666       PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    oversubscription is noteworthy: For the academic year 2000-
    01, approximately 82 percent of students selected one of the
    oversubscribed schools as their first choice, while only about
    18 percent picked one of the undersubscribed high schools as
    their first choice. Only when oversubscription occurs does the
    District become involved in the assignment process.
       If a high school is oversubscribed, all students applying for
    ninth grade are admitted according to a series of four tie-
    breakers, applied in the following order: First, students who
    have a sibling attending that school are admitted. In any given
    oversubscribed school, the sibling tiebreaker accounts for
    somewhere between 15 to 20 percent of the admissions to the
    ninth grade class.
       Second, if an oversubscribed high school is racially imbal-
    anced — meaning that the racial make up of its student body
    differs by more than 15 percent from the racial make up of the
    students of the Seattle public schools as a whole — and if the
    sibling preference does not bring the oversubscribed high
    school within plus or minus 15 percent of the District’s demo-
    graphics, the race-based tiebreaker is “triggered” and the race
    of the applying student is considered. (For the purposes of the
    race-based tiebreaker, a student is deemed to be of the race
    specified in his or her registration materials.) Thus, if a school
    has more than 75 percent nonwhite students (i.e., more than
    15 percent above the overall 60 percent nonwhite student pop-
    ulation) and less than 25 percent white students, or when it
    has less than 45 percent nonwhite students (i.e., more than 15
    percent below the overall 60 percent nonwhite student popula-
    tion) and more than 55 percent white students, the school is
    considered racially imbalanced.
       Originally, schools that deviated by more than 10 percent
    were deemed racially imbalanced. For the 2001-02 school
    year, however, the triggering number was increased to 15 per-
    cent, softening the effect of the tiebreaker.6 For that year, the
      6
       Although the record reflects the general effects of the tiebreaker in
    2001-02, it does not include the specific number of students affected by
                PARENTS INVOLVED v. SEATTLE SCHOOL DIST.              14667
    race-based tiebreaker was used in assigning entering ninth
    grade students only to three oversubscribed schools — Bal-
    lard, Franklin and Nathan Hale. Accordingly, in seven of the
    10 public high schools in 2001-02, race was not relevant in
    making admissions decisions.
       The race-based tiebreaker is applied to both white and non-
    white students. For example, in the 2000-01 school year —
    when the trigger point was still plus or minus 10 percent —
    89 more white students were assigned to Franklin than would
    have been assigned absent the tiebreaker, 107 more nonwhite
    students were assigned to Ballard than would have been
    assigned absent the tiebreaker, 82 more nonwhite students
    were assigned to Roosevelt than would have been assigned
    absent the tiebreaker and 27 more nonwhite students were
    assigned to Nathan Hale than would have been assigned
    absent the tiebreaker.7 These assignments accounted for about
    10 percent of admissions to Seattle’s high schools as a whole.
    That is, of the approximately 3,000 incoming students enter-
    ing Seattle high schools in the 2000-01 school year, approxi-
    mately 300 were assigned to an oversubscribed high school
    based on the race-based tiebreaker.
    
       In addition to changing the trigger point for the 2001-02
    school year to plus or minus 15 percent, the District also
    developed a “thermostat,” whereby the tiebreaker is applied to
    the entering ninth grade student population only until it comes
    within the 15 percent plus or minus variance. Once that point
    
    the tiebreaker in the three oversubscribed schools where the tiebreaker
    applied. The record, however, does include these numbers for the 2000-01
    school year. Although the tiebreaker operated differently in 2000-01, and
    applied to four schools rather than three, the 2000-01 numbers illustrate
    the general operation of the tiebreaker.
       7
         The Board’s decision to change the trigger point for use of the tie-
    breaker from plus or minus 10 percent to plus or minus 15 percent, how-
    ever, had the effect of rendering Roosevelt High School neutral for
    desegregation purposes. Thus, the tiebreaker did not factor into assign-
    ments to Roosevelt High School in the 2001-02 school year.
    14668      PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    is reached, the District “turns-off” the race-based tiebreaker,
    and there is no further consideration of a student’s race in the
    assignment process. The tiebreaker does not apply, and race
    is not considered, for students entering a high school after the
    ninth grade (e.g., by transfer).
    
       As demonstrated in the chart below, the District estimates
    that without the race-based tiebreaker, the nonwhite popula-
    tions of the 2000-01 ninth grade class at Franklin would have
    been 79.2 percent, at Hale 30.5 percent, at Ballard 33 percent
    and at Roosevelt 41.1 percent. Using the race-based tie-
    breaker, the actual nonwhite populations of the ninth grade
    classes at the same schools respectively were 59.5 percent,
    40.6 percent, 54.2 percent and 55.3 percent.
    
              2000-01 DIFFERENCE IN PERCENTAGES OF
               NONWHITE STUDENTS IN NINTH GRADE
                 WITH AND WITHOUT TIEBREAKER
    
    
    
        SCHOOL         WITHOUT           WITH           PERCENT
                      TIEBREAKER      TIEBREAKER       Difference
    
        FRANKLIN          79.2            59.5            −19.7
    
     NATHAN HALE          30.5            40.6            +10.1
    
        BALLARD           33.0            54.2            +21.2
    
       ROOSEVELT          41.1            55.3            +14.2
    
    
       In the third tiebreaker, students are admitted according to
    distance from the student’s home to the high school. Distance
    between home and school is calculated within 1/100 of a mile,
    with the closest students being admitted first. In any given
    oversubscribed school, the distance-based tiebreaker accounts
    for between 70 to 75 percent of admissions to the ninth grade.
    
      In the fourth tiebreaker, a lottery is used to allocate the
    remaining seats. Because the distance tiebreaker serves to
                 PARENTS INVOLVED v. SEATTLE SCHOOL DIST.                14669
    assign nearly all the students in the District, a lottery is virtu-
    ally never used.
    
      C. Procedural History
    
       Parents Involved in Community Schools (“Parents”), a
    group of parents whose children were not, or might not be,
    assigned to the high schools of their choice under the Plan,
    claimed that the District’s use of the race-based tiebreaker for
    high school admissions is illegal under the Washington Civil
    Rights Act (“Initiative 200”),8 the Equal Protection Clause of
    the Fourteenth Amendment9 and Title VI of the Civil Rights
    Act of 1964.10
    
       Both Parents and the District moved for summary judgment
    on all claims. In a published opinion dated April 6, 2001, the
    district court upheld the use of the racial tiebreaker under both
    state and federal law, granting the District’s motion. Parents
    Involved in Cmty. Schs. v. Seattle Sch. Dist. No. 1, 137 F.
    Supp. 2d 1224, 1240 (W.D. Wash. 2001) (“Parents I”). Par-
    ents timely appealed, and on April 16, 2002, a three-judge
    panel of this court issued an opinion reversing the district
    court’s decision, holding that the Plan violated Washington
    state law and discussing federal law only as an aid to constru-
       8
         Wash. Rev. Code § 49.60.400 (“The state shall not discriminate
    against, or grant preferential treatment to, any individual or group on the
    basis of race, sex, color, ethnicity, or national origin in the operation of
    public employment, public education, or public contracting.”).
       9
         U.S. Const. amend. XIV, § 1 (“No state shall . . . deny to any person
    within its jurisdiction the equal protection of the laws.”).
       10
          42 U.S.C. § 2000d (“No person in the United States shall, on the
    ground of race, color, or national origin, be excluded from participation in,
    be denied the benefits of, or be subjected to discrimination under any pro-
    gram or activity receiving Federal financial assistance.”). Because “dis-
    crimination that violates the Equal Protection Clause of the Fourteenth
    Amendment committed by an institution that accepts federal funds also
    constitutes a violation of Title VI,” we address the twin challenges to the
    racial tiebreaker simultaneously. Gratz, 539 U.S. at 276 n.23.
    14670       PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    ing state law. Parents Involved in Cmty. Schs. v. Seattle Sch.
    Dist. No. 1, 
    285 F.3d 1236
     (9th Cir. 2002) (“Parents II”). The
    panel subsequently withdrew its opinion and certified the state
    law question to the Washington Supreme Court. Parents
    Involved in Cmty. Schs. v. Seattle Sch. Dist. No. 1, 
    294 F.3d 1084
    , 1085 (9th Cir. 2002) (“Parents III”). The Washington
    Supreme Court disagreed with the panel’s decision, holding
    that the open choice plan did not violate Washington law.
    Parents Involved in Cmty. Schs. v. Seattle Sch. Dist. No. 1, 
    72 P.3d 151
    , 166 (Wash. 2003) (“Parents IV”) (holding that
    Washington law “does not prohibit the Seattle School Dis-
    trict’s open choice plan tie breaker based upon race so long
    as it remains neutral on race and ethnicity and does not pro-
    mote a less qualified minority applicant over a more qualified
    applicant”). Thereafter, a majority of the three-judge panel of
    this court held that although the District demonstrated a com-
    pelling interest in achieving the benefits of racial diversity,
    the Plan violated the Equal Protection Clause because it was
    not narrowly tailored. Parents Involved in Comty. Schs. v.
    Seattle Sch. Dist., No. 1, 
    377 F.3d 949
     (9th Cir. 2004)
    (“Parents V”). We granted en banc rehearing and now affirm
    the district court.11
    
                               II. Discussion
    
      A. Strict Scrutiny
    
       [1] We review racial classifications under the strict scrutiny
    standard, which requires that the policy in question be nar-
    rowly tailored to achieve a compelling state interest. See
    Johnson v. California, 
    125 S. Ct. 1141
    , 1146 (2005); Grutter,
    539 U.S. at 326; Adarand Constructors, Inc. v. Pena, 
    515 U.S. 200
    , 226-27 (1995).12 The strict scrutiny standard is not
      11
          We review the district court’s resolution of cross-motions for sum-
    mary judgment de novo. United States v. City of Tacoma, 
    332 F.3d 574
    ,
    578 (9th Cir. 2003).
       12
          Judge Kozinski’s concurrence makes a powerful case for adopting a
    less stringent standard of review here because the Plan does not attempt
                 PARENTS INVOLVED v. SEATTLE SCHOOL DIST.                  14671
    “strict in theory, but fatal in fact.” Adarand, 515 U.S. at 237
    (internal quotation marks omitted). “Although all governmen-
    tal uses of race are subject to strict scrutiny, not all are invali-
    dated by it.” Grutter, 539 U.S. at 326-27. We employ strict
    
    to “benefit[ ] or burden[ ] any particular group;” therefore it “carries none
    of the baggage the Supreme Court has found objectionable” in earlier
    equal protection cases. Kozinski, J., concurring, infra at 14712 and 14716.
    Recognizing the importance of context in the Supreme Court’s equal pro-
    tection jurisprudence, Judge Kozinski proposes “robust and realistic”
    rational basis rather than strict scrutiny review. Id. at 4. Cf. Coalition for
    Economic Equity v. Wilson, 
    122 F.3d 692
    , 708 n.16 (9th Cir. 1997) (“We
    have recognized . . . that ‘stacked deck’ programs trench on Fourteenth
    Amendment values in ways that ‘reshuffle’ programs do not. Unlike racial
    preference programs, school desegregation programs are not inherently
    invidious, do not work wholly to the benefit of certain members of one
    group and correspondingly to the harm of certain members of another
    group, and do not deprive citizens of rights.”) (internal quotation marks,
    alterations and citations omitted).
       Nonetheless, the Supreme Court in Johnson v. California, 
    125 S. Ct. 1141
     (2005), rejected the argument that a California Department of Cor-
    rections (“CDC”) policy in which all inmates were segregated by race
    should be subjected to relaxed scrutiny because the policy “neither bene-
    fits nor burdens one group or individual more than any other group or
    individual.” Id. at 1147 (internal quotation marks omitted); see also id. at
    1146 (noting that all racial classifications “raise special fears that they are
    motivated by an invidious purpose” and that “[a]bsent 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” (internal
    quotation marks and citation omitted)). As Judge Kozinski aptly notes,
    Johnson is not entirely analogous to the instant case because the CDC seg-
    regated inmates on the basis of race, whereas the District’s use of race is
    aimed at achieving the opposite result — attaining and maintaining inte-
    grated schools. Kozinski, J., concurring, infra. at 14712. Nevertheless, like
    the First and Sixth Circuits — the only other circuits to rule, post-Grutter
    and Gratz, on the constitutionality of a voluntary plan designed to achieve
    the benefits of racial diversity in the public secondary school setting — we
    conclude that the Plan must be reviewed under strict scrutiny. See Comfort
    v. Lynn School Committee, 
    418 F.3d 1
    , 6 (1st Cir. 2005) (en banc);
    McFarland v. Jefferson County Public Schools, 
    416 F.3d 513
    , 514 (6th
    Cir. 2005) (per curiam).
    14672      PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    scrutiny to “smoke out” impermissible uses of race by ensur-
    ing that the government is pursuing a goal important enough
    to warrant use of a highly suspect tool. Id. at 327 (internal
    quotation marks omitted). This heightened standard of review
    provides 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.
    Smith v. Univ. of Washington, 
    392 F.3d 367
    , 372 (9th Cir.
    2004). In evaluating the District’s Plan under strict scrutiny,
    we also bear in mind the Court’s directive that “[c]ontext mat-
    ters when reviewing race-based governmental action under
    the Equal Protection Clause.” Grutter, 539 U.S. at 326.
    
      B. Compelling State Interest
    
       Under strict scrutiny, a government action will not survive
    unless motivated by a “compelling state interest.” See id. at
    325, 327. Because strict scrutiny requires us to evaluate the
    “fit” between the government’s means and its ends, Wygant
    v. Jackson Bd. of Educ., 
    476 U.S. 267
    , 280 n.6 (1986), it is
    critical to identify precisely the governmental interests — the
    ends — to which the government’s use of race must fit. See
    United States v. Paradise, 
    480 U.S. 149
    , 171 (1987) (stating
    that, in order to determine whether an order was narrowly tai-
    lored, “we must examine the purposes the order was intended
    to serve”).
    
       [2] Although the Supreme Court has never decided a case
    involving the consideration of race in a voluntarily imposed
    school assignment plan intended to promote racially and eth-
    nically diverse secondary schools, its decisions regarding
    selective admissions to institutions of higher learning demon-
    strate that one compelling reason for considering race is to
    achieve the educational benefits of diversity. The compelling
    interest that the Court recognized in Grutter was the promo-
    tion of the specific educational and societal benefits that flow
    from diversity. See Grutter, 539 U.S. at 330 (noting that the
    law school’s concept of critical mass must be “defined by ref-
                 PARENTS INVOLVED v. SEATTLE SCHOOL DIST.                   14673
    erence to the educational benefits that diversity is designed to
    produce”). In evaluating the relevance of diversity to higher
    education, the Court focused principally on two benefits that
    a diverse student body provides: (1) the learning advantages
    of having diverse viewpoints represented in the “robust
    exchange of ideas” that is critical to the mission of higher
    education, id. at 329-30; and (2) the greater societal legiti-
    macy that institutions of higher learning enjoy by cultivating
    a group of national leaders who are representative of our
    country’s diversity, id. at 332-33. The Court also mentioned
    the role of diversity in challenging stereotypes. Id. at 330,
    333. The Court largely deferred to the law school’s educa-
    tional judgment not only in determining that diversity would
    produce these benefits, but also in determining that these ben-
    efits were critical to the school’s educational mission. Id. at
    328-33.13
    
       Against this background, we consider the specific interests
    that the District’s Plan seeks to advance. These interests are
    articulated in the “Board Statement Reaffirming Diversity
    Rationale” as:
    
         Diversity in the classroom increases the likelihood
         that students will discuss racial or ethnic issues and
         be more likely to socialize with people of different
       13
          The Court also heeded the judgment of amici curiae — including edu-
    cators, business leaders and the military — that the educational benefits
    that flow from diversity constitute a compelling interest. Grutter, 539 U.S.
    at 330 (“The Law School’s claim of a compelling interest is further bol-
    stered by its amici, who point to the educational benefits that flow from
    student body diversity.”); see also id. (“These benefits are not theoretical
    but real, as major American businesses have made clear that the skills
    needed in today’s increasingly global marketplace can only be developed
    through exposure to widely diverse people, cultures, ideas, and view-
    points.”); id. at 331 (“[H]igh-ranking retired officers and civilian leaders
    of the United States military assert that, ‘[b]ased on [their] decades of
    experience,’ a ‘highly qualified, racially diverse officer corps . . . is essen-
    tial to the military’s ability to fulfill its principle mission to provide
    national security.’ ”).
    14674      PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
        races. Diversity is thus a valuable resource for teach-
        ing students to become citizens in a multi-racial/
        multi-ethnic world.
    
        Providing students the opportunity to attend schools
        with diverse student enrollment also has inherent
        educational value from the standpoint of education’s
        role in a democratic society . . . . Diversity brings
        different viewpoints and experiences to classroom
        discussions and thereby enhances the educational
        process. It also fosters racial and cultural understand-
        ing, which is particularly important in a racially and
        culturally diverse society such as ours.
    
        The District’s commitment to the diversity of its
        schools and to the ability to voluntarily avoid
        racially concentrating enrollment patterns also helps
        ensure that all students have access to those schools,
        faculties, course offerings, and resources that will
        enable them to reach their full potential.
    
        Based on the foregoing rationale, the Seattle School
        District’s commitment is that no student should be
        required to attend a racially concentrated school. The
        District is also committed to providing students with
        the opportunity to voluntarily choose to attend a
        school to promote integration. The District provides
        these opportunities for students to attend a racially
        and ethnically diverse school, and to assist in the
        voluntary integration of a school, because it believes
        that providing a diverse learning environment is edu-
        cationally beneficial for all students.
    
       The District’s interests fit into two broad categories: (1) the
    District seeks the affirmative educational and social benefits
    that flow from racial diversity; and (2) the District seeks to
    avoid the harms resulting from racially concentrated or iso-
    lated schools.
                 PARENTS INVOLVED v. SEATTLE SCHOOL DIST.                14675
      1. Educational and Social Benefits that Flow from
      Diversity
    
       [3] The District has established that racial diversity pro-
    duces a number of compelling educational and social benefits
    in secondary education. First, the District presented expert
    testimony that in racially diverse schools, “both white and
    minority students experienced improved critical thinking
    skills — the ability to both understand and challenge views
    which are different from their own.”
    
       [4] Second, the District demonstrated the socialization and
    citizenship advantages of racially diverse schools. School
    officials, relying on their experience as teachers and adminis-
    trators, and the District’s expert all explained these benefits
    on the record. According to the District’s expert, the social
    science research “clearly and consistently shows that, for both
    white and minority students, a diverse educational experience
    results in improvement in race-relations, the reduction of prej-
    udicial attitudes, and the achievement of a more . . . inclusive
    experience for all citizens . . . . The research further shows
    that only a desegregated and diverse school can offer such
    opportunities and benefits. The research further supports the
    proposition that these benefits are long lasting.” (Emphasis
    added.) Even Parents’ expert conceded that “[t]here is general
    agreement by both experts and the general public that integra-
    tion is a desirable policy goal mainly for the social benefit of
    increased information and understanding about the cultural
    and social differences among various racial and ethnic
    groups.”14 That is, diversity encourages students not only to
    think critically but also democratically.
       14
          Academic research has shown that intergroup contact reduces preju-
    dice and supports the values of citizenship. See Derek Black, Comment,
    The Case for the New Compelling Government Interest: Improving Educa-
    tional Outcomes, 
    80 N.C. L
    . Rev. 923, 951-52 (2002) (collecting academic
    research demonstrating that interpersonal interaction in desegregated
    schools reduces racial prejudice and stereotypes, improving students’ citi-
    zenship values and their ability to succeed in a racially diverse society in
    their adult lives).
    14676        PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
       Third, the District’s expert noted that “research shows that
    a[ ] desegregated educational experience opens opportunity
    networks in areas of higher education and employment . . .
    [and] strongly shows that graduates of desegregated high
    schools are more likely to live in integrated communities than
    those who do not, and are more likely to have cross-race
    friendships later in life.”15
    
       The District’s interests in the educational and social bene-
    fits of diversity are similar to those of a law school as articu-
    lated in Grutter. The contextual differences between public
    high schools and universities, however, make the District’s
    interests compelling in a similar but also significantly differ-
    ent manner. See Grutter, 539 U.S. at 330 (noting that the
    compelling state interest in diversity is judged in relation to
    the educational benefits that it seeks to produce).
    
       [5] The Supreme Court in Grutter noted the importance of
    higher education in “preparing students for work and citizen-
    ship.” 539 U.S. at 331. For a number of reasons, public sec-
    ondary schools have an equal if not more important role in
    this preparation. First, underlying the history of desegregation
    in this country is a legal regime that recognizes the principle
    that public secondary education serves a unique and vital
    socialization function in our democratic society. As the Court
    explained in Plyler v. Doe, “[w]e have recognized the public
    schools as a most vital civic institution for the preservation of
       15
          The District’s compelling interests in diversity have been endorsed by
    Congress. In the Magnet Schools Assistance Act, Congress found that “It
    is in the best interests of the United States — (A) to continue the Federal
    Government’s support of local educational agencies that are voluntarily
    seeking to foster meaningful interaction among students of different racial
    and ethnic backgrounds, beginning at the earliest stages of such students’
    education; (B) to ensure that all students have equitable access to a high
    quality education that will prepare all students to function well in a tech-
    nologically oriented and a highly competitive economy comprised of peo-
    ple from many different racial and ethnic backgrounds.” 20 U.S.C.
    § 7231(a)(4) (emphasis added).
                PARENTS INVOLVED v. SEATTLE SCHOOL DIST.                14677
    a democratic system of government, and as the primary vehi-
    cle for transmitting the values on which our society rests.”
    
    457 U.S. 202
    , 221 (1982) (internal quotation marks and cita-
    tions omitted); see Bethel Sch. Dist. No. 403 v. Fraser, 
    478 U.S. 675
    , 683 (1986) (stating that the inculcation of civic val-
    ues is “truly the work of the schools”) (internal quotation
    marks omitted)); Plyler, 457 U.S. at 221-23 (noting that pub-
    lic education perpetuates the political system and the eco-
    nomic and social advancement of citizens and that “education
    has a fundamental role in maintaining the fabric of our soci-
    ety”); Ambach v. Norwick, 
    441 U.S. 68
    , 76-77 (1979) (observ-
    ing that public schools transmit to children “the values on
    which our society rests,” including “fundamental values nec-
    essary to the maintenance of a democratic political system”);
    Brown v. Bd. of Educ., 
    347 U.S. 483
    , 493 (1954)
    (“[Education] is required in the performance of our most basic
    public responsibilities . . . . It is the very foundation of good
    citizenship. Today it is a principal instrument in awakening
    the child to cultural values, in preparing him for later profes-
    sional training, and in helping him to adjust normally to his
    environment.”). Under Washington law, such civic training is
    mandated by the state constitution: “Our constitution is
    unique in placing paramount value on education for citizen-
    ship.” Parents IV, 72 P.3d at 158.
    
       [6] Second, although one hopes that all students who gradu-
    ate from Seattle’s public schools would have the opportunity
    to attend institutions of higher learning if they so desire, a
    substantial number of Seattle’s public high school graduates
    do not attend college.16 For these students, their public high
    school educational experience will be their sole opportunity to
    reap the benefits of a diverse learning environment. We reject
    the notion that only those students who leave high school and
      16
         According to the Seattle Times’ School Guide submitted by Parents,
    for the year 2000, on average 34 percent of Seattle’s high school graduates
    attend four-year colleges after graduation and 38.2 percent attend two-year
    colleges, although percentages vary from high school to high school.
    14678      PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    enter the elite world of higher education should garner the
    benefits that flow from learning in a diverse classroom.
    Indeed, it would be a perverse reading of the Equal Protection
    Clause that would allow a university, educating a relatively
    small percentage of the population, to use race when choosing
    its student body but not allow a public school district, educat-
    ing all children attending its schools, to consider a student’s
    race in order to ensure that the high schools within the district
    attain and maintain diverse student bodies.
    
       [7] Third, the public school context involves students who,
    because they are younger and more impressionable, are more
    amenable to the benefits of diversity. See Comfort, 418 F.3d
    at 15-16 (“In fact, there is significant evidence in the record
    that the benefits of a racially diverse school are more compel-
    ling at younger ages.”); Comfort v. Lynn School Committee,
    
    283 F. Supp. 2d 328
    , 356 (D. Mass. 2003) (noting expert testi-
    mony describing racial stereotyping as a “ ‘habit of mind’ that
    is difficult to break once it forms” and explaining 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”); see also Goodwin
    Liu, Brown, Bollinger, and Beyond, 47 How. L.J. 705, 755
    (2004) (“[I]f ‘diminishing the force of [racial] stereotypes’ is
    a compelling pedagogical interest in elite higher education, it
    can only be more so in elementary and secondary schools —
    for the very premise of Grutter’s diversity rationale is that
    students enter higher education having had too few opportuni-
    ties in early grades to study and learn alongside peers from
    other racial groups.”) (citing Grutter, 539 U.S. at 333)
    (emphasis added)).
    
       The dissent insists that racial diversity in a public high
    school is not a compelling interest, arguing that Grutter
    endorsed a law school’s compelling interest in diversity only
    in some broader or more holistic sense. Bea, J., dissenting,
    infra. at 14726-27. To attain this broader interest, the dissent
    contends, the District may only consider race along with other
               PARENTS INVOLVED v. SEATTLE SCHOOL DIST.        14679
    attributes such as socioeconomic status, ability to speak multi-
    ple languages or extracurricular talents. We read Grutter,
    however, to recognize that racial diversity, not some proxy for
    it, is valuable in and of itself. 539 U.S. at 330 (discussing the
    “substantial” benefits that flow from a racially diverse student
    body and citing several sources that detail the impact of racial
    diversity in the educational environment).
    
      [8] In short, the District has demonstrated that it has a com-
    pelling interest in the educational and social benefits of racial
    diversity similar to those articulated by the Supreme Court in
    Grutter as well as the additional compelling educational and
    social benefits of such diversity unique to the public second-
    ary school context.
    
      2. Avoiding the Harms Resulting               from    Racially
      Concentrated or Isolated Schools
    
       [9] The District’s interest in achieving the affirmative bene-
    fits of a racially diverse educational environment has a flip
    side: avoiding racially concentrated or isolated schools. In
    particular, the District is concerned with making the educa-
    tional benefits of a diverse learning environment available to
    all its students and ensuring that “no student should be
    required to attend a racially concentrated school.” See “Board
    Statement Reaffirming Diversity Rationale,” quoted supra p.
    14674. Research regarding desegregation has found that
    racially concentrated or isolated schools are characterized by
    much higher levels of poverty, lower average test scores,
    lower levels of student achievement, with less-qualified teach-
    ers and fewer advanced courses — “[w]ith few exceptions,
    separate schools are still unequal schools.” See Erica Franken-
    berg et al., A Multiracial Society with Segregated Schools:
    Are We Losing the Dream? 11 (The Civil Rights Project, Har-
    vard Univ. Jan. 2003), at http://www.civilrightsproject.
    harvard.edu/research/reseg03/AreWeLosingtheDream.pdf)
    (hereinafter “Civil Rights Project”) (last visited October 11,
    14680        PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    2005) (cited in Grutter, 539 U.S. at 345 (Ginsburg, J., concur-
    ring)).
    
       [10] In Seattle, the threat of having to attend a racially con-
    centrated or isolated school is not a theoretical or imagined prob-
    lem.17 As the district court found, the District “established that
    housing patterns in Seattle continue to be racially concentrat-
    ed,” and would result in racially concentrated or isolated
    schools if school assignments were based solely on a stu-
    dent’s neighborhood or proximity to a particular high school.
    Parents I, 137 F. Supp. 2d at 1235. Accordingly, the District’s
    Plan strives to ensure that patterns of residential segregation
    are not replicated in the District’s school assignments. Cf.
    Comfort, 418 F.3d at 29 (“The problem is that in Lynn, as in
    many other cities, minorities and whites often live in different
    neighborhoods. Lynn’s aim is to preserve local schools as an
    option without having the housing pattern of de facto segrega-
    tion projected into the school system.”) (Boudin, C.J., concur-
    ring). Although Parents make much of the fact that “Seattle
    has never operated a segregated school system,” and allege
    that “this is not a school desegregation case,” each court to
    review the matter has concluded that because of Seattle’s
    housing patterns, high schools in Seattle would be highly seg-
    regated absent race conscious measures. See Parents I, 137 F.
    Supp. 2d at 1237; Parents II, 285 F.3d at 1239-40; Parents
    III, 294 F.3d at 1088; Parents IV, 72 P.3d at 153.
    
       [11] The district court found that, “[t]he circumstances that
    gave rise to the court-approved school assignment policies of
    the 1970s [e.g., Seattle’s segregated housing patterns] con-
    tinue to be as compelling today as they were in the days of the
      17
         The prospect of children across the nation being required to attend
    racially concentrated or isolated schools is a crisis that school boards, dis-
    tricts, teachers and parents confront daily. See Civil Rights Project 4 (“At
    the beginning of the twenty-first century, American public schools are
    now twelve years in the process of continuous resegregation. The desegre-
    gation of black students, which increased continuously from the 1950s to
    the late 1980s, has now receded to levels not seen in three decades.”).
                 PARENTS INVOLVED v. SEATTLE SCHOOL DIST.                14681
    district’s mandatory busing programs . . . . [I]t would defy
    logic for this court to find that the less intrusive programs of
    today violate the Equal Protection Clause while the more
    coercive programs of the 1970s did not.” Parents I, 137 F.
    Supp. 2d at 1235. Thus, it concluded that “[p]reventing the re-
    segregation of Seattle’s schools is . . . a compelling interest.”
    Id. at 1237; see id. at 1233-35. Several other courts have also
    conceived of a school district’s voluntary reduction or preven-
    tion of de facto segregation as a compelling interest. See
    Comfort, 418 F.3d at 14 (holding that the “negative conse-
    quences of racial isolation that Lynn seeks to avoid and the
    benefits of diversity that it hopes to achieve” constituted com-
    pelling interests); Brewer v. W. Irondequoit Cent. Sch. Dist.,
    
    212 F.3d 738
    , 752 (2d Cir. 2000) (holding that “a compelling
    interest can be found in a program that has as its object the
    reduction of racial isolation and what appears to be de facto
    segregation”), superseded on other grounds as stated in Zer-
    vos v. Verizon N.Y., Inc., 
    252 F.3d 163
    , 171 n.7 (2d Cir.
    2001); Parent Ass’n of Andrew Jackson High Sch. v. Ambach,
    
    738 F.2d 574
    , 579 (2d Cir. 1984) (“[W]e held that the Board’s
    goal of ensuring the continuation of relatively integrated
    schools for the maximum number of students, even at the cost
    of limiting freedom of choice for some minority students, sur-
    vived strict scrutiny as a matter of law.”) (citing Parent Ass’n
    of Andrew Jackson High Sch. v. Ambach, 
    598 F.2d 705
    , 717-
    20 (2d Cir. 1979)); McFarland v. Jefferson County Pub. Sch.,
    
    330 F. Supp. 2d 834
    , 851 (W.D. Ky. 2004) (concluding that
    voluntary maintenance of the desegregated school system was
    a compelling state interest and the district could consider race
    in assigning students to comparable schools), aff’d, 
    416 F.3d 513
     (6th Cir. 2005).18 We join these courts in recognizing that
      18
         Like the District, none of the school districts in the above-cited cases
    was subject to a court-ordered desegregation decree nor, with the excep-
    tion of Andrew Jackson, did the schools face an imminent threat of litiga-
    tion to compel desegregation. Like the District, they may have been
    vulnerable to litigation in decades past, but the districts’ voluntary deseg-
    regation measures would make it difficult today to make the required
    showing that the districts intended to create segregated schools. See, e.g.,
    Comfort, 283 F. Supp. 2d at 390 (explaining that the district’s vulnerabil-
    ity to litigation had been “headed off by the very Plan in contention here”).
    14682       PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    school districts have a compelling interest in ameliorating
    real, identifiable de facto racial segregation.
    
       [12] The dissent, however, contends first that the District is
    not “desegregating” but rather is engaged in racial balancing.
    Bea, J., dissenting, infra. at 14718. Further, for the dissent,
    segregation requires a state actor intentionally to separate the
    races; and in the absence of such offensive state conduct, the
    Supreme Court cases detailing the remedies for Fourteenth
    Amendment violations are of no relevance. Bea, J., dissent-
    ing, infra. at 14738-39, n.17. Thus, without a court finding of
    de jure segregation the elected school board members of the
    District may not take voluntary, affirmative steps towards cre-
    ating a racially diverse student body. We disagree. The fact
    that de jure segregation is particularly offensive to our Consti-
    tution does not diminish the real harms of separation of the
    races by other means. “Segregation of white and colored chil-
    dren in public schools has a detrimental effect upon the col-
    ored children. The impact is greater when it has the sanction
    of law. . . .” Brown v. Bd. of Educ., 
    347 U.S. 483
    , 494 (1954)
    (emphasis added). The benefits that flow from integration (or
    desegregation) exist whether or not a state actor was responsi-
    ble for the earlier racial isolation. Brown’s statement that “in
    the field of public education . . . [s]eparate educational facili-
    ties are inherently unequal” retains its validity today. Id. at
    495. The District is entitled to seek the benefits of racial inte-
    gration and avoid the harms of segregation even in the
    absence of a court order deeming it a violator of the U.S. Con-
    stitution.
    
       [13] Support for this conclusion comes from statements in
    the Supreme Court’s school desegregation cases, which
    repeatedly refer to the voluntary integration of schools as
    sound educational policy within the discretion of local school
    officials.19 See Swann v. Charlotte-Mecklenburg Bd. of Educ.,
      19
        The dissent correctly notes that these decisions were rendered in the
    context of de jure segregation. But their import is also significantly com-
                PARENTS INVOLVED v. SEATTLE SCHOOL DIST.                14683
    
    402 U.S. 1
    , 16 (1971) (stating that school authorities “are tra-
    ditionally charged with broad power to formulate and imple-
    ment educational policy and might well conclude . . . that in
    order to prepare students to live in a pluralistic society each
    school should have a prescribed ratio of Negro to white stu-
    dents reflecting the proportion for the district as a whole”);
    N.C. State Bd. of Educ. v. Swann, 
    402 U.S. 43
    , 45 (1971)
    (“[A]s a matter of educational policy school authorities may
    well conclude that some kind of racial balance in the schools
    is desirable quite apart from any constitutional require-
    ments.”); Bustop, Inc. v. Bd. of Educ. of Los Angeles, 
    439 U.S. 1380
    , 1383 (1978) (denying a request to stay implemen-
    tation of a voluntary desegregation plan and noting that there
    was “very little doubt” that the Constitution at least permitted
    its implementation); Keyes v. Sch. Dist. No. 1, 
    413 U.S. 189
    ,
    242 (1973) (Powell, J., concurring in part and dissenting in
    part) (“School boards would, of course, be free to develop and
    initiate further plans to promote school desegregation . . . .
    Nothing in this opinion is meant to discourage school boards
    from exceeding minimal constitutional standards in promoting
    the values of an integrated school experience.”); Washington
    v. Seattle Sch. Dist. No. 1, 458 U.S. at 480, 487 (holding
    unconstitutional the state initiative that blocked the Seattle
    School District’s use of mandatory busing to remedy de facto
    segregation).
    
       [14] In sum, we hold that the District’s interests in obtain-
    ing the educational and social benefits of racial diversity in
    secondary education and in avoiding racially concentrated or
    isolated schools resulting from Seattle’s segregated housing
    pattern are clearly compelling.
    
    pelling in the context of de facto segregation, as in Seattle. Indeed, in
    Swann, the Court further stated, “Our objective in dealing with the issues
    presented by these cases is to see that school authorities exclude no pupil
    of a racial minority from any school, directly or indirectly, on account of
    race. . . .” 402 U.S. at 23 (emphasis added).
    14684       PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
      C. Narrow Tailoring
    
       We must next determine whether the District’s use of the
    race-based tiebreaker is narrowly tailored to achieve its com-
    pelling interests. See Grutter, 539 U.S. at 333. The narrow
    tailoring inquiry is intended to “ ‘smoke out’ illegitimate uses
    of race” by ensuring that the government’s classification is
    closely fitted to the compelling goals that it seeks to achieve.
    Richmond v. J.A. Croson Co., 
    488 U.S. 469
    , 493 (1989).
    Here, our analysis is framed by the Court’s narrow tailoring
    analysis in Grutter and Gratz, which, though informed by
    considerations specific to the higher education context, sub-
    stantially guides our inquiry. See Grutter, 539 U.S. at 334
    (stating that the narrow tailoring inquiry is context-specific
    and must be “calibrated to fit the distinct issues raised” in a
    given case, taking “relevant differences into account”) (inter-
    nal quotation marks omitted).
    
      In Gratz, the Court held unconstitutional the University of
    Michigan’s undergraduate admissions program, which auto-
    matically assigned 20 points on the admissions scale to an
    applicant from an underrepresented racial or ethnic minority
    group. 539 U.S. at 255, 272. In Grutter, by contrast, the Court
    upheld the University of Michigan Law School’s admissions
    policy, which took race into account as one of several vari-
    ables in an individual’s application. 539 U.S. at 315-16, 340.
    The law school’s policy also attempted to ensure that a “criti-
    cal mass” of underrepresented minority students would be
    admitted in order to realize the benefits of a diverse student
    body.20 Id. at 316.
    
       [15] In its analysis, the Court identified five hallmarks of
      20
         The Court explained that “critical mass” was defined by the law
    school as “meaningful numbers” or “meaningful representation,” or “a
    number that encourages underrepresented minority students to participate
    in the classroom and not feel isolated.” Grutter, 539 U.S. at 318 (internal
    quotation marks omitted).
               PARENTS INVOLVED v. SEATTLE SCHOOL DIST.       14685
    a narrowly tailored affirmative action plan: (1) individualized
    consideration of applicants; (2) the absence of quotas; (3)
    serious, good-faith consideration of race-neutral alternatives
    to the affirmative action program; (4) that no member of any
    racial group was unduly harmed; and (5) that the program had
    a sunset provision or some other end point. Smith v. Univ. of
    Washington, 
    392 F.3d 367
    , 373 (9th Cir. 2004); Comfort, 418
    F.3d at 17 (characterizing Grutter as outlining a “four-part
    narrow tailoring inquiry”).
    
       Hallmarks two through five are applicable here despite sig-
    nificant differences between the competitive admissions plans
    at issue in Gratz and Grutter and the District’s high school
    assignment Plan. The first hallmark, however, is less relevant
    to our analysis because of the contextual differences between
    institutions of higher learning and public high schools.
    
      1. Individualized, Holistic Consideration of Applicants
    
      a. An applicant’s qualifications
    
       In the context of university admissions, where applicants
    compete for a limited number of spaces in a class, the Court
    in Grutter and Gratz focused its inquiry on the role race may
    play in judging an applicant’s qualifications. The Court’s
    underlying concern was that the “admissions policy is flexible
    enough to consider all pertinent elements of diversity in light
    of the particular qualifications of each applicant, and to place
    them on the same footing for consideration, although not nec-
    essarily according them the same weight.” Grutter, 539 U.S.
    at 337 (emphasis added) (internal quotation marks omitted);
    see Adarand, 515 U.S. at 211 (“The injury in cases of this
    kind is that a discriminatory classification prevent[s] the
    plaintiff from competing on an equal footing.”) (emphasis
    added) (internal quotation marks omitted). The focus on fair
    competition is due, in part, to the stigma that may attach if
    some individuals are viewed as unable to achieve success
    without special protection. See Regents of Univ. of Cal. v.
    14686        PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    Bakke, 
    438 U.S. 265
    , 298 (1978) (Powell, J., concurring)
    (“preferential programs may only reinforce common stereo-
    types holding that certain groups are unable to achieve suc-
    cess without special protection based on a factor having no
    relationship to individual worth”); Croson, 488 U.S. at 493
    (“Classifications based on race carry a danger of stigmatic
    harm. Unless they are strictly reserved for remedial settings,
    they may in fact promote notions of racial inferiority and lead
    to a politics of racial hostility.”).
    
       In Grutter and Gratz, in order to prevent race from being
    used as a mechanical proxy for an applicant’s qualifications,
    the Court required individualized, holistic consideration of
    each applicant across a broad range of factors (of which race
    may be but one). Grutter, 539 U.S. at 336-37; see Gratz, 539
    U.S. at 272 (holding that the undergraduate admissions policy
    was not narrowly tailored because the “automatic distribution
    of 20 points has the effect of making ‘the factor of race . . .
    decisive’ for virtually every minimally qualified underrepre-
    sented minority applicant”) (emphasis added). This focus on
    an applicant’s qualifications — whether these qualifications
    are such things as an applicant’s test scores, grades, artistic or
    athletic ability, musical talent or life experience — is not
    applicable when there is no competition or consideration of
    qualifications at issue.
    
       [16] All of Seattle’s high school students must and will be
    placed in a Seattle public school.21 Students’ relative qualifi-
    cations are irrelevant because regardless of their academic
    achievement, sports or artistic ability, musical talent or life
      21
         Parents do not claim that their children have a right to attend a particu-
    lar school, nor could they. See Bustop Inc., 439 U.S. at 1383 (rejecting any
    legally protected right to have children attend their nearest school). In any
    case, under the current Plan, all students can attend a school close to their
    home. Because there are multiple schools in the north and south of Seattle,
    students for whom proximity is a priority may elect as their first choice
    one of the schools in their residential area that is not oversubscribed and
    be guaranteed an assignment to that school.
                   PARENTS INVOLVED v. SEATTLE SCHOOL DIST.             14687
    experience, any student who wants to attend Seattle’s public
    high schools is entitled to an assignment; no assignment to
    any of the District’s high schools is tethered to a student’s
    qualifications. Thus, no stigma results from any particular
    school assignment.22 Accordingly, the dangers that are present
    in the university context — of substituting racial preference
    for qualification-based competition — are absent here. See
    Comfort, 418 F.3d at 18 (“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 ‘cer-
    tain groups are unable to achieve success without special pro-
    tection.’ ”) (quoting Bakke, 438 U.S. at 298).
    
      b. Differences in compelling interests
    
       The Court’s requirement of individualized, holistic review
    in Grutter is also more relevant to the compelling interest
    advanced by the law school (“the robust exchange of ideas”
    fostered by viewpoint diversity) than it is to the District’s
    (racial diversity and avoiding racially concentrated or isolated
    schools). See Grutter, 539 U.S. at 337. The Court noted that
    the law school did not “limit in any way . . . the broad range
    of qualities and experiences that may be considered valuable
    contributions to student body diversity.” Id. at 338. To this
    end, the law school’s policy made clear that “[t]here are many
    possible bases for diversity admissions, and provide[d] exam-
    ples of admittees who have lived or traveled widely abroad,
    are fluent in several languages, have overcome personal
      22
        In Bakke, Justice Powell noted:
           Respondent’s position is wholly dissimilar to that of a pupil
           bused from his neighborhood school to a comparable school in
           another neighborhood in compliance with a desegregation decree.
           Petitioner did not arrange for respondent to attend a different
           medical school in order to desegregate Davis Medical School;
           instead, it denied him admission and may have deprived him alto-
           gether of a medical education.
    438 U.S. at 301 n.39.
    14688        PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    adversity and family hardship, have exceptional records of
    extensive community service, and had successful careers in
    other fields.” Id. (internal quotation marks and citations omit-
    ted). These multiple bases for diversity ensure the “classroom
    discussion is livelier, more spirited, and simply more enlight-
    ening and interesting when the students have the greatest pos-
    sible variety of backgrounds.” Id. at 330 (internal citations
    omitted).
    
       Although the District’s Plan, like the plan in Grutter, is
    designed to achieve the educational and social benefits of
    diversity, including bringing “different viewpoints and experi-
    ences to classroom discussions,” see “Statement Reaffirming
    Diversity Rationale,” viewpoint diversity in the law school
    and high school contexts serves different albeit overlapping
    ends. In the law school setting, viewpoint diversity fosters the
    “robust exchange of ideas.” Grutter, 539 U.S. at 324; see
    Comfort, 418 F.3d at 16 (“[L]ively classroom discussion is a
    more central form of learning in law schools (which prefer the
    Socratic method) than in a K-12 setting.”). In the high school
    context, viewpoint diversity fosters racial and civic understand-
    ing.23 For example, Eric Benson, the principal of Nathan Hale
    High School, one of the District’s most popular schools, testi-
    fied that as a result of racial diversity in the classroom, “stu-
    dents of different races and backgrounds tend to have
    significant interactions both in class, and outside of class.
    When I came to Nathan Hale, there were racial tensions in the
       23
          The dissent believes that “the educational benefits from diversity, if
    any, are much greater at the higher educational level because such benefits
    are greatly magnified by the learning that takes place outside the class-
    room . . . .” Bea, J., dissenting, infra. at 14737. This belittles the substan-
    tial role of high school classroom discussions in contributing to the
    educational development of our young citizens. “The [high school] class-
    room is peculiarly the marketplace of ideas. The Nation’s future depends
    upon leaders trained through wide exposure to that robust exchange of
    ideas which discovers truth out of a multitude of tongues.” Tinker v. Des
    Moines Independent Community Sch. Dist., 
    393 U.S. 503
    , 512 (1969)
    (internal quotation marks omitted).
               PARENTS INVOLVED v. SEATTLE SCHOOL DIST.        14689
    school, reflected in fighting and disciplinary problems. These
    kind of problems have, to a large extent, disappeared.”
    
       In addition, the law school takes other diversity factors,
    besides race and ethnicity, into consideration in order to
    achieve its other compelling interest — cultivating a group of
    national leaders. For example, extensive travel, fluency in for-
    eign languages, extensive community service and successful
    careers in other fields demonstrate that a candidate is some-
    how exceptional or out of the ordinary. cf. Gratz, 539 U.S. at
    273 (disapproving of the undergraduate admissions plan, in
    part, because of its failure to consider whether an applicant
    was extraordinary and noting that “[e]ven if [a] student[’s]
    ‘extraordinary artistic talent’ rivaled that of Monet or Picasso,
    the applicant would receive, at most, five points” as opposed
    to the automatic 20 points given to an applicant from an
    underrepresented minority). In contrast, the District is
    required to educate all high school age children, both the
    average and the extraordinary, regardless of individual leader-
    ship potential.
    
       [17] The District also has a second compelling interest that
    is absent from the university context — ensuring that its
    school assignments do not replicate Seattle’s segregated hous-
    ing patterns. The holistic review necessary to achieve view-
    point diversity in the university context, across a broad range
    of factors (of which race may be but one), is not germane to
    the District’s compelling interest in preventing racial concen-
    tration or racial isolation. Because race itself is the relevant
    consideration when attempting to ameliorate de facto segrega-
    tion, the District’s tiebreaker must necessarily focus on the
    race of its students. See Comfort, 418 F.3d at 18 (holding that
    when racial diversity is the compelling interest — “[t]he only
    relevant criterion, then, is a student’s race; individualized con-
    sideration beyond that is irrelevant to the compelling inter-
    est”); Brewer v. W. Irondequoit Cent. Sch. Dist., 212 F.3d at
    752 (“If reducing racial isolation is — standing alone — a
    constitutionally permissible goal, . . . then there is no more
    14690        PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    effective means of achieving that goal than to base decisions
    on race.”). We therefore conclude that if a noncompetitive,
    voluntary student assignment plan is otherwise narrowly tai-
    lored, a district need not consider each student in a individual-
    ized, holistic manner.24
    
       The dissent insists that absent such individualized consider-
    ation, the District’s plan cannot serve a compelling interest
    and is not narrowly tailored to protect individuals from group
    classifications by race. Bea, J., dissenting, infra. at 14740.
    This is a flawed reading of the Fourteenth Amendment.25 The
    District’s compelling interest is to avoid the harms of racial
    isolation for all students in the Seattle school district. As we
    have explained, to accomplish that objective the District may
    look to the racial consequences of honoring the preferred
    choices of individual students (and their parents). It is true
    that for some students their first choice of school, based on
    geographical proximity, will be denied because other stu-
    dents’ choices are granted in order to advance the overall
    interest in maintaining racially diverse school enrollments.
    The Fourteenth Amendment in this context does not preclude
    the District from honoring racial diversity at the expense of
      24
          The dissent calculates that individualized consideration would be
    administratively feasible because only 300 students would need to be con-
    sidered holistically. Though it is true that 300 students were subject to the
    race-based tiebreaker, it does not follow that only those 300 would require
    individualized consideration. Under the dissent’s view of the way the Dis-
    trict should operate, all 3,000 students would have to be subject to holistic
    consideration to determine their proper school assignment. Whether or not
    this is administratively feasible is not clear in the record, but we believe
    it is ultimately irrelevant because individualized consideration is not
    required in the context presented here.
       25
          Reliance on group characteristics is not necessarily constitutionally
    infirm under Fourteenth Amendment jurisprudence. See, e.g., Kimel v.
    Florida Bd. of Regents, 
    528 U.S. 62
    , 84 (2000) (“Under the Fourteenth
    Amendment, a State may rely on age as a proxy for other qualities, abili-
    ties, or characteristics that are relevant to the State’s legitimate interests.
    The Constitution does not preclude reliance on such generalizations. That
    age proves to be an inaccurate proxy in any individual case is irrelevant.”)
                 PARENTS INVOLVED v. SEATTLE SCHOOL DIST.                14691
    geographical proximity. We must not forget that “race unfor-
    tunately still matters,” Grutter, 539 U.S. at 333, and it is race
    that is the relevant consideration here.
    
       In sum, the contextual differences between public high
    schools and selective institutions of higher learning make the
    first of the Grutter hallmarks ill-suited for our narrow tailor-
    ing inquiry.26 The remaining hallmarks, however, are relevant
    and control our analysis.
    
      2. Absence of Quotas
    
       In Grutter, the Court approved the law school’s plan, in
    part, because it did not institute a quota, whereby a fixed
    number of slots are reserved exclusively for minority groups,
    thereby insulating members of those groups from competition
    with other candidates.27 539 U.S. at 335. Although the law
      26
          The dissent’s alternative proposals to achieve the District’s interests
    in diversity illustrate the difficulty of individualized consideration in the
    high school context. For example, the dissent offers socioeconomic status
    as a more narrowly tailored and acceptable form of diversifying the Dis-
    trict’s schools. However, socioeconomic status does nothing more than
    substitute a number from a family’s tax return for race. There is no holis-
    tic, individualized consideration under such an approach.
       27
          Much like the rationale underlying the Court’s requirement of individ-
    ualized, holistic review, the rationale underlying the Court’s prohibition of
    quotas does not apply to the race-based tiebreaker. In paradigmatic affir-
    mative action settings — employment and admissions to institutions of
    higher learning — the Court disapproves of quotas because they are
    viewed as insulating minority candidates from competition with nonmi-
    nority candidates for scarce government resources usually awarded on the
    basis of an applicant’s qualifications — jobs, promotions or places in a
    law school class. See Bakke, 438 U.S. at 317 (opinion of Powell, J.). This
    is objectionable because no “matter how strong their qualifications,” non-
    minority candidates are never afforded the chance to compete with appli-
    cants from the preferred groups for the set-aside. Id. at 319. Because
    noncompetitive assignment to Seattle’s public high schools is not based on
    a student’s relative qualifications, the dangers that are presented by a
    quota — of substituting racial preference for qualification-based competi-
    tion — are absent here.
    14692        PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    school’s plan did not seek to admit a set number or percentage
    of minority students, during the height of the admission’s sea-
    son, the law school would consult “daily reports” that kept
    track of the racial composition of the incoming class. Id. at
    318. The Court held that this attention to numbers did not
    transform the law school plan into a quota, but instead dem-
    onstrated that the law school sought to enroll a critical mass
    of minority students in order “to realize the educational bene-
    fits of a diverse student body.” Id. Similarly, we conclude that
    the District’s 15 percent plus or minus variance is not a quota
    because it does not reserve a fixed number of slots for stu-
    dents based on their race, but instead it seeks to enroll a criti-
    cal mass of white and nonwhite students in its oversubscribed
    schools in order to realize its compelling interests.28
    
      a. No fixed number of slots
    
       The District’s race-based tiebreaker does not set aside a
    fixed number of slots for nonwhite or white students in any
    of the District’s schools. The tiebreaker is used only so long
    as there are members of the underrepresented race in the
    applicant pool for a particular oversubscribed school. If the
    number of students of that race who have applied to that
    school is exhausted, no further action is taken, even if the 15
    percent variance has not been satisfied. That is, if the appli-
    cant pool has been exhausted, no students are required or
    recruited to attend a particular high school in order to bring
    it within the 15 percent plus or minus range for that year.
      28
         Although the dissent contends that the “tiebreaker aims for a rigid,
    predetermined ratio of white and nonwhite students,” we believe it is more
    appropriately viewed as a “permissible goal.” Such a goal “requires only
    a good faith effort . . . to come within a range demarcated by the goal
    itself.” Grutter, 539 U.S at 334 (internal quotation marks and citation
    omitted). The tiebreaker’s broad, 30% range and the District’s willingness
    to turn off the use of the tiebreaker after the ninth grade are consistent with
    a goal as opposed to a rigid ratio.
                 PARENTS INVOLVED v. SEATTLE SCHOOL DIST.                14693
       Moreover, the number of white and nonwhite students in
    the high schools is flexible and varies from school to school
    and from year to year.29 This variance in the number of non-
    white and white students throughout the District’s high
    schools is because, under the Plan, assignments are based on
    students’ and parents’ preferences.30 The tiebreakers come
    into play in the assignment process only when a school is
    oversubscribed. As Morgan Lewis, the Manager of Enroll-
    ment Planning, Technical Support and Demographics, testi-
    fied, “If all the parents . . . don’t pick [a] school in a massive
    number, then everyone gets in. And so it’s . . . a case where
    the choice patterns, the oversubscription . . . [is] the reason
    the [tiebreaker] kicks in . . . . Everything happens when more
    people want the seats. And why they want the seats some-
    times we don’t know.”
    
      b. Critical mass
    
       Within this flexible system, where parental and student
    choices drive the assignments to particular schools, the Dis-
    trict seeks to enroll and maintain a relatively stable critical
    mass of white and nonwhite students in each of its oversub-
    scribed high schools in order to achieve its compelling inter-
    est in racial diversity and to prevent the assignments from
      29
          Notably, the District’s percentage of white and nonwhite enrollment
    is significantly more varied than the percentage of underrepresented
    minorities admitted to the University of Michigan’s Law School, which
    remained relatively consistent. From 1995 to 1998, the percentage of
    minority students enrolled in the law school was 13.5 percent, 13.8 per-
    cent, 13.6 percent and 13.8 percent. Grutter, 539 U.S. at 389-90 (Ken-
    nedy, J., dissenting). In contrast, the District’s percentage of white and
    nonwhite enrollment encompasses a wide range. For example, for the
    2000-01 school year, the percentage of nonwhite students in the ninth
    grade classes of the four oversubscribed public high schools after the
    racial tiebreaker was applied, varied from 54.2 percent at Ballard, to 59.5
    percent at Franklin, to 40.6 percent at Nathan Hale to 55.3 percent at Roo-
    sevelt.
       30
          Slightly more than 80 percent of all entering ninth grade students were
    assigned to their first choice school.
    14694       PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    replicating Seattle’s segregated housing patterns. Faced with
    the question of what constituted a critical mass of students in
    this particular context, the District determined that a critical
    mass was best achieved by adopting the 15 percent plus or
    minus variance tied to demographics of students in the Seattle
    public schools. Thus, when an oversubscribed high school has
    more than 75 percent nonwhite students (i.e., more than 15
    percent above the overall 60 percent nonwhite student popula-
    tion) and less than 25 percent white students, or when it has
    less than 45 percent nonwhite students (i.e., more than 15 per-
    cent below the overall 60 percent nonwhite student popula-
    tion) and more than 55 percent white students, the school is
    considered racially concentrated or isolated, meaning that it
    lacks a critical mass of students needed “to realize the educa-
    tional benefits of a diverse student body.”
    
       Parents attack the District’s use of the 15 percent plus or
    minus variance tied to the District’s school population demo-
    graphics because they believe that the District cannot use race
    at all in its assignment process. We have rejected this argu-
    ment, however, applying Grutter and Gratz. See supra Part
    II.B. Alternatively, Parents contend that the District’s goal of
    enrolling between 75 and 45 percent nonwhite students and
    between 25 and 55 percent white students in its oversub-
    scribed schools establishes a quota, not a critical mass. They
    note that the critical mass sought by the law school in Grutter
    was smaller, consisting of between 12 and 20 percent of
    underrepresented minority students in each law school class.
    
       Parents’ argument, however, ignores Grutter’s admonition
    that the narrow tailoring inquiry be context-specific. First, like
    the District’s enrollment goals, which are tied to the demo-
    graphics of the Seattle schools’ total student population, the
    law school’s goal of enrolling between 12 to 20 percent of
    underrepresented minorities in a given year was tied to the
    demographics of its applicant pool.31 Second, in tying the use
      31
        For example, in 1995, 662 (approximately 16 percent) of the 4147 law
    school applicants were underrepresented minorities; in 1996, 559 (approx-
                PARENTS INVOLVED v. SEATTLE SCHOOL DIST.             14695
    of the tiebreaker to the District’s demographics with a 15 per-
    cent plus or minus trigger point, the District adopted a com-
    mon benchmark in the context of voluntary and court-ordered
    school desegregation plans. As the District’s expert testified,
    
         Most of the cases I’ve participated in . . . generally
         worked with numbers that reflect the racial composi-
         tion of the school district but, at the same time,
         tr[ied] to allow the district sufficient flexibility so
         that it would not have to regularly and repeatedly
         move students on a short-term basis simply to main-
         tain some specific number. That’s why we see
         ranges of plus or minus 15 percent in most cases of
         school desegregation.
    
    Even Parents’ expert testified that school districts throughout
    the country determine whether a district is sufficiently deseg-
    regated by looking to the “population of the district” in ques-
    tion. See also Comfort, 418 F.3d at 21 (holding that a
    “transfer policy conditioned on district demographics (+/- 10-
    15%)” was not a quota because it “reflects the defendants’
    efforts to obtain the benefits of diversity in a stable learning
    environment”); Belk v. Charlotte-Mecklenburg Bd. of Educ,
    
    233 F.3d 232
    , 287-88 (4th Cir. 2000) (Traxler, J., dissenting)
    (citing to a book written by David J. Armor, Parents’ expert,
    Forced Justice: School Desegregation and the Law 160
    (1995), which observed that over 70 percent of the school dis-
    tricts with desegregation plans use a variance of plus or minus
    15 percent or greater); cf. 34 C.F.R. § 280.4(b) (defining “mi-
    nority group isolation” as a “condition in which minority
    group children constitute more than 50 percent of the enroll-
    ment of [a] school”). Given this empirically and time-tested
    
    imately 15 percent) of the 3677 law school applicants were underrepre-
    sented minorities; in 1997, 520 (approximately 15 percent) of the 3429
    law school applicants were underrepresented minorities. See Grutter, 539
    U.S. at 384 (Rehnquist, C.J., dissenting).
    14696      PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    notion of critical mass in the public high school desegregation
    context, it would make little sense to force the District to uti-
    lize the same percentages that constituted a critical mass in
    the elite law school context to determine what constitutes a
    critical mass for Seattle public high schools. See Grutter, 539
    U.S. at 336 (“[S]ome attention to numbers, without more,
    does not transform a flexible admissions system into a rigid
    quota.”) (internal quotation marks and citations omitted).
    
       [18] Accordingly, we conclude that the District’s 15 per-
    cent plus or minus trigger point tied to the demographics of
    the Seattle school population is not a quota. It is a context-
    specific, flexible measurement of racial diversity designed to
    attain and maintain a critical mass of white and nonwhite stu-
    dents in Seattle’s public high schools.
    
      3. Necessity of the Plan and Race-Neutral Alternatives
    
       Narrow tailoring also requires us to consider the necessity
    of the race-based plan or policy in question and whether there
    are equally effective, race-neutral alternatives.
    
      a. Necessity of the Plan
    
       The District argues that the compelling interests that it
    seeks are directly served by the race-based tiebreaker. The tie-
    breaker allows the District to balance students’ and parents’
    choices among high schools with its broader compelling inter-
    ests — achieving the educational and social benefits of diver-
    sity and the benefits specific to the secondary school context,
    and discouraging a return to enrollment patterns based on
    Seattle’s racially segregated housing pattern.
    
        i. Need for race-based tiebreaker
    
       When the District moved from its controlled choice plan to
    the current Plan, see supra Part I.A, it predicted that families
    would tend to choose schools close to their homes. Indeed,
                 PARENTS INVOLVED v. SEATTLE SCHOOL DIST.                14697
    this feature was seen as a positive way to increase parental
    involvement. However, unfettered choice — especially with
    tiebreakers based on neighborhood or distance from a school
    — created the risk that Seattle’s high school enrollment
    would again do no more than reflect its segregated housing
    patterns. See supra Part II.C.2.
    
       It is this de facto residential segregation across a white/
    nonwhite axis that the District has battled historically and that
    it seeks to ameliorate by making the integration tiebreaker a
    part of its open choice Plan.32 The District, mindful of both
    Seattle’s history and future, appropriately places its focus
    here. In the 2001-02 school year, the integration tiebreaker
    operated in three high schools (that is, three high schools were
    oversubscribed and deviated by more than 15 percent from
    the ratio of white to nonwhite students district-wide). The
    integration tiebreaker served to alter the imbalance in the
    schools in which it operated in a minimally intrusive manner.
    The tiebreaker, therefore, successfully achieved the District’s
    compelling interests.
    
           ii. White/Nonwhite distinction
    
       Parents argue that the District paints with too broad a brush
    by distinguishing only between white and nonwhite students,
    without taking into account the diversity within the “non-
    white” group. However, the District’s choice to increase
    diversity along the white/nonwhite axis is rooted in Seattle’s
    history and current reality of de facto segregation resulting
    from Seattle’s segregated housing patterns. The white/
    nonwhite distinction is narrowly tailored to prioritize move-
      32
        Although we characterize it as de facto residential segregation, we are
    mindful of Justice Marshall’s dissent in Board of Education v. Dowell,
    “The . . . conclusion that the racial identity of the northeast quadrant now
    subsists because of ‘personal preference[s]’ pays insufficient attention to
    the roles of the State, local officials, and the Board in creating what are
    now self-perpetuating patterns of residential segregation.” 
    498 U.S. 237
    ,
    263 (1991) (internal citation omitted).
    14698       PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    ment of students from the north of the city to the south of the
    city and vice versa. This white/nonwhite focus is also consis-
    tent with the history of public school desegregation measures
    throughout the country, as reflected in a current federal regu-
    lation defining “[m]inority group isolation” as “a condition in
    which minority group children constitute more than 50 per-
    cent of the enrollment of the school,” without distinguishing
    among the various categories included within the definition of
    “minority group.” 34 C.F.R. § 280.4(b); see Grutter, 539 U.S.
    at 316 (noting that the law school sought to enroll a critical
    mass of “minority students,” a category that included African
    Americans, Hispanics and Native Americans); Comfort, 418
    F.3d at 22 (“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 sub-
    group as well.”) (emphasis added).
    
      b. Race-neutral alternatives
    
       In Grutter, the Court explained that narrow tailoring “re-
    quire[s] serious, good faith consideration of workable race-
    neutral alternatives that will achieve the diversity the univer-
    sity seeks.” 539 U.S. at 339 (emphasis added). On the other
    hand, “[n]arrow tailoring does not require exhaustion of every
    conceivable race-neutral alternative.” Id. Furthermore, the
    Court made clear that the university was not required to adopt
    race-neutral measures that would have forced it to sacrifice
    other educational values central to its mission. Id. at 340.
    Implicit in the Court’s analysis was a measure of deference
    toward the university’s identification of those values.33 See id.
      33
         The Supreme Court repeatedly has shown deference to school officials
    at the intersection between constitutional protections and educational pol-
    icy. See generally Wendy Parker, Connecting the Dots: Grutter, School
    Desegregation, and Federalism, 45 Wm. & Mary L. Rev. 1691 (2004).
    The theme of local control over public education has animated Supreme
    Court jurisprudence. See, e.g., Brown, 349 U.S. at 299 (directing local
                 PARENTS INVOLVED v. SEATTLE SCHOOL DIST.                 14699
    at 328, 340. Here, the record reflects that the District reason-
    ably concluded that a race-neutral alternative would not meet
    its goals.
    
         i. Using poverty as an alternative measure of diversity
    
       The record demonstrates that the School Board considered
    using a poverty tiebreaker in place of the race-based tie-
    breaker. It concluded, however, that this proxy device would
    not achieve its compelling interest in achieving racial diver-
    sity, and had other adverse effects. Although there was no for-
    mal study of the proposal by District staff, Board members’
    testimony revealed two legitimate reasons why the Board
    rejected the use of poverty to reach its goal of racial diversity.
    First, the Board concluded that it is insulting to minorities and
    often inaccurate to assume that poverty correlates with minor-
    ity status. Second, for the group of students for whom poverty
    would correlate with minority status, the implementation
    would have been thwarted by high school students’ under-
    
    school officials, with court oversight, to devise remedies for segregation
    in the light of “varied local school problems”); Milliken v. Bradley, 
    418 U.S. 717
    , 741-42 (1974) (“No single tradition in public education is more
    deeply rooted than local control over the operation of schools; local auton-
    omy has long been thought essential both to the maintenance of commu-
    nity concern and support for public schools and to quality of the
    educational process.”); Freeman, 503 U.S. at 490 (“As we have long
    observed, ‘local autonomy of school districts is a vital national tradi-
    tion.’ ” (quoting Dayton Bd. of Educ. v. Brinkman, 
    433 U.S. 406
    , 410
    (1977)); see also Bethel Sch. Dist. v. Fraser, 
    478 U.S. 675
    , 683 (1986)
    (“The determination of what manner of speech in the classroom or in the
    school assembly is inappropriate properly rests with the school board.”);
    Lavine v. Blaine School District, 
    257 F.3d 981
    , 988 (9th Cir. 2001) (“In
    the school context, we have granted educators substantial deference as to
    what speech is appropriate.”) (citing and quoting Hazelwood Sch. Dist. v.
    Kuhlmeier, 
    484 U.S. 260
    , 267 (1988)). These Supreme Court decisions
    suggest that secondary schools occupy a unique position in our constitu-
    tional tradition. For this reason, we afford deference to the District’s judg-
    ment similar to that which Grutter afforded the university. See Grutter,
    539 U.S. at 328-29.
    14700     PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    standable reluctance to reveal their socioeconomic status to
    their peers.
    
       Because racial diversity is a compelling interest, the Dis-
    trict may permissibly seek it if it does so in a narrowly tai-
    lored manner. We do not require the District to conceal its
    compelling interest of achieving racial diversity and avoiding
    racial concentration or isolation through the use of “some
    clumsier proxy device” such as poverty. See Comfort, 418
    F.3d at 29 (Boudin, C.J., concurring).
    
        ii. The Urban League plan
    
       Parents also assert that the District should have more for-
    mally considered an Urban League proposal, which did not
    eliminate the integration tiebreaker but merely considered it
    after other factors. The Urban League plan was a comprehen-
    sive plan seeking to enhance the quality of education in Seat-
    tle’s schools by focusing on educational organization, teacher
    quality, parent-teacher interaction, raising curricular stan-
    dards, substantially broadening the availability of specialized
    and magnet programs (which could attract a broader cross-
    section of students to undersubscribed schools) and support-
    ing extra-curricular development. The plan proposed decreas-
    ing the School District’s reliance on race in the assignment
    process by pairing neighborhoods with particular schools and
    creating a type of neighborhood/regional school model. Under
    the Urban League plan, preference initially would be given to
    students choosing a school in their paired region, and the
    existing racial tiebreaker would be demoted from second to
    third in the process of resolving any remaining oversubscrip-
    tion. The plan also suggested adding an eleventh high school.
    
       Board members testified that they rejected the plan because
    of the high value the District places on parental and student
    choice. Moreover, given Seattle’s segregated housing pat-
    terns, by prioritizing a neighborhood/regional school model
    where students are assigned to schools close to their homes,
               PARENTS INVOLVED v. SEATTLE SCHOOL DIST.        14701
    the Urban League plan did not sufficiently ensure the achieve-
    ment of the District’s compelling interests in racial diversity
    and avoidance of racial concentration or isolation. As one
    member of the School Board testified, “[it] would become
    Controlled Choice all over again. That’s basically what Con-
    trolled Choice was, [ ] a regional plan; it controlled your
    options by using regions or geography.” It was therefore per-
    missible for the District to reject a plan that neither comported
    with its priorities nor achieved its compelling interests.
    
        iii. Lottery
    
       Parents additionally contend in this court that the District
    should have considered using a lottery to assign students to
    the oversubscribed high schools. As an initial matter, we note
    that Parents did not argue before the district court that a lot-
    tery was a workable race-neutral alternative that would
    achieve the Districts’ compelling interests. Parents now argue
    on appeal, however, that a lottery would achieve the District’s
    compelling interests without having to resort to the race-based
    tiebreaker. They ask us to assume that because approximately
    82 percent of all students want to attend one of Seattle’s over-
    subscribed schools, the makeup of this 82 percent, as well as
    that of the applicant pool for each school, mirrors the demo-
    graphics of the District (60 percent white and 40 percent non-
    white). Employing this assumption, Parents also ask us to
    assume that a random lottery drawing from this pool would
    produce a student body in each of the oversubscribed schools
    that falls within the District’s 15 percent plus or minus vari-
    ance. These assumptions, however, are not supported —
    indeed, are undercut — by the factual record. For example,
    Superintendent Olchefske explained that District patterns
    indicate that more people choose schools close to home. That
    would mean that the pool of applicants would be skewed in
    favor of the demographic of the surrounding residential area.
    That is, the applicant pool for the north area oversubscribed
    high schools would have a higher concentration of white stu-
    dents and the applicant pool for the south area oversubscribed
    14702      PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    high school would have a higher concentration of nonwhite
    students. Thus, random sampling from such a racially skewed
    pool would produce a racially skewed student body. As one
    Board member testified, a lottery was not a viable alternative
    because “[i]f applicants are overwhelmingly majority and you
    have a lottery, then your lottery — the pool of your lottery
    kids are going to be overwhelmingly majority. We have a
    diversity goal.”
    
       Although the District has the burden of demonstrating that
    its Plan is narrowly tailored, see Gratz, 539 U.S. at 270, it
    need not “exhaust[ ] every conceivable race-neutral alterna-
    tive.” Grutter, 539 U.S. at 339. Parents’ belated and bald
    assertion that a lottery could achieve the District’s compelling
    interests, without any evidence to support their claim, fails to
    demonstrate that a lottery is a viable race-neutral alternative.
    See id. at 340 (dismissing the race-neutral alternative of “per-
    centage plans,” advocated by the United States in an amicus
    brief, because the “United States [did] not . . . explain how
    such plans could work for graduate and professional
    schools”); Comfort, 418 F.3d at 23 (noting that Lynn rejected
    the use of a lottery in place of the race-based tiebreaker and
    holding that “Lynn must keep abreast of possible alternatives
    as they develop . . . but it need not prove the impracticability
    of every conceivable model for racial integration”) (internal
    citation omitted).
    
      c. The District’s use of race
    
       The dissent posits variables the District could use instead
    of race, for example, embracing the San Francisco school dis-
    trict’s approach as a possible model for integration that would
    meet the dissent’s criteria. Bea, J., dissenting, infra. at 14750-
    51, n.24. Perhaps San Francisco has experienced success
    (however that school district defines it) in its multi-variable
    plan — the details and evaluations of which are not in the
    record. The District is free to consider the San Francisco
    model when it engages in the annual review of its own Plan.
               PARENTS INVOLVED v. SEATTLE SCHOOL DIST.       14703
    However, even assuming that San Francisco’s plan is work-
    ing, that does not mean that it must be used by other cities in
    other states. Much can be gained from the various states
    employing locally appropriate means to achieve desirable
    ends. In our system, where states are considered laboratories
    to be used to experiment with myriad approaches to resolving
    social problems, we certainly should not punish one school
    district for not adopting the approach of another. Justice Bran-
    deis said it well,
    
        There must be power in the States and the Nation to
        remould, through experimentation, our economic
        practices and institutions to meet changing social
        and economic needs . . . . To stay experimentation in
        things social and economic is a grave responsibility.
        Denial of the right to experiment may be fraught
        with serious consequences to the Nation. It is one of
        the happy incidents of the federal system that a sin-
        gle courageous State may, if its citizens choose,
        serve as a laboratory; and try novel social and eco-
        nomic experiments without risk to the rest of the
        country.
    
    New State Ice Co. v. Liebmann, 
    285 U.S. 262
    , 311 (1932)
    (Brandeis, J., dissenting).
    
       [19] In sum, the District made a good faith effort to con-
    sider feasible race-neutral alternatives and permissibly
    rejected them in favor of a system involving a sibling prefer-
    ence, a race-based tiebreaker and a proximity preference.
    Over the long history of the District’s efforts to achieve
    desegregated schools, it has experimented with many alterna-
    tives, including magnet and other special-interest programs,
    which it continues to employ, and race-conscious districting.
    But when a racially diverse school system is the goal (or
    racial concentration or isolation is the problem), there is no
    more effective means than a consideration of race to achieve
    the solution. Even Parents’ expert conceded that, “if you don’t
    14704        PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    consider race, it may not be possible to offer an integrated
    option to students. . . . [I]f you want to guarantee it you have
    to consider race.” As Superintendent Olchefske stated, “when
    diversity, meaning racial diversity, is part of the educational
    environment we wanted to create, I think our view was you
    took that issue head on and used — you used race as part of
    the structures you developed.” The logic is self-evident:
    When racial diversity is a principal element of the school dis-
    trict’s compelling interest, then a narrowly tailored plan may
    explicitly take race into account.34 Cf. Hunter v. Regents of
    Univ. of Cal., 
    190 F.3d 1061
    , 1067 (9th Cir. 1999) (upholding
    as narrowly tailored the admissions policy of an elementary
    school — operated as a research laboratory — that explicitly
    considered race in pursuit of a racially balanced research sam-
    ple).
    
      4. Undue Harm
    
       A narrowly tailored plan ensures that no member of any
    racial group is unduly harmed. Grutter, 539 U.S. at 341. Par-
    ents argue that every student who is denied his or her choice
    of schools because of the integration tiebreaker suffers a con-
    stitutionally significant burden. We agree with the Supreme
    Court of Washington, however, in its assessment that the Dis-
    trict’s Plan imposes a minimal burden that is shared equally
    by all of the District’s students. Parents IV, 72 P.3d at 159-60
    (noting that the burden of not being allowed to attend one’s
    preferred school is shared by all students equally). As that
    court noted, it is well established that “there [is] no right
    under Washington law to attend a local school or the school
    of the student’s choice.” Id. at 159.35 Indeed, public schools,
      34
          The dissent urges, “The way to end discrimination is to stop discrimi-
    nating by race.” Bea, J., dissenting, infra. at 14762. More properly stated,
    the way to end segregation is to stop separation of the races. The Seattle
    school district is attempting to do precisely that.
       35
          Subject to federal statutory and constitutional requirements, structur-
    ing public education has long been within the control of the states as part
    of their traditional police powers. See Barbier v. Connolly, 
    113 U.S. 27
    ,
    31-32 (1884) (describing the states’ traditional police powers).
                 PARENTS INVOLVED v. SEATTLE SCHOOL DIST.                14705
    unlike universities, have a tradition of compulsory assign-
    ment. See Bazemore v. Friday, 
    478 U.S. 385
    , 408 (1986)
    (White, J., concurring) (noting that “school boards customar-
    ily have the power to create school attendance areas and oth-
    erwise designate the school that particular students may
    attend”). When an applicant’s qualifications are not under
    consideration at all, there is no notion that one student is enti-
    tled to a place at any particular school. See Comfort, 418 F.3d
    at 20 (“The denial of a transfer under the [District’s] Plan is
    . . . markedly different from the denial of a spot at a unique
    or selective educational institution.”).
    
       Moreover, it is undisputed that the race-based tiebreaker
    does not uniformly benefit one race or group to the detriment
    of another. At some schools, white students are given prefer-
    ence over nonwhite students, and, at other schools, nonwhite
    students are given preference over white students. For exam-
    ple, in the 2000-01 school year, 89 more white students were
    assigned to Franklin, one of Seattle’s most popular schools,
    than would have been assigned absent the tiebreaker; 107
    more nonwhite students were assigned to Ballard, another of
    Seattle’s most popular schools, than would have been
    assigned absent the tiebreaker; 27 more nonwhite students
    were assigned to Nathan Hale than would have been assigned
    absent the tiebreaker; and 82 more nonwhite students were
    assigned to Roosevelt than would have been absent the tie-
    breaker.36
    
       [20] In sum, because (1) the District is entitled to assign all
    students to any of its schools, (2) no student is entitled to
    attend any specific school and (3) the tiebreaker does not uni-
    formly benefit any race or group of individuals to the detri-
      36
        As detailed earlier, the Board’s decision to change the trigger point for
    use of the tiebreaker from plus or minus 10 percent to plus or minus 15
    percent had the effect of rendering Roosevelt High School neutral for
    desegregation purposes. Thus, the tiebreaker did not factor into assign-
    ments to Roosevelt High School in the 2001-02 school year.
    14706        PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    ment of another, the tiebreaker does not unduly harm any
    students in the District.
    
      5. Sunset Provision
    
       [21] A narrowly tailored plan must be limited not only in
    scope, but also in time. Grutter, 539 U.S. at 342. 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 Dis-
    trict’s Plan includes such reviews. It revisits the Plan annually
    and has demonstrated its ability to be responsive to parents’
    and students’ choice patterns and to the concerns of its con-
    stituents. For example, in 2000, when a higher than normal
    number of students selected the same schools, the Board
    responded by increasing the race-based trigger from 10 per-
    cent to a 15 percent deviation from the school population,
    adopting the thermostat that turns off the tiebreaker as soon
    as the school has come within the 15 percent plus or minus
    trigger point and by using the tiebreaker solely for the incom-
    ing ninth grade class.
    
       [22] With respect to the dissent’s concern for a “logical end
    point,” Bea, J., dissenting, infra. at 14755, like Justice
    O’Connor this court shares in the hope that “25 years from
    now, the use of racial preferences will no longer be necessary
    to further the interest approved today.” Grutter, 539 U.S. at
    343. We expect that the District will continue to review its
    Plan, and we presume, as did the Court in Grutter, that school
    officials will demonstrate a good faith commitment to moni-
    toring the continued need for the race-based tiebreaker and
    terminating its use when that need ends.37 See 539 U.S. at 343.
      37
        It is worth noting that plans like the District’s may actually contribute
    to achieving the Court’s vision in Grutter that racial preferences will no
    longer be necessary in 25 years — or even sooner. As Justice Ginsburg
    observed, “As lower school education in minority communities improves,
    an increase in the number of [highly qualified and competitive] students
    may be anticipated.” Grutter, 539 U.S. at 346 (Ginsburg, J., concurring).
              PARENTS INVOLVED v. SEATTLE SCHOOL DIST.      14707
                          III. Conclusion
    
      [23] For the foregoing reasons, we hold that the Plan
    adopted by the Seattle School District for high school assign-
    ments is constitutional and the use of the race-based tie-
    breaker is narrowly tailored to achieve the District’s
    compelling interests. Accordingly, we AFFIRM the district
    court’s judgment.
    
      AFFIRMED.
    PARENTS INVOLVED v. SEATTLE SCHOOL DIST.   14709
                                       Volume 2 of 2
    14710      PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    
    
    
    KOZINSKI, Circuit Judge, concurring:
    
       My colleagues in the majority and the dissent have written
    extensively and well. Given the exacting standard they are
    attempting to apply, I cannot say that either is clearly wrong.
    But there is something unreal about their efforts to apply the
    teachings of prior Supreme Court cases, all decided in very
    different contexts, to the plan at issue here. I hear the thud of
    square pegs being pounded into round holes. Ultimately, nei-
    ther analysis seems entirely persuasive.
    
       I start as did our eminent colleague Chief Judge Boudin of
    the First Circuit, in commenting on a highly-analogous plan
    adopted by the city of Lynn, Massachusetts:
    
        [The] plan at issue in this case is fundamentally dif-
        ferent from almost anything that the Supreme Court
        has previously addressed. It is not, like old-fashioned
        racial discrimination laws, aimed at oppressing
        blacks, e.g., Brown v. Bd. of Educ., 
    347 U.S. 483
        (1954); Strauder v. West Virginia, 
    100 U.S. 303
               PARENTS INVOLVED v. SEATTLE SCHOOL DIST.        14711
        (1880); nor, like modern affirmative action, does it
        seek to give one racial group an edge over another
        (either to remedy past discrimination or for other
        purposes). E.g., Adarand Constructors, Inc. v. Pena,
        
    515 U.S. 200
     (1995). By contrast to Johnson v. Cali-
        fornia, 
    125 S. Ct. 1141
     (2005), the plan does not seg-
        regate persons by race. See also Loving v. Virginia,
        
    388 U.S. 1
     (1967). Nor does it involve racial quotas.
        E.g., Regents of the Univ. of Cal. v. Bakke, 
    438 U.S. 265
     (1978).
    
    Comfort v. Lynn Sch. Comm., 
    418 F.3d 1
    , 27 (1st Cir. 2005)
    (Boudin, C.J., concurring).
    
       These are meaningful differences. When the government
    seeks to use racial classifications to oppress blacks or other
    minorities, no conceivable justification will be sufficiently
    compelling. See, e.g., Yick Wo v. Hopkins, 
    118 U.S. 356
    , 374
    (1886). Similarly, when lawyers use peremptory challenges to
    exclude jurors of a particular race, thereby denying them the
    right to participate in government service, they must justify
    their challenges based on objective, non-racial considerations;
    justifications based on race will be rejected out of hand, no
    matter how compelling they might seem. See Batson v. Ken-
    tucky, 
    476 U.S. 79
    , 85-88 (1986). When government seeks to
    segregate the races, as in Johnson, the courts will look with
    great skepticism at the justifications offered in support of such
    programs, and will reject them when they reflect assumptions
    about the conduct of individuals based on their race or skin
    color. See Johnson, 125 S. Ct. at 1154 (Stevens, J., dissenting)
    (concluding that California’s policy of racially segregating
    inmates “supports the suspicion that the policy is based on
    racial stereotypes and outmoded fears about the dangers of
    racial integration”). When the government engages in racial
    gerrymandering, it not only keeps the races apart, but exacer-
    bates racial tensions by making race a proxy for political
    power. See Shaw v. Reno, 
    509 U.S. 630
    , 648 (1993) (“When
    a district obviously is created solely to effectuate the per-
    14712      PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    ceived common interests of one racial group, elected officials
    are more likely to believe that their primary obligation is to
    represent only the members of that group, rather than their
    constituency as a whole.”). Programs seeking to help minori-
    ties by giving them preferences in contracting, see, e.g., Ada-
    rand, and education, see, e.g., Bakke, benign though they may
    be in their motivations, pit the races against each other, and
    cast doubts on the ability of minorities to compete with the
    majority on an equal footing.
    
       The Seattle plan suffers none of these defects. It certainly
    is not meant to oppress minorities, nor does it have that effect.
    No race is turned away from government service or services.
    The plan does not segregate the races; to the contrary, it seeks
    to promote integration. There is no attempt to give members
    of particular races political power based on skin color. There
    is no competition between the races, and no race is given a
    preference over another. That a student is denied the school
    of his choice may be disappointing, but it carries no racial
    stigma and says nothing at all about that individual’s aptitude
    or ability. The program does use race as a criterion, but only
    to ensure that the population of each public school roughly
    reflects the city’s racial composition.
    
       Because the Seattle plan carries none of the baggage the
    Supreme Court has found objectionable in cases where it has
    applied strict scrutiny and narrow tailoring, I would consider
    the plan under a rational basis standard of review. By rational
    basis, I don’t mean the standard applied to economic regula-
    tions, where courts shut their eyes to reality or even invent
    justifications for upholding government programs, see, e.g.,
    Williamson v. Lee Optical of Okla., Inc., 
    348 U.S. 483
     (1955),
    but robust and realistic rational basis review, see, e.g., City of
    Cleburne v. Cleburne Living Ctr., Inc., 
    473 U.S. 432
     (1985),
    where courts consider the actual reasons for the plan in light
    of the real-world circumstances that gave rise to it.
    
       Under this standard, I have no trouble finding the Seattle
    plan constitutional. Through their elected officials, the people
               PARENTS INVOLVED v. SEATTLE SCHOOL DIST.        14713
    of Seattle have adopted a plan that emphasizes school choice,
    yet tempers such choice somewhat in order to ensure that the
    schools reflect the city’s population. Such stirring of the melt-
    ing pot strikes me as eminently sensible.
    
       The record shows, and common experience tells us, that
    students tend to select the schools closest to their homes,
    which means that schools will reflect the composition of the
    neighborhood where they are located. Neighborhoods, how-
    ever, do not reflect the racial composition of the city as a
    whole. In Seattle, “as in many other cities, minorities and
    whites often live in different neighborhoods.” Comfort, 418
    F.3d at 29 (Boudin, C.J., concurring). To the extent that stu-
    dents gravitate to the schools near their homes, the schools
    will have the same racial composition as the neighborhood.
    This means that student patterns of interacting primarily with
    members of their own race that are first developed by living
    in racially isolated neighborhoods will be continued and exac-
    erbated by the school experience.
    
       It is difficult to deny the importance of teaching children,
    during their formative years, how to deal respectfully and col-
    legially with peers of different races. Whether one would call
    this a compelling interest or merely a highly rational one
    strikes me as little more than semantics. The reality is that
    attitudes and patterns of interaction are developed early in life
    and, in a multicultural and diverse society such as ours, there
    is great value in developing the ability to interact successfully
    with individuals who are very different from oneself. It is
    important for the individual student, to be sure, but it is also
    vitally important for us as a society.
    
       It may be true, as the dissent suggests, that students are
    influenced far more by their experiences in the home, church
    and social clubs they attend outside of school. But this does
    not negate the fact that time spent in school and on school-
    related activities, which may take up as much as half of a stu-
    dent’s waking hours, nevertheless has a significant impact on
    14714      PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    that student’s development. The school environment forces
    students both to compete and cooperate in the classroom, as
    well as during extracurricular activities ranging from football
    to forensics. Schoolmates often become friends, rivals and
    romantic partners; learning to deal with individuals of differ-
    ent races in these various capacities cannot help but foster the
    live-and-let-live spirit that is the essence of the American
    experience. I believe this is a rational objective for an educa-
    tional system—every bit as rational as teaching the three Rs,
    advanced chemistry or driver’s education. Schools, after all,
    don’t simply prepare students for further education, though
    they certainly can and should do that; good schools prepare
    students for life, by instilling skills and attitudes that will
    serve them long after their first year of college.
    
       To borrow Judge Boudin’s words once again, the plan here
    is “far from the original evils at which the Fourteenth Amend-
    ment was addressed. . . . This is not a case in which, against
    the background of core principles, all doubts should be
    resolved against constitutionality.” Comfort, 418 F.3d at 29
    (Boudin, C.J., concurring). I am acutely mindful of the
    Supreme Court’s strong admonition only last Term that any
    and all racial classifications must be adjudged under the strict
    scrutiny standard of review. See Johnson, 125 S. Ct. at 1146
    (citing cases). But the Supreme Court’s opinions are necessar-
    ily forged by the cases presented to it; where the case at hand
    differs in material respects from those the Supreme Court has
    previously decided, I would hope that those seemingly cate-
    gorical pronouncements will not be applied without consider-
    ation of whether they make sense beyond the circumstances
    that occasioned them.
    
       When the Supreme Court does review the Seattle plan, or
    one like it, I hope the justices will give serious thought to
    bypassing strict—and almost always deadly—scrutiny, and
    adopt something more akin to rational basis review. Not only
    does a plan that promotes the mixing of races deserve support
    rather than suspicion and hostility from the judiciary, but
               PARENTS INVOLVED v. SEATTLE SCHOOL DIST.       14715
    there is much to be said for returning primacy on matters of
    educational policy to local officials. Long past is the day
    when losing an election or a legislative vote on a hotly con-
    tested issue was considered the end of the matter—at least
    until the next election when the voters might “throw the ras-
    cals out.” Too often nowadays, an election or a vote is a mere
    precursor to litigation, with the outcome of the dispute not
    known until judges decide the case many years later.
    
       Whatever else the strict scrutiny standard of review may
    do, it most certainly encourages resort to the courts and often
    delays implementation of a program for years. The more com-
    plex and exacting the standard of review, the more uncertain
    the outcome, and the greater are the incentives for the parties
    to bloat the record with depositions, expert reports, exhibits,
    documents and various other materials they hope will catch
    the eye of the judges who ultimately decide the issue. This is
    a perfectly fine example, the litigation having taken over five
    years so far, generating 11 published opinions from the 24
    judges who have considered the matter in the federal and state
    courts. In the meantime, the plan was put on hold, and at least
    one class has entered and will have completed its entire high
    school career without ever being affected by it.
    
       While it’s tempting to adopt rules of law that give us the
    ultimate say on hotly contested political questions, we should
    keep in mind that we are not infallible, nor are we the reposi-
    tory of ultimate wisdom. Elected officials, who are much
    closer to ground zero than we are—and whose political power
    ebbs and flows with the approval of the voters—understand
    the realities of the situation far better than we can, no matter
    how many depositions and expert reports we may read in the
    quiet of our chambers. It therefore behooves us to approach
    issues such as those presented here with a healthy dose of
    modesty about our ability to understand the past or predict the
    future. It should make us chary about use of the strict scrutiny
    standard of review, which proclaims us the ultimate arbiters
    of the issue and gives those who oppose the policy in question
    14716       PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    every incentive to turn litigation, to paraphrase Clausewitz,
    into a continuation of politics by other means.
    
       To resort to Chief Judge Boudin’s words one last time, “we
    are faced with a local experiment, pursuing plausible goals by
    novel means that are not squarely condemned by past
    Supreme Court precedent. The problems that the . . . plan
    addresses are real, and time is more likely than court hearings
    to tell us whether the solution is a good one . . . .” Comfort,
    418 F.3d at 29 (Boudin, C.J., concurring). I share Judge
    Boudin’s preference for resolving such difficult issues by trial
    and error in the real world, rather than by experts jousting in
    the courtroom. When it comes to a plan such as this—a plan
    that gives the American melting pot a healthy stir without
    benefitting or burdening any particular group—I would leave
    the decision to those much closer to the affected community,
    who have the power to reverse or modify the policy should it
    prove unworkable. It is on this basis that I would affirm the
    judgment of the district court.
    
    
    
    BEA, Circuit Judge, with whom Circuit Judges KLEINFELD,
    TALLMAN and CALLAHAN join dissenting:
    
      I respectfully dissent.
    
       At the outset, it is important to note what this case is not
    about. The idea that children will gain social, civic, and per-
    haps educational skills by attending schools with a proportion
    of students of other ethnicities and races, which proportion
    reflects the world in which they will move, is a notion
    grounded in common sense. It may be generally, if not univer-
    sally, accepted.1 But that is not the issue here. The issue here
    is whether this idea may be imposed by government coercion,
    rather than societal conviction; whether students and their par-
      1
       For a dissenting view, see infra pp. 14731-33.
                 PARENTS INVOLVED v. SEATTLE SCHOOL DIST.                 14717
    ents may choose, or whether their government may choose for
    them.2
    
       In the Seattle School District (“District”), some schools are
    oversubscribed and in higher demand than others, so the Dis-
    trict uses a tiebreaker to assign some ninth-grade students, and
    not others, to those schools. The tiebreaker operates solely on
    the basis of the student’s race. In fact, rather than differentiat-
    ing between African-American, Asian-American, Latino,
    Native American, or Caucasian students, the tiebreaker classi-
    fies students only as “white” or “nonwhite.”3 The District
    seeks a racially balanced student body of 40% white, 60%
    nonwhite children; the tiebreaker excludes white or nonwhite
    students from an oversubscribed school if their admission will
    not further that preferred ratio.
    
       Notwithstanding the majority’s fervent defense of that plan,
    the District is engaged in simple racial balancing, which the
    Equal Protection Clause forbids. The majority can arrive at
    the opposite conclusion only by applying a watered-down
      2
         Because of our country’s struggle with racial division and the injustices
    of compelled government de jure segregation, we must be especially
    suspicious of any compulsive government program based upon race, even
    when such a program is supposedly beneficial. Good intentions cannot
    insulate the government’s use of race from the commands of the Equal
    Protection Clause; history is rife with examples of well-intentioned gov-
    ernment programs which later caused grievous harm to society and indi-
    viduals. See Adarand Constructors v. Pena, 
    515 U.S. 200
    , 226 (1995)
    (“More than good motives should be required when government seeks to
    allocate its resources by way of an explicit racial classification system.”);
    Olmstead v. United States, 
    277 U.S. 438
    , 479 (1928) (Brandeis, J., dissent-
    ing) (“Experience should teach us to be most on our guard to protect lib-
    erty when the Government’s purposes are beneficent. . . . The greatest
    dangers to liberty lurk in insidious encroachment by men of zeal, well-
    meaning but without understanding.”).
       3
         This makes all the more puzzling the majority’s assertion that “that the
    District has a compelling interest in securing the educational and social
    benefits of racial (and ethnic) diversity.” Majority op. 14660 (emphasis
    added). There simply is no ethnic tiebreaker.
    14718        PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    standard of review—improperly labeled “strict scrutiny”—
    which contains none of the attributes common to our most
    stringent standard of review. I respectfully disagree with the
    majority’s gentle endorsement of the racial tiebreaker and
    would instead hold the District violates the Equal Protection
    Clause whenever it excludes a student from a school solely on
    the basis of race.
    
                                         I.
    
       As an introductory note, I call attention to the majority’s
    frequent misuse of the terms “segregation,” “segregated
    schools,” and “segregated housing patterns.” See, e.g., Major-
    ity op. at 14660, 14661. As a perfectly understandable rhetori-
    cal ploy, the majority continually uses those charged terms
    when there has been no such segregation in the Seattle
    schools in any textual or legal sense.4 Throughout the desegre-
    gation cases, the U.S. Supreme Court stated that only the
    remediation of de jure segregation justified the use of racial
    classifications. Freeman v. Pitts, 
    503 U.S. 467
    , 494 (1992).
    “[T]he differentiating factor between de jure segregation and
    so-called de facto segregation . . . is purpose or intent to seg-
    regate.” Keyes v. School Dist. No. 1, 
    413 U.S. 189
    , 208 (1973)
    (emphasis in original); see Swann v. Charlotte-Mecklenburg
    Bd. of Educ., 
    402 U.S. 1
    , 17 (1971) (“ ‘Desegregation’ means
    the assignment of students to public schools and within such
    schools without regard to their race, color, religion, or
    national origin, but ‘desegregation’ shall not mean the assign-
    ment of students to public schools in order to overcome racial
    imbalance.”) (emphasis added).
    
      “Segregate” is a transitive verb. It requires an actor to do
    an act which effects segregation. See OXFORD ENGLISH DICTIO-
      4
       Remediation of de jure segregation is not at issue here; the parties con-
    cede the District’s schools have never been de jure segregated. No one
    even suggests that Seattle’s housing market has ever been affected by de
    jure segregation.
                PARENTS INVOLVED v. SEATTLE SCHOOL DIST.             14719
    NARY  (2d ed. 1989) (“segregate, v. 1. a. trans.: To separate (a
    person, a body or class of persons) from the general body, or
    from some particular class; to set apart, isolate, seclude”).5
    Instead of de jure segregation, what the majority describes is
    racial imbalance in the District’s schools and Seattle’s resi-
    dential makeup.
    
       Of course, it is much easier to argue for measures to end
    “segregation” than for measures to avoid “racial imbalance.”
    Especially is this so in view of the U.S. Supreme Court’s fre-
    quent pronouncements that “racial balancing” violates the
    Equal Protection Clause. See Grutter v. Bollinger, 
    539 U.S. 306
    , 330 (2003) (“[O]utright racial balancing . . . is patently
    unconstitutional.”); Freeman, 503 U.S. at 494 (“Racial bal-
    ance is not to be achieved for its own sake.”); Regents of the
    Univ. of Calif. v. Bakke, 
    438 U.S. 265
    , 307 (1978) (Powell,
    J.) (“If petitioner’s purpose is to assure within its student body
    some specified percentage of a particular group merely
    because of its race or ethnic origin, such a preferential pur-
    pose must be rejected not as insubstantial but as facially
    invalid. Preferring members of any one group for no reason
    other than race or ethnic origin is discrimination for its own
    sake. This the Constitution forbids.”).
    
      It should be remembered by the reader of the majority opin-
    ion that one can no more “segregate” without a person
    actively doing the segregation than one can separate an egg
    without a cook.
    
       Like Judge Boudin,6 in his concurring opinion Judge Koz-
    inski tries to distinguish past Supreme Court cases involving
    racial discrimination by focusing on the effects of the discrim-
    ination, rather than the fact of the discrimination.
      5
         Indeed, the term “de facto segregation” is somewhat of an oxymoron.
    That is perhaps why the Supreme Court preceded the term with the quali-
    fier “so-called.” See Keyes, 413 U.S. at 208.
       6
         See Comfort v. Lynn Sch. Comm., 
    418 F.3d 1
    , 27 (1st Cir. 2005)
    (Boudin, C.J., concurring).
    14720        PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
       This creates for them two categories different from the
    effects of the Seattle plan: (1) the effects of other race dis-
    crimination plans were much worse than Seattle’s and (2) the
    effects were visited on certain races.
    
       But the difference reflected in these two categories are
    irrelevant. “[T]here is no de minimis exception to the Equal
    Protection Clause. Race discrimination is never a ‘trifle.’ ”
    Monterey Mechanical Co. v. Wilson, 
    125 F.3d 702
    , 712 (9th
    Cir. 1997). Second, the Fourteenth Amendment protects indi-
    vidual rights, not the rights of certain races or groups.
    
      Further, that a “plan does not segregate persons by race”7
    does not justify it in refusing school admission to a qualified
    scholar because he does not belong to a particular race. There
    was no segregation by race at Cal Davis medical school, when
    Bakke was improperly refused admission. See Bakke, 
    438 U.S. 265
    .
    
       Also, it is quite accurate to say the Seattle plan does not
    “involve racial quotas.”8 The numerical quota is the percent-
    age by which the school in question’s racial composition dif-
    fers from the school district’s target.9 Not calling it a quota,
    does not make it something other. “A rose by any other name
    . . . etc.”
    
       Perhaps the Supreme Court will adopt a “rational relation”
    basis for review of race-based discrimination by government,
    based on the concurrence’s view of what is “realistic” or what
    are “real-world circumstances.”10 As indicated above, how-
      7
         Id.
      8
         Concurrence at 14710-11 (citing Comfort, 418 F.3d at 27 (Boudin,
    C.J., concurring).
       9
         See infra pp. 14745-48 (discussion of why the racial tiebreaker used by
    Seattle is a quota).
       10
          What is “the reality” or “realistic” or “real-world” is usually a rhetori-
    cal tool for dressing up one’s own view as objective and impartial, and
    therefore, more presentable.
                PARENTS INVOLVED v. SEATTLE SCHOOL DIST.             14721
    ever, it certainly has given no such indication.11 But if it does,
    one doubts that it will do so based on a “melting pot” meta-
    phor.
    
       Up to now, the American “melting pot” has been made up
    of people voluntarily coming to this country from different
    lands, putting aside their differences and embracing our com-
    mon values. To date it has not meant people who are told
    whether they are white or non-white, and where to go to
    school based on their race.
    
       The suggestion that local political forces should decide
    when to employ racial discrimination in the allocation of gov-
    ernmental resources is certainly nothing new in American his-
    tory. Such “local option” discrimination was adopted in the
    Missouri Compromise of 1820, which established the Mason-
    Dixon line, and the Compromise of 1850. But since then, the
    Civil War, the post-war Amendments to the Constitution and
    Brown v. Bd. of Ed. of Topeka, Shawnee County, Kan., 
    347 U.S. 483
     (1954) have made racial discrimination a matter of
    national concern and national governance.
    
      As noted in the opening lines of this dissenting opinion, it
    certainly is rational to believe that racial balancing in schools
    achieves better racial socialization and, as a result, better citi-
    zens. The issue is not that, but whether what is rational can
    be achieved by compulsory racial discrimination by the State.
    
                                      II.
    
       I agree with the majority that the District’s use of the racial
    tiebreaker is a racial classification, and all racial classifica-
    tions are subject to “strict scrutiny” review under the Equal
    Protection Clause. See Majority op. at 14672. Yet the majority
      11
        See e.g. Adarand, 515 U.S. at 224, Gratz v. Bollinger, 
    539 U.S. 244
    ,
    270 (2003), Johnson v. California, 
    125 S. Ct. 1141
    , 1146 (2005). On this
    point, the majority agrees. See Majority op. pp. 14670-71 n.12.
    14722      PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    conceives of strict scrutiny as some type of relaxed, deferen-
    tial standard of review. I view it differently.
    
       “No State shall . . . deny to any person within its jurisdic-
    tion the equal protection of the laws.” U.S. Const., amend.
    XIV, § 1. The right to equal protection is an individual one,
    and so where federal or state governments classify a person
    according to race—“a group classification long recognized as
    in most circumstances irrelevant and therefore prohibited”—
    we review such state action under the most “detailed judicial
    inquiry”—that is, under strict scrutiny. Grutter, 539 U.S. at
    326; see Miller v. Johnson, 
    515 U.S. 900
    , 911 (1995) (“At the
    heart of the Constitution’s guarantee of equal protection lies
    the simple command that the Government must treat citizens
    as individuals, not as simply components of a racial, religious,
    sexual or national class.)” (internal quotation marks omitted).
    
       The right to equal protection is held equally among all indi-
    viduals. “[A]ll racial classifications reviewable under the
    Equal Protection Clause must be strictly scrutinized.” Ada-
    rand, 515 U.S. at 224 (1995) (emphasis added). Strict scrutiny
    applies regardless whether the racial classifications are invidi-
    ous or benign and “is not dependent on the race of those bur-
    dened or benefited by a particular classification.” Gratz, 539
    U.S. at 270; see Johnson, 125 S. Ct. at 1146 (“We have
    insisted on strict scrutiny in every context, even for so-called
    ‘benign’ racial classifications, such as race-conscious univer-
    sity admissions policies, race-based preferences in govern-
    ment contracts, and race-based districting intended to improve
    minority representation.”) (internal citations omitted). We
    require such a demanding inquiry “to ‘smoke out’ illegitimate
    uses of race by assuring that the legislative body is pursuing
    a goal important enough to warrant use of a highly suspect
    tool.” Adarand, 515 U.S. at 226.
    
       The right to equal protection provides a liberty; it repre-
    sents freedom from government coercion based upon racial
    classifications. See Miller, 515 U.S. at 904 (the Equal Protec-
               PARENTS INVOLVED v. SEATTLE SCHOOL DIST.        14723
    tion Clause’s “central mandate is racial neutrality in govern-
    mental decisionmaking”). Thus, under strict scrutiny, all
    racial classifications by the government, regardless of pur-
    ported motivation, are “inherently suspect,” Adarand, 515
    U.S. at 223, and “presumptively invalid,” Shaw v. Reno, 
    509 U.S. 630
    , 643-44 (1993). They are permissible only where the
    government proves their use is “narrowly tailored to further
    compelling governmental interests.” Grutter, 539 U.S. at 326.
    
       It follows, then, that the government carries the burden of
    proving that its use of racial classifications satisfies strict
    scrutiny. Johnson, 125 S. Ct. at 1146 n.1 (“We put the burden
    on state actors to demonstrate that their race-based policies
    are justified.”); Gratz, 539 U.S. at 270; W. States Paving Co.,
    Inc. v. Wash. State Dep’t of Transp., 
    407 F.3d 983
    , 990 (9th
    Cir. 2005) (“The burden of justifying different treatment by
    ethnicity . . . is always on the government.”) (quoting Monte-
    rey Mech. Co. v. Wilson, 
    125 F.3d 702
    , 713 (9th Cir. 1997).
    
       Despite this formidable standard of review, the majority
    does not hesitate to endorse the District’s use of the racial tie-
    breaker. Rather than recognizing the protections of the indi-
    vidual against governmental racial classifications, the
    majority instead endorses a rigid racial governmental group-
    ing of high school students for the purpose of attaining racial
    balance in the schools. For the reasons expressed below, I do
    not share in the majority’s confidence that such a plan is con-
    stitutionally permissible.
    
                                   III.
    
       I consider first whether the District has asserted a “compel-
    ling governmental interest,” the first element of the strict scru-
    tiny test. The District contends it has a valid compelling
    governmental interest in using racial balancing to achieve “the
    educational and social benefits of racial . . . diversity” within
    its high schools and avoid “racially concentrated” schools.
    The District argues its interest will enhance student discussion
    14724      PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    of racial issues in high school and will foster cross-racial
    socialization and understanding, both in school and later in
    the students’ lives.
    
       The U.S. Supreme Court has “declined to define compel-
    ling interest or to tell [the lower courts] how to apply that
    term.” Hunter v. Regents of the Univ. of Calif., 
    190 F.3d 1061
    ,
    1070 n.9 (9th Cir. 1999) (Beezer, J., dissenting); Mark R. Kil-
    lenbeck, Pushing Things Up to Their First Principles: Reflec-
    tions on the Values of Affirmative Action, 87 Calif. L. Rev.
    1299, 1349 (1999) (the definition of a compelling interest “is
    admittedly imprecise. The Supreme Court has never offered
    a workable definition of the term . . . and is unlikely ever to
    do so, preferring to approach matters on a case-by-case
    basis”).
    
       The majority is correct in noting the U.S. Supreme Court
    has never endorsed “racial balancing” as a “compelling inter-
    est.” Indeed, throughout the history of strict scrutiny, the
    Supreme Court has rejected as invalid all such asserted com-
    pelling interests, save for two exceptions. With respect, the
    majority errs in creating a third.
    
                                  A.
    
       The Court has endorsed two race-based compelling govern-
    mental interests in the public education context. First, the
    Court has allowed racial classifications to remedy past racial
    imbalances in schools resulting from past de jure segregation.
    Freeman, 503 U.S. at 494. Second, the Court has allowed
    undergraduate and graduate universities to consider race as
    part of an overall, flexible assessment of an individual’s char-
    acteristics to attain student body diversity. Grutter, 539 U.S.
    at 328; Gratz, 539 U.S. at 268-69.
    
       Besides those two valid compelling interests, the Court has
    struck down every other asserted race-based compelling inter-
    est that has come before it. See Shaw v. Hunt, 
    517 U.S. 899
    ,
                 PARENTS INVOLVED v. SEATTLE SCHOOL DIST.                14725
    909-12 (1996) (rejecting racial classifications to “alleviate the
    effects of societal discrimination” in the absence of findings
    of past discrimination, and to promote minority representation
    in Congress); Richmond v. J.A. Croson Co., 
    488 U.S. 469
    ,
    511 (1989) (plurality) (rejecting racial classifications in the
    awarding of public construction contracts in the absence of
    findings of past discrimination); Wygant v. Jackson Bd. of
    Educ., 
    476 U.S. 267
    , 274-76 (1986) (rejecting racial classifi-
    cations in a school district’s teacher layoff policy when
    offered as a means of providing minority role models for its
    minority students and as a means of alleviating past societal
    discrimination); Bakke, 438 U.S. at 310-11 (Powell, J.)
    (rejecting the application of race-conscious measures to
    improve “the delivery of health-care services to communities
    currently underserved”). A crucial guiding point here—and
    one elided entirely by the majority—is the Court’s consistent
    reiteration that “outright racial balancing . . . is patently
    unconstitutional.” See, e.g., Grutter, 539 U.S. at 330.
    
       Thus, we face a landscape littered with rejected asserted
    “compelling interests” requiring race-based determinations,
    but with two exceptions still standing. The first exception is
    inapplicable here because the Seattle schools have never been
    de jure segregated. See Freeman, 503 U.S. at 494.
    
       The second exception is also inapplicable, albeit not so
    directly acknowledged. At oral argument, the District con-
    ceded that it is not asserting the Grutter “diversity” interest;
    the majority recognizes this in stating the District’s asserted
    interest is “significantly different” in some ways from the
    interest asserted in Grutter. Majority op. at 14676. Nonethe-
    less, the majority concludes those differences are inconse-
    quential because of the different “context”12 between high
       12
          The majority cites often to Grutter’s statement that “context matters”
    in reviewing racial classifications under the Equal Protection Clause. See
    539 U.S. at 327 (“Context matters when reviewing race-based governmen-
    tal action under the Equal Protection Clause.”). There, the Court counseled
    that strict scrutiny was to take “relevant differences” into account. Id.
    14726        PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    schools and universities, and the District’s asserted interest is
    a compelling governmental interest in its own right.
    
       Not so. The very differences between the Grutter “diversi-
    ty” interest and the District’s asserted interest illustrate why
    the latter violates the Equal Protection Clause as opposed to
    the former. The Grutter “diversity” interest focuses upon the
    individual, of which race plays a part, but not the whole. The
    District’s asserted interest, however, focuses only upon race,
    running afoul of equal protection’s focus upon the individual.
    
                                         B.
    
      In Grutter and Gratz, the Court made clear that the valid
    compelling interest in “diversity” does not translate into a
    
       Indeed, “context” does matter; context always matters in the application
    of general rules of law to varied factual settings. See Gomillion v. Light-
    foot, 
    364 U.S. 339
    , 343-44 (1960) (“Particularly in dealing with claims
    under broad provisions of the Constitution, which derive content by an
    interpretive process of inclusion and exclusion, it is imperative that gener-
    alizations, based on and qualified by the concrete situations that gave rise
    to them, must not be applied out of context in disregard of variant control-
    ling facts.”). In Grutter, the “context” was a public law school’s race-
    conscious, individualized consideration of applicants for purposes of
    admissions, designed to achieve diversity. Here, the context is different;
    we consider a rigid racial tiebreaker, which considers only race, designed
    to avoid racial imbalance in the schools. And so, as we do for all cases,
    we look to general principles of law and apply them through the correct
    standard of review, cognizant of the different results reached in other cases
    because of different facts and the “context” in which the cases arose. But
    what must be remembered is that a different “context” does not change the
    general rules of law, nor does a different “context” change the applicable
    standard of review (at least for government-imposed racial classifications).
       Yes, “context” matters, but the mention of “context” should not be a tal-
    isman to banish further enquiry. The “context” of the Michigan Law
    School is different from the District’s schools. But the difference is in the
    age of the students, their number and the obligation of the District to admit
    all students. Does that change the fact that some students are sent to cer-
    tain schools solely because of their races? How does “context” change
    that? Let us not succumb to the use of an abstraction (“context”) to invoke
    “sensitivity” to “nuances,” thus to attempt to change the bald fact of selec-
    tion based on race.
                 PARENTS INVOLVED v. SEATTLE SCHOOL DIST.              14727
    valid compelling interest in “racial diversity.” The “diversity”
    interest
    
           is not an interest in simple ethnic diversity, in which
           a specified percentage of the student body is in effect
           guaranteed to be members of selected ethnic groups
           . . . . Rather, the diversity that furthers a compelling
           state interest encompasses a far broader array of
           qualifications and characteristics of which racial or
           ethnic origin is but a single though important ele-
           ment.
    
    Grutter, 539 U.S. at 324-25 (emphasis added); see Gratz, 539
    U.S. at 272-73 (“[T]he critical criteria [in a permissible race-
    conscious admissions program] are often individual qualities
    or experiences not dependent upon race but sometimes associ-
    ated with it.”).
    
       The Grutter “diversity” interest focuses upon the individ-
    ual, which can include the applicant’s race, but also includes
    other factors, such as the applicant’s family background, her
    parent’s educational history, whether she is fluent in other
    languages, whether she has overcome adversity or hardship,
    or whether she has unique athletic or artistic talents. See 539
    U.S. at 338. Such a focus is consistent with the Equal Protec-
    tion Clause, which protects the individual, not groups.
    
       But here, the District’s operation of the racial tiebreaker
    does not consider the applicant as an individual. To the con-
    trary, the racial tiebreaker considers only whether the student
    is white or nonwhite. While the Grutter “diversity” interest
    pursues genuine diversity in the student body (of which race
    is only a single “plus” factor), the District pursues an interest
    which considers only racial diversity, i.e., a predefined group-
    ing of races in the District’s schools.13 Such an interest is not
      13
        The majority fails to recognize this distinction. For example, compar-
    ing the District’s claimed interest with those endorsed in Grutter, the
    14728        PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    a valid compelling interest; it is simple racial balancing, for-
    bidden by the Equal Protection Clause. See id. at 330 (stating
    a government institution’s interest “to assure within its stu-
    dent body some specified percentage of a particular group
    merely because of its race . . . would amount to outright racial
    balancing, which is patently unconstitutional”).
    
       Grutter emphasized the dangers resulting from lack of an
    individualized consideration of each applicant. Observing that
    the Michigan Law School sought an unquantified “critical
    mass” of minority students to avoid only token representation,
    rather than some defined balance, id. at 330, the Court rea-
    soned the law school’s individualized focus on students form-
    ing that “critical mass” would avoid perpetuating the
    stereotype that all “minority students always . . . express some
    characteristic minority viewpoint on any issue,” id. at 333.
    
       But here, the District’s concept of racial diversity is a pre-
    determined, defined ratio of white and nonwhite children. The
    racial tiebreaker works to exclude white students from schools
    that have a 50-55% white student body (depending on the tie-
    breaker trigger used in a particular year), and works to
    exclude nonwhite students from schools with a 70-75% non-
    white student body (depending on the tiebreaker trigger used).
    Thus, the District’s concept of racial diversity does not permit
    
    majority reasons high schools “have an equal if not more important role”
    in preparing students for work and citizenship, and concludes “it would be
    a perverse reading of the Equal Protection Clause that would allow a uni-
    versity, educating a relatively small percentage of the population, to use
    race when choosing its student body but not allow a public school district,
    educating all children attending its schools, to consider a student’s race in
    order to ensure that the high schools within the district attain and maintain
    diverse student bodies.” Majority op. at 14676, 14678. Yet Grutter did not
    allow universities to consider race in admissions to achieve racial balanc-
    ing. The whole point of Grutter and Gratz was that universities may con-
    sider race, but only as part of the overall individual. I see nothing perverse
    in recognizing the Equal Protection Clause to be the protector of the indi-
    vidual, whether he be among the few at an elite law school, or among the
    many in a public high school.
                 PARENTS INVOLVED v. SEATTLE SCHOOL DIST.                 14729
    a school with a student body that is too white, or a school with
    a student body that is too nonwhite.
    
       The District argues its concept of racial diversity is neces-
    sary to foster classroom discussion and cross-racial socializa-
    tion. That argument, however, is based on the stereotype that
    all white children express traditional white viewpoints and
    exhibit traditional white mannerisms; all nonwhite children
    express opposite nonwhite viewpoints and exhibit nonwhite
    mannerisms, and thereby white and nonwhite children will
    better understand each other. Yet there is nothing in the racial
    tiebreaker to ensure such viewpoints and mannerisms are rep-
    resented within the preferred student body ratio. As noted in
    Grutter, the only way to achieve diverse viewpoints and man-
    nerisms is to look at the individual student. White children
    have different viewpoints and backgrounds than other white
    children; the same goes for nonwhite children; and some
    white children have the same viewpoints and backgrounds as
    some nonwhite children. The assumption that there is a differ-
    ence between individuals just because there is a difference in
    their skin color is a stereotype in itself, nothing more.14
    
       The District also claims it must use the racial tiebreaker to
    avoid racially imbalanced schools, which may result in
    schools with large white or nonwhite student bodies and in
    which the supposed benefits from the District’s concept of
    racial diversity will not occur. This theory, however, presents
    another racial stereotype, which assumes there is something
    wrong with a school that has a heavy nonwhite student body
    population, or something better about a school that has a
    heavy white student body population. See Missouri v. Jenkins,
    
    515 U.S. 70
    , 122 (1995) (Thomas, J., concurring) (“After all,
      14
        Again, there is nothing illegal in freely choosing to believe in this ste-
    reotype and to act upon it as a private citizen in sending one’s child to a
    particular school. The case changes when such racial stereotype is
    accepted by the state, and is the basis for the imposition of racial discrimi-
    nation.
    14730        PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    if separation itself is a harm, and if integration therefore is the
    only way that blacks can receive a proper education, then
    there must be something inferior about blacks. Under this the-
    ory, segregation injures blacks because blacks, when left on
    their own, cannot achieve. To my way of thinking, that con-
    clusion is the result of a jurisprudence based upon a theory of
    black inferiority.”).
    
       Besides the District’s reliance on racial stereotypes, there is
    good reason categorically to forbid racial balancing. The pro-
    cess of classifying children in groups of color, rather than
    viewing them as individuals, encourages “notions of racial
    inferiority” in both white and nonwhite children and incites
    racial hostility. See Grutter, 539 U.S. at 328. Indeed, those
    risks are particularly great here because of the blunt nature of
    the racial tiebreaker. The District’s racial grouping of stu-
    dents, either as white or nonwhite, assumes that each minority
    student is the same, regardless whether he is African-
    American, Asian-American, Latino, or Native American; the
    only difference noted by the District is that the minority stu-
    dent is not white.15 The District thus “conceives of racial
    diversity in simplistic terms as a dichotomy between white
    and nonwhite, as if to say all nonwhites are interchangeable.”
    Parents Involved in Cmty. Schs. v. Seattle Sch. Dist., No. 1,
    
    72 P.3d 151
    , 169 n.5 (Wash. 2003) (Sanders, J., dissenting).
    I join my colleague on the Washington Supreme Court in
    observing that “[a]s a theory of racial politics, this view is
    patently offensive and as a policy to promote racially diverse
    schools, wholly inadequate.” Id.
      15
        The majority notes that for purposes of the racial tiebreaker, “a student
    is deemed to be of the race specified in his or her registration materials.”
    Majority op. at 14666. That generalization declines to note a particularly
    overbearing facet of the racial tiebreaker. Although the District encourages
    the students’ parents to identify the race of their student in the registration
    materials, if a parent or student chooses to follow the example of Tiger
    Woods and refuses to identify his or her race, the District then engages in
    a visual inspection of the student or parent and will decide the child’s
    color notwithstanding the parent’s or student’s choice.
               PARENTS INVOLVED v. SEATTLE SCHOOL DIST.        14731
       Unlike a voluntary decision by parents to expose their chil-
    dren to individuals of different races or background, the Dis-
    trict classifies each student by skin color and excludes certain
    students from particular schools—solely on the basis of race
    —to ensure those schools remain racially balanced. Even if
    well-intentioned, the District’s use of racial classifications in
    such a stark and compulsory fashion risks perpetuating the
    same racial divisions which have plagued this country since
    its founding:
    
        Race is perhaps the worst imaginable category
        around which to organize group competition and
        social relations more generally. At the risk of bela-
        boring the obvious, racial categories in law have
        played an utterly pernicious and destructive role
        throughout human history. This incontrovertible fact
        should arouse wonder . . . at the hubris of those who
        imagine that we can distinguish clearly enough
        between invidious and benign race discrimination to
        engrave this distinction into our constitutional order.
        Vast human experience mocks this comforting illu-
        sion, as does the fact that most Americans, including
        many minorities, think racial preferences are invidi-
        ous, not benign. Whether benignly intended or not,
        using the category of race—which affirmative action
        proponents oddly depict as socially constructed and
        primordial and immutable—to distribute advantage
        and disadvantage tends to ossify the fluid, forward-
        looking political identities that a robust democratic
        spirit inspires and requires.
    
    Peter H. Schuck, Affirmative Action: Past, Present, and
    Future, 20 Yale L. & Pol’y Rev. 1, 92-93 (2002).
    
       We should not minimize these shadows that are cast over
    the supposed benefits of the District’s asserted interest. The
    District’s stark racial classifications not only offend intrinsic
    notions of individuality, they even suggest principles opposite
    14732      PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    to what the District claims to seek. Although the District con-
    tends it uses the racial tiebreaker for good, i.e., to foster cross-
    racial socialization and understanding, the District’s concept
    of racial diversity also suggests other principles which many
    may find objectionable, especially when taught to children:
    
        While a public law preference does express a certain
        kind of compassion for and commitment to the pre-
        ferred groups, other signals dominate its message -
        among them, that American society thinks it just to
        group people by race and ethnicity, to treat those
        groups monolithically, and to allocate precious
        resources and opportunities accordingly; that it holds
        equal treatment and individual merit as secondary,
        dispensable ideals; that the preferred groups cannot
        succeed without special public favors; that such
        favors do not stigmatize them in the minds of fair-
        minded others; that those who oppose preferences
        thereby oppose the aspirations of the preferred
        groups; and that society can assuage old injustices by
        creating new ones. When public law says such
        things, it speaks falsely, holds out vain promises, and
        brings itself into disrepute.
    
    Id. at 87-88.
    
       The District’s asserted interest may be supported by noble
    goals. But the stereotypes on which it is based, and the risks
    that it presents, make that interest far from compelling.
    
                                    C.
    
       The sociological evidence presented by the District, relied
    upon strongly by the majority, does not change my view. The
    majority discusses much of the evidence that supports the
    District’s position that racially balanced schools foster cross-
    racial socialization and understanding in school and later in
    the students’ lives. Majority op. at 14675-76. Yet the majority
               PARENTS INVOLVED v. SEATTLE SCHOOL DIST.        14733
    puts aside the other evidence suggesting there is no definitive
    agreement as to the beneficial effects of racial balance in K-
    12 schools, that the benefits attributed to racially balanced
    schools are often weak, and that any benefits do not always
    have a direct correlation to racial balance. Yet again, a private
    citizen is free to accept one body of opinion and reject another
    in deciding to send his child to a particular school. Is the state
    similarly privileged when required to determine that its
    claimed goal is a “compelling interest”? One would think that
    to be “compelling” there would be no room for doubt of the
    need for the measure. That is certainly not the case here.
    
       For example, a source provided by the District states that
    “family background has a significantly stronger effect on stu-
    dent achievement than any other single school factor or con-
    stellation of school factors, including school racial and ethnic
    composition.” [SER 182.] Another source presented by the
    District states that court-ordered desegregation (i.e., a court-
    ordered breakup of a de jure segregated student body)
    resulted in only minimal benefits:
    
        [R]esearch suggests that desegregation has had some
        positive effect on the reading skills of African Amer-
        ican youngsters. The effect is not large, nor does it
        occur in all situations, but a modest measurable
        effect does seem apparent. Such is not the case with
        mathematical skills, which seem generally unaf-
        fected by desegregation. Second, there is some evi-
        dence that desegregation may help to break what can
        be thought of as a generational cycle of segregation
        and racial isolation. Although research on this topic
        is scant and often marred by unavoidable flaws, evi-
        dence has begun to accumulate that desegregation
        may favorably influence such adult outcomes as col-
        lege graduation, income, and employment patterns.
        The measured effects are often weak . . . .
    
    [SER 205, 207-208.]
    14734        PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
       That source concludes that “[t]he evidence regarding the
    impact of desegregation on intergroup relations is generally
    held to be inconclusive and inconsistent.” [SER 208.]. See
    Grutter, 539 U.S. at 364-65 (Thomas, J., dissenting) (collect-
    ing studies suggesting black students perform at higher levels
    of achievement at historically black colleges); David I.
    Levine, Public School Assignment Methods after Grutter and
    Gratz: The View from San Francisco, 30 Hastings Const. L.Q.
    511, 536 (2003) (noting that a high school’s focus on racial
    balance misses the “key element” in the context of education,
    i.e., that “the life chances of students are improved only with
    economic integration”).16
    
       The serious risks presented by racial classifications coun-
    teract the marginal benefits provided by racial balancing.
      16
         See also David J. Armor & Christine H. Rossell, Desegregation and
    Resegregation in the Public Schools, in Beyond the Color Line: New Per-
    spectives on Race and Ethnicity in America 251 (Abigail Thernstrom &
    Stephan Thernstrom eds., 2002) (“[R]acial composition by itself has little
    effect on raising the achievement of minority students or on reducing the
    minority white achievement gap. Some studies show that there is no rela-
    tionship at all between black achievement and racial composition . . . , and
    other studies show that there is no relationship between the black-white
    achievement gap and racial composition. In either case, though there is
    some evidence here that achievement can be affected by programmatic
    changes, there is no evidence that it responds to improved racial balance
    by itself.”); id. at 252 (“The evidence on the benefit of school desegrega-
    tion for race relations is probably the weakest of all. Indeed, there are
    more studies showing harmful effects than studies showing positive
    effects. This led to another and more recent reviewer of the race relations
    literature to conclude, somewhat generously: “ ‘In general, the reviews of
    desegregation and intergroup relations were unable to come to any conclu-
    sion about what the probable effects of desegregation were. . . . Virtually
    all of the reviewers determined that few, if any, firm conclusions about the
    impact of desegregation on intergroup relations could be drawn. The
    reluctance of reviewers to draw conclusions about the benefits of school
    desegregation for race relations or self-esteem only reinforces our conclu-
    sion that the psychological harm theory of de facto segregation and the
    social benefit theory of desegregation are clearly wrong, at least when
    applied to desegregation as a racial balance policy.’ ”).
               PARENTS INVOLVED v. SEATTLE SCHOOL DIST.        14735
    Courts have long recognized racial classifications promote
    “notions of racial inferiority and lead to a politics of racial
    hostility.” See Grutter, 539 U.S. at 328; Michael Perry, Mod-
    ern Equal Protection, 79 Colum. L. Rev. 1023, 1048 (1979)
    (“Affirmative action “ ‘inevitably foments racial resentment
    and thereby strains the effort to gain wider acceptance for the
    principle of moral equality of the races.’ ”). Other studies sug-
    gest that where racial classifications are a means of achieving
    racial balance, academic achievement by minorities is hin-
    dered, and racial tensions are riled:
    
        In a culture that ardently affirms the principles of
        individual freedom, merit, and equality of opportu-
        nity, [the] demoralization and anger [precipitated by
        being victim to government-imposed racial classifi-
        cations] must be counted as a very large social cost.
        It is no less a cost because it is borne by whites, and
        often less privileged whites at that. If these princi-
        ples make it unfair to impose this cost, the fact that
        the unfairness is spread across a large group of peo-
        ple may not make it any more palatable. In fact, dif-
        fusing the unfairness in this way will simply increase
        the number of people who feel themselves
        aggrieved.
    
    Schuck, supra, at 69.
    
       But despite the inconsistencies in the sociological evidence
    and the vivid risks of the District’s asserted interest, the
    majority implicitly defers to the District’s position. Grutter
    took a similar approach, emphasizing that its endorsement of
    the “diversity” interest relied in large part upon deference to
    the educational judgment of the Michigan Law School. 539
    U.S. at 330.
    
      Yet perhaps to steal a line from the majority, the “context”
    here is different. We are not faced with a university’s “aca-
    demic freedom,” which arises from “a constitutional dimen-
    14736      PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    sion, grounded in the First Amendment, of educational
    autonomy,” and which includes the freedom to select its stu-
    dent body. Id. We instead consider a public high school’s
    admissions plan which admits or excludes students from par-
    ticular schools solely on the basis of their race. For several
    reasons, we should not defer to such a plan.
    
       First, other than for race-conscious university admissions
    based on holistic diversity, deference to a government actor
    is inconsistent with strict scrutiny. See Johnson, 125 S. Ct. at
    1146 n.1 (stating generally that “deference [by the courts in
    applying strict scrutiny] is fundamentally at odds with our
    equal protection jurisprudence”); id. at 1150 (stating the
    Supreme Court has “refused to defer to state officials’ judg-
    ments on race . . . where those officials traditionally exercise
    substantial discretion.”). In Grutter, the Court deferred to the
    Michigan Law School’s “diversity” interest because of the
    law school’s “academic freedom”—grounded in the First
    Amendment and including the law school’s freedom to select
    its own student body—and the law school’s asserted need for
    diversity to achieve a “robust exchange of ideas” within its
    classrooms, a vital part of the law school’s mission. 539 U.S.
    at 330.
    
       None of those same issues are implicated here. The “aca-
    demic freedom” of a university allows it “to determine for
    itself on academic grounds who may teach, what may be
    taught, how it shall be taught, and who may be admitted to
    study.” Bakke, 438 U.S. at 312 (Powell, J.). High schools do
    not have such similar freedoms. They cannot determine who
    may teach, at least when that determination is based upon
    racial grounds. See Wygant, 476 U.S. at 274-76. They also
    cannot determine who may be admitted to study; when the
    government chooses to provide public education in secondary
    schools, it “must be made available to all on equal terms.” See
    Plyler v. Doe, 
    457 U.S. 202
    , 221-23 (1982). Further, there is
    no comparable line of U.S. Supreme Court cases affording
    high schools the special “[A]cademic freedom[s]” granted to
               PARENTS INVOLVED v. SEATTLE SCHOOL DIST.        14737
    universities by the First Amendment. See United States v.
    Fordice, 
    505 U.S. 717
    , 728-29 (1992) (“a state university sys-
    tem is quite different in very relevant respects from primary
    and secondary schools.”); Jay P. Lechner, Learning From
    Experience: Why Racial Diversity Cannot Be a Legally Com-
    pelling Interest in Elementary and Secondary Education, 32
    SW. U. L. Rev. 201, 215 (2003) (stating the Supreme Court
    “has been less deferential to the discretion of elementary or
    secondary school officials in Equal Protection cases, in part
    because the Court has viewed school desegregation as serving
    social rather than educational goals. The Court has acknowl-
    edged that even the most important, delicate, and highly dis-
    cretionary functions of state educators are subject to the limits
    of the Bill of Rights and subordinate to the Constitutional
    freedoms of the individual. Moreover, the educational bene-
    fits from diversity, if any, are much greater at the higher edu-
    cational level because such benefits are greatly magnified by
    the learning that takes place outside the classroom—in dormi-
    tories, social settings, and extracurricular activities—as stu-
    dents must learn to live and work with persons of other races
    and ethnic backgrounds.”) (internal quotation marks omitted).
    
       Moreover, there is a crucial difference between the “robust
    exchange of ideas” theory referenced in Grutter and the Dis-
    trict’s claim that its interest “brings different viewpoints and
    experiences to classroom discussions and thereby enhances
    the educational process.” [ER 237.] The District applies the
    racial tiebreaker only to entering ninth-grade students. [ER
    253, 308.] It is self-evident that classroom discussion plays a
    significantly more vital role in universities with their typical
    dialectic or Socratic teaching method, than in ninth-grade
    high school courses with their typical didactic or rote teaching
    method.
    
       Last, the District’s claim that its asserted interest helps to
    foster cross-racial socialization and understanding later in the
    students’ lives is a sociological judgment outside the expertise
    of the District’s educators. Those external benefits are diffuse,
    14738        PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    manifest long after students leave the classroom, and cannot
    be measured with skills possessed uniquely by educators.
    Unlike Grutter, which deferred to the Law School on the
    basis that diversity in the classroom was vital to its educa-
    tional mission during the three-year law school curriculum,
    here, the District’s asserted interest depends upon benefits
    only loosely linked to the District’s educational mission and
    to take effect years after its schooling of the children, or
    entirely outside the expertise of its educators. Here, high
    school administrators and teachers are predicting what sociol-
    ogists will find years later.
    
       Strict scrutiny cannot remain strict if we defer to judgments
    not even within the particular expertise or observation of the
    party being scrutinized. Hence, deference is not due to the
    District regarding the benefits the District contends are attrib-
    utable to its claimed interest.17
      17
         The majority states that Swann v. Charlotte-Mecklenburg Bd. of
    Educ., 
    402 U.S. 1
     (1971), supports the proposition that the District has
    broad discretion to engage in racial balancing as an “educational policy.”
    In Swann, the Supreme Court stated: “School authorities are traditionally
    charged with broad power to formulate and implement educational policy
    and might well conclude, for example, that in order to prepare students to
    live in a pluralistic society each school should have a prescribed ratio of
    Negro to white students reflecting the proportion for the district as a
    whole. To do this as an educational policy is within the broad discretion-
    ary powers of school authorities; absent a finding of a constitutional viola-
    tion, however, that would not be within the authority of a federal court.”
    Id. at 16; see also North Carolina State Bd. of Educ. v. Swann, 
    402 U.S. 43
    , 45 (1971) (same, citing Swann, 402 U.S. at 16).
       Swann’s passage seems to provide powerful language for the majority’s
    position, but alas, the majority takes the passage out of context. Swann
    considered the remedies available to a federal court to combat past de jure
    segregation. The Court never considered whether a school district could
    use racial classifications to achieve racial balance absent de jure segrega-
    tion. Indeed, the Court stated: “We are concerned in these cases with the
    elimination of the discrimination inherent in the dual school systems, not
    with myriad factors of human existence which can cause discrimination in
    a multitude of ways on racial, religious, or ethnic grounds. . . . Our objec-
                 PARENTS INVOLVED v. SEATTLE SCHOOL DIST.                 14739
       In the absence of deference to the District’s sociological
    evidence, the faults of the District’s asserted interest come
    into sharper focus. It has none of the saving graces present in
    the Grutter holistic diversity interest. It perpetuates racial ste-
    reotypes and risks fomenting racial hostility. Last, the District
    enforces the interest through government compulsion in the
    starkest black and white terms, espousing the principle that
    race trumps the individual.
    
       The sociological evidence presented by the District sug-
    gests that some benefits will accrue from racial balancing. To
    me, evidence of some benefits does not satisfy the District’s
    burden of proving a compelling governmental interest, espe-
    cially in light of the Supreme Court’s frequent pronounce-
    ments that racial balancing itself is unconstitutional. Thus,
    viewed under the lens of strict scrutiny, and without the defer-
    ence invoked in Grutter, the District’s interest is simply not
    a compelling governmental interest. Hence, I would hold that
    the District’s operation of the racial tiebreaker is an impermis-
    sible racial classification and violates the Equal Protection
    Clause.
    
    tive in dealing with the issues presented by these cases is to see that school
    authorities exclude no pupil of a racial minority from any school, directly
    or indirectly, on account of race; it does not and cannot embrace all the
    problems of racial prejudice, even when those problems contribute to dis-
    proportionate racial concentrations in some schools.” Id. at 22-23
    (emphasis added).
       Swann was also decided decades before the Court resolved the issue of
    the level of scrutiny to apply to “benign” racial classifications, vis-a-vis
    “invidious” racial classifications. Thus, Swann’s dictum cannot shelter the
    District’s use of the racial tiebreaker from the searching inquiry required
    by strict scrutiny.
       The majority similarly errs in relying on Washington v. Seattle Sch.
    Dist. No. 1, 
    458 U.S. 457
     (1982). There, the Court also specifically stated
    it did not reach the issue of the constitutionality of “race-conscious student
    assignments for the purpose of achieving integration, even absent a find-
    ing of prior de jure segregation.” Id. at 472 n.15.
    14740      PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
                                  IV.
    
       Even if the District’s asserted interest were a compelling
    governmental interest, the means used by the District must
    still be narrowly tailored to serve that interest. See Grutter,
    539 U.S. at 333. For argument’s sake, I here assume, without
    conceding, the District has asserted a valid compelling gov-
    ernmental interest in using racial balancing to achieve “the
    educational and social benefits of racial . . . diversity” within
    its high schools and to avoid “racially concentrated” schools.
    Yet even under that assumption, the District’s use of the racial
    tiebreaker is not narrowly tailored to serve that interest.
    
       The majority notes that Grutter set forth “five hallmarks of
    a narrowly tailored affirmative action plan: (1) individualized
    consideration of applicants; (2) the absence of quotas; (3)
    serious, good-faith consideration of race-neutral alternatives
    to the affirmative action program; (4) that no member of any
    racial group was unduly harmed; and (5) that the program had
    a sunset provision or some other end point.” Majority op. at
    14684-85. I agree with that general formulation. Yet the
    majority’s application of those factors again evinces an
    improper deference to the District; such deference is ill suited
    for the searching inquiry needed under the narrow-tailoring
    prong of strict scrutiny. See Johnson, 125 S. Ct. at 1146 n.1.
    I consider below whether the District’s use of the racial tie-
    breaker is narrowly tailored to its asserted interest, and con-
    clude that racial tiebreaker is not narrowly tailored.
    
                                   A.
    
       The first narrow-tailoring factor requires the District to
    engage in an individualized consideration of each applicant’s
    characteristics and qualifications. See Grutter, 539 U.S. at
    337. The importance of this factor is self-evident; individual-
    ized consideration serves the primary purpose of the Equal
    Protection Clause, which protects the individual from group
    classifications, especially those by race. See id. at 326.
                 PARENTS INVOLVED v. SEATTLE SCHOOL DIST.         14741
       Yet the majority concludes that individualized consider-
    ation of each applicant is irrelevant here “because of the con-
    textual differences between institutions of higher learning and
    public high schools.” Majority op. at 14685. I could not dis-
    agree more.18 By removing consideration of the individual
    from the narrow tailoring analysis, the majority threatens to
    read the Equal Protection Clause out of the Constitution. It is
    the very nature of equal protection to require individualized
    consideration when the government uses racial classifications:
    “the Fourteenth Amendment “protects persons, not groups.”
    Grutter, 539 U.S. at 326 (quoting Adarand, 515 U.S. at 227)
    (emphasis in original). Grutter emphasized the importance of
    the individualized consideration of each applicant: in the con-
    text of a race-conscious university admissions program, such
    consideration
    
           must remain flexible enough to ensure that each
           applicant is evaluated as an individual and not in a
           way that makes an applicant’s race or ethnicity the
           defining feature of his or her application. The impor-
           tance of this individualized consideration in the con-
           text of a race-conscious admissions program is
           paramount.
    
    Id. at 337 (emphasis added). The differences between univer-
    sity and secondary education do not justify denial of individu-
    alized equal protection of the law to secondary school
    students.
    
       Individualized consideration of an applicant does not
    require an admissions program to be oblivious to race; the
    program may consider race, but in doing so, it must remain
    “flexible enough to consider all pertinent elements of diver-
    sity in light of the particular qualifications of each applicant,
    and to place them on the same footing for consideration,
      18
        See supra pp. 14670-71 n.12 (explaining why the talismanic use of
    “context” can not alter the fact of racial discrimination).
    14742      PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    although not necessarily according them the same weight.” Id.
    at 334. There can be “no policy, either de jure or de facto, of
    automatic acceptance or rejection based on any single ‘soft’
    variable . . . [such as the awarding of] mechanical, predeter-
    mined diversity ‘bonuses’ based on race or ethnicity.” Id. at
    337.
    
       Here, the racial tiebreaker works to admit or exclude high
    school students from certain oversubscribed schools solely on
    the basis of their skin color. No other consideration affects the
    operation of the racial tiebreaker; when it operates, it operates
    to admit or exclude either a white or nonwhite student,
    depending upon how the admission will affect the preferred
    balance at the oversubscribed school. Such a program is pre-
    cisely what Grutter warned against, and what Gratz held
    unconstitutional: a mechanical, predetermined policy “of
    automatic acceptance or rejection based on a[ ] single ‘soft’
    variable,” that being the student’s skin color. See id.
    
       The racial tiebreaker’s overbroad classification of students
    as “white” or “nonwhite” also runs counter to the required
    individualized consideration of each applicant. The District
    does not even consider the student’s actual race. Instead, the
    District presumably places all Caucasian students into the
    “white” category, and then places all African-American,
    Latino, Asian-American, Pacific Islander and Native Ameri-
    cans into the “nonwhite” category. This puts aside the catego-
    rization of any individuals whose skin color does not correlate
    directly with the classifications. Although parents and stu-
    dents may identify their particular group on the registration
    materials, if they do not, the District will make the racial iden-
    tification itself through visual inspection of the parent or stu-
    dent. Thus, a fair-skinned minority may wind up in the
    “white” category, or a darker-skinned Caucasian may wind up
    in the “nonwhite” category.
    
      Courts have often recognized that the inclusion of all
    minorities within a “nonwhite” classification suggests the
               PARENTS INVOLVED v. SEATTLE SCHOOL DIST.        14743
    operation of a racial classification is not narrowly tailored.
    See Wygant, 476 U.S. at 284 n.13 (noting the “definition of
    minority to include blacks, Orientals, American Indians, and
    persons of Spanish descent further illustrates the undifferenti-
    ated nature of the plan”); Monterey Mech. Co., 125 F.3d at
    714 (noting the inclusion of all minority races within a broad
    “minority” category serves as a “red flag[ ] signaling that the
    statute is not, as the Equal Protection Clause requires, nar-
    rowly tailored”). At the very least, a narrowly tailored pro-
    gram would require an individualized focus which would
    separate the student according to his or her correct race, rather
    than as a process of simple pigmental matching.
    
       The majority concludes, however, that individualized con-
    sideration of each applicant is unnecessary because the Dis-
    trict does not exclude any student from a public education by
    operation of the racial tiebreaker. The majority reasons that
    because all students are entitled to a public education in one
    of the District’s schools, there is no competition in the District
    for admission to any of those schools, and thus no racial
    stigma could attach when a student is excluded from admis-
    sion to one of the schools on the basis of his race. Majority
    op. at 14685-87.
    
       Yet the majority offers no explanation why, in the 2000-01
    school year, 82% of the students selected one of the oversub-
    scribed schools (i.e., the schools subject to the racial tie-
    breaker) as their first choice, while only 18% picked one of
    the undersubscribed schools as their first choice. Majority op.
    at 14665-66. Clearly, the students’ and their parents’ “market”
    appraise some of the schools as providing a better education
    than the others. Even the District’s superintendent confirmed
    that the students’ parents considered some of the schools to be
    of higher quality. [ER 534.]
    
       It is common sense that some public schools are better than
    others. Parents often move into areas offering better school
    districts, and ubiquitous research guides compare the quality
    14744       PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    of public schools according to standardized test scores, pro-
    gram offerings, and the sort. It may be that soothing, if self-
    interested, bureaucratic voices sing a lullaby of equal educa-
    tional quality in the District’s schools. But the facts show that
    parents and children have voted with their feet in choosing
    some schools rather than others. The verdict of that “market”
    makes a hash out of such assurances by the District.
    
       Thus, the District’s operation of the racial tiebreaker in
    reality does limit access to a governmental benefit among cer-
    tain students. The District insulates applicants belonging to
    certain racial groups from competition for admission to those
    schools perceived to be of higher quality. A narrowly tailored
    race-conscious admissions program “cannot insulate each cat-
    egory of applicants with certain desired [racial] qualifications
    from competition with all other applicants.” Grutter, 539 U.S.
    at 334. The racial tiebreaker fails that test.
    
      Yet the majority insist that because the District seeks to
    avoid racially concentrated schools, “the District’s tiebreaker
    must necessarily focus on the race of its students.” Majority
    op. at 14689. Again, the majority misses the crucial protection
    provided by the Equal Protection Clause. The District’s
    narrow-tailoring obligation does not prohibit it from consider-
    ing race; it just cannot consider only race. The constitutional
    guarantee of equal protection requires the District to focus
    upon the individual’s whole make up, rather than just a
    group’s skin color; this protects each student’s right to equal
    protection under the law. See Grutter, 539 U.S. at 326.
    
       The counter-argument, of course, is that administrative
    inconveniences would prohibit the District from examining
    each student’s file for individual characteristics, of which race
    may be a part. To the contrary, the record shows such an
    effort is certainly feasible.
    
      First, thirteen- or fourteen-year-old students19 are not so
      19
        As noted, the District applies the racial tiebreaker only to entering
    ninth-grade students (presumably around thirteen to fourteen years old).
                PARENTS INVOLVED v. SEATTLE SCHOOL DIST.              14745
    young that they have not yet developed unique traits to set
    themselves apart from other students and add greater diversity
    to the student body. The students’s race is a factor in assess-
    ing the student as an individual, but the student may also
    speak English as a second language, come from a different
    socioeconomic stratum than other students, have overcome
    adversity, be a talented baseball player, musician, or have par-
    ticipated in community service.
    
       Second, as noted by the majority, in the 2000-01 school
    year, approximately 3,000 students entered the District’s high
    schools as ninth graders. Ten percent of those students were
    subject to the racial tiebreaker. Majority op. at 14667. Thus,
    under an individualized approach, the District would have had
    to examine only three hundred applications to determine who
    to admit to the oversubscribed schools. Instead, the District
    grouped those three hundred students into white and nonwhite
    categories and allowed a computer to select their assignment
    based solely upon their race.20
    
       Thus, rather than providing an individualized consideration
    of applicants, the District is engaged in a “de jure [policy] of
    automatic acceptance or rejection based on a[ ] single ‘soft’
    variable.” See Grutter, 539 U.S. at 337. Such inflexibility
    shows the racial tiebreaker is not “narrowly tailored to any
    goal, except perhaps outright racial balancing.” See Croson,
    488 U.S. at 507 (plurality).
    
                                       B.
    
      The second narrow-tailoring factor prohibits the use of quo-
      20
         Three hundred applications seem like only a minor administrative
    challenge, but the Supreme Court’s admonition bears repeating nonethe-
    less: “[T]he fact that the implementation of a program capable of provid-
    ing individualized consideration might present administrative challenges
    does not render constitutional an otherwise problematic system.” Grutter,
    539 U.S. at 275.
    14746       PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    tas based upon race. Grutter, 539 U.S. at 334. A quota is
    defined as “a program in which a certain fixed number or pro-
    portion of opportunities are reserved exclusively for certain
    minority groups. Quotas impose a fixed number or percentage
    which must be attained, or which cannot be exceeded.” Id. at
    335 (internal quotation marks and citations omitted).
    
       Here, when a District school is oversubscribed and “inte-
    gration positive”—i.e., the white or nonwhite student body of
    the school deviates by plus or minus 10% or 15% (depending
    on the school year)21 of the preferred 40% white/60% non-
    white ratio—the District uses the racial tiebreaker to admit
    students whose presence will move the overall student body
    closer to the preferred ratio. Using the 2000-2001 school year
    as an example, the District would employ the racial tiebreaker
    to exclude white students and admit nonwhite students where
    the white student body population exceeded 50%. The District
    would also employ the racial tiebreaker to exclude nonwhite
    students and admit white students where the nonwhite student
    body population in a particular school exceeded 70%.
    
       By its nature, the tiebreaker aims for a rigid, predetermined
    ratio of white and nonwhite students, and thus operates to
    reach “a fixed number or percentage.” (emphasis supplied).
    Gratz specifically rejected such a plan as not narrowly tai-
    lored. See 539 U.S. at 270 (“[T]he University’s policy, which
    automatically distributes [20%] . . . of the points needed to
    guarantee admission, to every single ‘underrepresented
    minority’ applicant solely because of race, is not narrowly tai-
    lored . . . .”); id. at 271-72 (“The only consideration that
    accompanies this distribution of points is a factual review of
    an application to determine whether an individual is a mem-
    ber of one of these minority groups.”).
      21
        In 2000-01, the District used a 10% deviation trigger, but increased
    the trigger to 15% for the 2001-02 school year.
                 PARENTS INVOLVED v. SEATTLE SCHOOL DIST.                 14747
       Yet the majority argues no quota exists here because the
    racial tiebreaker “does not set aside a fixed number of slots
    for nonwhite or white students,” nor is the 10 or 15% variance
    always satisfied (generally because there are insufficient num-
    bers of white or nonwhite students needed to balance the
    school). Majority op. at 14692.22 With respect, the majority
    misses the point. A quota does not become less of a quota
    because there are an insufficient number of whites or non-
    whites to fill the preselected spots. The District created a
    quota when it established the predetermined, preferred ratio of
    white and nonwhite students. In Bakke, the medical school
    argued that it did not operate a quota in its admissions system
    because it did not always fill the preselected seats; thus, its
    admissions system only had a “goal.” Justice Powell rejected
    that argument, stating that regardless of whether the preselec-
    ted seats were a “quota” or a “goal,” such a
    
           semantic distinction is beside the point: The special
           admissions program is undeniably a classification
           based on race and ethnic background. To the extent
           that there existed a pool of at least minimally quali-
           fied minority applicants to fill the 16 special admis-
           sions seats, white applicants could compete only for
           84 seats in the entering class, rather than the 100
           open to minority applicants. Whether this limitation
           is described as a quota or a goal, it is a line drawn
           on the basis of race and ethnic status.
      22
         Although the majority concludes a quota does not exist here, it also
    concludes “the rationale underlying the . . . prohibition of quotas does not
    apply” here. Majority op. at 14691 n.27. The majority reasons that because
    there is no competition in assignment to the District’s schools, the dangers
    presented by a quota—i.e., insulating applicants from competition on the
    basis of race—are absent here. Majority op. at 14691 n.27. But saying it
    does not make it so, whether it is said by the District or by the majority.
    As explained above, there is clearly a “market” for higher quality schools
    in the District, and there is competition for the schools the parents and stu-
    dents view to be the better schools.
    14748      PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    Bakke, 438 U.S. at 289 (Powell, J.).
    
       The majority makes a further attempt to avoid Grutter’s
    admonition against quotas by attempting to classify the Dis-
    trict’s predetermined ratio as a “critical mass.” The District’s
    preferred ratio could not be further from the definition of a
    “critical mass.” Grutter recognized that a “critical mass” had
    no quantified definition; instead, it was generally referred to
    as “meaningful numbers” or “meaningful representation” of
    minorities. 539 U.S. at 318. The Court expressly stated that a
    “critical mass” was not a means “simply to assure within its
    student body some specified percentage of a particular group
    merely because of its race or ethnic origin.” Id. at 329 (inter-
    nal quotation marks omitted).
    
       But unlike the unquantified “critical mass” from Grutter,
    the District’s preferred ratio is firmly set at 40% white, 60%
    nonwhite. When the 15% deviation trigger is used with the
    racial tiebreaker, the District seeks to enroll between 75% and
    45% nonwhite students and 25% to 55% of white students.
    The District’s admissions plan clearly seeks to assure a speci-
    fied percentage of white or nonwhite students in its schools;
    rather than seeking a “critical mass,” the District instead seeks
    racial balance. Thus, the District’s operation of the racial tie-
    breaker fails this factor as well.
    
                                   C.
    
      The third narrow-tailoring factor requires the District to
    have engaged in a “serious, good-faith consideration of work-
    able race-neutral alternatives.” See id. at 339. The majority
    concludes the District made such an effort. Majority op. at
    14698. For several reasons, I disagree.
    
       First, the District’s superintendent flatly admitted the Dis-
    trict did not engage in a serious, good-faith consideration of
    race-neutral alternatives. When asked whether the District
    “g[a]ve any serious consideration to the adoption of a plan for
                 PARENTS INVOLVED v. SEATTLE SCHOOL DIST.                14749
    the assignment of high school students that did not use racial
    balancing as a factor or goal,” the District’s superintendent
    stated: “I think the general answer to that question is no . . .
    I don’t remember a significant body of work being done. I
    mean it’s possible informally ideas were floated here or there,
    but I don’t remember any significant staff work being done.”
    [ER 521.]
    
       The record supports this concession. The District never
    asked its demographer to conduct any analysis regarding the
    effect of using a race-neutral lottery. [ER 483.] The District
    also never asked its demographer to conduct any analysis
    regarding a diversity program with non-racial indicia such as
    a student’s eligibility for free lunch or the students’s socio-
    economic background.23 [ER 481-82.]
       23
          The majority makes the conclusory statement that the District’s
    “white/nonwhite distinction is narrowly tailored to prioritize movement of
    students from the north of the city to the south of the city and vice versa”
    as an effort to combat Seattle’s racially imbalanced residential patterns.
    Majority op. at 14697-98. Yet the District’s attempt to balance students
    from north Seattle and south Seattle strongly suggests a less-restrictive,
    race-neutral approach to achieve such balancing: socioeconomic balanc-
    ing. As the majority notes, the northern Seattle area contains a majority of
    “white” students and is “historically more affluent.” Majority op. at
    14660. This would mean the southern Seattle area is less affluent. Thus,
    moving more affluent students south, and less affluent students north,
    could possibly provide a more diverse student body. At the very least, seri-
    ous consideration would have been warranted into this race-neutral alter-
    native. See Levine, supra, at 536 (noting the key element to successfully
    integrating students of different backgrounds and race is not racial bal-
    ance, but “economic integration”).
      Yet the majority accepts the District’s rejection of the use of socioeco-
    nomic factors, reasoning that “[a]lthough there was no formal study of the
    proposal by District staff, Board members’ testimony revealed two legiti-
    mate reasons” for rejecting the socioeconomic alternative: (1) “it is insult-
    ing to minorities and often inaccurate to assume that poverty correlates
    with minority status;” and (2) students would be reluctant to reveal their
    socioeconomic status to their peers. Majority op. at 14699-700. Such anal-
    ysis seems far from the “serious, good-faith consideration of workable
    14750        PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
       Also, in 2000, the Urban League of Metropolitan Seattle
    presented a high school assignment plan to the District. The
    plan proposed that each neighborhood region in Seattle would
    have a designated high school. Students would still be able to
    apply to any high school in Seattle, but when oversubscription
    occurred, students living in the designated “reference area”
    would first be assigned to their regional high school ahead of
    those who did not. To avoid racial concentration in the
    schools, the plan proposed “merit-based academic, avoca-
    tional and vocational magnet programs.” These programs
    “will help each school address racial diversity issues by
    encouraging students to travel outside of their communities to
    participate in a specific magnet program.”24
    
    race-neutral alternatives” demanded by Grutter. See 539 U.S. at 339. First,
    without formal studies (or indeed any earnest consideration of the alterna-
    tives), we have no way of knowing whether the District actually seriously
    considered, and rejected for valid reasons, less-restrictive race-neutral
    alternatives. In Croson, the Court emphasized the importance of a satisfac-
    tory record to determine whether race-neutral alternatives were consid-
    ered. See Croson, 488 U.S. at 498-511 (plurality) (detailing the
    government actor’s failure to document the basis for its use of a racial
    quota and stressing the need to do so). Second, the majority’s insistence
    that the District’s consideration of poverty would be “insulting” ignores
    the demeaning—and indeed, constitutionally objectionable—effect of
    placing persons into groups solely by their skin color for the purpose of
    receiving or being denied a governmental benefit. See Loving v. Virginia,
    
    388 U.S. 1
    , 11 (1967) (“[T]his Court has consistently repudiated distinc-
    tions between citizens solely because of their ancestry as being odious to
    a free people whose institutions are founded upon the doctrine of equali-
    ty.”). Even if a sole focus on poverty might be insulting to some minori-
    ties, socioeconomic considerations need not inquire only into poverty
    status; eligibility for free lunch, the parents’ levels of education, or
    whether English is a second language for the child are also relevant deter-
    minations in evaluating diversity. Third, there is no reason students would
    have to reveal their socioeconomic status to their peers; the District could,
    of course, keep such information confidential.
       24
          Similar race-neutral alternatives are common throughout the United
    States. For example, the San Francisco, California public school district
    employs a program focused on enhancing diversity in the classrooms. The
                 PARENTS INVOLVED v. SEATTLE SCHOOL DIST.                14751
       Despite the majority’s assertion, the record suggests the
    District did not seriously consider this plan. The District did
    not ask its demographer to conduct any analysis as to the
    effect or workability of the plan [ER 504]; one District board
    member stated the District “didn’t deal with” the plan [ER
    514]; another board member stated the District didn’t consider
    the plan [ER 643]; and last, another board member stated he
    refused to read the proposal because he would “rather play
    with my bass lunker fishing game.” [ER 573.]
    
       Of course, “[n]arrow tailoring does not require exhaustion
    of every conceivable race-neutral alternative,” Grutter, 539
    U.S. at 339, but it does require an earnest, good-faith consid-
    eration of the alternatives. Here, the District made no such
    attempt, and thus the District’s use of the racial tiebreaker
    fails this narrow-tailoring factor.25
    
    program allows students to choose any school within the district. When a
    school is oversubscribed, the program first assigns students with siblings
    to the same school, and then accommodates students with specialized
    learning needs. After that, the “Diversity Index” handles further assign-
    ments. “Under the Diversity Index process, the school district calculates
    a numerical profile of all student applicants. The current Diversity Index
    is composed of six binary factors: socioeconomic status, academic
    achievement status, mother’s educational background, language status,
    academic performance index, and home language.” David I. Levine, Pub-
    lic School Assignment Methods after Grutter and Gratz: The View from
    San Francisco, 30 Hastings Const. L.Q. 511, 528-31 (2003). Notably, the
    San Francisco system “does not use race as an express criterion for school
    assignments” and thus avoids the sharp focus of strict scrutiny. Id. at 531.
       25
          In assessing whether the District seriously considered race-neutral
    alternatives, the majority applies deference to the District’s consideration
    (or lack thereof) and rejection of the various alternatives. Majority op. at
    14698-99 n.33. With respect, the majority errs in two respects. First, as
    previously noted, deference to local officials’ use of race is generally
    barred in the application of strict scrutiny. See Johnson, 125 S. Ct. at 1146
    n.1. Second, if the majority is attempting to apply the deference invoked
    in Grutter, the Court there applied deference in determining whether the
    Law School asserted a compelling governmental interest, not whether the
    14752        PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
                                         D.
    
       The fourth narrow-tailoring factor requires that the Dis-
    trict’s use of the racial tiebreaker “must not unduly burden
    individuals who are not members of the favored racial and
    ethnic groups.” See Grutter, 539 U.S. at 341. The majority
    adjusts this test slightly to consider “any racial group,” rather
    than just members of the disfavored group. Majority op. at
    14684. Because the racial tiebreaker disadvantages both white
    and nonwhite children, I agree that the modification is valid.
    But unlike the majority, I conclude the District’s operation of
    the racial tiebreaker fails this factor as well.
    
       The racial tiebreaker unduly burdens thirteen- and fourteen-
    year-old school children by (1) depriving them of their choice
    of school, and (2) imposing on them tedious cross-town com-
    mutes, solely upon the basis of their race.
    
       First, as recognized above, the “good” schools in Seattle
    are a limited government benefit. Thus, the racial tiebreaker
    burdens white or nonwhite students, and often deprives them
    of the opportunity to enroll at what are considered the better
    schools, solely on the basis of race.
    
      Second, the children of plaintiff members Jill Kurfurst and
    Winnie Bachwitz were denied admission to Ballard High
    School based on their race and instead were forced to attend
    Ingraham, a school on the other side of Seattle from their
    home. To attend that school, the two white students faced a
    
    means used to achieve that interest were narrowly tailored. See 539 U.S.
    at 328 (“The Law School’s educational judgment that such diversity is
    essential to its educational mission is one to which we defer.”).
       The pattern now established by the majority seems suspicious. Out of
    five narrow-tailoring factors, the majority has concluded two are inappli-
    cable, and now a third is entitled to deference. I find it difficult to under-
    stand how such analysis could truly be considered strict scrutiny as to the
    narrowing requirement.
              PARENTS INVOLVED v. SEATTLE SCHOOL DIST.       14753
    daily multi-bus round-trip commute of over four hours. The
    parents instead enrolled their children in private schools.
    Those children were not only deprived of the school of their
    choice, they were effectively denied a public education
    (surely at much lower cost than private tuition), based on
    nothing but their race.
    
       A look at the operation of the tiebreaker provides further
    evidence of the injury the District inflicts on both white and
    nonwhite students. As noted by the majority, in the 2000-01
    school year, 89 more white students were assigned to Franklin
    than would have occurred absent the tiebreaker; 107 more
    nonwhite students were admitted to Ballard; 82 more non-
    white students were admitted to Roosevelt; and Twenty-seven
    more nonwhite students were admitted to Nathan Hale.
    Majority op. at 14667. To place the racial tiebreaker into
    proper perspective, in the 2000-01 school year, 89 nonwhite,
    minority students were denied admission to Franklin, and had
    to attend what to them was a less desirable school, solely
    because of their skin color. One hundred-seven white students
    were denied admission to Ballard, and had to attend what to
    them was a less desirable school, solely because of their skin
    color. Eighty-two white students were denied admission to
    Roosevelt, and had to attend what to them was a less desirable
    school, solely because of their skin color. Twenty-seven white
    students were denied admission to Nathan Hale, and had to
    attend what to them was a less desirable school, solely
    because of their skin color.
    
       Yet the majority discounts the burdens imposed by the
    racial tiebreaker, concluding that (1) the “minimal burden” of
    the tiebreaker is shared equally among white and nonwhite
    students; (2) no student is entitled to attend any specific
    school in any event; and (3) the tiebreaker does not uniformly
    benefit one race over the other because the tiebreaker operates
    against both whites and nonwhites. Majority op. at 14706-07.
    Regarding the first point, the U.S. Supreme Court has long
    rejected the notion that a racial classification which burdens
    14754      PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    races equally is any less objectionable under the Equal Protec-
    tion Clause. In Loving v. Virginia, 
    388 U.S. 1
     (1967), the U.S.
    Supreme Court held a Virginia statute criminalizing interra-
    cial marriages was unconstitutional under the Equal Protec-
    tion Clause. Id. at 12. The Court rejected the state’s argument
    that the miscegenation statute did not discriminate on the
    basis of race because it “punish[ed] equally both the white
    and the Negro participants in an interracial marriage.” Id. at
    8. The Court reasoned: “In the case at bar . . . we deal with
    statutes containing racial classifications, and the fact of equal
    application does not immunize the statute from the very heavy
    burden of justification which the Fourteenth Amendment has
    traditionally required of state statutes drawn according to
    race.” Id. at 9. Hence, it is irrelevant whether the racial tie-
    breaker disadvantages both races equally.
    
       Second, I think I have already disposed of the majority’s
    argument that no student is entitled to attend any specific Dis-
    trict school. The students and parents clearly value some of
    the District’s schools above the others, and limiting access to
    those higher quality schools on the basis of race is just the
    same as any other preferential racial classification.
    
       Third, I agree the tiebreaker does not uniformly benefit one
    race over the other and can exclude both white and nonwhite
    students from the preferred schools. Yet that does not lessen
    the injury of being subject to a racial classification. Equal pro-
    tection is an individual right, and whenever the District tells
    one student, whether white or nonwhite, he or she cannot
    attend a particular school on the basis of race, that action
    works an injury of constitutional proportion. See Adarand,
    515 U.S. at 230 (“[A]ny individual suffers an injury when he
    or she is disadvantaged by the government because of his or
    her race, whatever that race may be.”); Monterey Mech. Co.,
    125 F.3d at 712 (“Race discrimination is never a ‘trifle.’ ”).
    
       The District’s use of the racial tiebreaker thus unduly bur-
    dens members of the disfavored class, and the tiebreaker fails
    this narrow-tailoring factor as well.
               PARENTS INVOLVED v. SEATTLE SCHOOL DIST.        14755
                                   E.
    
       The fifth and final narrow-tailoring factor requires the Dis-
    trict’s use of the racial tiebreaker to “be limited in time,” and
    “have a logical end point.” See Grutter, 539 U.S. at 342. A
    workable “sunset” provision within any government-operated
    racial classification is vital:
    
        [A] core purpose of the Fourteenth Amendment was
        to do away with all governmentally imposed dis-
        crimination based on race. . . . The requirement that
        all race-conscious admissions programs have a ter-
        mination point assures all citizens that the deviation
        from the norm of equal treatment of all racial and
        ethnic groups is a temporary matter, a measure taken
        in the service of the goal of equality itself.
    
    Id. at 341-42 (internal quotation marks and alterations omit-
    ted).
    
       Citing Grutter, the majority contends the racial tiebreaker
    satisfies this factor because “this durational requirement can
    be met by periodic reviews to determine whether racial pref-
    erences are still necessary to achieve student body diversity,”
    and the District engages in such periodic reviews. Majority
    op. at 14706. Yet citing Grutter in full shows that “the dura-
    tional requirement can be met by sunset provisions in race-
    conscious admissions policies and periodic reviews to deter-
    mine whether racial preferences are still necessary to achieve
    student body diversity.” 539 U.S. at 342 (emphasis added).
    Periodic reviews are not enough; there must be some “dura-
    tional requirement,” some “logical end point,” to the racial
    classifications.
    
       The District argued the end point is in the “thermostat” to
    the tiebreaker, in which the District ceases to use the racial
    tiebreaker at any school for the year once its use had brought
    the school into racial balance. Yet it is undisputed that the
    14756       PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    District has never been segregated by law; the racial imbal-
    ance in its schools results from Seattle’s racially imbalanced
    housing patterns. If Seattle’s children were simply assigned to
    the high schools nearest their homes, those schools suppos-
    edly would tend to reflect such imbalance.
    
       Because there is no reason—much less evidence—to con-
    clude Seattle’s housing patterns will change, or that the Dis-
    trict’s student assignment program will affect such patterns, I
    must respectfully disagree that such a provision satisfies the
    “sunset provision” requirement enunciated in Grutter. Pre-
    sumably, where the District employs the racial tiebreaker, the
    schools will become racially balanced, that is 40% white,
    60% nonwhite (plus or minus a few percentage points,
    depending on the particular percentage deviation triggering
    the tiebreaker that year). Pursuant to the “thermostat,” the
    District would then stop using the racial tiebreaker. But
    because Seattle’s residential makeup is racially imbalanced26
    (and remains so despite the use of the racial tiebreaker),
    assignment to the oversubscribed schools would then occur
    only with use of (1) the sibling tiebreaker; and (2) the distance
    tiebreaker. Assuming that not every student also has a sibling
    attending one of the District’s schools, the schools will inevi-
    tably become racially imbalanced again because of the
    racially imbalanced residential makeup, thus rendering the
    thermostat useless as a “sunset provision.”
    
       One could argue, then, that this result supports the need for
    use of the racial tiebreaker. Not necessarily so. If the racial
    imbalance in the schools is caused not by the students, but by
    the choices of the parents as to where to live, then why not
    put the onus of remedying that imbalance on the parents
    rather than the students? Seattle’s city council could create
      26
        About 70% of the residents of Seattle, Washington are white, and 30%
    of the residents are nonwhite. Sixty-six percent of white students live in
    the northern part of Seattle, while 75% of nonwhite students live in the
    southern part of Seattle.
              PARENTS INVOLVED v. SEATTLE SCHOOL DIST.       14757
    “incentives” for whites to move into nonwhite areas, and for
    nonwhites to move into white areas. And if incentives do not
    accomplish the task, well, why not use compulsion, as the
    District does to high school children? The city council could
    take measures to prevent new persons taking up residence in
    Seattle from living in areas where their presence might other-
    wise alter the sought-after racial balance. This would protect
    the racial balance within the schools and squarely put the bur-
    den of remedying the racial imbalance upon the parents,
    rather than the students.
    
       Of course, less political resistance can be expected from
    choosing students for social engineering experiments in racial
    balancing, than in telling everyone—including voters—into
    which neighborhood they can move. Further, regulation of
    residence by race might run afoul of Shelley v. Kraemer, 
    334 U.S. 1
     (1948), although it is difficult to distinguish why the
    “compelling interest” of socialization among the races could
    not as easily be pressed in housing regulation as it is in
    schooling regulation.
    
      The simple truth is that some people choose to live near
    members of their own ethnic or racial group.
    
        There is no denying that American blacks often live
        in their own residential enclaves, especially in our
        big cities. But the same is true of whites and of every
        other racial and ethnic group—Jews, Chinese, Cam-
        bodians, Cubans, Arabs. Such racial and ethnic clus-
        tering means that a third of non-Asian minorities
        attend schools that are less than 10-percent white.
        And even though whites constitute just over 60 per-
        cent of the nation’s schoolchildren, the average
        white student goes to a school that is 80-percent
        white.
    
           But why should we expect identical proportions of
        blacks and whites to live in each and every neighbor-
    14758        PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
           hood? People like to live near others with whom
           they identify, and the schools mirror their choices.
           When asked about their residential preferences, only
           about 5 percent of blacks said they wished to live on
           an entirely or almost entirely white block. The vast
           majority preferred neighborhoods that were half or
           more than half African-American—in other words,
           neighborhoods in which the black concentration was
           “disproportionately” high. According to the 2000
           census, this happens to correspond closely to the
           actual distribution of black city-dwellers.
    
              In a complex, heterogenous society, it is only nat-
           ural that people should sort themselves out in urban
           space along lines of race as well as of religion and
           social class. This pattern was firmly established in
           the U.S. by the European immigrants who landed in
           the cities of the North in the 19th and early 20th cen-
           turies. The sociologists who studied these settle-
           ments recognized the important social functions
           served by “Little Italies” and “Poletowns.
    
    Abigail Thernstrom27 & Stephan Thernstrom, Have We Over-
    come?, Commentary, Nov. 2004, at 51-52.28
    
       Of course, the continuing racial imbalance in some residen-
    tial areas is in significant part a byproduct of past efforts to
      27
          Mrs. Thernstrom is presently the Vice Chair of the U.S. Commission
    on Civil Rights.
       28
          Further evidence that such self-selection results is submitted by this
    year’s Nobel Laureate, Thomas C. Schelling, by application of game the-
    ory in chapter four of his book Micromotives and Macrobehavior (1978).
    Schelling employs an exercise using coins to demonstrate how an inte-
    grated neighborhood can become largely segregated as long as each resi-
    dent desires at least one third of his or her neighbors to be of his or her
    race. When one person moves to get a preferred set of neighbors, it causes
    a chain reaction which settles down only when the neighborhood is effec-
    tively segregated.
              PARENTS INVOLVED v. SEATTLE SCHOOL DIST.      14759
    exclude minority groups from predominately white areas. Yet
    as racial tolerance and enforcement of civil rights laws have
    increased, neighborhoods are becoming more racially bal-
    anced. Id. In 1960, 15% of African-Americans lived in sub-
    urbs. In 2004, 36% live in suburbs. Id. African-Americans
    account for 9% of the total suburban population, “surprisingly
    close to proportionality for a group that constitutes only 12
    percent of the American population.” Id. Moreover, from
    1960 to 2000, the proportions of African-Americans living in
    census tracts that were over 80% black fell from 47% to under
    30%. Id. During that same period, the proportion residing in
    census tracts that were over 50% black fell from 70% to 50%.
    Id. Most importantly, this balancing takes place without any
    government coercion, except perhaps by the enforcement of
    fair housing laws which prevent racial discrimination such as
    California’s Unruh Civil Rights Act, Cal. Civ. Code § 51
    (West 2001).
    
        No one who understands what makes America great
        can quarrel with ethnic pride. At home, on the week-
        end, in the family and the neighborhood, Jews will
        be Jews, Italians Italian - and there is no reason
        blacks should be any different. Religion and ethnic-
        ity are essential parts of our lives, and government
        should not curtail how we express them in the pri-
        vate sphere. But when it comes to public life, even
        the benevolent color coding of recent decades has
        proved a recipe for alienation and resentment. Soci-
        ety need not be color-blind or color-less, but the law
        cannot work unless it is color-neutral, and the gov-
        ernment should not be in the business of abetting or
        paying for the cultivation of group identity.
    
    Schuck, supra, at 88 (quoting Tamar Jacoby, Someone Else’s
    House: America’s Unfinished Struggle for Integration 541
    (1998)) (internal alteration omitted).
    
      The racial imbalance in Seattle’s schools results not from
    de jure segregation nor from any invidious exclusion of non-
    14760        PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
    white minorities from the schools. Instead, it results from
    racially imbalanced residential housing patterns, an issue
    which the District does not even contend it can alter. Hence,
    the method chosen by the District to impose racially balanced
    schools is fatally flawed. Because it does not respond to the
    racial imbalances in Seattle’s residential makeup, and instead
    only attempts to fix it within the schools, there will be no sun-
    set to the use of the racial tiebreaker. See Grutter, 539 U.S.
    at 343 (“It would be a sad day indeed were America to
    become a quota-ridden society, with each identifiable minor-
    ity assigned proportional representation in every desirable
    walk of life. But that is not the rationale for programs of pref-
    erential treatment; the acid test of their justification will be
    their efficacy in eliminating the need for any racial or ethnic
    preferences at all.”). Thus, the District’s operation of the
    racial tiebreaker fails this factor as well.
    
                                        V.
    
      As pointed out in the majority opinion, other courts have
    concluded that a school district’s use of a racial tiebreaker in
    search of racial balance in the student body passes muster
    under the Equal Protection Clause.29 I respectfully disagree.
    The District’s use of the racial tiebreaker to achieve racial bal-
    ance in its high schools infringes upon each student’s right to
       29
          Cf. Comfort v. Lynn Sch. Comm., No. 03-2415, 
    2005 WL 1404464
    , at
    *1 (1st Cir. June 16, 2005) (en banc) (holding a public high school district
    had a compelling interest, in the absence of de jure segregation, in using
    race-based assignments to “secur[e] the educational benefits of racial
    diversity,” and the means used to serve that interest were narrowly tai-
    lored); McFarland v. Jefferson County Pub. Sch., 
    330 F. Supp. 2d 834
    ,
    850 (W.D. Ky. 2004) (holding a public high school district had a compel-
    ling interest in using race-based assignments to maintain racially inte-
    grated schools, and the means used to serve that interest were narrowly
    tailored), aff’d, No. 04-5897, 
    2005 WL 1693700
     (6th Cir. July 21, 2005);
    Brewer v. W. Irondequoit Central Sch. Dist., 
    212 F.3d 738
    , 752 (2d Cir.
    2000) (holding a public middle school district had a compelling interest,
    in the absence of de jure segregation, in using race-based assignments to
    reduce “racial isolation” in its schools).
               PARENTS INVOLVED v. SEATTLE SCHOOL DIST.       14761
    equal protection and tramples upon the unique and valuable
    nature of each individual. We are not different because of our
    skin color; we are different because each one of us is unique.
    That uniqueness incorporates our opinions, our background,
    our religion (or lack thereof), our thought, and our color.
    Grutter attempted to strike a balance between the individual
    protections of equal protection and being conscious of race
    even when looking at the individual. The District’s use of the
    racial tiebreaker, however, attempts no such balance; it
    instead classifies each ninth-grade student solely by race.
    Because of that, I must conclude such a program violates the
    Equal Protection Clause.
    
       The majority’s decision risks unfortunate repercussions. On
    the short-term, the specter of “white flight” (a recurring issue
    in the aftermath of the elimination of de jure desegregation)
    manifests itself here. The racial balancing of students will
    require busing and long-distance transportation to schools
    outside of some students’ neighborhoods. Parental involve-
    ment in those distant schools (such as with the PTA) will
    undoubtedly decrease. Parents who can afford private educa-
    tion (such as those in the more affluent northern part of Seat-
    tle) may very well choose to pull their children from the
    District schools and enroll them elsewhere, much like the
    Kurfurst and Bachwitz children. On the long-term, such an
    exodus could result in a decreased tax base and public support
    for the District schools and may result in the exact opposite
    the District hopes to achieve—a loss of white students from
    their school campuses.
    
       One of the greatest stains upon the history of our country
    is our struggle with race discrimination. Perhaps that stain
    would not be so deep had we chosen a different approach to
    our equal protection jurisprudence, an approach often-quoted:
    
        Our Constitution is color-blind, and neither knows
        nor tolerates classes among citizens. In respect of
        civil rights, all citizens are equal before the law. The
    14762      PARENTS INVOLVED v. SEATTLE SCHOOL DIST.
        humblest is the peer of the most powerful. The law
        regards man as man, and takes no account of his sur-
        roundings or of his color when his civil rights as
        guaranteed by the supreme law of the land are
        involved.
    
    Plessy v. Ferguson, 
    163 U.S. 537
    , 559 (1896) (Harlan, J., dis-
    senting).
    
      Or, as more recently said by the late Justice Stanley Mosk
    of the California Supreme Court:
    
        Racism will never disappear by employing devices
        of classifying people and of thus measuring their
        rights. Rather, wrote Professor Van Alstyne, ‘one
        gets beyond racism by getting beyond it now: by a
        complete, resolute, and credible commitment [n]ever
        to tolerate in one’s own life or in the life or practices
        of one’s government the differential treatment of
        other human beings by race. Indeed, that is the great
        lesson for government itself to teach: in all we do in
        life, whatever we do in life, to treat any person less
        well than another or to favor any more than another
        for being black or white or brown or red, is wrong.
        Let that be our fundamental law and we shall have
        a Constitution universally worth expounding.’
    
    Price v. Civil Serv. Comm., 
    604 P.2d 1365
    , 1391 (Cal. 1980)
    (Mosk, J., dissenting) (quoting William Van Alstyne, Rites of
    Passage: Race, the Supreme Court, and the Constitution, 46
    U. Chi. L. Rev. 775, 809-10 (1979)).
    
      The way to end racial discrimination is to stop discriminat-
    ing by race.
    
      For the reasons expressed above, I respectfully dissent and
    would reverse the judgment of the district court, holding the
    District’s use of the racial tiebreaker in its high school admis-
              PARENTS INVOLVED v. SEATTLE SCHOOL DIST.       14763
    sions program violates the equal protection rights of each stu-
    dent excluded from a particular school solely on the basis of
    that student’s race.