Opinion

In this decade-long school desegregation case, the trial court has issued an “amendment to mandate” in response to the remand of the case by the California Supreme Court. The San Bernardino City Unified School *72District has appealed. We reverse, because the trial court incorrectly followed the directions of the Supreme Court.

I.

Believing that the San Bernardino City Unified School District had not done enough to alleviate racial segregation and imbalance in its schools, the National Association for the Advancement of Colored People (NAACP) filed suit against the district in April 1972, on behalf of the minority students attending the district’s schools. After a two-stage trial, the trial court found that certain schools in the district were segregated and that “[vjoluntary desegregation [which is what the district had been utilizing] . . . cannot and will not achieve ultimate desegregation . . . .” The district was ordered to “proceed forthwith to eliminate or alleviate segregation . . . and to place into operation a plan reasonably calculated to do so by September of 1974. ”1

The Supreme Court reviewed the judgment in National Assn. for the Advancement of Colored People v. San Bernardino City Unified Sch. Dist. (1976) 17 Cal.3d 311 [130 Cal.Rptr. 744, 551 P.2d 48] (hereinafter San Bernardino). The court there held that the trial court had used an improper definition of “segregation,” but nonetheless affirmed the determination that segregation exists within the district.

Concerning the trial court’s error, the Supreme Court stated: “[T]he trial court used a number of alternative formulas in defining ‘segregation, ’ but all of the formulas identified a ‘segregated school’ on the basis of the school’s deviation from a ‘racial balance’ norm; after the trial court’s order was filed, our court upheld the constitutionality of the portion of Proposition 21 that repealed the ‘racial balance’ provisions upon which the trial court had relied. [2] (See Santa Barbara Sch. Dist. v. Superior Court (1975) 13 Cal.3d 315, 328-332 [118 Cal.Rptr. 637, 530 P.2d 605].) In light of this repeal, California school districts bear no statutory obligation to achieve racially balanced schools.

“Moreover, as we explain in Crawford [v. Board of Education (1976) 17 Cal.3d 280 (130 Cal.Rptr. 724, 551 P.2d 28)]: ‘[T]he constitutional mandate *73articulated in Jackson [v. Pasadena City School Dist. (1963) 59 Cal.2d 876 (31 Cal.Rptr. 606, 382 P.2d 878)] and reaffirmed today is not a constitutional command that each school in a district must reflect the racial composition of the district as a whole .... Our decisions, instead, require only that school districts take reasonable and feasible steps to eliminate segregated schools, i.e., schools in which the minority student enrollment is so disproportionate as realistically to isolate minority students from other students and thus deprive minority students of an integrated educational experience. ... It is such segregated schools which traditionally have resulted in the inherently unequal educational opportunities condemned in Brown [v. Board of Education (1954) 347 U.S. 483 [98 L.Ed. 873, 74 S.Ct. 686, 38 A.L.R.2d 1180]].’ (Ante, at pp. 302-303.)” (San Bernardino, supra, 17 Cal.3d at pp. 325-326, original italics, fn. omitted.)

The trial court was instructed to apply the “Crawford criteria” and “identify all the schools in the district to which the board’s constitutional duty [to alleviate segregation] relates.” (San Bernardino, supra, 17 Cal.3d at pp. 326-327.) In a footnote, the Supreme Court added the following caveat: “In making such a determination, the court should heed our caution in Crawford that ‘in determining whether a particular school is “segregated” for constitutional purposes, we do not believe set racial or ethnic percentages can be established, either in absolute terms or in terms of the racial composition of a particular district’s student population. . . . “[w]hat is or is not a segregated school will necessarily depend on the facts of each particular case. In addition to the racial and ethnic composition of a school’s student body, other factors, such as the racial and ethnic composition of faculty and staff[,] and the community and administration attitudes toward the school, must be taken into consideration.’” (Ante, at pp. 303-304.)” (San Bernardino, supra, 17 Cal.3d at p. 327, fh. 18.)

Despite the trial court’s use of an improper definition of “segregation,” the Supreme Court was able to uphold the determination that there were segregated schools in the district on the basis of “the court’s specific statistical findings.” (San Bernardino, supra, 17 Cal.3d at p. 326.) The Supreme Court stated, “it is evident. . . that, as of September 1972, a number of schools in the San Bernardino school district were ‘segregated’ within the meaning of the Crawford decision. Thus, in a district whose student population is 62.9 percent Caucasian and 37.1 percent minority, a number of schools had minority enrollments of close to 100 percent.” (Ibid.)

On remand, the trial court was asked to identify all of the segregated schools in the district and to “evaluate the adequacy of the district’s current desegregation plan in light of the guidelines set out in Crawford. ” (San Bernardino, *74supra, 17 Cal.3d at p. 327.) The judgment was then affirmed, except for the “segregation” definition, and the case was remanded for proceedings consistent with the Supreme Court opinion.

n.

Despite the clear map which the Supreme Court provided in its San Bernardino opinion, on remand the trial court and the parties struck out on the wrong course. Instead of identifying the segregated schools in the district and evaluating the adequacy of the district’s desegregation plan, the trial court held evidentiary hearings “to determine an appropriate definition of ‘segregation’ or racial isolation as it applied to the . . . district.” From the extensive evidence presented, the trial court found that “ethnic isolation and consequent harms” are “likely” to occur when minority students constitute more than 50 percent of a school’s student body and are “virtually certain” to occur when minority students constitute more than 70 percent of a school’s student body.3 The trial court concluded that when between 51 and 69 percent of a school’s student body were minority students, there was a rebuttable presumption that the school was segregated. When 70 percent or more of a school’s student body were minority students, the presumption of segregation was conclusive.

In conformance with its findings of fact and conclusions of law, the trial court ordered the district to evaluate each school with a minority student enrollment of between 50 and 70 percent “to determine if such school is a racially or minority isolated school, and if the students therein suffer, or are likely to suffer, the harms occasioned thereby.”4 The trial court further ordered that, for those schools which the district concluded were racially isolated and were causing minority children to suffer the harms of such isolation, the district submit “a plan to eliminate the isolation giving rise to those harms. ” For those schools with a minority student enrollment of over 70 percent, the court ordered the district to “immediately begin the development of a plan to reduce the minority enrollment. . . .”

*75m.

Two things are apparent about the trial court’s ruling after remand. First, the directions of the Supreme Court have not been followed. Second, the “definition” fashioned by the trial court is in conflict with the principles stated in the desegregation cases of the Supreme Court.

The Supreme Court expressly directed the trial court to, “applying the Crawford criteria, . . . identify all the schools in the district to which the board’s constitutional duty [to alleviate segregation] relates.” (San Bernardino, supra, 17 Cal.3d at pp. 326-327.) Except for ruling that all schools with a minority student body of over 70 percent are segregated, which ruling we later explain is erroneous, the trial court has not made the required identification. The trial court’s order has shifted the responsibility for identifying segregated schools to the district. In light of the language of the Supreme Court, it is clear that this duty was placed on the trial court and is nondelegable. What we believe the Supreme Court envisioned as the final product of the trial court in this regard is simply a list of the segregated schools in San Bernardino.5

The trial court also failed to “evaluate the adequacy of the district’s current desegregation plan . . . .” {San Bernardino, supra, 17 Cal.3d atp. 327.) The Supreme Court instructed, “Although from the present record the trial court may well conclude that the district’s desegregation plan is . . . inadequate under Crawford because it does not promise meaningful progress in the alleviation of segregated schools, we believe that the trial court should explicitly resolve this question on remand, with full knowledge of the Crawford decision.” {Ibid., final italics added.) The trial court has not made this explicit finding. Instead, the district has been ordered to submit and develop plans to eliminate segregation. Such an order should not have been made unless and un*76til the trial court evaluated the district’s current plan and expressly found it lacking.6

We also conclude that, when this case is once again remanded, the trial court should abandon its “definition” of segregation. As we have stated, the rebut-table presumption of segregation for schools with between 50 and 70 percent minority student bodies is improper because, in this case, the Supreme Court specifically instructed the trial court to identify San Bernardino’s segregated schools. Further, the rebuttable presumption, and the conclusive presumption of segregation for 70 percent or more minority schools, violate the Supreme Court’s command against establishing set racial or ethnic percentages in determining whether a school is segregated. Factors other than the racial composition of a school’s student body must be taken into account except in the most extreme circumstances.

As we noted earlier, the Supreme Court in this case did base its conclusion that a number of San Bernardino schools were segregated, solely on student enrollment data. These were schools with “minority enrollments of close to 100 percent” in a district with only a 37.1 percent minority student population.7 (San Bernardino, supra, 17 Cal.3d at p. 326.) The trial court apparently took this finding as an invitation to set a specific percentage for a conclusive presumption of segregation. We do not believe the Supreme Court intended this at all. The trial court has focused on an isolated finding in the San Bernardino opinion and has failed to recognize the pattern that has been developed throughout the Supreme Court’s desegregation opinions. The crux of those opinions is that student population should rarely be the only factor determining segregation, and that when student enrollment data is examined, precise numerical formulas should be avoided. A more comprehensive approach is required. These principles, announced so clearly in Crawford and San Bernardino, were recently reaffirmed in McKinny v. Board of Trustees (1982) 31 Cal.3d 79, 95 [181 Cal.Rptr. 549, 642 P.2d 460],

*77Sometimes, as the Supreme Court’s opinion in this case demonstrates, the racial imbalance in a school’s student population may be so great that segregation can be found regardless of other factors, but it remains imprudent to select a specific percentage as a dividing line. We may look at a 93-year-old man and, knowing no more than his age, properly conclude that he is “old”; but no one can accurately identify the exact day that he became so. So too a school of 1,000 students should not be automatically classified as segregated with 700 minority students if it might be considered not segregated with 699 minority students. This is an exercise in artificial and illusory precision.

A more fundamental reason why the trial court should no longer use its segregation definition is that the Supreme Court has already provided the applicable definition. Segregated schools are those “ ‘in which the minority student enrollment is so disproportionate as realistically to isolate minority students from other students and thus deprive minority students of an integrated educational experience. ’” (San Bernardino, supra, 17 Cal.3d at p. 326.) No further definition is necessary.8

On remand, we commend to the trial court the methodology adopted by a school board and approved by the Supreme Court in McKinny v. Board of Trustees, supra, 31 Cal.3d 79. There a 22-person advisory committee examined 5 criteria for each school and decided whether each criterion did or did not indicate segregation at that school.9 The committee then made an overall determination for each school of whether segregation existed. An evaluation by the trial court using similar criteria would be profitable in this case and would comport with the instructions of the Supreme Court in San Bernardino.10

*78The judgment is reversed and the case is remanded for proceedings consistent with this opinion and the Supreme Court opinion in this case.

Kaufman, J., and McDaniel, J., concurred.