[784]*784Opinion
BROUSSARD, J.
The community of Yorba Linda has its own elementary school district, but is part of the Fullerton Joint Union High School District (Fullerton HSD). The State Board of Education (State Board) approved a plan (the Plan) to create a new Yorba Linda Unified School District1 and transfer responsibility for high school education of Yorba Linda students from the Fullerton HSD to the new district. The State Board directed that the proposal be submitted for approval in an election limited to the residents of Yorba Linda.
Fullerton HSD petitioned for mandate to prevent the election. The trial court upheld portions of the State Board’s decision, but held invalid other portions, including the limitation of the franchise to Yorba Linda residents, and accordingly barred the election. Both parties appealed.
The appeal presents three issues: (1) Did the State Board comply with the requirements of Education Code section 4200, which specifies the conditions which must be met before the State Board authorizes the creation of a new school district? (2) Is the State Board’s decision invalid for noncompliance with the California Environmental Quality Act (Pub. Resources Code, § 21050 et seq.)? (3) Does the State Board’s decision limiting the vote to Yorba Linda residents deny the equal protection of the laws to other residents of the Fullerton HSD?
With respect to the first issue, the parties agree that a reviewing court may inquire only whether the State Board’s decision was arbitrary, capricious, or entirely lacking in evidentary support; applying that standard of review, we find that the State Board’s decision complied with section 4200. We conclude, however, that the State Board’s failure to undertake a threshold environmental study violated the California Environmental Quality Act, and that its limitation of the electoral franchise was unconstitutional.
1. Statement of facts.
The Yorba Linda Elementary School District is completely surrounded by two unified school districts: Brea-Olinda to the north, and Placentia to the west, south and east. The territory included in the Yorba Linda Elementary School District is also a part of Fullerton HSD, but being surrounded by two other districts, it is not contiguous to the remainder of Fullerton HSD. Because there is no high school in Yorba Linda approximately 1,200 Yorba Linda high school [785]*785students must be bused 5 to 7 miles, across a portion of Placentia Unified School District, to Troy High School, the nearest high school in the Fullerton HSD system.
The isolation of the Yorba Linda portion of Fullerton HSD from the remainder of the district created a number of problems. High school students had to make a lengthy round trip each day to and from school. The distance also imposed burdens on students who wanted to participate in extracurricular ac- ' tivities. In view of these facts, the Orange County Committee on School District Organization (County Committee) prepared the Plan to create a Yorba Linda Unified High School District. In effect, the Yorba Linda portion of Fullerton HSD would “secede” and become unified with Yorba Linda Elementary.
Section 4200 of the Education Code2 requires that a school organization plan must substantially meet the following conditions; (a) the new districts will have adequate enrollment; (b) the new districts will be adequate in terms of financial ability; (c) the new districts will each have a substantial community identity; (d) the proposal will result in an equitable division of property and facilites of the original district; and (e) the formation of the new district will not promote racial or ethnic discrimination or segregation.
The County Committee found that all statutory requirements were met and submitted the Plan to the State Board for approval. In addition, the County Committee determined that the election be held in the area proposed for unification only (Yorba Linda), and not throughout the entire Fullerton HSD.
Before approving the Plan, the State Board must find substantial compliance with the statutory requirements of section 4200, and it must also review the propriety of the County Committee’s designation of the territory in which the election will be held. To assist the State Board, the State Department of Education submitted a report which analyzed each of the statutory conditions, found substantial compliance, and approved limiting the election to the Yorba Linda area. At its first hearing the State Board requested further information on the racial impact of removing Yorba Linda from the Fullerton HSD, but after considering that information it voted unanimously to approve the Plan.
Fullerton HSD, which opposed the unification proposal, filed a petition for writ of mandate in the superior court, contending that the Plan did not meet the statutory criteria and that the election was improperly restricted to Yorba Linda. The trial court ruled that the Plan substantially met the requirements of ade[786]*786quate enrollment, adequate financial ability, community identity, and equitable division of the property. However, the trial court held that the State Board had abused its discretion and acted arbitrarily and capriciously because (1) the Plan did not intend to promote racial or ethnic segregation or discrimination; (2) the State Board did not comply with the California Environmental Quality Act; and (3) for a variety of reasons, the voters of the entire Fullerton HSD should vote on the Plan, and their exclusion from the election violated their fundamental right to vote.
2. Compliance with section 4200.
The parties initially agree that the action in the trial court was a traditional mandate proceeding pursuant to Code of Civil Procedure section 1085 and not an administrative mandamus pursuant to Code of Civil Procedure section 1094.5. They further agree that the exercise by the State Board of the authority to approve proposals for the formation of unified school districts pursuant to the legislative directive of sections 4200-4419 is a “quasi-legislative” act.3
In reviewing such quasi-legislative decisions, the trial court does not inquire whether, if it had power to act in the first instance, it would have taken the action taken by the administrative agency. The authority of the court is limited to determining whether the decision of the agency was arbitrary, capricious, entirely lacking in evidentiary support, or unlawfully or procedurally unfair. (Pitts v. Perluss (1962) 58 Cal.2d 824, 833 [27 Cal.Rptr. 19, 377 P.2d 83]; Brock v. Superior Court (1952) 109 Cal.App.2d 594, 605-607 [241 P.2d 283].)
Applying this test of review, we examine the State Board’s finding that the Plan substantially complies with the requirements of section 4200. That section provides in relevant part as follows:
“The State Board of Education may approve proposals for the formation of districts based upon the division of the territory of existing high school districts, provided that the board has determined, with respect to the proposal and the resulting new districts that the following conditions are substantially met:
[787]*787“(a) That the new districts will be adequate in terms of number of pupils enrolled.
“(b) That the new districts will be adequate in terms of financial ability. For purposes of determining financial ability, consideration shall be given to revenue limits per pupil, assessed valuation per pupil, and tax rates. . . .
“(c) That the new districts are each organized on the basis of a substantial community identity.
“(d) That the proposal will result in an equitable division of property and facilities of the original district.
“(e) That the proposal and the formation of the new districts will not promote racial or ethnic discrimination or segregation. ...”
As we have noted, the State Board found that all five criteria had been substantially met. The trial court ruled that the last condition, i.e., that the Plan must not promote racial segregation, was not met. On appeal, the State Board contends that the Plan does not promote racial segregation, while Fullerton HSD contends that it does not comply with the conditions of adequate enrollment, financial ability and equitable division of the property.
(a.) Specific Findings
Preliminarily, Fullerton HSD argues that the State Board action was invalid because it did not specifically state how the Plan substantially complied with the conditions enumerated in section 4200, subdivisions (a)-(e). Fullerton HSD relies on Code of Civil Procedure section 1094.5, the administrative mandamus provision, which impliedly requires the administrative agency in adjudicatory decisions to make some findings “to bridge the analytic gap between the raw evidence and ultimate decision or order.” (Topanga Assn. for a Scenic Community v. County of Los Angeles (1974) 11 Cal.3d 506, 515 [113 Cal.Rptr. 836, 522 P.2d 12].) However, as we have stated, in approving the Plan the State Board was exercising a “quasi-legislative” function reviewable by traditional mandamus. (Code Civ. Proc., § 1085.) Because the decision approving the Plan to create Yorba Linda Unified School District was not an adjudicatory decision made as the result of a proceeding in which by law a hearing is required to be given and evidence required to be taken, Code of Civil Procedure section 1094.5 and cases dealing with such adjudicatory hearings are inapplicable.
[788]*788Furthermore, even where findings are required, the administrative agency is not required to make formal findings as would be required in a court of law. An administrative agency may adopt the findings prepared by another body. (Carmel Valley View, Ltd. v. Board of Supervisors (1976) 58 Cal.App.3d 817, 823 [130 Cal.Rptr. 249].) The State Board in effect adopted the findings of the County Committee as contained in the proposed Plan.4 These findings are fully adequate to allow meaningful judicial review of the quasi-legislative action of the State Board for the purpose of determining whether the action was arbitrary, capricious or without evidentiary support.
(b.) Adequate Enrollment
Fullerton HSD contends that the State Board acted arbitrarily in finding that the proposed Yorba Linda Unified School District would be adequate in terms of number of pupils enrolled.
Section 4200, subdivision (a) does not itself specify what is meant by “adequate in terms of number of pupils enrolled.” Pursuant to statutory mandate, the State Board adopted regulations (Cal. Admin. Code, tit. 5, §§ 18570-18574) to serve as guidelines for determining whether a proposal complies with section 4200.5 Those guidelines advise that each proposed school district will, at the time the proposal becomes effective, be adequate if it substantially complies with the following enrollment standards;
Elementary District.............................................................3,500
High School District............................................................1,500
Unified District..................................................................5,000
(Cal. Admin. Code, tit. 5, § 18573.) The Plan states that attendance in Yorba Linda during the 1976-1977 school year was 2,511 in grades K-8, and 1,200 in grades 9-12, for a total of 3,711.
Fullerton HSD objects to the 3,711 figure because it refers to enrollment figures at the time the Plan was approved and not at the time the proposal becomes effective. Consequently, Fullerton HSD argues, the Plan is defective [789]*789because it does not show that the proposed Yorba Linda Unified School District would have an enrollment of 5,000 in 1979-1980 when the Plan was intended to take effect. The evidence in fact shows that projected enrollment for 1979-1980 would be approximately 2,050 in grades K-8 and approximately 1,050 in grades 9-12, for a total enrollment of 3,100; Fullerton HSD urges that 3,100 is not substantial compliance with the guideline of 5,000.
Although the projected enrollment is less than specified in the guidelines, the State Board may well have determined, in the exercise of its discretion, that the relatively low enrollment should not deprive the Yorba Linda area residents of the opportunity to form a unified school district. Yorba Linda Elementary, although somewhat under-enrolled, already existed as a separate district. In view of the special geographic and community factors and of the fact that the relatively low enrollment is largely due to the preexisting elementary school district, we cannot say that the State Board acted arbitrarily in finding that the projected enrollment substantially met the guidelines for adequate enrollment.6
(c.) Financial Ability
At the time the Plan was approved, California Administrative Code, title 5, section 18573, subdivision (b)(2) provided that “Each district affected will be adequate in terms of financial ability if either the revenue limit per unit of average daily attendance of the proposed district does not vary from the revenue limit per unit of average daily attendance on the assessed valuation per unit of average daily attendance in all of the affected districts by more than 15%.”7
Fullerton HSD argues that both the revenue limit per unit of average daily attendance (ADA) and the assessed valuation per ADA for the proposed Yorba Linda Unified School District vary by more than 15 percent from those for Fullerton HSD. The Plan shows that the assessed value per ADA varies by 29.42 percent. It also shows that the revenue limit per ADA deviates by only 5.27 percent, but Fullerton HSD points out that incorrect figures were used to arrive at that conclusion. It was inaccurate to determine the deviation of Yorba [790]*790Linda Unified School District’s revenue limit from the average revenue limit of Fullerton HSD and the elementary school districts within its boundaries. The deviation of the proposed Yorba Linda Unified School District revenue limit from the Fullerton HSD revenue limit—excluding the elementary school districts—in fact is 17.19 percent.8
[789]*789“(2) Each district affected will be adequate in terms of financial ability if:
“(A) The revenue limit per unit of average daily attendance of the proposed district does not vary from the revenue limit per unit of average daily attendance in all of the affected districts by more than 15 %, or
“(B) The proposal does not increase costs to the State for the affected territory by more than 10%.”
[790]*790The issue before us, however, is not whether the Plan conforms to the administrative guidelines. It is whether the State Board acted arbitrarily, capriciously, or without evidentiary support in concluding that the Plan substantially complied with the statutory requirements. Despite the deviation from the figures set by the administrative regulations, the record does not demonstrate that the State Board acted arbitrarily, capriciously, or without evidentiary support when it found that the proposed Yorba Linda Unified School District would have adequate financial ability to finance its educational program.9
(d.) Division of Property
The Plan proposes to appraise all real property, improvements, personal property and funds of Fullerton HSD and to divide the value of all such property on the basis of ADA ratio between the Yorba Linda portion and the remaining portion of Fullerton HSD. The real property and improvements will remain in that portion of the district in which they are located, but a monetary credit will be given to a portion of the district if the property located there was less than its proportional share based on ADA ratio. The Plan further divides bonded indebtedness of Fullerton HSD according to ADA ratio, but provides that one district may choose to accept a greater portion of the bonded indebtedness to offset charges against that district for an imbalance in real property, improvements or personal property received by that district.
Fullerton HSD complains that the Plan proposes an inequitable distribution of property in violation of section 4200, subdivision (d). Since section 4123 specifies that real property shall remain the property of the district in which it is located, it argues that the State Board should not [791]*791take into account the value of real property. Section 4370, however, authorizes the County Committee in drafting the proposal for division of the property and funds to consider “the value and location of the school property.” Moreover, section 4200 mandates an equitable division of the property and facilities of Fullerton HSD. If section 4123 were read to preclude any consideration of the value of the real property, the State Board would necessarily fail in its attempt to divide the property equitably.
Yorba Linda Unified School District, under the interpretation urged by Fullerton HSD, would receive the one parcel of land owned by Fullerton HSD located in the Yorba Linda area and a 7.8 percent share of funds. All other real property, all the improvements, and all the personal property of the district would remain with Fullerton HSD. Such a distribution would be manifestly inequitable. Thus, the State Board did not act arbitrarily or capriciously in approving the Plan’s provision to divide the values of the property, even though the property itself remains in the portion of the district where it is located.
As to the bonded indebtedness, the Plan provides that one portion of the district may opt to accept a greater share of the bonded indebtedness than its proportional share in order to offset the amount owed, if any, on account of the division of the value of the other assets. Fullerton HSD contends that this provision of the Plan violates article XVI, section 18, of the California Constitution, which provides in pertinent part that: “No . . . board of education, or school district, shall incur any indebtedness . . . without the assent of two-thirds of the qualified electors thereof, voting at an election to be held for that purpose, ...”
The answer to this argument is that the total amount of bonded indebtedness has already been voted on by the qualified voters of Fullerton HSD. By a change of boundaries of a school district, the district does not lose its identity or cease to be the same legal entity it was before. (Hughes v. Ewing, supra, 93 Cal. 414, 419; see also Bates v. Gregory (1891) 89 Cal. 387, 395 [26 P. 891].) In any event, the Plan does not require that a nonpro rata distribution be made; Fullerton HSD is free to accept its proportional share of the bonded indebtedness only.
Finally, Fullerton HSD argues that the Plan does not provide for an equitable division of the property because the distribution of cash equivalent to Yorba Linda for Yorba Linda’s proportional share of the real property and improvements would financially impoverish Fullerton HSD. Yorba Linda’s proportional share of the property (7.8 percent based on ADA ratios) is estimated at $4 million. Fullerton HSD argues that, even if it assumed all the bonded in-[792]*792debtedness and gave Yorba Linda all the personal property, the balance owed to Yorba Linda would wipe out its liquid assets.
On the present record, the actual effect of the property division is quite uncertain, depending upon the values eventually assigned to the real property acquired by the new Yorba Linda District as against the value of that retained by the Fullerton HSD. If it turns out that the division imposes substantial burdens on Fullerton HSD, the two districts are free to make arrangements for payment that take account of the economic realities. All that the State Board did was to divide the property so that Yorba Linda would receive its proportional share of the district assets, and such a division clearly complies with the requirement of section 4200, subdivision (d), that the property be divided equitably.
(e.) Racial or Ethnic Segregation
Section 4200, subdivision (e), provides that a school reorganization plan must “not promote racial or ethnic discrimination or segregation.” The State Board determined that the proposal at issue in this case met this criterion, but the trial court ruled to the contrary, finding that the Plan would promote segregation “by removing an area of primarily white students from an area that is increasingly becoming more integrated.”
Statistics presented to the State Board and the trial court showed that in the 1976-1977 school year the Fullerton HSD had a white enrollment of 84.6 percent, and that the proposed Yorba Linda unified district would have a white enrollment of 91.6 percent. Fullerton HSD presented testimony predicting that minority enrollment in most of the Fullerton HSD would steadily increase, but that the racial and ethnic composition of Yorba Linda would remain relatively stable. Additional evidence presented to this court verified this prediction; as of November 1980 the white percentage in the Fullerton HSD as a whole has declined to 80 percent, but that of Yorba Linda only to 90.9 percent.10 If Yorba Linda seceded from the Fullerton HSD, the remaining portion of the district would have a white percentage of 66 percent.
The statutory language, that the proposal must not “promote racial or ethnic discrimination or segregation,” has not received judicial definition. We believe it imposes on the State Board a more stringent obligation than its constitutional duty—a duty which would exist in the absence of section 4200—not to approve [793]*793the creation of a district with segregated schools.11 The State Board should also inquire into the effect of a secession Plan upon the ability of both the new district and the remaining portion of the original district to adapt to the present and anticipated racial and ethnic composition of its students, and to implement voluntary integration programs (cf. Crawford v. Board of Education, supra, 17 Cal.3d 280, 294; San Francisco Unified School Dist. v. Johnson (1971) 3 Cal.3d 937, 951-952 [92 Cal.Rptr. 309, 479 P.2d 669]).
Measured against this view of the statutory purpose of section 4200, the decision of the State Board is questionable. The Plan will carve out from an increasingly integrated district a small enclave which is, and will remain, 90 percent white.12 Although we do not question the motives of those who propose the Plan, many in the area may see it as an attempt to isolate the white students of Yorba Linda from contact with a significant number of minority students. Any future efforts of Yorba Linda school district to obtain a more racially balanced student body, and to avoid community identification as a “white” high school, will likely be frustrated by small size and stable demography of the area. Moreover, the removal of the stable white enrollment from Yorba Linda will make it more difficult for Fullerton HSD to adapt to a changing racial and ethnic population, which difficulty may eventually result in some schools in that district becoming predominantly minority schools.
[794]*794As we have said, however, the issue before us is limited to whether the State Board acted arbitrarily, capriciously, or without evidentiary support. This standard severely limits the scope of our review. The State Board’s finding is supported by evidence that both Yorba Linda and the remaining portion of the Fullerton HSD are presently predominantly white; if we look to the present composition of the districts and dismiss future changes as speculative, we could not classify the State Board’s decision as arbitrary or capricious. We therefore uphold the State Board’s finding on this issue.
3. Compliance with the California Environmental Quality Act.
Under the California Environmental Quality Act (CEQA) (Pub. Resources Code, § 21050 et seq.), any public agency directly undertaking a project which may have a significant effect on the environment must first conduct a threshold study of such impact. If the study shows that the project will not have a significant effect, the agency may so declare in a brief negative declaration; if it demonstrates that the project may have a significant effect, the agency must prepare an environment impact report. (No Oil, Inc. v. City of Los Angeles (1974) 13 Cal.3d 68, 74 [118 Cal.Rptr. 34, 529 P.2d 66].)
Implementation of the secession Plan in the present case involves the possibility of a significant impact. Secession will likely require the construction of a new high school in Yorba Linda and may result in abandonment of some facilities in the remaining portion of the Fullerton HSD.13 It will change bus routes and schedules, and affect traffic patterns. Although it is uncertain whether the total impact will.be significant enough to require an environmental impact report, it is clear that it is sufficient to require at least an initial study to inquire into the need for such a report.
The State Board, however, ruled that it had no duty to undertake such a study before approving the Plan. Judicial review of its ruling is governed by Public Resources Code section 21168.5, which limits judicial inquiry into “whether there was a prejudicial abuse of discretion,” and provides that “[a]buse of discretion is established if the agency has not proceeded in a manner required by law or if the determination or decision is not supported by substantial evidence.”14 In the present case, the principal controversy concerns whether [795]*795the State Board’s approval of the Plan constitutes a “project” within the purview of CEQA.15 This is an issue of law which can be decided on undisputed data in the record on appeal; thus the present appeal presents no question of deference to agency discretion or review of substantiality of evidence.
We explained CEQA’s concept of a “project” requiring an environmental study in Bozung v. Local Agency Formation Com. (1975) 13 Cal.3d 263 [118 Cal.Rptr. 249, 529 P.2d 1017]. Bozung was concerned with a Local Agency Formation Commission (LAFCO) decision to approve an annexation proposal. The commission argued that although the development of the land following annexation might have an environmental effect, the mere approval of the proposal had no such effect. We explained, however, that “[t]he notion that the project itself must directly have such an effect was effectively scotched in Friends of Mammoth. [Friends of Mammoth v. Board of Supervisors (1972) 8 Cal.3d 247 (104 Cal.Rptr. 761, 502, P.2d 1049).] The granting of a conditional use permit—a piece of paper—does not directly affect the environment any more than an annexation approval—another piece of paper. Friends of Mammoth, of course, said that the word “project” appears to emphasize activities culminating in physical changes to the environment, . . .’ (Id., at p. 265. Italics added.) In response to that concept, the Guidelines refer to ‘physical impact on the environment, directly or ultimately. ’ (Cal. Admin. Code, tit. 14, § 15037. Italics added.)” (13 Cal.3d at p. 279.) We then held in Bozung that approval of the annexation—a necessary step in a chain of events which would culminate in physical impact on the environment—required an environmental impact report.
The State Board, consequently, cannot argue that its approval of the secession Plan is not a project merely because further decisions must be made before schools are actually constructed, bus routes changed, and pupils reassigned. It does, however, argue that its approval is not a project because it simply submits the issue to the voters, and because the Plan as approved is not sufficiently definite to allow an adequate environmental study.
[796]*796The State Board first points to California Administative Code, title 14, section 15037, which in subdivision (b) states that “[p]roject does not include . . . (4) The submittal of proposals to a vote of the people of the State or of a particular community.”
It is clear, however, that Board’s approval of the Plan is not exempt from CEQA merely because that approval must be ratified by the voters. As the court explained in People ex rel. Younger v. Local Agency Formation Com. (1978) 81 Cal.App.3d 464 [146 Cal.Rptr. 400], in holding LAFCO approval of a deannexation proposal was subject to CEQA, “[t]he ‘project’ here is more than the ‘submittal of proposals to a vote of the people. ’ Rather, it is but the first step required by statute on a deannexation proposal with consequent substantial impact on the physical and human environment.” (81 Cal.App.3d at p. 479.)
The State Board also relies on Simi Valley Recreation & Park Dist. v. Local Agency Formation Com. (1975) 51 Cal.App.3d 648 [124 Cal.Rptr. 635], which concerned LAFCO approval of an election to detach an area from a recreation and park district. Noting that the asserted environmental impact was simply the replacement of one group of managers by others who might hold different views on the future use of the land in question, the court held that under these circumstances the LAFCO approval was not a “project” under CEQA. Distinguishing Bozung v. Local Agency Formation Com., supra, 13 Cal.3d 263, Simi Valley stated that the earlier case “dealt only with the situation where LAFCO approval was a necessary step in the development and in effect constituted an entitlement for use for such development. . . . [f] The evaluation process contemplated by CEQA relates to the effect of proposed changes in the physical world which a public agency is about to either make, authorize or fund, not to every change of organization or personnel which may affect future determinations relating to the environment.” (51 Cal.App.3d at pp. 665-666.)16
The distinction set out in Simi Valley is not between approval of a proposal which requires an election and approval of a proposal which does not. It is be[797]*797tween governmental approval which constitutes an essential step culminating in action which may affect the environment (Bozung) and approval of a reorganization which portends no particular action affecting the environment (Simi Valley).
The present case more closely resembles Bozung v. Local Agency Formation Com., supra, 13 Cal.3d 263. The anticipated environmental impact here stems from the construction of a new high school in Yorba Linda and consequent changes in the remaining portion of the Fullerton HSD. Although it is conceivable that Fullerton HSD itself could build a high school in Yorba Linda, it is apparent that it does not intend to do so; the future Yorba Linda Unified School District, on the other hand, must construct such a facility. Thus, as a practical matter State Board approval of the secession Plan is an essential step leading to ultimate environmental impact; it is therefore under the reasoning of Bozung and Simi Valley a “project” within the scope of CEQA. (See People ex rel. Younger v. Local Agency Formation Com., supra, 81 Cal.App.3d 464, 478.)
A closer question is presented by the State Board’s contention that an environmental study before the election would be premature. The timing of an environmental study can present a delicate problem. “ ‘Statements must be written late enough in the development process to contain meaningful information, but they must be written early enough so that whatever information is contained can practically serve as an input into the decision making process.’ ” (No Oil, Inc. v. City of Los Angeles, supra, 13 Cal.3d 68, 77 fn. 5, quoting Scientists’ Inst, for Pub. Info., Inc. v. Atomic Energy Com’n. (D.C.Cir. 1973) 481 F.2d 1079, 1094.)
Specific plans have not yet been formulated for construction of a new high school in Yorba Linda or for changes in the education program in the remaining portion of the Fullerton HSD. Thus delay of an environmental study until after the election might result in a more specific and useful study. The problem with such a delay, however, is that as a practical matter it precludes the alternative of continuing the status quo. Once the voters approve the secession Plan the new Yorba Linda Unified School District will have to build a high school, and Fullerton HSD will have to adjust to the loss of the Yorba Linda students.
The fundamental purpose of CEQA is to ensure “that environmental considerations play a significant role in governmental decision-making” (Friends of Mammoth v. Board of Supervisors (1972) 8 Cal.3d 247, 263 [104 Cal.Rptr. 761, 502 P.2d 1049]). Consequently, it is desirable that environmental information be furnished the decision-maker “at the earliest possible stage’’ (Bozung v. Local Agency Formation Com., supra, 13 Cal.3d 263, 282). [798]*798In the present setting, the State Board and the voters are the decision-makers; they must decide whether to approve the proposed secession, an approval which necessarily entails building a new high school and other actions which may have an environmental effect. In making that decision, the State Board and the voters should have the benefit of relevant environmental data and analysis. (See People ex rel. Younger v. Local Agency Formation Com., supra, 81 Cal.App.3d 464,481.)17 We conclude that the initial environmental study of the effect of detaching Yorba Linda from the Fullerton HSD should have been undertaken before the State Board approved the Plan and submitted it to the voters, and that in failing to undertake that study the State board violated the requirements of CEQA.18
4. Limitation of the election to residents of Yorba Linda.
Fullerton HSD contends that the decision of the State Board to approve limiting the election to residents of Yorba Linda denies the equal protection of the laws to the other residents of Fullerton HSD.
The first step in evaluating this contention is to determine the applicable level of judicial review.19 In Curtis v. Board of Supervisors (1972) 7 Cal.3d 942 [104 Cal.Rptr. 297, 501 P.2d 537], a voting rights case similar in many respects to the present case, we explained that: “[T]his court and the United States Supreme Court apply a two-level test. [Citation.] In the typical equal protection [799]*799case the classification need only bear a rational relationship to a conceivable legitimate state purpose; ‘[on] the other hand, in cases involving “suspect classifications” or touching on “fundamental interest,” ... the court has adopted an attitude of active and critical analysis, subjecting the classification to strict scrutiny. [Citations omitted.] Under the strict standard applied in such cases, the state bears the burden of establishing not only that it has a compelling interest which justifies the law but that the distinctions drawn by the law are necessary to further its purpose.’” (7 Cal.3d at pp. 951-952, quoting Westbrook v. Mihaly (1970) 2 Cal.3d 765, 784-785 [87 Cal.Rptr. 839, 471 P.2d 487].)
In Hawn v. County of Ventura (1977) 73 Cal.App.3d 1009, 1019-1020 [141 Cal.Rptr. 111], the Court of Appeal explained the application of these principles to state action limiting the franchise: “[T]he strict standard of review has long been held to apply to voting legislation which excludes certain potential voters from participation. (Kramer v. Union School District (1969) 395 U.S. 621 [23 L.Ed.2d 583, 89 S.Ct. 1886]; Cipriano v. City of Houma (1969) 395 U.S. 701 [23 L.Ed.2d 647, 89 S.Ct. 1897]; Curtis v. Board, of Supervisors [supra] 1 Cal.3d 942.) There has been a subsequent refinement of this general principle as it applies to analysis of legislative voting classifications: an exclusion of voters invoking strict constitutional scrutiny has been held to mean an ‘identifiable class’ of voters (Weber v. City Council (1973) 9 Cal.3d 950 [109 Cal.Rptr. 553, 513 P.2d 601]; Gordon v. Lance (1971) 403 U.S. 1 [29 L.Ed.2d 273, 91 S.Ct. 1889]) and, ‘[although not every classification created ... is subject to strict scrutiny, the “compelling interest” measure must be applied if a classification has a “real and appreciable impact” upon the equality, fairness and integrity of the electoral process.’ (Choudhry v. Free (1976) 17 Cal.3d 660, 664 [131 Cal.Rptr. 654, 552 P.2d 438]; see Bullock v. Carter (1972) 405 U.S. 134, 144 [31 L.Ed.2d 92, 92 S.Ct. 849].) For a legislative classification relating to the elective process to avoid the strict scrutiny test of equal protection, it must have ‘only minimal, if any, effect on the fundamental right to vote.’ (Gould [v. Grubb (1975)] 14 Cal.3d 661, 670 [122 Cal.Rptr. 377, 536 P.2d 1337].)” It is important to note that it is the impact of the classification on the electoral process that triggers strict scrutiny. That heightened mode of analysis is not limited to cases involving suspect or invidious classifications.
It is clear that the classification in the present case does not involve a mere incidental or marginal effect on a fundamental right (compare Califano v. Jobst (1977) 434 U.S. 47, 53-54 [54 L.Ed.2d 228, 234-235, 98 S.Ct. 95]; In re Flodilin (1979) 25 Cal.3d 561, 568 [159 Cal.Rptr. 327, 601 P.2d 559]); it excludes all residents of Fullerton HSD outside Yorba Linda from voting on the proposed measure, an exclusion which may well affect the outcome of the election. The case is thus analogous to those Supreme Court decisions (Phoenix v. [800]*800Kolodziesjki (1970) 399 U.S. 204 [26 L.Ed.2d 523, 90 S.Ct. 1990]; Kramer v. Union School District (1969) 395 U.S. 621 [23 L.Ed.2d 583, 89 S.Ct. [1886]) which have applied strict scrutiny to invalidate restrictions on the franchise in municipal and school district elections, and to our decision in Curtis v. Board of Supervisors, supra, 7 Cal.3d 942, applying that test to an election to incorporate a city.
The State Board, however, contends that geographical classifications constitute an exception to the rule, that exclusion of voters because of their place or residence requires strict scrutiny only if the exclusion serves some impermissible end such as racial discrimination (see Gomillion v. Lightfoot (1960) 364 U.S. 339 [5 L.Ed.2d 110, 81 S.Ct. 125]). They rely for this proposition on two decisions of the United States Supreme Court, Lockport v Citizens for Community Action (1977) 430 U.S. 259 [51 L.Ed.2d 313, 97 S.Ct. 1047], and Holt Civic Club v. Tuscaloosa (1978) 439 U.S. 60 [58 L.Ed.2d 292, 99 S.Ct. 383]. As we shall explain, the exception established by those cases is not as broad as the State Board claims—and not broad enough to permit the classification at issue here to escape strict scrutiny.
Lockport upheld a New York law requiring that adoption of a new county charter be approved by separate majorities of city and noncity residents.20 When a proposed charter was approved by a majority of city residents, but defeated by the less numerous noncity residents, proponents of the charter claimed the statute violated the one-man, one-vote principles by giving noncity voters an excessive voice.
The court, however, noted that one-man, one-vote principles had little application to a single-issue referendum since the state could “determine whether its adoption or rejection will have a disproportionate impact on an identifiable group of voters” (p. 266 [51 L.Ed.2d at p. 321]) and shape its election requirements accordingly. The court observed that if city and noncity voters had identical interests in the adoption or rejection of the charter, “any distinction . . . between voters drawn on the basis of residence” would be unconstitutional. (P. 268 [51 L.Ed.2d at p. 322].) Conversely, if the interests of one group far outweighed the other, the court implied that the state could limit the franchise to the former group. (See p. 266 [51 L.Ed.2d at p. 321], citing Salyer Land Co. v. Tulare Water District (1973) 410 U.S. 719 [35 L.Ed.2d 659, 93 S.Ct. 1224].) The challenged statute in Lockport, however, presented an in[801]*801termediate situation: city and noncity residents each had substantial but different interests in the controversy.
Such differences, the court held, justified the state’s decision to require concurrent majorities of city and noncity residents. The court explained that if the constitutional “. . . question were posed in the context of annexation proceedings, the fact that the residents of the annexing city and the residents of the area to be annexed formed sufficiently different constituencies with sufficiently different interests could be readily perceived. The fact of impending union alone would not so merge them into one community of interest as constitutionally to require that their votes be aggregated in any referendum to approve annexation. Cf. Hunter v. Pittsburgh, 207 U.S. 161. Similarly, a proposal that several school districts join to form a consolidated unit could surely be subject to voter approval in each constituent school district. [1] . . . [T|n terms of recognizing constituencies with separate and potentially opposing interests, the structural decision to annex or consolidate is similar in impact to the decision to restructure county government in New York. In each case, separate voter approval requirements are based on the perception that the real and long-term impact of a restructuring of local government is felt quite differently by the different county constituent units that in a sense compete to provide similar governmental services. Voters in these constituent units are directly and differentially affected by the restructuring of county government, which may make the provider of public services more remote and less subject to the voters’ individual influence, [f] The provisions of New York law here in question no more than recognize the realities of these substantially differing electoral interests. Granting to these provisions the presumption of constitutionality to which every duly enacted state and federal law is entitled, we are unable to conclude that they violate the Equal Protection Clause of the Fourteenth Amendment.” (Fn. omitted; 430 U.S., at pp. 271-272 [51 L.Ed.2d, at pp. 324-325].)
The quoted language from Lockport makes it clear that the state can recognize that residents of different areas may have different interests and, in a single-issue referendum, can constitutionally require concurrent majorities. But nothing in Lockport endorses measures which deny the vote entirely to residents of one of the areas, nor permits such measures to escape strict judicial scrutiny.
This reading of Lockport gains support from the only California decision relating to geographical limits on voting rights. In Hawn v. County of Ventura, supra, 73 Cal.App.3d 1009, the county enacted an ordinance providing that any decision to establish a comity airport located in whole or in part within an incorporated city must be approved by a majority of city voters. The county undertook to acquire a former Air Force base located partially within the City of Camarillo and partially beyond the city limits. The Court of Appeal applied [802]*802strict scrutiny to the ordinance limiting the franchise to city residents and, finding both city and county residents directly interested in the issue of airport location, held the restricted franchise invalid.
Distinguishing the Lockport decision, the Court of Appeal observed that “[i]t is readily apparent that the New York law involved in Lockport is quite different from the airport initiative ordinance before us. Lockport did not deal with a statute which excluded voters based on residence. On the contrary, the statute in Lockport gave equal recognition to the interests of the noncity voters in a new county charter as it did to the interests of the city voters. The rationale of Lockport lends support to the contention that the airport ordinance before us constitutes an invidious discrimination against noncity residents and voters of Ventura County . . . .” (P. 1021.)
We conclude that Lockport does not support the proposition that geographical restrictions on the franchise are immune from strict scrutiny. We therefore turn to the second case relied on by the State Board, Holt Civic Club v. Tuscaloosa, supra, 439 U.S. 60. That case concerned Alabama statutes which permitted cities to extend their police regulation and licensing authority to unincorporated areas within three miles of the city limits. Nonresidents, however, were immune from city taxes (except business licenses, charged at one-half the rate assessed within the city limits), city zoning authority, and many other powers of city government. Although appellants, who lived within the three-mile area beyond the city limits, did not directly seek the right to vote in municipal elections, they claimed that because they were denied the vote the city’s exercise of extraterritorial jurisdiction was unconstitutional.
The court reviewed the voting rights cases in which it had applied strict scrutiny, and noted that the cases involved a common characteristic: “The challenged statute in each case denied the franchise to individuals who were physically resident within the geographic boundaries of the governmental entity concerned.” (P. 68 [58 L.Ed.2d at p. 300].) The geographic entity involved in the case at hand was the City of Tuscaloosa, and since the statute did not deny the vote to anyone within that line, strict scrutiny was not required; “[t]he line . . . marked by this Court’s voting qualifications decisions coincides with the geographical boundary of the governmental unit at issue, and we hold that appellants’ case, like their homes, falls on the farther side.” (P. 70 [58 L.Ed.2d at p. 302].) Applying a less rigorous test—whether the state plan bore a rational relationship to a legitimate state purpose—the court upheld the challenged statutes.
The State Board reads Holt as holding that geographical distinctions, unlike all other voting classifications, never require strict scrutiny. We read the case [803]*803differently. In our view, it holds only that distinctions which coincide with the boundaries of the governmental entity concerned, and exclude no one physically resident within those boundaries, do not require strict scrutiny. If, on the other hand, the state attempts to carve up the existing political entity and deny the vote to some of its residents—if, for example, in Holt the statute had denied the vote to some persons who lived within the Tuscaloosa city limits—then strict scrutiny is still required.
Accordingly, we cannot avoid the requirement of strict scrutiny on the ground that the decision in question here imposes geographical restrictions on voting rather than the property or residency limitations examined in prior cases. Instead, we must determine the constitutionally relevant boundaries, “the geographic boundaries of the governmental entity concerned” (Holt Civic Club v. Tuscaloosa, supra, 439 U.S. at p. 68 [58 L.Ed.2d at p. 300]), and subject to strict scrutiny any measure which limits voting within those boundaries.
In the present case, the State Board could argue that Yorba Linda, not the Fullerton HSD, is the relevant geographic area, and that the exclusion of voters outside Yorba Linda thus is not subject to strict scrutiny. The Fullerton HSD, however, is not merely a neighboring district incidentally affected by the secession of Yorba Linda, but the existing entity with legal authority over high school education in Yorba Linda; the impact of the secession is in large part the consequence of the present status and authority of the Fullerton district. For example, under the plan approved by the State Board, Fullerton HSD may incur a substantial debt to the new Yorba Linda Unified School District, a debt which will have to be paid by the residents remaining in the Fullerton HSD. Moreover, having built and staffed a high school to educate the children of Yorba Linda, among others, the withdrawal of Yorba Linda will leave the district with excess capacity and affect its educational programs. None of these effects affect neighboring districts; the impact on the Fullerton HSD and its residents is unique. We conclude that the entire present Fullerton HSD is the relevant geographic area for the purpose of applying constitutional requirements of equal protection.21
The State Board relies primarily on another exception to the requirement for strict scrutiny—that in an election involving a special district of limited powers whose activities disproportionately affect members of a particular group, a measure limiting the franchise to that group does not encounter strict scrutiny. The State Board relies in particular on Salyer Land Co. v. Tulare Water [804]*804District, supra, 410 U.S. 719, the leading case establishing this exception to the strict scrutiny requirement.
Salyer upheld a franchise restricted to landowners for the election of directors of a water storage district. The decision relied on the limited powers of the district, the disproportionate effect of its activities on landowners as a group, and the financing of district activities solely by levies on the landowners. California courts have applied a similar analysis to elections concerning a recreation and park district (Simi Valley Recreation & Park Dist. v. Local Agency Formation Com., supra, 51 Cal.App.3d 648), a reclamation district (Philippart v. Hotchkiss Tract Reclamation Dist. 799 (1976) 54 Cal.App.3d 797 [127 Cal.Rptr. 42]) and a small irrigation district (Schindler v. Palo Verde Irrigation Dist. (1969) 1 Cal.App.3d 831 [82 Cal.Rptr. 61]).
Our decision in Choudhry v. Free, supra, 17 Cal.3d 660, explored the limits of this exception. Choudry concerned a statute which required that directors of the Imperial Irrigation District be freeholders. The opinion noted the size of the district—the largest irrigation district in the state—its substantial payroll, its power to acquire and operate electrical power and flood control facilities, and its ability to finance operations through charges for such services. Distinguishing Salyer by virtue of the larger size and impact on nonlandowners of the Imperial Irrigation District, the court held that the restriction must be tested by a standard of strict scrutiny.
Mindful of the broad powers possessed by school districts, and the impact of district actions upon the life of the community, courts have consistently applied the same standards to school district elections as they apply to municipal and state elections. In Kramer v. Union School District, supra, 395 U.S. 621, which involved a statute limiting the franchise in school board elelctions, the Supreme Court stated that the “ ‘alleged infringement of the right of citizens to vote must be carefully and meticulously scrutinized.’ ” (P. 626 [23 L.Ed.2d at p. 589], quoting Reynolds v. Sims (1964) 377 U.S. 533, 562 [12 L.Ed.2d 506, 527, 84 S.Ct. 1362].) Our decision in Curtis v. Board of Supervisors, supra, 7 Cal.3d 942, reviewed Kramer and other high court decisions, and concluded that the standards established in those cases, including strict judicial scrutiny of voting restrictions, applied to “the government of [the] state, city, county, school district, and other agencies of general governmental power.” (P. 960). (Italics added.)
We note in particular the decision of the United States Supreme Court in Hadley v. Junior College District (1970) 397 U.S. 50 [25 L.Ed. 2d 45, 90 S.Ct. 791], which held that the “one man, one vote” requirement applied in an election of trustees of a junior college district. The opinion explained that “It is [805]*805of course possible that there might be some case in which a State elects certain functionaries whose duties are so far removed from normal governmental activities and so disproportionately affect different groups that a popular election in compliance with [the one man, one vote requirement] might not be required, but certainly we see nothing in the present case that indicates that the activities of these trustees fit in that category. Education has traditionally been a vital governmental function, and these trustees ... are governmental officials in every relevant sense of that term.” (P. 56 [25 L.Ed.2d at p. 51].) In Salyer Land Co. v. Tulare Water District, supra, 410 U.S. 719, the court quoted the foregoing language from Hadley to justify its decision upholding a limited franchise in a water district election. (Pp. 727-728 [35 L.Ed.2d at pp. 665-666].) Thus it appears that Salyer, the precedent on which the State Board relies, itself recognizes that school district elections do not fall within the Salyer exception for districts of limited authority and impact.
We conclude that the decision of the State Board is subject to strict judicial scrutiny, and cannot be sustained unless justified by a compelling state interest. The State Board, in defense of that decision, argues that the residents of Yorba Linda have a greater interest in the question of establishing a separate high school district in that community, but that their preference might be swamped by the “selfish” vote of the less interested, but more numerous, residents of the remaining Fullerton HSD. The State Board does not explicitly claim that justification constitutes a compelling state interest, but it is nevertheless our task to inquire into that matter.
It is, of course, clear that the state cannot claim a compelling interest in excluding voters because of how they may vote. (Carrington v. Rash (1965) 380 U.S. 89, 94 [13 L.Ed.2d 675, 679, 85 S.Ct. 775].) To the extent that the State Board seeks to exclude some Fullerton voters because their vote might defeat the proposed secession, the State Board’s justification is constitutionally impermissible. If we rephrase the justification, however, as one of excluding uninterested voters in order to protect the interests of persons vitally concerned, we encounter a state interest which might, in an appropriate case, achieve compelling character.
This is not such a case. Without denigrating the interests of Yorba Linda residents, it is nonetheless apparent that the remaining Fullerton HSD residents also have a substantial interest in the election. As we noted earlier, the secession will affect the present and future racial composition of the Fullerton HSD, and may make it more difficult for the district to cope with an increasing number of minority students. We also note that the Plan approved by the State Board may impose a substantial debt on the Fullerton HSD which might require stringent economies or a tax increase. Although neither consideration is, in our [806]*806opinion, enough in itself to compel us to overturn the State Board’s approval of the secession proposal, the racial and financial impact of that Plan are matters of great concern to the residents who will remain in the Fullerton HSD. The record further indicates that the withdrawal of the Yorba Linda students from Troy High School in Fullerton May necessitate closing that school, or at least a substantial curtailment of its facilities and curriculum. Such measures will seriously affect the quality of education offered by the Fullerton HSD, as well as the financial burden on its residents.22
In sum, we do not view the clasification at issue here as one which separates interested and uninterested voters, but one which divides two groups, each with a substantial although different interest in the election. In such a case the state has no compelling interest to grant the franchise to one group and deny it to the other. We therefore conclude that the decision of the State Board, to the extent that it excluded Fullerton HSD residents from voting at the election to approve the proposed secession of Yorba Linda from the Fullerton HSD, denied such residents the equal protection of the laws.
5. Conclusions respecting the trial court judgement.
Having found that the State Board’s approval of the Plan violated Education Code section 4200, the trial court issued mandate prohibiting state and local officials from taking any action whatsoever on the secession Plan. Since we conclude that the Board’s approval of the Plan was not arbitrary, capricious, or unsupported by evidence, we cannot agree with the trial court’s order barring any further action. The State Board, in our opinion, may proceed to consider the Plan, so long as it conforms to the requirements of CEQA and does not exclude any Fullerton HSD residents from voting in any election to approve the Plan.
The trial court further prohibited the relevant officials “from calling, noticing, or conducting an election on March 7, 1978, or at any other time whatsoever” for the purposes of submitting the Plan to the voters. This portion of the order below is overbroad, and should be modified to prohibit only an election limited to the residents of less than the entire Fullerton HSD.
Finally, the trial court directed the State Board to vacate its approval of the Plan and prohibited any election conducted before the State Board had complied with the requirements of CEQA. Compliance with CEQA, however, is re[807]*807quired not only before an election takes place, but before the State Board grants final approval to the Plan. (See Pub. Resources Code, § 21050.) The judgment should be modified accordingly.
The judgment of the superior court is reversed and the cause remanded with directions to modify the judgment in conformity with the views expressed herein. Since the modified judgment will afford Fullerton HSD substantially the relief it seeks, it shall recover its costs on appeal.
Bird, C. J., and Mosk, J., concurred.