Southern Cal. Rapid Transit Dist. v. Bolen (1992)

Annotate this Case
[No. S015986. Jan 30, 1992.]

SOUTHERN CALIFORNIA RAPID TRANSIT DISTRICT, Plaintiff and Respondent, v. HELEN M. BOLEN, Defendant and Appellant; ATCHISON, TOPEKA AND SANTA FE RAILWAY COMPANY et al., Interveners and Appellants.

(Superior Court of Los Angeles County, No. C656503, Billy G. Mills, Judge.)

(Opinion by Arabian, J., with Lucas, C. J., Panelli, Baxter and George, JJ., concurring. Separate dissenting opinion by Kennard, J., with Mosk, J., concurring.)

COUNSEL

Marilyn L. Garcia, Brobeck, Phleger & Harrison, John J. Wasilczyk and Earle Miller for Defendant and Appellant.

MacDonald, Halsted & Laybourne, John R. Shiner, Stephanie Berrington McNutt, Lisa Winfield Liberatore, Hill, Farrer & Burrill, William M. Bitting, Vincent C. Page, Kevin H. Brogan and Dean E. Dennis for Interveners and Appellants.

Bird, Marella, Boxer, Wolpert & Matz, Vincent J. Marella, Dorothy Wolpert, Mark T. Drooks and Diane P. Shakin for Plaintiff and Respondent.

De Witt W. Clinton, County Counsel (Los Angeles), David B. Kelsey, Assistant County Counsel, Nossaman, Guthner, Knox & Elliott, James C. Powers and Alvin S. Kaufer as Amici Curiae on behalf of Plaintiff and Respondent.

OPINION

ARABIAN, J.

The principal question before us is whether the equal protection guarantees of the state and federal Constitutions are violated by a statute conditioning the right to vote on the ownership of real property and alloting votes on the basis of its assessed value. We hold that, under the narrow circumstances presented, such a voting scheme is reasonably related to the objectives of the statute and thus survives constitutional scrutiny.

I

A

The Southern California Rapid Transit District (hereafter SCRTD or transit district) is the lead agency for the construction, financing, and operation of "a comprehensive mass rapid transit system in the southern California area, and particularly in Los Angeles County." (Pub. Util. Code, § 30001, subd. (a).) The initial segment of the transit system, an 18.6- mile [1 Cal. 4th 660] subway line connecting the central business district of downtown Los Angeles with North Hollywood and known as "Metro Rail," was approved by SCRTD in 1983. The first operating element of Metro Rail, a 4.4- mile rapid transit line extending from Union Station to Wilshire Boulevard and Alvarado Street, will cost an estimated $1.25 billion. fn. 1 Financing for the project comes from a combination of sources-federal, state, and local-and, as a condition of federal funds, from the private sector. To help defray part of the enormous cost of the project, the Legislature has authorized SCRTD to establish "special benefit assessment districts" surrounding planned Metro Rail subway stations along the rapid transit corridor. (Pub. Util. Code, §§ 33000-33020.)

Although the transit district is statutorily authorized to establish assessment districts without voter approval, a referendum must be held if requested by the "owners of at least 25 percent of the assessed value of real property" within a proposed assessment district. (Pub. Util. Code, §§ 33002.1, 33002.2.) The statute limits voting at such a referendum, however, to those owners of real property who will be subject to assessment in the event an assessment district is approved. (Pub. Util. Code, § 33002.3.) Because in the present case residential property has been exempted from assessment, participation in any referendum is denied to non-property-owning residents and residential property owners alike within the proposed districts.

In addition to the foregoing franchise restrictions, the statute directs that votes in any referendum be alloted on the basis of the assessed value of the real property for ad valorem tax purposes. Qualified owners are allotted one vote for each $1,000 of assessed value of their real property. (Pub. Util. Code, § 33002.3, subd. (b).) Although votes are alloted according to assessed value, the statute requires that any assessments actually levied by the transit district be calculated on the basis of the parcel or floor area, depending on the condition of the real property. (Pub. Util. Code, § 33002, subd. (a).)

B

[1] We have described special or local assessments of the sort authorized by the Legislature here as a "compulsory charge placed by the state upon real property within a pre-determined district, made under express legislative authority for defraying in whole or in part the expense of a permanent public improvement therein ...." ( San Marcos Water Dist. v. San Marcos Unified [1 Cal. 4th 661] School Dist. (1986) 42 Cal. 3d 154, 161 [228 Cal. Rptr. 47, 720 P.2d 935] (quoting Spring Street Co. v. City of Los Angeles (1915) 170 Cal. 24, 29 [148 P. 217], internal quotation marks omitted); their use as a means of financing the cost of municipal improvements has a lengthy pedigree in the law of public finance. (See, generally, 14 McQuillin, The Law of Municipal Corporations (3d ed. 1987 rev.) §§ 38.01- 38.338.) Analogous to but differing in important respects from the power of taxation, the essential feature of the special assessment is that the public improvement financed through it confers a special benefit on the property assessed beyond that conferred generally. (See, e.g., Norwood v. Baker (1898) 172 U.S. 269 [43 L. Ed. 443, 19 S. Ct. 187]; Spring Street Co. v. City of Los Angeles, supra, 170 Cal. 24, 30-31.)

The uses of the special assessment as a financing device for public improvements are as manifold as the forms of such improvements themselves. It has been employed to finance such variegated public improvements as the construction of drains and sewers (Dawson v. Town of Los Altos Hills (1976) 16 Cal. 3d 676 [129 Cal. Rptr. 97, 547 P.2d 1377] (Dawson)), residential subdivisions (Burrey v. Embarcadero Mun. Improvement Dist. (1971) 5 Cal. 3d 671 [97 Cal. Rptr. 203, 488 P.2d 395] (Burrey)), gas distribution works (County of Riverside v. Whitlock (1972) 22 Cal. App. 3d 863 [99 Cal. Rptr. 710] (Whitlock)), flood control projects (City of Larkspur v. Marin County Flood Control etc. Dist. (1985) 168 Cal. App. 3d 947 [214 Cal.Rptr. 689]), the redevelopment of blighted areas (In re Redevelopment Plan for Bunker Hill (1964) 61 Cal. 2d 21 [37 Cal. Rptr. 74, 389 P.2d 538]), and the construction of a transit tunnel (Larsen v. San Francisco (1920) 182 Cal. 1 [186 P.2d 757]).

In recent years, the special assessment has found a new field of application in the context of the construction of massive urban rapid transit systems, the extraordinary capital costs of which have heightened pressure for the exploration of novel forms of financing. Grounded in the fact that the installation of a transit station and related facilities at a point along a rapid transit corridor enhances the value of real property in the immediate vicinity by generating intensified commercial activity, the special assessment has been promoted as a means by which transit authorities can recoup some of the value added to the surrounding real property and the resulting "windfall" enrichment of property owners. (See Note, Rapid Transit Financing: Use of the Special Assessment (1977) 29 Stan.L.Rev. 795.)

As applied to rapid transit financing, the mechanics of the assessment process do not differ materially from its use in conjunction with other public improvements. Briefly, the authorized public entity adopts a resolution of its intention to impose an assessment, determines the boundaries of the planned [1 Cal. 4th 662] assessment district, and computes the proposed levy. It is then generally required to submit the issue to voter approval through some form of referendum or protest procedure. On approval, bonds are issued to finance construction of the public improvement, the principal and interest being paid and the bonds retired from income obtained through annual assessments.

C

In 1984, as part of its implementation of the assessment district revenue device, the SCRTD board appointed a benefit assessment policy task force, composed of representatives of a cross-section of property owners along the planned Metro Rail transit corridor, and charged it with the task of conducting a study and making recommendations for structuring the benefit assessment districts. Following a series of public hearings, the task force submitted its recommendations to the SCRTD board in 1985. In substance, it found that the property within the proposed assessment districts would benefit from the installation of Metro Rail transit stations through enhanced land values, higher lease rates and occupancy levels, increased retail sales, easier visitor access, reduced parking costs, and the intensification of land development. Following public notice and additional hearings, the SCRTD board adopted the resolutions necessary to proceed with the establishment of the benefit districts.

After submission of the resolutions to the Los Angeles City Council as required by statute and modification by that body to exempt residential property owners within the proposed districts from assessment, the transit district took final action on the matter in July of 1985 by establishing two benefit districts, one covering real property within the one-half mile radii of the four Metro Rail transit stations planned for the downtown central business district, and a second covering the real property within a one-third mile radius of the Wilshire-Alvarado transit station.

The SCRTD board's final resolutions established an initial assessment rate of $.30 per square foot of assessable property, with a maximum rate of $.42 per square foot-a rate calculated to raise a total of $130.3 million in capital funding over a multiyear assessment period, or roughly 11 percent of the total cost of the 4.4-mile segment. The board also acceded to the exemption of residential property within the two benefit districts, required as a condition for Los Angeles City Council approval, and directed that the assessments otherwise be levied uniformly throughout the two districts and that they expire in the year 2008 or earlier. Although assessment notices for the 1986-1987 year were mailed to commercial property owners within the two districts, at the request of affected property owners the board subsequently deferred collection of any assessments until 1992. [1 Cal. 4th 663]

Soon after the foregoing administrative action was taken, this suit was filed. Brought by SCRTD against its secretary, Helen M. Bolen, the complaint sought a writ of mandate directing her to certify additional resolutions of the SCRTD board authorizing the sale of $200 million in municipal bonds secured by revenues from the benefit assessments, the object of the suit being to judicially validate the actions of the board in establishing the two benefit districts before the bonds were issued. fn. 2 (Cf. Whitlock, supra, 22 Cal. App. 3d 863, 868 & fn. 6; Code Civ. Proc., § 860 et seq.; Sts. & Hy. Code, § 10601 et seq.) Several commercial property owners within the two proposed assessment districts and one tenant under a commercial lease were granted leave by the superior court to intervene in the action on the side of defendant Bolen.

After reviewing the administrative record supporting the task force's recommendations and taking additional evidence, the superior court upheld the property-based voting scheme at issue against claims that it violated the equal protection guarantees of the state and federal Constitutions. The trial court likewise validated the assessment districts in all other respects against the nonconstitutional objections of interveners and Bolen, entered judgment for the transit district on its claims, issued its writ directing defendant Bolen to certify the board resolutions authorizing issuance of the revenue bonds, and denied interveners any relief.

The Court of Appeal reversed. In its view, the property-based voting scheme was constitutionally flawed in two respects. First, it violated equal protection by invidiously discriminating against nonproperty owners. According to the Court of Appeal's reasoning, public transportation affects all citizens, not merely property owners, and the financing, operation, and maintenance of Metro Rail will impact all segments of the population of greater Los Angeles. Since the benefits and the burdens of the assessments fall indiscriminately on property owners and nonproperty owners alike, the principle of "one person, one vote" is triggered, requiring a compelling state interest to justify the exclusion of nonproperty owners from the franchise. Finding none, the Court of Appeal pronounced that part of the statutory voting scheme void.

In addition, the Court of Appeal concluded that the differing methods adopted by the Legislature in alloting votes and imposing assessments under the statute-the former linked to the ad valorem tax value of the property, the latter tied to parcel size-were fundamentally unfair because of the [1 Cal. 4th 664] absence of proportionality between the allotment of votes and the burden of paying assessments. In the Court of Appeal's view, even if the principle of one person, one vote were inapplicable, in order to pass equal protection scrutiny the voting scheme was required to allot the most votes to those paying the largest assessments. After finding that these constitutionally defective referendum provisions could not be severed from the remainder of the statutory scheme, and ruling that the transit district lacked the power to exempt residential property owners from the assessment, the Court of Appeal invalidated the entire benefit assessment legislation.

We reverse the judgment of the Court of Appeal.

II

It is important to underline at the outset what is not before us in this case. The question at issue is not whether Metro Rail should be built and, if so, how its costs should be distributed. Neither are we asked to consider directly the desirability or usefulness of special assessment districts as a financing device to recapture some of the economic value added to a commercial area as a result of locating a rapid transit station within it. Questions concerning metropolitan transportation policy in greater Los Angeles, and the economic, environmental and aesthetic aspects of Metro Rail and of benefit assessment districts, important as they undoubtedly are as public issues, are not presented by this case, having been debated, sometimes litigated (see, e.g., Rapid Transit Advocates, Inc. v. Southern Cal. Rapid Transit Dist. (1986) 185 Cal. App. 3d 996 [230 Cal.Rptr. 225]), and resolved in other forums over the past decade.

What is at issue is the substantially narrower question whether the principle of "one person, one vote," laid down by the United States Supreme Court in Reynolds v. Sims (1964) 377 U.S. 533 [12 L. Ed. 2d 506, 84 S. Ct. 1362] (Reynolds), applies to the property-based assessment district voting scheme described above, or whether the circumstances qualify as those in which the Constitution does not stand as a "roadblock[] in the path of innovation, experiment, and development among units of local government." fn. 3 (See Avery v. Midland County (1968) 390 U.S. 474, 485 [20 L. Ed. 2d 45, 53-54, 88 S. Ct. 1114] (Avery).) As with much of the jurisprudence of equal protection, the answer to that question is significantly affected by the threshold selection of the appropriate level of judicial scrutiny. [1 Cal. 4th 665]

If the principle of one person, one vote applies to voting in the assessment district referenda, the state is placed under a substantial burden of demonstrating a compelling justification for an exclusion from the voting franchise resting on the ownership of real property. If, however, this case qualifies as an exception to the principle of Reynolds, the constitutional test is the less demanding one of whether the voting scheme is either "wholly irrelevant" or "reasonably related" to the statutory objectives. As we shall explain, the crucial task in the case of property- based voting schemes is to identify the constitutionally relevant factual basis for selecting one level of scrutiny over the other.

A

The solution to the equal protection problem posed by the statutory classification in this case lies in a trio of post-Reynolds decisions of the high court which recognize an exception to the principle of one person, one vote. [2a] In substance, these cases hold that the right protected by Reynolds-equality at the ballot box-is not fundamental under limited circumstances. fn. 4 As the court formulated the exception in Avery, supra, 390 U.S. 474, 483-484 [20 L. Ed. 2d 45, 53], these circumstances consist of "a special- purpose unit of government assigned the performance of functions affecting definable groups of constituents more than other[s] ...." Where these two conditions jointly occur, the strict demands of Reynolds, supra, 377 U.S. 533, do not apply and voting power "may be apportioned in ways which give greater influence to the citizens most affected by the organization's functions" (Avery, supra, at p. 484 [20 L.Ed.2d at p. 53]) without violating the guarantee of equal protection provided that the resulting classification is reasonably related to the statutory objective.

The first condition focuses on the extent to which the public entity involved is vested with governmental powers. In Avery, supra, 390 U.S. 474, the court held that Reynolds applied to the election of a countywide "commissioners court," a local body exercising "general governmental powers over the entire geographic area served by [that] body." (390 U.S. at p. 485 [1 Cal. 4th 666] [20 L.Ed.2d at p. 53].) In so ruling, the court said that if the governmental entity involved makes "a large number of decisions having a broad range of impacts on all the citizens" within its jurisdiction, the principle of one person, one vote applies presumptively. (Id. at pp. 483-485 [20 L.Ed.2d at pp. 52-54].) In Hadley v. Junior College District (1970) 397 U.S. 50, 54 [25 L. Ed. 2d 45, 49, 90 S. Ct. 791], the high court elaborated on this requirement, applying the presumption whenever "important governmental functions," including, as in Hadley, those as "vital" as education are involved. Thus, "as a general rule, whenever a state or local government decides to select persons by popular election to perform governmental functions, the Equal Protection Clause of the Fourteenth Amendment requires that each qualified voter must be given an equal opportunity to participate in that election ...." (Id. at p. 56 [25 L.Ed.2d at pp. 50-51].)

The second branch of the equal protection analysis looks to the impact of the election outcome on voters and nonvoters. It focuses on the extent to which a contested statutory voting classification is supported by "genuine difference[s] in the relevant interests" of those enfranchised and those excluded by a given voting scheme. (See Lockport v. Citizens for Community Action (1977) 430 U.S. 259, 268 [51 L. Ed. 2d 313, 322-323, 97 S. Ct. 1047] (Lockport) [upholding concurrent majority voting requirements for city and non city residents in county wide referendum].) If the class enfranchised by the scheme at issue is "primarily affected" or "primarily interested" in the election and "those excluded are in fact substantially less interested or affected than those the statute includes" (Kramer, supra, 395 U.S. at p. 632 [23 L.Ed.2d at p. 592]), the voting scheme does not deny equal protection so long as the statutory classification is not "wholly irrelevant" to the achievement of the statute's objectives. (Kotch v. Pilot Comm'rs (1947) 330 U.S. 552, 556 [91 L. Ed. 1093, 1096-1097, 67 S. Ct. 910]; McGowan v. Maryland (1961) 366 U.S. 420, 425-426 [6 L. Ed. 2d 393, 398-399, 81 S. Ct. 1101].)

B

In Salyer Land Co. v. Tulare Water District (1973) 410 U.S. 719 [35 L. Ed. 2d 659, 93 S. Ct. 1224] (Salyer), the conditions required to support an exception to the principle of Reynolds materialized in the form of a contested election of the board of directors of a California water storage district. In ruling that a property-based voting scheme was not subject to the principle of one person, one vote, the court invoked both the "special-purpose unit of government" and the "primarily affected or interested" criteria to anchor its result.

Comprising 193,000 acres of intensively cultivated farmland, the water storage district was populated by only 60 adults, most of whom were [1 Cal. 4th 667] employed by one of the 4 corporations that farmed 85 percent of the land within it. The statutory voting scheme at issue in Salyer limited voting in the general election for the district board of directors to owners of land within the water storage district, apportioning votes according to the assessed value of district land. (Salyer, supra, 410 U.S. 719, 724-725 [35 L.Ed.2d at pp. 664-665].) In upholding the statutory limitation on voting against claims by resident nonproperty owners that it violated equal protection by failing to comply with the principle of one person, one vote, the court first pointed out that although the district exercised some typical governmental powers, these were incidental to its limited authority and primary purpose of "provid[ing] for the acquisition, storage, and distribution of water for farming in the Tulare Lake Basin." (Id. at p. 728 [35 L.Ed.2d at p. 666].) Moreover, the district's activities affected landowners disproportionately since "[a]ll of the costs of district projects are assessed against land ... in proportion to the benefits received [and] ... charges for services rendered are collectible from persons receiving their benefit in proportion to the services." (Id. at p. 729 [35 L.Ed.2d at p. 667].)

Because there was "no way that the economic burdens of district operations [could] fall on residents qua residents, and the operation of the district[] primarily affect[s] land within [its] boundaries," the court concluded that "the popular election requirements enunciated in Reynolds ... and succeeding cases are inapplicable to elections such as the general election of [the] Water District." (Salyer, supra, 410 U.S. at pp. 729-730 [35 L.Ed.2d at p. 667].) The equal protection question, the court said, was simply "whether the State's decision to deny the franchise to residents of the district while granting it to landowners was 'wholly irrelevant to achievement of the [statute's] objectives,' [citation]." (Id. at p. 730 [35 L.Ed.2d at pp. 667-668].)

Salyer, supra, 410 U.S. 719, is analytically linked to the court's subsequent decision in Ball v. James (1981) 451 U.S. 355 [68 L. Ed. 2d 150, 101 S. Ct. 1811] (Ball). On facts strikingly different from those of Salyer, the court reached an identical result, ruling that the principle of Reynolds, supra, 377 U.S. 533, did not apply to an election for the directors of the Salt River District, a water reclamation district encompassing 236,000 acres in central Arizona whose governing board was elected by those owning land within the district, voting power being apportioned according to the number of acres owned. (Ball, supra, 451 U.S. at pp. 370-372 [68 L.Ed.2d at pp. 162-164].)

Like the water storage district in Salyer, the primary purpose of the Salt River District was the conservation and distribution of water owned by its landowning membership. But unlike the district in Salyer, the Salt River [1 Cal. 4th 668] District had constructed dams and other public works to generate hydro-electric power. By 1980, the district was one of the largest public utilities in the state. It provided power to almost half of the population of Arizona, including much of metropolitan Phoenix, met most of its capital and operating costs through electric power revenues, and had over $3 billion in long-term debt; 40 percent of the water stored by the district was delivered to urban areas for nonagriculatural uses.

Despite recognition that these activities were "more diverse and affected far more people" than those in Salyer, supra, 410 U.S. 719, the court held that "these distinctions do not amount to a constitutional difference." (Ball, supra, 451 U.S. at pp. 365-366 [68 L.Ed.2d at pp. 159-160].) This was so, the majority reasoned, because despite its manifold activities, the Salt River District did "not exercise the sort of governmental powers that invoke the strict demands of Reynolds." (Id. at p. 366 [68 L.Ed.2d at p. 160].) Its powers did not run the gamut of those typical of a general government, and those that it did exercise were incidental to and in the service of its relatively narrow mission of storing, conserving, and distributing water to its landowner members. (Id. at pp. 368-369 [68 L.Ed.2d at pp. 161-162].) "The constitutionally relevant fact is that all water delivered by the Salt River District ... is distributed according to land ownership, and the District does not and cannot control the use to which the landowners who are entitled to the water choose to put it." (Id. at pp. 367-368 [68 L.Ed.2d at p. 161], fn. omitted.)

Coordinate with its limited purpose and functions, the reclamation district's activities fell disproportionately on the specific class which the statutory voting scheme enfranchised-its landowner membership. Only they were subject to the acreage-based taxing power of the district, only they had committed capital to the district through assessments, and only their land was subject to liens to secure district bonds. (Ball, supra, 451 U.S. at p. 370 [68 L.Ed.2d at pp. 162-163].) Since the district's operations had a disproportionate effect on those enfranchised, "the voting scheme ... [was] constitutional because it bears a reasonable relationship to its statutory objectives." (Id. at p. 371 [68 L.Ed.2d at p. 163].)

No one reviewing this area of the high court's equal protection jurisprudence can fail to be impressed with the result in Ball-not because the opinion represents an analytical advance over the principles developed in Salyer, but because it illustrates the majority's steadfast willingness to [1 Cal. 4th 669] adhere to the Salyer analysis in the face of a record presenting such compelling, if "constitutionally irrelevant," facts. fn. 5 Clearly, in light of Ball, as far as the governmental function analysis is concerned, the constitutionally decisive fact is that the voting scheme at issue reflects the "narrow primary purpose for which the [public entity] is created." fn. 6 (Ball, supra, 451 U.S. at p. 369 [68 L.Ed.2d at p. 162].)

[3] Viewing the record in this case through the lens of that insight, one conclusion seems evident. Manifestly, the benefit districts at issue here are not invested with and do not exercise powers remotely similar to the "general governmental powers" to which the principle of Reynolds, supra, 377 U.S. 533, presumptively applies. And unlike the substantial but narrowly directed complement of powers exercised by the water districts in Salyer and Ball, the benefit assessment districts lack virtually any of the incidents of government. In fact, they are little more than formalistic, geographically defined perimeters whose raison d'etre is to serve as the conceptual medium for the recognition of economic benefits conferred and the imposition of a corresponding fiscal burden. (Compare Whitlock, supra, 22 Cal. App. 3d 863, 874 [assessment district "simply denotes the land area benefited by the proposed improvements and to be assessed for the costs thereof"], and Dawson, supra, 16 Cal. 3d 676, 683 [special assessment district "is not a legal entity with officers and corporate rights and duties"], with Burrey, supra, 5 Cal. 3d 671, 677 [Reynolds applied to municipal improvement [1 Cal. 4th 670] district "because it is the sole local governmental unit ... carrying out the basic municipal functions which directly affect and benefit each of the district's residents"], and Curtis v. Board of Supervisors (1972) 7 Cal. 3d 942, 959 & fn. 20 [104 Cal. Rptr. 297, 501 P.2d 537] [statutory limitations on voting for municipal incorporation unconstitutional where city would exercise "general governmental powers"].)

The transit district itself, of course, is invested with and exercises substantial governmental powers; indeed, it has been described as having "virtual autonomy in self-governance" and "a regional governmental body with statewide concerns." (Rapid Transit Advocates, Inc. v. Southern Cal. Rapid Transit Dist., supra, 185 Cal. App. 3d 996, 1000.) But SCRTD is not the governmental body implicated by the voting scheme at issue and the range of governmental powers exercised by it are irrelevant to our inquiry.

The political entities constitutionally relevant to the challenged voting scheme are the limited-purpose benefit districts, the organizing principle of which is the recoupment of some of the added economic value conferred on commercial property resulting from its proximity to the transit stations. The narrow purpose for which the districts are established is reflected in a voting scheme that limits the franchise to those who will directly and primarily enjoy the benefits of transit station siting and shoulder the reciprocal burden of assessments-owners of commercial property within the two proposed districts. In light of that congruence, we are satisfied that the governmental units at issue lack the indicia of "general governmental powers" and, on this leg of the equal protection inquiry, qualify as the sort of "special-purpose units of government" that are not subject to the strict requirements of Reynolds, supra, 377 U.S. 533.

C

[4a] We turn now to the cognate question whether the challenged voting classification is supported by a "genuine difference in the relevant interests" of those enfranchised and those excluded. (Lockport, supra, 430 U.S. 259, 268 [51 L. Ed. 2d 313, 322-323].) The analysis is advanced by a recognition of the "single-shot" nature of referenda. "In a referendum, the expression of voter will is direct, and there is no need to assure that the voters' views will be adequately represented through their representatives .... The policy impact of a referendum is also different in kind from the impact of choosing representatives .... [T]he referendum puts one discrete issue to the voters. That issue is capable ... of being analyzed to determine whether its adoption or rejection will have a disproportionate impact on an identifiable group of voters." (Id. at p. 266 [51 L.Ed.2d at p. 321].) At the outset, then, [1 Cal. 4th 671] we must recognize that the issue-specific nature of referenda in general and of the assessment district elections in particular reduces somewhat the prominence equal protection values would assume if representational interests were at stake in the challenged election.

Resolution of this part of the equal protection inquiry also requires us to make careful distinctions in identifying constitutionally relevant facts. While certain broad criteria are clear, the analysis ineluctably implicates a measure of constitutional line drawing. [2b] The touchstone of the high court's "primarily affected or interested" doctrine is the extent of the impact of the election on those within and those outside the challenged voting classification. As noted, two complementary estimates must be made: whether the class of eligible voters enfranchised is disproportionately affected by the election issue, and whether those excluded are in fact substantially less interested in its outcome. Importantly, the court has stressed that, in weighing the statutory classification against the factual record, absolute distinctions between affected classes are not constitutionally compelled. The inquiry is the relative one of identifying differences sufficiently substantial to sustain the classification; the fact that some of those excluded from voting in the election may be "affected" by its outcome is not in itself fatal: "[c]onstitutional adjudication cannot rest on any such 'house that Jack built' foundation ...." (Salyer, supra, 410 U.S. 719, 731 [35 L.Ed.2d at p. 668].)

From the undeniably correct premise that public transportation is an issue affecting all citizens, the Court of Appeal reasoned that, the development of an urban mass transit system being crucial to the orderly growth of a metropolitan area, its financing, construction and operation necessarily affected all segments of the population. Specifically noting the interests of commercial lessees within the assessment districts who are excluded from voting in referenda, but championing as well what it termed the "same" interests of "other residents, both within and without the [benefit districts] who will be impacted" by Metro Rail, the Court of Appeal concluded that, along an axial line, this case was factually closer to two municipal bond referendum cases decided by the high court, Cipriano v. City of Houma (1969) 395 U.S. 701 [23 L. Ed. 2d 647, 89 S. Ct. 1897] (Cipriano), and Phoenix v. Kolodziejski (1970) 399 U.S. 204 [26 L. Ed. 2d 523, 90 S. Ct. 1990] (Phoenix), than it was to Salyer or Ball. It was thus one in which the principle of one person, one vote applied, the court concluded. Our analysis of these cases and the facts surrounding the benefit districts impels us to disagree with the Court of Appeal.

In the two cases relied upon by the Court of Appeal to support its result, the high court concluded that the differences "between the interests of [1 Cal. 4th 672] property owners and the interests of nonproperty owners [were] not sufficiently substantial to justify excluding the latter from the franchise." (Phoenix, supra, 399 U.S. 204, 209 [26 L. Ed. 2d 523, 527].) In Cipriano, a municipally owned utility called an election to approve the issuance of $10 million in revenue bonds to finance improvements to the utility. Under state law, only "property taxpayers" were qualified to vote in such a referendum. The court invalidated the voting limitation as underinclusive, concluding that those excluded from voting were not " 'in fact substantially less interested or affected [by the outcome] than those the statute includes.' " (395 U.S. at p. 704 [23 L.Ed.2d at p. 651], quoting Kramer, supra, 395 U.S. 621, 632 [23 L. Ed. 2d 583, 592].) Since the operations of the public utility affected virtually every resident of the governmental unit (the city)-the rates of all users would be directly affected by the utility's debt service requirements, and the bonds were paid out of utility revenues rather than property taxes-the impact of the revenue bond issue on those permitted to vote had little to do with their status as property owners. The challenged scheme thus excluded qualified voters who were "as substantially affected and directly interested in the matter voted upon as ... those ... permitted to vote." (Cipriano, supra, 395 U.S. at p. 706 [23 L.Ed.2d at pp. 651-652].)

Similarly, in Phoenix, supra, 399 U.S. 204, the court invalidated a statutory voting scheme restricting the franchise to real property taxpayers in a referendum called by the city to approve general obligation bonds, the proceeds of which were to be used to finance multiple municipal improvements-"the city sewer system, parks and playgrounds, police and public safety buildings, and libraries." (Id. at p. 206 [26 L.Ed.2d at p. 526].) In holding that the Reynolds (supra, 377 U.S. 533) principle of one person, one vote applied to the bond referendum, the court identified three reasons why the challenged voting classification was not supported by sufficiently substantial differences between those permitted to vote and those excluded.

First, it was plain that all the residents of the municipality had a substantial and indistinguishable interest in the public facilities and services to be financed by the bond issue and thus would be substantially affected by the election outcome. (Phoenix, supra, 399 U.S. at p. 209 [26 L.Ed.2d at pp. 527-528].) Second, although under state law the city theoretically had recourse to a real property tax levy to support its bond issue, historically it had financed more than half of its debt requirements by revenues from nonproperty taxes, taxes paid by property and nonproperty owners alike. (Id. at pp. 209-210 [26 L.Ed.2d at pp. 527- 528].) Finally, the court recognized that a significant portion of those property taxes paid by owners to finance the bonds would ultimately be passed on to lessees in the form of higher rents, and to the general public as increases in the cost of goods and services, [1 Cal. 4th 673] thereby dispersing the financial burden beyond the limited class enfranchised. (Id. at p. 210 [26 L.Ed.2d at p. 528].)

D

[4b] Applying the teaching of these two cases to the circumstances presented by the record here, we reach a conclusion contrary to that of the Court of Appeal. First and foremost, unlike Cipriano, supra, 395 U.S. 701, and Phoenix, supra, 399 U.S. 204, in which it could be said that virtually all of the residents of the governmental unit had a beneficial interest in the outcome of the bond referendum-the prospect of improved utility service for resident consumers in the one case and an array of civic improvements and services in the other-nonvoting residents of the assessment districts have no specific beneficial interest in the proceeds of the assessments distinguishable from that of every other resident of the multicounty area comprising the transit district.

Although nonproperty owning residents of the assessment districts are "affected" by the outcome of the referendum, they are no more affected than any other resident of the greater Los Angeles metropolitan area served by Metro Rail. It is not contended by any party that the state or federal Constitution requires that assessment district referenda be open to all qualified electors in the entire geographical area served by Metro Rail. Since the beneficial impact of the assessment districts on nonvoting residents is indistinguishable from the impact on residents of the Metro Rail service area outside of the benefit districts, that impact is insufficient to require extension of the franchise to the excluded class.

Moreover, again unlike the circumstances in Cipriano, supra, 395 U.S. 701, and Phoenix, supra, 399 U.S. 204, the economic burden of the assessments does not fall indiscriminately on property- and non-property- owning residents of the benefit districts alike. If the benefit districts are approved, the levy will fall directly on precisely that limited class enfranchised by the statutory scheme-owners of commercial property within the district; nonvoting residents of the districts will bear no discernably direct financial burden as a result of the assessments, nor will those residing outside the benefit districts.

And while one possible result of the failure to gain voter approval of the benefit districts may be to force SCRTD to look elsewhere to make up for lost funding, the possibility of a future search for alternative revenues that could conceivably result in some financial impact on the taxpayers of much of Southern California in the form of additional levies does not mean that all [1 Cal. 4th 674] qualified voters residing within the benefit districts must be permitted to vote. Again, the economic effect is remote and diluted rather than direct and substantial, and for that reason does not implicate interests sufficient to trigger the principle of Reynolds, supra, 377 U.S. 533. As in Salyer, supra, 410 U.S. 719, 729 [35 L. Ed. 2d 659, 667], "there is no way that the economic burdens" of the assessment districts "can fall on residents qua residents."

Third, unlike both Cipriano, supra, 395 U.S. 701, and Phoenix, supra, 399 U.S. 204, this is not a case in which those residents of the governmental unit denied the franchise will in fact contribute to the burden of financing Metro Rail as directly as those enfranchised. The constitutionally decisive fact is that, under the assessment scheme, only commercial real property within the two districts will be assessed-an immediate and tangible economic burden that is confined to commercial property and does not implicate directly any economic interest of either non-property-owning residents or residential property owners within the assessment districts.

As in Salyer, supra, 410 U.S. 719, and Ball, supra, 451 U.S. 355, the "activities" of the assessment districts-the raising of revenue to defray in part the cost of Metro Rail-will affect disproportionately owners of commercial property within them; it is they who will most directly feel both the beneficial economic effects of the transit station locations and bear the financial burden of the annual assessments. Likewise, all district "costs" (the assessments themselves) "are assessed against land"; district "operations" (again, the raising of revenue to finance public improvements directly benefiting the enfranchised class) "primarily affect the land within [district] boundaries." (410 U.S. at p. 729 [35 L.Ed.2d at p. 667].)

This is not to deny that some within the benefit districts who are not permitted to vote will be more "affected" by the proposed assessments than those outside. Notably, we may assume, as did the court in Salyer, supra, 410 U.S. 719, that a subclass of those within the two benefit districts whose economic interests are analogous to those of commercial property owners-commercial lessees in this case-will be affected secondarily by the assessments because of "pass through" clauses in their lease agreements. But so, too, are consumers of goods and services retailed by this subclass "affected," as the financial consequences of the assessments are presumably distributed throughout the region. In the specific factual context of this case, however, we do not find this prospect sufficiently substantial to invoke the demands of Reynolds, supra, 377 U.S. 533.

As in Salyer, supra, 410 U.S. 719, we think that recognition of the indirect and secondary affect on this limited class smacks too much of "house that [1 Cal. 4th 675] Jack built" casuistry to support constitutional determinations. As mentioned, the equal protection inquiry is one of constitutional line drawing; the cases do not teach that the limited class enfranchised "must be the only parties at all affected ... or that their entire economic well-being must depend on that [special-purpose unit of government]." (Ball, supra, 451 U.S. 355, 371 [68 L. Ed. 2d 150, 163].) Voting power may constitutionally be apportioned "to give greater influence to the constituent groups found to be most affected by the governmental unit's functions." (Lockport, supra, 430 U.S. 259, 266 [51 L. Ed. 2d 313, 321], italics added.) In short, we are satisfied that those excluded from voting under the statutory scheme at issue in this case are not "as substantially affected and directly interested in the matter voted upon as are those who are permitted to vote." (Cipriano, supra, 395 U.S. 701, 706 [23 L. Ed. 2d 647, 651- 652].)

Finally, we cannot fail to note the apparent inconsistency of interveners on this issue. Although urging that the principle of Reynolds, supra, 377 U.S. 533, applies, they nevertheless acknowledge the superior fairness of a voting scheme that would offend the principle of one person, one vote, by contending that the proper voting classification would include commercial property owners and tenants as well as residents of the benefit districts. Were the Reynolds principle to prevail here, however, it is district residents, precisely that class least affected by the assessment scheme, who would be enfranchised, to the exclusion of nonresident corporate property owners and probably most commercial tenants of the two districts. As in Salyer, supra, 410 U.S. 719, 730 [35 L. Ed. 2d 659, 667], to sustain interveners' contention "would not result merely in the striking down of an exclusion from what was otherwise a delineated class, but would instead engraft onto the statutory scheme a wholly new class of voters in addition to those enfranchised by the statute."

[5] We conclude, therefore, that neither the "special-purpose unit of government" nor the "primarily affected or interested" analysis yields a requirement that the principle of one person, one vote must be applied to benefit assessment district referenda.

E

[6] Of course, respondents are entitled to have their claims adjudicated under the equal protection requirement applicable in this case, namely, that the statutory voting scheme not be " 'wholly irrelevant' " to the " 'achievement of the [statutory] objectives,' [citation]." (Salyer, supra, 410 U.S. 719, 730 [35 L. Ed. 2d 659, 668].) The question whether the voting classification meets that constitutional standard is one that we examine in the abstract; it is [1 Cal. 4th 676] not whether we would authorize the same exclusions were we the Legislature, but whether "any state of facts reasonably may be conceived to justify" a voting scheme limiting the franchise to owners of record of commercial real property located within the benefit districts. (Id. at p. 732 [35 L.Ed.2d at pp. 668-669], citing McGowan v. Maryland, supra, 366 U.S. 420, 426 [6 L. Ed. 2d 393, 399].)

The challenged scheme denies the vote to two identifiable subgroups within the benefit districts-residents, whether lessees or owners, of non- commercial real property, and lessees of commercial property. The Legislature reasonably could have permitted the exclusion of the former on the obvious ground that limiting voting to those who will directly bear the cost of the assessments is demonstrably fairer or more equitable than including those whose affirmative vote carries no personal financial consequences or risk. As did the court in Salyer, supra, 410 U.S. 719, 731 [35 L. Ed. 2d 659, 668], we conclude that nothing in the equal protection clause precludes the total exclusion of "those who merely reside within the district."

The case with respect to commercial lessees is only slightly less evident. As noted, interveners' claim with respect to this class is that, by virtue of "pass through" clauses in commercial lease agreements, they (or some of them) will bear a financial impact as a result of the assessments that is so closely analogous to that affecting the class enfranchised as to be indistinguishable for equal protection purposes. Here again, however, it was for the Legislature to draw the line. And again, as in Salyer, supra, 410 U.S. 719, 732 [35 L. Ed. 2d 659, 668-669], we think that it reasonably could have drawn the line that it did in light of the significant administrative difficulties that foreseeably would have arisen had the vote been extended to the "pass through" class of commercial tenants.

As the transit district points out, including this class within those enfranchised would require those administering the referenda to determine a multitude of discrete voter qualification issues-identifying those commercial leases with "pass through" provisions and those without, the percentage of the assessment passed to particular tenants, whether partial "pass throughs" are permitted, the duration of commercial leases, and other electoral minutiae which have not occurred to us. Although imperfect, the "rough accommodation" to practicality and administrative convenience chosen passes constitutional muster. (Salyer, supra, 410 U.S. 719, 732 [35 L. Ed. 2d 659, 668-669]; Dandridge v. Williams (1970) 397 U.S. 471, 485-487 [25 L. Ed. 2d 491, 501-503, 90 S. Ct. 1153]; Wood v. Public Utilities Commission (1971) 4 Cal. 3d 288, 295, fn. 2 [93 Cal. Rptr. 455, 481 P.2d 823]; United States Steel Corp. v. Public Utilities Com. (1981) 29 Cal. 3d 603, 613-614 [1 Cal. 4th 677] [175 Cal. Rptr. 169, 629 P.2d 1381].) We cannot conclude that the Legislature was unreasonable in drawing the line so as to deny the franchise to this class.

III

The conclusion that the distribution of the elective franchise in benefit district referenda is not subject to the strict demands of Reynolds, supra, 377 U.S. 533, also yields the answer to interveners' ancillary claim that the manner in which the Legislature chose to apportion votes and levy assessments among those enfranchised is constitutionally defective. [7a] The specific contention is that principles of equal protection require that those who pay the most in assessments be alloted the most votes in the referenda. Absent this substantial "proportionality" between voting power and financial burden, interveners argue, the statutory voting scheme lacks a rational basis.

As authority for this argument, interveners rely entirely on a sentence in Ball, supra, 451 U.S. 355, and language in the opinion of the Court of Appeal in Whitlock, supra, 22 Cal. App. 3d 863. Interveners seize on the statement in Ball that an acreage-based vote allotment scheme is rational because it "reasonably reflects the relative risks ... incurred [by] landowners and the distribution of the benefits and burdens of the District's water operations." (451 U.S. at p. 371 [68 L.Ed.2d at p. 163], fn. omitted.) They also invoke language in Whitlock sustaining as rational a statutory provision for the termination of assessment proceedings on the protest of owners of more than one-half of the land affected. fn. 7

We cannot regard these brief remarks as definitive on the issue, however, impliedly condemning alternative vote allotment schemes. In Salyer, supra, 410 U.S. 719, 734 [35 L. Ed. 2d 659, 669-670], the court upheld a statutory scheme that alloted votes according to the tax assessed value of real property and imposed assessments on a project-specific basis according to the benefit conferred, a divided arrangement not unlike the statutory scheme challenged in this case. The high court declined to declare such a scheme not rationally based, given the rough proportionality between voting strength and assessment burden. (Ibid.) As in Salyer, we think that the Legislature was entitled [1 Cal. 4th 678] to assume that within these relatively small benefit assessment districts, assessed valuation bears a rough relationship to lot or building size and that the vote allotment and assessment formulas are thus not unreasonably disproportionate for equal protection purposes. fn. 8

Alternatively, the conclusion that the circumstances surrounding the benefit district referenda satisfy the constitutional criteria for an exception to the principle of one person, one vote, necessarily means that the proponent of the challenged statutory voting scheme must demonstrate only that it is not "wholly irrelevant" to the objectives of the statute. (Salyer, supra, 410 U.S. 719, 730 [35 L. Ed. 2d 659, 668].) The equal protection issue raised by interveners' proportionality claim, therefore, is simply whether the divided arrangement chosen by the Legislature bears some reasonable relation to the statutory scheme. That is, we perceive no constitutional basis independent of Reynolds, supra, 377 U.S. 533, for what is, in substance, simply an argument for a scheme of weighted voting different from that chosen by the Legislature, one that gives greater voting power to some property owners rather than others. fn. 9

That said, it is not difficult to imagine legitimate concerns that might have motivated the Legislature to adopt differing formulas for the allotment of votes and the calculation of assessments. It might, for example, have adopted [1 Cal. 4th 679] the working assumption that building or lot size bears a rough but reasonably direct relationship to the enhanced economic benefits resulting from the location of Metro Rail facilities, a supposition that would rationally justify basing assessment calculations on square footage or parcel size. It may be true, as interveners point out, that such an assumption is faulty when applied to warehouse space, for example, but at this level of equal protection analysis we must deal in imperfect generalities. Given the interests at stake here, the Legislature has wide latitude within which to draw lines before a reviewing court can say it has crossed over into the zone of complete irrelevancy.

Likewise, the Legislature's direction that current ad valorem tax assessment rolls be used to identify qualified voters and allot votes has the merit of accuracy, simplicity and administrative convenience. At least where the countervailing interests are not fundamental in the constitutional sense or otherwise entitled to special solicitude, these are virtues that are not lightly abandoned. (Salyer, supra, 410 U.S. 719, 732-733 [35 L.Ed.2d at pp. 668-669].) It was hardly unreasonable for the Legislature to direct that votes be alloted by means of existing tax assessment rolls-an accurate, convenient, and verifiable basis for notice and one that permits benefit districts to be approved at referenda before undertaking the administratively detailed, prolonged, and costly task of calculating and verifying the square footage of each parcel subject to assessment.

[8] (See fn. 10.), [7b] Whatever the case, although both alloting votes and calculating assessments according to square footage might be more "equitable" in that it would tend to equalize the burden of assessments and voting power, we discern no constitutional basis for compelling the Legislature to adopt such an arrangement, on pain of having its voting scheme invalidated as lacking a rational basis. fn. 10

IV

A final matter requires our attention. [9] As noted, in addition to holding that the referendum voting scheme violated principles of equal [1 Cal. 4th 680] protection, the Court of Appeal also held that SCRTD lacked legal authority to exempt from assessment residential property within the two benefit districts. It reached this conclusion on nonconstitutional grounds, concluding that only the Legislature possessed the power to grant exemptions from special assessments and that it had neither done so here nor delegated that authority to SCRTD. We disagree. The Legislature expressly authorized the relevant local government to amend the transit district board's resolution creating the assessment district in various ways, including exempting residential property from assessment, and empowered SCRTD to adopt the resolution as amended.

Specifically, Public Utilities Code section 33001.5 requires that, before being established, any benefit district proposed by SCRTD be submitted to the relevant "governing body," defined by section 33001.5, subdivision (d), as "the city council of a city in which the proposed benefit district is located" or county board of supervisors if not within a city. The statute goes on to empower the governing body, after a public hearing, to "approve, or amend and approve, as amended, or disapprove the geographic boundaries of the benefit district and the method of assessment," following which the SCRTD board may by a two-thirds vote either establish the benefit district on the terms approved by the governing body or forgo its establishment. (Pub. Util. Code, § 33001.5, subds. (b) & (c).)

Relying on the rule that exemptions from special assessments "should ... be based on express statutory authority" (Hollywood Cemetery Assn. v. Powell (1930) 210 Cal. 121, 135 [291 P. 397, 71 A.L.R. 310]), and that a presumption, founded on distributive equity, extends assessment to all those who are beneficially affected by the improvement absent affirmative indications supporting an exemption (Cedars of Lebanon Hosp. v. County of L. A. (1950) 35 Cal. 2d 729, 747-748 [221 P.2d 31, 15 A.L.R.2d 1045]), interveners claim that the text of the statute fails to demonstrate the requisite clarity of legislative intent. The Court of Appeal agreed that the statute did not grant SCRTD any exemption power. We disagree and, for the reasons that follow, conclude that the exemption of residential property from the assessment districts at issue here is "based on express statutory authority." (Hollywood Cemetery Assn. v. Powell, supra, 210 Cal. 121, 135.)

Public Utilities Code section 33001.5, subdivision (b), expressly authorizes the City Council of Los Angeles to "amend and approve, as amended ... the geographic boundaries of the benefit district and the method of assessment." Although the term "method of assessment" is not defined, the Legislature previously has used this phrase to encompass the determination of the type of property upon which an assessment will be levied. (Stats. 1977, ch. 1218, §§ 3, 4, [1 Cal. 4th 681] pp. 4101-4103, Deering's Wat. Uncod. Acts (1991 pocket supp.) Act 7150, §§ 13, 13.5, pp. 9-10.) Accordingly, we conclude that by using this phrase in section 33001.5, the Legislature authorized Los Angeles to amend the method of assessment by excluding residential properties from the assessment district. Our conclusion is butressed by the fact that, in addition to receiving authorization to amend the method of assessment, the city was statutorily authorized to amend the geographic boundaries of the benefit district. It is clear, therefore, that the Legislature intended the city to have a significant voice in determining which properties would be subject to the special assessment.

Finally, subdivision (c) of section 33001.5 of the Public Utilities Code provides that once the city has acted, SCRTD must decide "whether to create the benefit district as approved by the [city]." The Legislature thus granted SCRTD authority to create the assessment district as amended by the city, including the exemption of residential properties from assessment.

Conclusion

Early in our history, the high court observed that "the science of government is ... the science of experiment." (Anderson v. Dunn (1821) 19 U.S. [6 Wheat.] 204, 226 [5 L. Ed. 242, 247].) Not long ago, in upholding an innovation in the government of a county school system, the court reminded us that "[v]iable local governments may need many innovations, numerous combinations of old and new devices, great flexibility in municipal arrangements to meet changing urban conditions." (Sailors v. Board of Education (1967) 387 U.S. 105, 110-111 [18 L. Ed. 2d 650, 654-655, 87 S. Ct. 1549].) In the circumstances of this case, combining an old device with a new setting, we see "nothing in the Constitution to prevent experimentation." (Id. at p. 111 [18 L.Ed.2d at p. 655].)

The judgment of the Court of Appeal is reversed.

Lucas, C. J., Panelli, J., Baxter, J., and George, J., concurred.

KENNARD, J.

I dissent.

The challenged electoral system, under which none but owners of commercial property may vote on the imposition of special benefit assessments to finance rail rapid transit stations, violates the equal protection guarantee of the Fourteenth Amendment of the United States Constitution. Because the construction of rapid transit stations is a matter of concern to all residents of the area surrounding the sites of the proposed stations, and because the [1 Cal. 4th 682] burden of the assessments will be passed on to other members of the community in the form of higher rents and higher prices for goods and services, the right to vote on a financing mechanism for the stations may not be restricted to owners of commercial property. The contrary conclusion reached by the majority is based on a faulty analysis derived from a misreading of the applicable decisions of the United States Supreme Court.

I

The Legislature established the Southern California Rapid Transit District (hereafter the SCRTD) to construct, operate, and maintain "a comprehensive mass rapid transit system in the southern California area, and particularly in Los Angeles County." (Pub. Util. Code, § 30001, subd. (a).) The governing body of the SCRTD is its board of directors (hereafter the Board). (Id., § 30200.)

Recognizing that "rail rapid transit facilities and services provide special benefits to parcels of land, and improvements thereon, in the vicinity of rail rapid transit stations" (Pub. Util. Code, § 33000, subd. (b)), the Legislature authorized the SCRTD to identify the area around each transit station that would be specially benefited by the station, to designate this area a benefit assessment district (or a zone within a benefit assessment district), and to levy special benefit assessments on property so designated (id., §§ 33000, subd. (a), 33001). Imposition of the special benefit assessments is "for the purpose of financing, in whole or in part, the acquisition, construction, development, joint development, operation, maintenance, or repair of one or more rail transit stations and rail transit related facilities located within the benefit districts." (Id., § 33001, subd. (a).) Revenue derived from a special benefit assessment, or from bonds secured by such assessment, may be used only for financing the facility for which it was levied. (Id., § 33002, subd. (d).)

The Board may establish the benefit districts and impose the assessments without an election (Pub. Util. Code, § 33002.1) unless presented with a petition for election "signed by the owners of at least 25 percent of the assessed value of real property within the benefit district." (Id., § 33002.2.) If a petition with the required signatures is presented, the SCRTD holds an election at which only owners of assessed property may vote and each such property owner "may cast one vote for each one thousand dollars ($1,000), or fraction thereof, worth of land or improvements owned by the voter in the benefit district as is shown on the most recent equalized assessment roll." (Id., § 33002.3, subd. (b).)

Using its statutory authority, the Board established two benefit districts (one of which contains four zones) consisting of land surrounding proposed [1 Cal. 4th 683] rail rapid transit stations. Although the Legislature had not expressly authorized the exemption of residential property from assessment, such property was exempted and the Board levied the assessment only on commercial property. The lawsuit now before this court challenges the electoral features of the existing scheme, under which only owners of commercial property can petition for and vote in an election on the benefit assessments.

II

Under the Fourteenth Amendment to the federal Constitution, no state may "deny to any person within its jurisdiction the equal protection of the laws." In the context of state and local voting systems, this guarantee means that the government may not impose a voting restriction other than residence, age, or citizenship, in elections of general interest, unless it can demonstrate that the restriction is necessary to promote a compelling state interest. (Hill v. Stone (1975) 421 U.S. 289, 297 [44 L. Ed. 2d 172, 178-179, 95 S. Ct. 1637].) The SCRTD does not contend that the voting restriction at issue here, which limits the franchise to owners of commercial property, can withstand this strict scrutiny. The primary question to be answered, therefore, is whether an election on the imposition of special benefit assessments to finance rail rapid transit stations is an election of general or special interest.

To support its conclusion that the restrictions at issue here are constitutionally permissible, the majority relies heavily on Salyer Land Co. v. Tulare Water District (1973) 410 U.S. 719 [35 L. Ed. 2d 659, 93 S. Ct. 1224] (hereafter Salyer), and Ball v. James (1981) 451 U.S. 355 [68 L. Ed. 2d 150, 101 S. Ct. 1811] (hereafter Ball). As I will explain, these cases are of limited relevance because they concern elections to select representatives rather than, as here, elections to decide discrete issues. To the extent these cases are relevant, they support the conclusion that voting restrictions in elections conducted by the SCRTD are not exempt from strict scrutiny.

In Salyer, supra, 410 U.S. 719, and Ball, supra, 451 U.S. 355, the issue before the United States Supreme Court was the validity, under the Fourteenth Amendment's equal protection guarantee, of a law limiting to property owners the right to vote in elections of the governing body of a water district. In upholding the property ownership restrictions, the high court explained that a water district, in the discharge of its primary function, does not exercise traditional governmental powers. Noting that a water district exists to acquire, store, and distribute water, the court commented that the district at issue in Salyer provided "no other general public services such as schools, housing, transportation, utilities, roads, or anything else of the type [1 Cal. 4th 684] ordinarily financed by a municipal body... and it does not have a fire department, police, buses, or trains." (Salyer, supra, at pp. 728-729 [35 L.Ed.2d at pp. 666-667], italics added.) Although the district at issue in Ball did provide a "general public service" by producing and distributing electrical power, this activity was merely incidental to its primary purpose and thus, the court concluded, the provision of this utility service could not change the district's character. (Ball, supra, at pp. 368-369 [68 L.Ed.2d at pp. 161-162].)

The United States Supreme Court also emphasized that the actions of water districts disproportionately affect landowners because district costs are assessed against landowners in proportion to the benefits received, with delinquencies becoming a lien on the land. (Salyer, supra, 410 U.S. 719, 729 [35 L.Ed.2d at p. 667].) Although the district at issue in Ball provided much of its water for nonagricultural uses, the court explained that the "constitutionally relevant fact" was that all of its water was distributed according to land ownership. (Ball, supra, 451 U.S. 355, 367 [68 L.Ed.2d at pp. 160-161].) The court concluded that the districts "remain essentially business enterprises, created by and chiefly benefiting a specific group of landowners." (Id. at p. 368 [68 L.Ed.2d at p. 161].)

The SCRTD's primary purpose and manner of operation contrast sharply with those of the water districts discussed in Salyer, supra, 410 U.S. 719, and Ball, supra, 451 U.S. 355. The SCRTD's primary purpose is to establish and operate an urban rail mass transit system. Providing urban mass transportation is a task traditionally performed by local government, as the United States Supreme Court recognized when it included transportation among the "general public services" that local governments have historically provided to their citizens. (Salyer, supra, at pp. 728-729 [35 L.Ed.2d at pp. 666-667]; see also, Cunningham v. Municipality of Metropolitan Seattle (W.D.Wash. 1990) 751 F. Supp. 885, 890 [applying one-person, one-vote principle to special district providing mass transit and water pollution abatement].)

If a public entity's primary purpose is to provide even a single traditional governmental service, the federal Constitution may require strict scrutiny of voting restrictions in elections of its governing body. (See, e.g., Hadley v. Junior College District (1970) 397 U.S. 50 [25 L. Ed. 2d 45, 90 S. Ct. 791] [applying one-person, one-vote principle to election for school board members]; Fumarolo v. Chicago Bd. of Educ. (1990) 142 Ill. 2d 54 [566 N.E.2d 1283, 1295].) The transportation service that the SCRTD provides is not distributed according to land ownership but is available to all who choose to ride its trains. Although the building of a rail mass transit system provides special benefits to some landowners, these benefits are incidental to the rapid [1 Cal. 4th 685] transit district's primary purpose. The chief beneficiaries of the transit system are those who use it for transportation, a class unrelated to land ownership. Finally, only a small portion of the SCRTD's revenues will be derived from assessments on land. Its primary sources of funding are grants from the federal and state governments and the fares to be collected from transit patrons. Given these many significant differences between the SCRTD and the water districts considered by the United States Supreme Court in Salyer, supra, 410 U.S. 719, and Ball, supra, 451 U.S. 355, this case is not controlled by the decisions in those two cases.

The majority opinion does not analyze the issue in these terms. Instead of comparing the water districts at issue in Salyer, supra, 410 U.S. 719, and Ball, supra, 451 U.S. 355, with the SCRTD, the majority compares them with the benefit assessment districts. This is fundamentally erroneous, and indeed senseless, because the SCRTD itself, not the benefit districts, imposes and collects the special benefit assessments and conducts any election that may be required. (Pub. Util. Code, § 33000 et seq.) Furthermore, although a rapid transit district and a benefit assessment district are both called "district," only one of them-the rapid transit district-exists as a legal entity. A benefit assessment district is merely a geographical area within the SCRTD's borders identified by the Board for the purpose of imposing the assessment. Such a "district" has no governing body, no employees, and no powers or responsibilities. It is not a public entity. (See Dawson v. Town of Los Altos Hills (1976) 16 Cal. 3d 676, 683 [129 Cal. Rptr. 97, 547 P.2d 1377].) Because it is not a unit of government, a benefit assessment district, unlike the SCRTD itself, cannot be meaningfully compared to the water districts at issue in Salyer and Ball.

The conclusion to be drawn from Salyer, supra, 410 U.S. 719, and Ball, supra, 451 U.S. 355, is that the SCRTD is a governmental entity of general powers. If the election at issue were for the purpose of selecting the SCRTD's governing body, it would be an election of general interest and restrictions on the franchise other than residence, age, or citizenship would therefore be subject to strict scrutiny. (Hill v. Stone, supra, 421 U.S. 289, 297 [44 L. Ed. 2d 172, 178- 179].) But the scope of the election is narrower. Its purpose is to determine whether special benefit assessments shall be imposed. To decide whether such an election, conducted by a governmental entity of general powers, is an election of general interest, it is appropriate to consider decisions of the United States Supreme Court concerning similar limited- purpose elections.

As the high court has emphasized, the equal protection principles applicable to "an election involving the choice of legislative representatives" have [1 Cal. 4th 686] only "limited relevance" in determining the validity of restrictions on the franchise in a " 'single-shot' referendum." (Lockport v. Citizens for Community Action (1977) 430 U.S. 259, 266 [51 L. Ed. 2d 313, 321, 97 S. Ct. 1047].) Unlike an election of legislative representatives, a referendum "puts one discrete issue to the voters," and the proposal can be analyzed "to determine whether its adoption or rejection will have a disproportionate impact on an identifiable group of voters." (Ibid.) If adoption or rejection of the proposal that is the subject of the referendum would have such a disproportionate impact, "the question then is whether a State can recognize that impact either by limiting the franchise to those voters specially affected or by giving their votes a special weight." (Ibid.)

Here, it cannot be questioned that the decision to impose a special benefit assessment will have a particular impact on an identifiable group of voters-owners of the property on which the assessment is imposed. But the existence of this special impact is not in itself sufficient to justify restriction of the franchise to the class specially affected in this manner. The relevant analysis is found in another United States Supreme Court decision, Phoenix v. Kolodziejski (1970) 399 U.S. 204 [26 L. Ed. 2d 523, 90 S. Ct. 1990] (hereafter Phoenix).

The high court held in Phoenix, supra, 399 U.S. 204, that in an election to approve a municipality's issuance of general obligation bonds, a state could not restrict the franchise to real property taxpayers, even though the municipality substantially relied on property taxes to service the bonds. The court concluded that "[t]he differences between the interests of property owners and the interests of nonproperty owners are not sufficiently substantial to justify excluding the latter from the franchise." (Id. at p. 209 [26 L.Ed.2d at p. 527].)

The court gave three reasons for this conclusion. First, all municipal residents had a substantial interest in the facilities and services financed by the bonds. "Presumptively, when all citizens are affected in important ways by a governmental decision subject to a referendum, the Constitution does not permit weighted voting or the exclusion of otherwise qualified citizens from the franchise." (Phoenix, supra, 399 U.S. 204, 209 [26 L.Ed.2d at p. 527].) Second, the municipality would not rely entirely on property taxes to service the bond debt, but would also rely in significant part on other local taxes paid by nonproperty owners. (Id. at pp. 209-210 [26 L.Ed.2d at pp. 527-528].) Finally, the landowners could redistribute the property tax burden to others in the community in the form of higher rents and, in the case of commercial property, in the form of higher prices for goods and services produced or sold on the taxed property. (Id. at pp. 210-211 [26 L.Ed.2d at pp. 528-529].) [1 Cal. 4th 687]

This analysis compels a similar conclusion here. As the United States Supreme Court recognized, a vote on a revenue measure cannot be divorced from the facilities and services to be financed by the revenue. Here, the issue addressed by the special benefit assessment election is not just the imposition of the assessment, but also the construction of the mass transit stations that the assessments will finance. The stations will generate an increase in commercial activity in the areas around the stations, as the majority acknowledges. This increase in commercial activity will make the areas more attractive for certain kinds of high volume businesses (e.g., fast-food outlets), and less attractive for other, more neighborhood-oriented businesses (e.g., laundromats). Inevitably, all residents of the benefit assessment area, not just the owners of commercial property, will be affected in important ways by the change in the commercial environment resulting from the location of the transit station.

The effects will not be confined to commercial activity. In most instances, residential property located near proposed rapid transit stations will increase in value in recognition of the convenience of ready access to the transit system. fn. 1 This increase in value will translate into higher rents for rented dwellings and higher sales prices for owner-occupied dwellings. These increases in housing costs will significantly affect the residents of the area immediately surrounding the proposed stations.

Because the construction of the transit stations affects all community residents in important ways, the exclusion of residents who own no commercial land is presumptively a violation of equal protection. Although the assessments will be levied initially on the owners of commercial property, they can redistribute the burden to other community residents. Rents charged to commercial tenants will certainly increase. Indeed, it is undisputed that most commercial leases in the benefit assessment areas contain "pass through" provisions under which the tenant assumes liability for any tax or assessment levied on the property. The occupant of the premises who pays the assessment, whether landowner or tenant, can recover the cost from consumers, many of whom will be local residents, by increases in the prices of goods and services produced or sold on the taxed property.

Under the test articulated in Phoenix, supra, 399 U.S. 204, which this court is required to employ, the election on the SCRTD's special benefit [1 Cal. 4th 688] assessments is an election of general interest, in which restrictions other than residence, age, and citizenship must be subjected to strict scrutiny. The restriction imposed, which limits the franchise to owners of commercial property, concededly cannot withstand such scrutiny. Accordingly, the existing system for the SCRTD special benefit assessment elections, by disenfranchising all but owners of commercial property subject to assessment, violates the equal protection guarantee of the Fourteenth Amendment to the United States Constitution.

III

Like the other members of this court, I am reluctant to accept a conclusion that might impede the construction of needed public facilities, and the need for a modern and efficient rapid transit system in the greater Los Angeles area cannot be denied. Yet, as the United States Supreme Court has noted, restrictions on the franchise that violate equal protection cannot be justified "on exigencies of history or convenience." (New York City Bd. of Estimate v. Morris (1989) 489 U.S. 688, 703, fn. 10 [103 L. Ed. 2d 717, 733, 109 S. Ct. 1433].) Thus, like the Court of Appeal, I conclude that, under controlling federal precedent, the existing electoral system for the SCRTD's special benefit assessments is invalid on its face.

Because the electoral system is invalid for the reasons I have stated, I find it unnecessary to consider the other bases on which that system has been challenged in this litigation. Having concluded that the existing electoral system violates the federal Constitution, I would affirm the judgment of the Court of Appeal.

Mosk, J., concurred.

FN 1. We are advised by the parties that the Metro Rail segment is now commonly called the "Red Line" and that the Los Angeles County Transportation Commission has recently assumed contractual responsibility for all uncompleted construction work on the first operating element.

FN 2. The difference of roughly $70 million between the $130.3 million for capital expenditure and the $200 million bond cap authorized by the board covered interest costs, bond issuance fees, and administrative costs associated with the assessment district program.

FN 3. The requirement of substantial equality in voting power is derived from the basic tenet of Reynolds that democratic government is fundamentally representative in character. "[R]epresentative government is in essence self-government through the medium of elected representatives .... Full and effective participation by all citizens in ... government requires, therefore, that each citizen have an equally effective voice in the election of [representatives]." (Reynolds, supra, 377 U.S. 533, 565 [12 L. Ed. 2d 506, 529].)

FN 4. Parenthetically, we reject at the outset the proposition that the principle of Reynolds, supra, 377 U.S. 533, is triggered simply because a limited class of those otherwise qualified to vote is enfranchised by the voting scheme in issue. Despite sweeping language in early post- Reynolds decisions of the high court suggesting that enfranchising any class of qualified voters is itself sufficient to invoke the principle of one person, one vote, and thus to require the state to demonstrate a compelling need for the classification (see, e.g., Kramer v. Union School District (1969) 395 U.S. 621, 627 [23 L. Ed. 2d 583, 589-590, 89 S. Ct. 1886] (Kramer)), later cases have undermined that proposition to such an extent that it fairly can be doubted to be the rule. Indeed, we cannot accept it as an accurate statement of the high court's views of the compulsions of equal protection in the distribution of the voting franchise. (See post, p. 667 et seq.)

FN 5. A third decision in the trilogy of high court opinions upholding property-based voter qualification schemes bears mention. In Associated Enterprises, Inc. v. Toltec District (1973) 410 U.S. 743 [35 L. Ed. 2d 675, 93 S. Ct. 1237], the court, in a brief per curiam ruling decided the same day as Salyer, supra, 410 U.S. 719, held that a Wyoming statute conditioning voting in a referendum authorizing the establishment of a watershed district on property ownership did not violate equal protection, the district being "a governmental unit of special or limited purpose whose activities have a disproportionate effect on landowners within the district." (410 U.S. at p. 744 [35 L.Ed.2d at p. 677].)

FN 6. Between Salyer and Ball, this court decided Choudhry v. Free (1976) 17 Cal. 3d 660 [131 Cal. Rptr. 654, 552 P.2d 438]. There we invalidated a provision of the Irrigation District Law (Wat. Code, § 21100), requiring directors of irrigation districts to be freeholders, as it applied to prevent a nonproperty owner's candidacy for director of the Imperial Irrigation District. In part, our decision in Choudhry was grounded in equal protection concerns arising from the statute's interference with ballot access and the "real and appreciable impact" test on the electoral process formulated in Bullock v. Carter (1972) 405 U.S. 134, 144 [31 L. Ed. 2d 92, 100, 92 S. Ct. 849]. (17 Cal. 3d 660, 664- 665; cf. Anderson v. Celebrezze (1983) 460 U.S. 780 [75 L. Ed. 2d 547, 103 S. Ct. 1564].) In other respects, our decision rested on a "governmental function/specially affected" analysis derived from Reynolds, supra, 377 U.S. 533, and its progeny. (17 Cal. 3d 660, 666-668.) We distinguished the scope of the governmental powers of the water storage district in Salyer from the more extensive powers of the Imperial Irrigation District and contrasted the limited impact of district activities in Salyer with those of the Imperial Irrigation District, the largest irrigation district in California and one the scope and effects of which were akin to the reclamation district in Ball, supra, 451 U.S. 355. (See 17 Cal. 3d 660, 663-664, 667-668.)

FN 7. The precise claim in Whitlock was that the protest scheme under review-a section of the Special Assessment Investigation, Limitation and Majority Protest Act of 1931 (Sts. & Hy. Code, § 2905)-discriminated against small landowners by permitting assessment proceedings to be aborted by those owning more than one-half of the total land area affected. The Court of Appeal rejected this argument in the following language: "Since only those landowners who are directly benefited are charged with the cost of the improvements in proportion to the benefit conferred and since land area bears some reasonable relationship to the amount of the assessment, there is a rational basis for making the governmental decision subject to landowners' protest and in measuring the sufficiency of the protest by the land area protested." (22 Cal.App.3d at p. 876.)

FN 8. It is true that Salyer was decided prior to the passage of Proposition 13, an event that, as interveners vigorously remind us, transformed the California real property tax landscape. By making the assessed value of real property for tax purposes essentially dependent on sale price, Proposition 13 has a "wild card" effect on property- based bifurcated vote allotment/assessment schemes such as the one at issue here. Since property held longer will be assessed for tax purposes at a lower rate than property recently transferred and reassessed (assuming continuing inflationary trends in real property values), a referendum scheme that allots votes according to tax assessment value may result in substantial disparities in the voting power and assessment burdens of real property of comparable market value and square footage. As interveners point out, under the statutory scheme, the voting power of comparably sized parcels within the same benefit district may vary by as much as (in one case) a factor of 20, depending on the date of acquisition.

We have upheld substantial inequalities in the assessed value of comparable properties for ad valorem tax purposes against equal protection challenge and validated as reasonable the acquisition-value approach to property assessment embodied in Proposition 13. (Amador Valley Joint Union High Sch. Dist. v. State Bd. of Equalization (1978) 22 Cal. 3d 208, 233-236 [149 Cal. Rptr. 239, 583 P.2d 1281].) The fact that such a constitutionally valid tax assessment system may have the marginal result of diluting voting power in an assessment referendum is an inequality that the Legislature can remedy by amending the voting scheme; it does not constitutionally invalidate the voting classification per se.

FN 9. Interveners make much of the fact that following passage of the benefit assessment legislation, the Legislature passed an amendatory measure which would have mandated both voting allotment and assessments on a square footage or parcel basis; the measure, however, was vetoed by the Governor. As noted in the main text, the issue before us is one of constitutional limitations on the scope of legislative classifications, not "fairness" simpliciter; for that reason, these post hoc events have little force.

FN 10. Defendant Bolen summarily suggests that the statutory voting scheme violates article I, section 22, of the California Constitution; that section provides that "the right to vote or hold office may not be conditioned by a property qualification." We have long since construed article I, section 22 to " 'refer to the qualification of electors entitling them to vote at the ordinary elections, local and general, held in the course of the usual functions of civil government.' [Citation.]" (Tarpey v. McClure (1923) 190 Cal. 593, 606 [213 P. 983] [water storage district]; see also Wheeler v. Herbert (1907) 152 Cal. 224, 232 [92 P. 353]; Potter v. Santa Barbara (1911) 160 Cal. 349, 355 [116 P. 1101] [road improvement district]; Martinelli v. Morrow (1916) 172 Cal. 472, 473 [156 P. 1017] [comparable constitutional provision applies to political subdivisions exercising "governmental functions," not to "limited purpose" government such as municipal water district].) Nothing in the record or the arguments of the parties in this case persuades us that we should revisit these holdings.

FN 1. There may be situations in which a rapid transit station would depress the value of adjacent residential property. For instance, this could occur in the unlikely event that the station were to be located in a neighborhood of expensive single-family residences. For present purposes, the essential point is that construction of a rapid transit station is virtually certain to have some effect, either positive or negative, on the value of nearby residential property.