California voters passed Proposition 209 in 1996, adding section 31 to article I of the state Constitution (section 31). The provision forbids public entities from discriminating against, “or granting] preferential treatment to, any individual or group on the basis of race, sex, color, ethnicity, or national origin in the operation of *342public employment, public education, or public contracting.” (§31, subd. (a), italics added.) This court later construed section 31 to ban not only preferences already prohibited by the federal equal protection clause (U.S. Const, 14th Amend., § 1), but also race- and sex-conscious remedial measures that the United States Constitution would otherwise permit. (Hi-Voltage Wire Works, Inc. v. City of San Jose (2000) 24 Cal.4th 537, 567 [101 Cal.Rptr.2d 653, 12 P.3d 1068] (Hi-Voltage).) We are now asked to resolve whether section 31 violates the federal Constitution.1
At the outset, I note several uncontroversial facts. Despite language forbidding discrimination (which added nothing to existing law), there can be little doubt Proposition 209 was enacted because of its effect on race- and sex-conscious preferences (and, given the ballot materials, primarily the former). Nor is there any serious uncertainty that the race- and sex-conscious preferences eliminated by section 31 primarily benefitted racial minorities and women and were so intended. Finally, it is uncontroverted that Proposition 209 did more than merely repeal existing programs; by amending the state Constitution, Proposition 209 requires those who would seek new race- and sex-conscious preferential programs to obtain the passage of another statewide initiative. By contrast, those seeking preferences on all other bases can resort to the usual, less burdensome, political process.2 These facts, I conclude, require section 31 be invalidated as unconstitutional under the United States Supreme Court’s decisions in Hunter v. Erickson (1969) 393 U.S. 385 [21 L.Ed.2d 616, 89 S.Ct. 557] (Hunter) and Washington v. Seattle School Dist. No. 1 (1982) 458 U.S. 457 [73 L.Ed.2d 896, 102 S.Ct. 3187] (Seattle).
In Hunter and Seattle, the high court established what has been called the political structure doctrine, a less familiar variant of equal protection analysis. In sum, the doctrine (and thus, the Constitution) is violated when a facially neutral law singles out a racial issue for special treatment and also alters the political process, entrenching the result and imposing unique structural burdens on minorities’ future ability to obtain beneficial legislation. (Seattle, supra, 458 U.S. at pp. 470-474, 479-480; Hunter, supra, 393 U.S. at pp. 390-391.) That is, I submit, precisely what section 31 accomplishes. The arguments offered to the contrary are thin gruel. (Maj. opn., ante, at pp. 329-332; cone. opn. of Corrigan, J., ante, at pp. 339-340.) Whatever the concerns about the ongoing vitality of the political structure doctrine (conc. *343opn. of Corrigan, J., ante, at pp. 340-341), unless the United States Supreme Court decides to overrule its decisions, our obligation is to follow its clearly applicable and controlling precedent. (Rodriguez de Quijas v. Shearson/Am. Exp., Inc. (1989) 490 U.S. 477, 484 [104 L.Ed.2d 526, 109 S.Ct. 1917].) Because the majority fails to do so, I must dissent.3
L
As the majority relates the facts and history of this case, there is no need to repeat them. However, the extent of the City and County of San Francisco’s (City) efforts to investigate whether discrimination was a problem in its public contracting and, if so, whether there was a basis for a legislative remedy bears brief mention.
Prior to enacting the 1989 version of the City’s Minority/Women/Local Business Utilization Ordinance (ordinance), the City “received, among other information, testimony from 42 witnesses, and written submittals from 127 minority, women, local, and other business representatives. Subsequently . . . the City held an additional ten public hearings, commissioned two statistical studies, and sought written submissions from the public.” (Associated General Contractors of California v. Coalition (9th Cir. 1991) 950 F.2d 1401, 1404.)
After the 1989 ordinance expired, the City’s board of supervisors and Human Rights Commission held an additional 14 public hearings, heard live testimony from 254 witnesses, videotaped testimony from numerous other witnesses, and considered additional statistical disparity studies and other documentary evidence pertinent to alleged discrimination and bidding irregularities. Minority contractors observed that, as compared with nonminority contractors, City inspectors imposed more onerous requirements on them, scrutinized their work more closely and treated them more harshly if they made mistakes. One minority contractor spoke of being harassed and subject to racist and derogatory remarks from City inspectors. Another complained of being subjected to more rigorous vetting despite his extensive qualifications and experience. Out of this process emerged the 1998 Ordinance.
The 1998 ordinance expired in 2003. The City conducted further investigations, including a disparity analysis and a Human Rights Commission report containing additional examples of continuing discrimination in public contracting. The City also conducted additional hearings at which 134 individuals testified. Minority contractors testified of ongoing discrimination *344in the contracting process, and the board of supervisors heard evidence that prime contractors tried to circumvent compliance with the Ordinance. The board of supervisors made extensive legislative findings, including that the small percentage of City contracts going to minority- and women-owned businesses was due to discrimination by the City and discrimination in the private sector, that the City was actively discriminating against women and minority groups in its contracting and passively participating in private sector discrimination, that the City’s contracting practices were in violation of federal law, and that the Ordinance was required to remedy the discrimination against minority- and women-owned businesses. The City thereafter enacted the 2003 ordinance, the version challenged by plaintiffs in this case.
II.
A. The Political Structure Doctrine
The City contends section 31 violates the political structure doctrine, the contours of which are set out in a trilogy of United States Supreme Court cases: Hunter, supra, 393 U.S. 385, and Seattle, supra, 458 U.S. 457, which established the basic tenets of the doctrine, and Crawford v. Los Angeles Board of Education (1982) 458 U.S. 527 [73 L.Ed.2d 948, 102 S.Ct. 3211] (Crawford), which described the doctrine’s outer limits. I begin by discussing each case in turn.
1. Hunter
In 1964, the city council in Akron, Ohio, enacted a fair housing ordinance prohibiting discrimination on the basis “ ‘of race, color, religion, ancestry or national origin.’ ” (Hunter, supra, 393 U.S. at p. 386.) The ordinance established a commission in the mayor’s office to enforce the measure’s provisions through conciliation or persuasion if possible, and by court order if necessary. (Ibid.) Hunter, an African-American woman, filed a complaint with the commission alleging a real estate agent had come to show her a fist of houses for sale but, upon meeting Hunter, said she could not show Hunter the homes “ ‘because all of the owners had specified they did not wish their houses shown to negroes.’ ” (Id. at p. 387.)
Hunter was told the ordinance provided no remedy because the city charter had been amended in a citywide election to provide that any ordinance enacted by the city council regulating real estate transactions “ ‘on the basis of race, color, religion, national origin or ancestry must first be approved by a majority of the electors voting on the question at a regular or general election before said ordinance shall be effective. Any such ordinance in effect at the time of the adoption of this section shall cease to be effective until approved *345by the electors as provided herein.’ [Citation.]” (Hunter, supra, 393 U.S. at p. 387.) The proposed charter amendment had been placed on the ballot after more than 10 percent of Akron’s voters had signed a petition, and the initiative was passed by a majority of voters. (Ibid.) Hunter sought a writ of mandamus in state court to enforce the ordinance, but the trial court ultimately concluded it had been rendered ineffective by the amended city charter, and the Supreme Court of Ohio affirmed. (Id. at pp. 387-388.) The high court reversed.
While suggesting the mere repeal of an ordinance would not violate the Fourteenth Amendment (Hunter, supra, 393 U.S. at p. 390, fn. 5), the Supreme Court noted the charter amendment went further by “not only suspending] the operation of the existing ordinance forbidding housing discrimination, but also requiring] the approval of the electors before any future ordinance could take effect.” (Id. at pp. 389-390, italics added.) In doing so, the charter amendment “drew a distinction between those groups who sought the law’s protection against racial, religious, or ancestral discriminations in the sale and rental of real estate and those who sought to regulate real property transactions in the pursuit of other ends.” (Id. at p. 390.) Under the amended city charter, it became substantially more difficult for religious and racial minorities to secure the enactment of ordinances forbidding housing discrimination than for those seeking other ordinances. (Ibid.)
Ordinances not subject to the charter amendment would become effective 30 days after passage by the city council (or immediately, if passed as an emergency measure) and would be subject to referendum only if 10 percent of the voters so requested by signing a petition. (Hunter, supra, 393 U.S. at p. 390.) By contrast, the city council’s approval would not be enough for those seeking ordinances to forbid racial and religious housing discrimination. (Ibid.) Those seeking ordinances to prevent housing discrimination would have to overcome the unique hurdle of obtaining the approval of a majority of voters in an election before such a city council-approved ordinance could be effective. (Ibid.)
That the charter amendment was facially neutral was of no moment, the high court explained. Even though it did not distinguish between various religious and racial groups, the amended charter “disadvantages those who would benefit from laws barring racial, religious, or ancestral discriminations as against those who would bar other discriminations or who would otherwise regulate the real estate market in their favor.” (Hunter, supra, 393 U.S. at *346pp. 390-391.)4 Moreover, the Supreme Court recognized that, “although the law on its face treats Negro and white, Jew and gentile in an identical manner, the reality is that the law’s impact falls on the minority.” (393 U.S. at p. 391, italics added.) Placing “special burdens on racial minorities within the governmental process ... is no more permissible than denying them the vote, on an equal basis with others. [(Cf. Gomillion v. Lightfoot (1960) 364 U.S. 339 [5 L.Ed.2d 110, 81 S.Ct. 125]; Reynolds v. Sims (1964) 377 U.S. 533 [12 L.Ed.2d 506, 84 S.Ct. 1362]; Avery v. Midland County (1968) 390 U.S. 474 [20 L.Ed.2d 45, 88 S.Ct. 1114].)]” (Hunter, at p. 391.) The high court’s citation to its voting rights decisions (and specifically its vote dilution decisions) is enlightening, as it underscores that the doctrine’s central focus is not on the attainment of particular legislative outcomes, but instead on ensuring minorities’ meaningful and equal access to the political process.
Because the charter amendment singled out an issue of particular import to racial minorities and effectively entrenched the result by imposing unique hurdles in front of minorities’ future efforts to achieve beneficial legislation, the court concluded it constituted a racial classification and applied strict scrutiny. (Hunter, supra, 393 U.S. at p. 392.) It ultimately found the asserted justifications for the amendment insufficient. (Ibid.) Of particular note, the court emphasized that “the implementation of this change through popular referendum [does not] immunize it. [Citation.] The sovereignty of the people is itself subject to those constitutional limitations which have been duly adopted and remain unrepealed.” (Ibid.) Having chosen a process for enacting legislation, the high court concluded, “the State may no more disadvantage any particular group by making it more difficult to enact legislation in its behalf than it may dilute any person’s vote or give any group a smaller representation than another of comparable size. [Citations.]” (Id. at p. 393, italics added.)
2. Seattle
In 1977, Seattle School District No. 1 enacted a resolution to combat de facto racial segregation in the school system resulting from the housing patterns in Seattle, Washington. (Seattle, supra, 458 U.S. at p. 459.) The school district initially employed race-neutral voluntary measures; the steps taken, however, actually led to increased racial imbalance in the schools. (Id. at p. 461.) The school district ultimately determined “that mandatory reassignment of students was necessary if racial isolation in its schools was to *347be eliminated” and thus implemented a program involving the extensive use of busing and mandatory reassignments in the elementary schools (the Plan). (Ibid.)
After a failed attempt to enjoin the implementation of the Plan, Seattle residents who opposed its remedial measures drafted a statewide ballot measure “designed to terminate the use of mandatory busing for purposes of racial integration.” (Seattle, supra, 458 U.S. at p. 462.) The proposed initiative, which ultimately passed with 66 percent of the vote, prohibited school boards from “ ‘requiring] any student to attend a school other than the school which is geographically nearest or next nearest the student’s place of residence . . . and which offers the course of study pursued by such student.’ ” (Id. at pp. 462-463.) While the initiative did not mention race or, “busing for racial purposes,” it contained a number of exceptions to its ban (id. at p. 463), such that school districts were permitted to “bus their students ‘for most, if not all,’ of the nonintegrative purposes required by their educational policies. [Citation.]” (Id. at p. 471.) In light of these exceptions, the Supreme Court observed, it was clear that the measure was aimed solely at eliminating the remedy of “desegregative busing in general, and . . . the . . . Plan in particular.” (Id. at p. 463.)5
Three school districts sued the State of Washington in federal court, arguing the initiative violated the equal protection clause. (Seattle, supra, 458 U.S. at p. 464.) The district court concluded the initiative constituted an impermissible racial classification in violation of the political structure doctrine because it forbade busing from being used as a remedy for racial segregation while permitting its use for all nonracial reasons. (Id. at p. 465.) The Ninth Circuit affirmed, and the state and state officers appealed. (Id. at p. 466.)
The high court began by noting the equal protection clause guarantees racial minorities “the right to full participation in the political life of the community.” (Seattle, supra, 458 U.S. at p. 467.) To that end, the Fourteenth Amendment not only protects minorities’ right to vote and to enter into the political process in a reliable and meaningful manner, but also “reaches ‘a political structure that treats all individuals as equals,’ [citation], yet more subtly distorts governmental processes in such a way as to place special burdens on the ability of minority groups to achieve beneficial legislation.” *348(Seattle, at p. 467, italics added.) This “political structure” principle, the high court noted, was expressed and applied in Hunter and Lee v. Nyquist (W.D.N.Y 1970) 318 F.Supp. 710, summarily affd. sub. nom. Nyquist v Lee (1971) 402 U.S. 935 [29 L.Ed.2d 105, 91 S.Ct. 1618].6 (Seattle, at pp. 467-469.) From these cases (as well as from Justice Harlan’s concurring opinion in Hunter, supra, 393 U.S. at pages 393-396), the Seattle court drew “a simple but central principle” (Seattle, at p. 469): “[L]aws structuring political institutions or allocating political power according to ‘neutral principles’—such as the executive veto, or the typically burdensome requirements for amending state constitutions—are not subject to equal protection attack, though they may ‘make it more difficult for minorities to achieve favorable legislation.’ [(Hunter, supra,] 393 U.S. at [p.] 394 [(conc. opn. of Harlan, J.).)] Because such laws make it more difficult for every group in the community to enact comparable laws, they ‘provid[e] a just framework within which the diverse political groups in our society may fairly compete.’ (Id., at p. 393.) Thus, the political majority may generally restructure the political process to place obstacles in the path of everyone seeking to secure the benefits of governmental action. But a different analysis is required when the State allocates governmental power nonneutrally, by explicitly using the racial nature of a decision to determine the decisionmaking process.” (Seattle, supra, 458 U.S. at p. 470.)
Applying the political structure doctrine to the ballot measure, the Supreme Court concluded that, rather than attempting “ ‘to allocate governmental power on the basis of any general principle’ [citation]” (Seattle, supra, 458 U.S. at p. 470), it impermissibly used “the racial nature of an issue to define the governmental decisionmaking structure, and thus impose[d] substantial and unique burdens on racial minorities.” (Ibid.)
In so holding, the court conducted a two-part inquiry. It first inquired whether the initiative, despite its facial neutrality, singled out a racial issue for special treatment. (Seattle, supra, 458 U.S. at p. 471.) While the initiative nowhere mentioned “race” or “integration,” the court had little difficulty finding it was nonetheless “effectively drawn for racial purposes.” (Ibid.)7 As evidence, the court pointed to the fact that the exceptions in the initiative meant that only desegregative busing was prohibited, which was consistent *349with the proponents’ statements during the campaign. (458 U.S. at p. 471.) “It is beyond reasonable dispute, then, that the initiative was enacted ‘ “because of,” not merely “in spite of,” its adverse effects upon’ busing for integration. [Citation.]” (Ibid.)
In addition, the court noted that busing for integration is a racial issue that, “at bottom inures primarily to the benefit of the minority, and is designed for that purpose.” (Seattle, supra, 458 U.S. at pp. 471-472, italics added.) While so concluding, the court acknowledged that racial minorities, as well as those in the majority, could be counted among both the proponents and opponents of the ballot measure, just as both racial minority and majority members benefited from diverse schools. (Id. at p. 472.) “But neither of these factors serves to distinguish Hunter, for we may fairly assume that members of the racial majority both favored and benefited from Akron’s fair housing ordinance. [Citations.]” (Ibid.) “[I]t is enough that minorities may consider busing for integration to be ‘legislation that is in their interest.’ [Citation.] Given the racial focus of [the initiative], this suffices to trigger application of the Hunter doctrine.” (Id. at p. 474, italics added.)
For the second part of its inquiry, the high court considered whether “the practical effect of [the ballot measure] is to work a reallocation of power of the kind condemned in Hunter.” (Seattle, supra, 458 U.S. at p. 474, italics added.) In concluding the initiative did so, the court noted that the ballot measure “removes the authority to address a racial problem—and only a racial problem—from the existing decisionmaking body, in such a way as to burden minority interests.” (Ibid.) After passage of the measure, those seeking to address school segregation had to seek relief from the state legislature or from the statewide electorate; by comparison, those wishing to effect any other nonracial school reassignment or educational policy needed only to petition their local school board. (Ibid.) Thus, the changed political structure “expressly requires those championing school integration to surmount a considerably higher hurdle than persons seeking comparable legislative action. As in Hunter, then, the community’s political mechanisms are modified to place effective decisionmaking authority over a racial issue at a different level of government.” (Seattle, at p. 474.)8
In concluding the initiative had worked an impermissibly nonneutral alteration of the political structure, the court considered, and rejected, arguments raised by Justice Powell’s dissent and the defendants. Justice Powell argued *350this case was unlike Hunter because “the political system [of Washington] has not been redrawn or altered.” (Seattle, supra, 458 U.S. at p. 498 (dis. opn. of Powell, J.).) The majority dismissed this distinction as facile, pointing out “[t]he evil condemned by the Hunter Court was not the particular political obstacle of mandatory referenda imposed by the Akron charter amendment; it was, rather, the comparative structural burden placed on the political achievement of minority interests.” (Id. at pp. 474-475, fn. 17, italics added.)9 In both cases, the power to enact racial legislation was relocated to a more distant level, the citywide electorate in Hunter and the statewide electorate or the state legislature in Seattle. (458 U.S. at pp. 474-475, fn. 17.)10
The court also rejected the defendants’ contention that the ballot measure merely constituted a permissible intervention by the state in its own school system. (Seattle, supra, 458 U.S. at pp. 475-476.) While it acknowledged that Washington had plenary authority over its education system, the court pointed out that “[t]he issue here ... is not whether Washington has the authority to intervene in the affairs of local school boards; it is, rather, whether the State has exercised that authority in a manner consistent with the Equal Protection Clause.” (Id. at p. 476.) Having previously chosen to vest decisionmaking authority of the type at issue here in local school boards (id. at pp. 477-480), the ballot measure “worked a major reordering of the State’s educational decisionmaking process. . . . After [the] passage of [the ballot measure], authority over all but one of those areas remained in the hands of the local board. By placing power over desegregative busing at the state level, then, [the initiative] plainly ‘differentiates between the treatment of problems involving racial matters and that afforded other problems in the same area.’ [Citation.]” (Id. at pp. 479-480, italics added.)11
*351Moreover, the court noted, while voters are free to merely repeal unpopular legislation at the ballot box, the initiative went further. (Seattle, supra, 458 U.S. at p. 483.) “It burdens all future attempts to integrate Washington schools in districts throughout the State, by lodging decisionmaking authority over the question at a new and remote level of government.” (Ibid., italics added.) This new political structure “imposes direct and undeniable burdens on minority interests. ‘If a governmental institution is to be fair, one group cannot always be expected to win,’ [citation]; by the same token, one group cannot be subjected to a debilitating and often insurmountable disadvantage.” (Id. at p. 484.)
The high court ultimately concluded strict scrutiny applied to the initiative because, “when the political process or the decisionmaking mechanism used to address racially conscious legislation—and only such legislation—is singled out for peculiar and disadvantageous treatment, the governmental action plainly ‘rests on “distinctions based on race.” ’ [Citation.]” (Seattle, supra, 458 U.S. at p. 485, fn. omitted.) It continued, “when the State’s allocation of power places unusual burdens on the ability of racial groups to enact legislation specifically designed to overcome the ‘special condition’ of prejudice, the governmental action seriously ‘curtail[s] the operation of those political processes ordinarily to be relied upon to protect minorities.’ [(United States v. Carolene Products Co. (1938) 304 U.S. 144, 153, fn. 4 [82 L.Ed. 1234, 58 S.Ct. 778].)]” (Seattle, at p. 486.) The court thus concluded the initiative violated the Fourteenth Amendment and was invalid. {Seattle, at p. 487.)12
3. Crawford
In 1963, minority students attending school in the Los Angeles Unified School District filed a class action in state court seeking to desegregate the district’s schools. (Crawford, supra, 458 U.S. at pp. 529-530.) In 1970, the trial court found substantial segregation in violation of the state and federal Constitutions and ordered the district to prepare a desegregation plan. (458 U.S. at p. 530.) Although the court had found de jure segregation in violation of the Fourteenth Amendment, we affirmed its ruling based solely upon California’s equal protection clause. (458 U.S. at p. 530, citing Crawford v. *352Board of Education (1976) 17 Cal.3d 280 [130 Cal.Rptr. 724, 551 P.2d 28].) We explained the state Constitution was broader than its federal counterpart, requiring districts to take reasonable steps to address segregation, “ ‘whether [it] be de facto or de jure in origin.’ [(Crawford v. Board of Education, supra, 17 Cal.3d at p. 290.)]” (Crawford, supra, 458 U.S. at pp. 530-531.)
On remand, the trial court considered possible desegregation plans, first approving a plan including mandatory busing and then considering alternatives. (Crawford, supra, 458 U.S. at p. 531.) In the meantime, California’s voters passed Proposition 1, amending article I, section 7 of the state Constitution. (Crawford, at pp. 531-532 & fn. 5.) The proposition conformed the power of state courts to order busing for any reason to federal courts’ power under the Fourteenth Amendment, but did not prohibit schools boards “ ‘from voluntarily continuing or commencing a school integration plan.’ ” (458 U.S. at p. 532 & fn. 6, quoting Cal. Const., art. I, § 7, as amended Nov. 6, 1979.)
Following Proposition l’s passage, the school district sought to halt all mandatory student reassignment and busing, but the trial court denied the request based on its prior finding of de jure segregation in violation of the federal Constitution. (Crawford, supra, 458 U.S. at p. 533.) The state Court of Appeal reversed. It concluded the trial court’s finding of de jure segregation was not supported, and thus Proposition 1 applied, so it barred the trial court from ordering mandatory student reassignment and busing. (458 U.S. at pp. 533-534.) The Court of Appeal also concluded Proposition 1 was constitutional under the Fourteenth Amendment because California was not obliged to provide a greater remedy against racial segregation than what was provided under the federal Constitution. (Crawford, at p. 534.) We denied review, and the high court granted certiorari. (Ibid.)
The Supreme Court held Proposition 1 did not violate the Fourteenth Amendment generally, or the political structure doctrine specifically, concluding the initiative merely repealed this court’s interpretation of California’s Constitution as imposing broader desegregation obligations than those imposed by the federal Constitution. (Crawford, supra, 458 U.S. at pp. 535-536, 537, fn. 14.) The high court distinguished Hunter and Seattle in three critical respects: first, it noted that, under Proposition 1, school districts remained free as before to adopt reassignment and busing plans based on race to combat segregation (458 U.S. at p. 536); second, it noted Proposition 1 did not single out the issue of desegregative student assignment and busing, but rather, it removed state courts’ power to order reassignment and busing for any reason, racial or otherwise (458 U.S. at p. 538, fn. 18); and third, it noted Proposition 1 constituted a mere repeal of a right by the same *353entity—the people—who created it in the first place (via adoption of the state Constitution) (Crawford, at p. 542; id. at p. 547 (conc. opn. of Blackmun, J.)).13
4. Summary of the Doctrine’s Scope
From these cases, the contours of the political structure doctrine can be delineated. For a law to violate the doctrine, and thus require application of heightened scrutiny, two conditions must be met.
First, the law must single out a racial issue for special treatment. (Seattle, supra, 458 U.S. at pp. 470-474; Hunter, supra, 393 U.S. at pp. 390-391.)14 That a law is facially neutral is of no consequence (Seattle, at p. 471; Hunter, at pp. 390-391); rather, we ask whether it was “effectively drawn for racial purposes.” (Seattle, at p. 471.) In speaking of a law’s “purpose,” we do not mean whether it was motivated by invidious intent. (Id. at pp. 484-486.) A racial purpose exists where the law was enacted because of, rather than in spite of, its effect upon the racial issue (id. at p. 471) or where the law’s “impact falls on the minority” (Hunter, at p. 391) because the repealed policy “inures primarily to [minorities’] benefit . . . and is designed for that purpose.” (Seattle, at pp. 471-472.)
Second, the law must restructure the political process in a nonneutral manner, imposing unique burdens on minorities’ future efforts to enact beneficial legislation. (Seattle, supra, 458 U.S. at pp. 474, 479-480; Hunter, supra, 393 U.S. at pp. 390-391.)15 In other words, we ask whether the law effectively entrenches the result by altering the process such that a “comparative structural burden [is] placed on the political achievement of minority interests” (Seattle, at p. 475, fn. 17) in contrast to the “ ‘treatment . . . afforded other problems in the same area’ ” (id. at pp. 479-480).
For a violation of the doctrine to be established, both conditions must be met; one alone will not suffice. For example, the repeal of a law advantaging *354racial minorities certainly singles out a racial issue for special treatment. However, without more, the mere repeal of such a law by the enacting governmental entity does not run afoul of the doctrine. (Crawford, 458 U.S. at p. 547 (conc. opn. of Blackmun, J.); Seattle, supra, 458 U.S. at pp. 483; Hunter, supra, 393 U.S. at p. 390 & fn. 5.)16 It simply “reflects the normal operation of the political process in which there are winners and losers. Repeal of legislation favorable to the interests of a racial minority simply indicates that a prior winner has now lost . . . [but] does not alter or distort the existing political process in any way.” (Amar & Caminker, Equal Protection, Unequal Political Burdens, and the CCRI (1996) 23 Hastings Const. L.Q. 1019, 1044 (Amar & Caminker); see Hunter, at p. 394 (conc, opn. of Harlan, J.).)
As another example, amending California’s Constitution to require that 25 percent of the electorate sign a petition before an initiative can qualify to be placed on the ballot (as opposed to the current, lower requirement in Cal. Const., art. II, § 8) would without a doubt “make it more difficult for minorities to achieve favorable legislation” via the initiative process. (Hunter, supra, 393 U.S. at pp. 393-394 (conc. opn. of Harlan, J.).) However, such an alteration would not violate the doctrine because such a change would be “grounded in neutral principle.” (Id. at p. 395; see Seattle, supra, 458 U.S. at p. 480, fn. 23.) Because such a change to the process would “make it more difficult for every group in the community to enact comparable laws,” there would continue to be a level playing field on which vying political groups could compete. (Seattle, at p. 470.)
Thus, only when a law singles out a racial issue for special treatment and alters the political process, imposing a unique structural burden on minorities’ future ability to achieve beneficial legislation, is the political structure doctrine violated and is the law subject to heightened scrutiny.
B. Proposition 209 and Hi-Voltage
Before explaining why section 31 violates the political structure doctrine, it is necessary to first briefly revisit Proposition 209, the ballot measure that enacted section 31, and Hi-Voltage, supra, 24 Cal.4th 537, in which this court construed section 31’s scope.
*3551. Proposition 209 and the November 5, 1996, Election
After Proposition 209 qualified as an initiative constitutional amendment, it was placed on the November 5, 1996, General Election ballot. Prior to the election, voters received an official ballot pamphlet prepared by the nonpartisan Legislative Analyst’s Office. (Coalition for Economic Equality v. Wilson (N.D.Cal. 1996) 946 F.Supp. 1480, 1493 (Coalition I).)17 The ballot pamphlet, which included an official description and analysis of each statewide initiative, described Proposition 209 as a measure that would eliminate race- and sex-conscious affirmative action programs in the areas of public employment, contracting, and education. (946 F.Supp. at p. 1493.) The Legislative Analyst’s brief summary explained “A YES vote on [Proposition 209] means: The elimination of those affirmative action programs for women and minorities run by the state or local governments in the areas of public employment, contracting, and education that give ‘preferential treatment’ on the basis of sex, race, color, ethnicity, or national origin.” “A NO vote on [Proposition 209] means: State and local government affirmative action programs would remain in effect to the extent they are permitted under the United States Constitution.” (Ballot Pamp., Gen. Elec. (Nov. 5, 1996) (Ballot Pamphlet) Legis. Analyst’s brief summary of Prop. 209, p. 6.)
In addition to its summary of the measure, the Legislative Analyst provided a more in-depth analysis, emphasizing that passage of the initiative would effectively abolish all race- and sex-conscious affirmative action programs. “The federal, state, and local governments run many programs intended to increase opportunities for various groups—including women and racial and ethnic minority groups. These programs are commonly called ‘affirmative action’ programs. . . . [f] . . . [][] [Proposition 209] would eliminate state and local government affirmative action programs in the areas of public employment, public education, and public contracting to the extent these programs involve ‘preferential treatment’ based on race, sex, color, ethnicity, or national origin.” (Ballot Pamp., supra, Legis. Analyst’s analysis of Prop. 209, p. 30.) The Legislative Analyst then discussed various race- and sex-conscious affirmative action programs in the areas of public employment, contracting, and education that would be banned with the passage of the initiative. (Id. at p. 31.)
In addition to the nonpartisan analysis by the Legislative Analyst, the Ballot Pamphlet contained partisan arguments submitted by proponents and *356opponents of the measure. These arguments further underscored that the central issue at stake in Proposition 209 was race- and sex-conscious affirmative action (and race-conscious measures in particular). The argument in favor of the measure stated, “ ‘REVERSE DISCRIMINATION’ BASED ON RACE OR GENDER IS PLAIN WRONG! [f] . . . [Students are being rejected from public universities because of their RACE. Job applicants are turned away because their RACE does not meet some ‘goal’ or ‘timetable.’ Contracts are awarded to high bidders because they are of the preferred RACE. [][]... [|] •• ■ Proposition 209 will stop the terrible programs . . . .” (Ballot Pamp., supra, argument in favor of Prop. 209, p. 32.) The argument against the initiative warned that “California law currently allows tutoring, mentoring, outreach, recruitment, and counseling to help ensure equal opportunity for women and minorities. Proposition 209 will eliminate affirmative action programs like these that help achieve equal opportunity for women and minorities in public employment, education and contracting.” (Id., argument against Prop. 209, p. 33.)
In explaining that Proposition 209 would eliminate race- and sex-conscious affirmative action programs, the Legislative Analyst, proponents, and opponents implicitly acknowledged preferential treatment on all other bases would be unaffected by the ballot measure. (See Hi-Voltage, supra, 24 Cal.4th at p. 566; id. at pp. 586-587 (conc. & dis. opn. of George, C. J.).) Indeed, this was explicitly recognized in the rebuttal to the argument against Proposition 209. “Affirmative action programs that don’t discriminate or grant preferential treatment [on the basis of race or sex] will be UNCHANGED. ...[][] Note that Proposition 209 doesn’t prohibit consideration of economic disadvantage. . . . The state must remain free to help the economically disadvantaged, but not on the basis of race or sex.” (Ballot Pamp., supra, rebuttal to argument against Prop. 209, p. 33.)
On November 5, 1996, Proposition 209 passed, with 54 percent voting in favor of the measure and 46 percent voting against. (Coalition I, supra, 946 F.Supp. at p. 1495.) White voters were the only racial group that, as a majority, voted in favor of Proposition 209, with 63 percent voting for passage. (946 F.Supp. at p. 1495, fn. 12.) Seventy-four percent of Black voters, 76 percent of Latino voters, and 61 percent of Asian-American voters opposed the measure. (Ibid.) Sixty-one percent of men voted in favor, while 52 percent of women voted against. (Ibid.)
2. Hi-Voltage and Our Construction of Section 31
In 2000, this court considered the validity of a program adopted by the City of San Jose to encourage the participation of minority-owned business enterprises and women-owned business enterprises in public works projects. *357(Hi-Voltage, supra, 24 Cal.4th at p. 542.) Contractors bidding for city projects were required to fulfill either an outreach or a participation component, and the plaintiffs argued that the program’s requirements violated section 31. (Hi-Voltage, at pp. 543-544.) We agreed, concluding the program’s “outreach option affords preferential treatment to [minority-/women-owned business enterprise] subcontractors on the basis of race or sex, and the participation option discriminates on the same bases against non- [minority-/women-owned business enterprise] subcontractors as well as general contractors that fail to fulfill either of the options when submitting their bids.” (Id. at p. 560, fn. omitted.)
In reaching this conclusion, we reviewed the language of section 31 as well as the ballot materials accompanying Proposition 209. (Hi-Voltage, supra, 24 Cal.4th at pp. 559-562.) We noted the measure’s language was not limited in any way and effectively prohibited all race- and sex-conscious affirmative action programs. (Id. at pp. 559-560; id. at pp. 591-592 (conc. & dis. opn. of George, C. J.).) Turning to the ballot materials, we concluded the partisan statements and the Legislative Analyst’s analysis provided further support for our construction. (Id. at pp. 560-562.) In particular, we noted the proponents’ statements focused on race- and sex-conscious affirmative action programs and evinced a clear desire to ban all such programs. (Id. at pp. 560-561.)
This court ultimately concluded the electorate intended to do “something more” than restate existing law prohibiting discrimination on the basis of race or sex. (Hi-Voltage, supra, 24 Cal.4th at p. 561 [citing Coalition I, supra, 946 F.Supp. at p. 1489].)18 That “something more,” the majority reasoned, was a repudiation of case law permitting race- and sex-conscious measures “formulated to remediate the lingering effects of past discrimination or conspicuous imbalance” in public contracting, employment, and education. (Hi-Voltage, supra, at p. 566.) We accordingly held that, following the adoption of section 31, “any action that discriminates or grants preferential treatment on the basis of race or sex would be forbidden,” including race- and sex-conscious measures that would otherwise be permitted by the equal protection clause. (Hi-Voltage, at pp. 566-567.)
Thus, the practical effect of section 31’s ban of race- and sex-conscious measures is limited to just one category of legislation. It has no independent effect on legislation that cannot survive heightened scrutiny because such legislation already violates the Constitution. (See Hi-Voltage, supra, 24 Cal.4th at p. 561.) It also has no effect on legislation required by the Constitution *358(U.S. Const., art. VI, cl. 2), when, for example, a public entity seeks to remedy its own past intentional discrimination (Hi-Voltage, at p. 568). Consequently, section 31’s only nonredundant effect is on race- and sex-conscious measures that are permitted, but not required, by the Constitution, i.e., legislation that, despite classifying on the basis of race or sex, can survive heightened scrutiny. For race-conscious programs to do so, they must be narrowly tailored and justified by a compelling interest. (McLaughlin v. Florida (1964) 379 U.S. 184, 191-192 [13 L.Ed.2d 222, 85 S.Ct. 283].) Under the high court’s decisions, this category of legislation has become increasingly narrow.
For example, in public contracting, a race-conscious program must be narrowly tailored to remedy the effects of past discrimination by the public entity itself or by private sector entities within its jurisdiction. (Richmond v. J. A. Croson Co. (1989) 488 U.S. 469, 498-506 [102 L.Ed.2d 854, 109 S.Ct. 706] (Croson).) The Supreme Court has rejected as insufficiently compelling the remedying of societal discrimination (id. at p. 505) or the attainment of racial balance in an industry (id. at p. 507). Similarly, in public employment, remedying the effects of past discrimination by the public entity is a compelling interest, while remedying societal discrimination is not. (Wygant v. Jackson Board of Education (1986) 476 U.S. 267, 274 [90 L.Ed.2d 260, 106 S.Ct. 1842] (plur. opn. of Powell, J.).) In the public education context, in addition to remedying the effects of past discrimination, the high court has found compelling the promotion of racial diversity in higher education. (Parents Involved in Community Schools v. Seattle School Dist. No. 1 (2007) 551 U.S. 701, 720 [168 L.Ed.2d 508, 127 S.Ct. 2738]; Grutter v. Bollinger (2003) 539 U.S. 306, 325 [156 L.Ed.2d 304, 123 S.Ct. 2325].) Thus, the only real change section 31 makes to existing law is to close an already narrow, albeit significant, window of constitutionally permissible remedial legislation. (Hi-Voltage, supra, 24 Cal.4th at p. 568.)
C. Application of the Political Structure Doctrine to Section 31
I now discuss why section 31 violates the political structure doctrine. The first step of the inquiry requires consideration of whether section 31, while facially neutral, nonetheless singles out a racial issue for special treatment. (Seattle, supra, 458 U.S. at p. 471.) The answer is plainly yes.
Section 31 prohibits preferential treatment “on the basis of race, sex, color, ethnicity, or national origin.” (§31, subd. (a).) Thus, as with the charter amendment in Hunter, section 31 is explicitly race conscious. In addition, it singles out a racial issue for special treatment inasmuch as it draws a distinction between groups seeking beneficial legislation on the basis of race and sex (i.e., racial minorities and women) and those seeking beneficial *359legislation on all other bases (e.g., veterans, the economically disadvantaged, in-state residents, local businesses, physically disabled, athletes, etc.). (Hunter, supra, 393 U.S. at pp. 390-391; see Ballot Pamp., supra, rebuttal to argument against Prop. 209, p. 33.) For example: California (along with many other entities) grants civil service preferences to veterans (see, e.g., Gov. Code, § 18978); cities and counties may grant contracting preferences to locally owned or economically disadvantaged businesses (see, e.g., S.F. Admin. Code, ch. 6, § 6.4; id., ch. 14A);19 some public school districts grant preferences to children of former students (see, e.g., Mehta, Public Schools Offer Legacy Admissions, L.A. Times (May 16, 2009) p. A3 [discussing school districts’ adoption of admission policy giving preference to out-of-district children of former students]). Put another way, section 31 does not purport to regulate the enactment of all preferential legislation in the operation of public employment, public education, or public contracting, but rather only race- and sex-conscious preferences in those arenas.
Moreover, there can be no serious doubt that section 31 was “enacted 1 “because of,” not merely “in spite of,” its adverse effects upon’ ” race- and sex-based affirmative action, and was therefore “effectively drawn for racial purposes.” (Seattle, supra, 458 U.S. at p. 471.) Indeed, the ballot summary and the partisan statements underscore that the measure’s central purpose was to effectively eliminate all race- and sex-based affirmative action. (E.g., Ballot Pamp., supra, Legis. Analyst’s brief summary of Prop. 209, p. 6; id., argument in favor of Prop. 209, p. 32.) We arrived at the same conclusion in Hi-Voltage. (Hi-Voltage, supra, 24 Cal.4th at pp. 566-567; id. at pp. 591-592 (conc. & dis. opn. of George, C. J.).)
Finally, affirmative action is a racial issue in the same way that the fair housing ordinance in Hunter and the desegregative busing in Seattle were racial issues. (Seattle, supra, 458 U.S. at pp. 471-472; Hunter, supra, 393 U.S. at p. 391.) That is, “although [section 31] on its face treats [Whites and racial minorities, men and women] in an identical manner, the reality is that the law’s impact falls on the minority.” (Hunter, at p. 391.) Affirmative action “inures primarily to the benefit of the minority, and is designed for that purpose.” (Seattle, at p. 472.) And while racial minorities and women could undoubtedly be counted among both Proposition 209’s opponents and supporters, that does not serve to distinguish Hunter or Seattle. (Seattle, at p. 474.) For our purposes, “it is enough that [racial minorities and women] may consider [race- and sex-conscious affirmative action] to be ‘legislation *360that is in their interest.’ [Citation.]” (Ibid.)20 In summary, based on section 31’s language, its different treatment of race- and sex-conscious legislation as compared to other legislation in the same area, its avowed purpose, and whom it primarily affects, I conclude section 31 without question singles out a racial issue for special treatment, triggering application of the political structure doctrine. (See Seattle, at p. 474.)21
The second step of the inquiry requires consideration of whether section 31 constitutes a nonneutral restructuring of the political process, placing higher hurdles in front of those seeking race- and sex-conscious measures as opposed to what those seeking beneficial legislation on all other bases face. (Seattle, supra, 458 U.S. at p. 474 & fn. 17.) As in Hunter and Seattle, the most obvious way to answer this question is to examine the impact of Proposition 209’s passage on the process for obtaining favorable legislation. Because the practical effect of section 31 is the erection of a steep hurdle in front of those seeking race- and sex-conscious preferential legislation (as compared with those seeking similar legislation on other bases), section 31 obviously “workfs] a reallocation of power of the kind condemned [by the high court].” (Seattle, at p. 474.)
Prior to Proposition 209, any person seeking beneficial legislation for any group in the areas of public contracting, employment, or education could petition their government representatives to adopt, amend, or retain such a program. {Coalition I, supra, 946 F.Supp. at p. 1498.) For example, in matters of public contracting, Public Contract Code section 2000 gave local agencies the discretion, subject to certain limitations, to give bidding preferences to contractors who had complied with affirmative action requirements. (See Monterey Mechanical Co. v. Sacramento Regional County Sanitation Dist. (1996) 44 Cal.App.4th 1391, 1400-1403 [52 Cal.Rptr.2d 395]; Domar Electric, Inc. v. City of Los Angeles (1995) 41 Cal.App.4th 810, 820-822 [48 Cal.Rptr.2d 822].) With the passage of Proposition 209, the decision to permit and create such programs has been withdrawn from the Legislature and from *361local agencies, and now the adoption of affirmative action programs can only be accomplished via amendment of the state Constitution. Thus, the political mechanisms for seeking race- and sex-conscious legislation have been “modified to place effective decisionmaking authority over [such legislation] at a different level of government.” (Seattle, supra, 458 U.S. at p. 474.) On the other hand, the City may continue to grant contracting preferences on other bases (see, e.g., S.F. Admin. Code, ch. 6, § 6.4; id., ch. 14A), and those seeking preferential legislation not involving race or sex continue to have the opportunity to lobby local governmental entities and the Legislature for the expansion of such preferences or the creation of new ones.
That section 31’s placement in the state Constitution erects a new and formidable barrier to those advocating for race- and sex-conscious affirmative action programs can hardly be doubted. Amending California’s Constitution can be accomplished via either of two methods, both of which impose heavy burdens. First, the state Constitution can be amended via passage of another initiative. (Cal. Const., art. II, § 8.) To qualify a ballot measure, sponsors would initially have to obtain signatures equal to 8 percent of the previous gubernatorial vote. {Ibid.) To put that in context, to qualify an initiative constitutional amendment for the 2010 ballot, sponsors would have had to collect 694,354 valid signatures. (Cal. Sect, of State, Initiative Guide (2010) <http://www.sos.ca.gov/elections/ballotmeasures/how-to-qualify-an-initiative.htm> [as of Aug. 2, 2010].) Because a number of signatures tend to be disqualified, approximately 50 percent more “raw” signatures than the threshold amount must be collected (Coalition I, supra, 946 F.Supp. at p. 1498), and they must be obtained within a 150-day period (Elec. Code, § 336).22 If enough valid signatures are gathered, the initiative is placed on the ballot at the next statewide election and must obtain the approval of a majority of voters. (Cal. Const., art. H, § 8.)
As a second method, the Legislature can amend the state Constitution by securing a two-thirds vote of approval of a proposed amendment by both houses. (Cal. Const., art. XVIII, § 1.) If the Legislature does so, the measure is placed on the ballot and must gamer a majority of votes. {Id., § 4.)
If sponsors are able to successfully navigate getting the proposed amendment on the ballot via either process, substantial funds will then be required to ran a statewide campaign. (Coalition I, supra, 946 F.Supp. at p. 1499.) Because of the size of California, such campaigns tend to be very expensive—as of October 1996, for example, the campaign in support of Proposition 209 had spent $3.1 million. (Coalition I, at p. 1499.) As a more recent *362example, in the November 2008 election, $75 million was spent in support of and against Proposition 8, the initiative to ban same-sex marriage. (McKinley & Goodstein, Bans in 3 States on Gay Marriage, N.Y. Times (Nov. 6, 2008) p. A20.)23
In light of the substantial hurdles that racial minorities and women (and only those individuals) must overcome in seeking remedial legislation, section 31’s ban on constitutionally permissible race- and sex-conscious measures impermissibly reallocates power.24 It “removes the authority to address [race- and sex-conscious beneficial legislation]—and only [such legislation]—from the existing decisionmaking body, in such a way as to burden minority interests.” (Seattle, supra, 458 U.S. at p. 474.) In the wake of Proposition 209, veterans, the economically disadvantaged, the physically disabled, children of alumni, in-state residents, etc., all may continue to seek, obtain, and benefit from preferential legislation as before.25 That is, they can approach and lobby their school board, city council, county government, state university, state legislator, or any other public entity to enact legislation or adopt policies. The same is no longer true for those seeking race- and sex-conscious legislation, even when the legislation is narrowly tailored to remedy the effects of past discrimination by a public entity. Thus, the practical effect of section 31 is to restructure the political process in a nonneutral fashion. (Seattle, at pp. 479-480.)26
Although I conclude that section 31’s ban of race- and sex-conscious affirmative action violates the political structure doctrine, this does not mean that, once adopted, affirmative action measures may never be abolished. As the Supreme Court has repeatedly explained, the doctrine is not violated when enacting governmental entities merely overturn previous policies or *363when voters simply repeal unpopular legislation at the ballot box. (Seattle, supra, 458 U.S. at p. 483; Hunter, supra, 393 U.S. at p. 390, fn. 5.)27 Section 31, however, works more than a mere repeal of beneficial legislation by the enacting entity. (Seattle, at p. 483.) It burdens all future attempts to obtain race- and sex-conscious remedies by relocating decisionmaking power over such legislation from the public entities that had previously wielded it to a remote level of government, the state Constitution. (Seattle, at p. 483.)28
Significantly, section 31’s most profound impact is felt in communities where affirmative action does not arouse substantial opposition. (Seattle, supra, 458 U.S. at p. 483; Hunter, supra, 393 U.S. at pp. 395-396 (conc. opn. of Harlan, J.).) Where a public entity’s decision to enact affirmative action would be unpopular, voters could have simply overturned the policy via referendum or election of new representatives-—in such instances, section 31’s impact is slight. Where affirmative action is relatively uncontroversial, however, section 31’s alteration of the process for enacting future legislation “imposes direct and undeniable burdens on minority interests. ‘If a governmental institution is to be fair, one group cannot always be expected to win,’ [citation]; by the same token, one group cannot be subjected to a debilitating and often insurmountable disadvantage.” (Seattle, at p. 484.)29
The majority does not contest that section 31 was enacted because of its effect on race- and sex-conscious affirmative action or that, as a result, section 31’s impact falls primarily on minorities. Nor does it dispute that section 31 works more than a mere repeal, but rather imposes significant burdens on future efforts to enact race- and sex-conscious preferential legislation. These facts, I submit, establish a clear violation of the political structure doctrine and require, under controlling United States Supreme Court precedent, invalidation of section 31. The majority concludes otherwise. Its main—indeed its only—argument is that the doctrine does not apply when race- and sex-conscious preferential legislation is at issue. (Maj. opn., ante, at *364pp. 330-332.)30 As this distinction finds no support in Hunter or Seattle, however attractive the majority’s holding might be to some as a matter of policy, it is incorrect as a matter of law. In reaching its conclusion, the majority misreads and misapplies the high court’s political structure doctrine cases, imposing requirements and finding exceptions that do not appear in, and are not supported by, Hunter or Seattle.
The majority contends the legislation invalidated in Hunter and Seattle is distinguishable from section 31 in that the former “obstructed] equal treatment” while the latter “ban[s] preferences.” (Maj. opn., ante, at p. 330.) In response to the City’s argument that race-conscious preferences are as much “ ‘beneficial legislation’ ”—the phrase used by the Seattle court—as the antidiscrimination ordinance in Hunter or the race-conscious busing policy in Seattle, the majority declares: “We do not think, however, that the term ‘beneficial legislation’ can bear the weight the City would place upon it.” (Maj. opn., ante, at p. 330.)31 The majority’s rejoinder is curious; it is difficult to see how race-conscious preferential legislation does not constitute “beneficial legislation” (Seattle, supra, 458 U.S. at p. 467) or “legislation that is in [minorities’] interest” (Hunter, supra, 393 U.S. at p. 395 (conc. opn. of Harlan, J.)). It is more reasonable to assume the high court understood that “beneficial legislation” is an expansive phrase and used it intending it be so construed. The majority points to nothing in either Hunter or Seattle to suggest the Supreme Court intended the cramped reading the majority now adopts. To the contrary, there is substantial evidence in those opinions supporting the notion that the high court did intend for its language to be read broadly.
For example, the high court discussed minorities’ ability “to enact legislation in [their] behalf’ (Hunter, supra, 393 U.S. at p. 393); “the ability of minority groups to achieve beneficial legislation” (Seattle, supra, 458 U.S. at p. 467); minorities’ ability “ ‘to achieve favorable legislation’ ” (id. at p. 470, *365citing Hunter, at p. 394 (cone. opn. of Harlan, J.)); the ability “to achieve legislation that is in [minorities’] interest” (Hunter, at p. 395 (cone. opn. of Harlan, J.)); the “burden [on] minority interests” (Seattle, at pp. 474, 484); the “political achievement of minority interests” (id. at p. 475, fn. 17); and “the ability of racial groups to enact legislation specifically designed to overcome the ‘special condition’ of prejudice” (id. at p. 486). That the Supreme Court repeatedly spoke in expansive terms bolsters the City’s argument—the political structure doctrine is concerned not with minorities’ actual legislative goals, but with their equal and meaningful access to the political process. (See, ante, at p. 346.)32 The majority’s reliance on the preferential nature of the legislation is therefore an inapt basis for distinguishing Hunter and Seattle.
Moreover, the majority’s efforts to distinguish “between initiatives obstructing equal treatment and initiatives banning preferences” (maj. opn., ante, at p. 330) elides the fact that section 31’s only practical effect is on the narrow group of laws that can survive heightened constitutional scrutiny. (See, ante, at p. 335.) Given that reality, the distinction between the race-conscious affirmative action here and the antidiscrimination law in Hunter, supra, 393 U.S. at page 386, is illusory.33 Before adopting race-conscious preferential legislation, a public entity must document in some detail the existence of discrimination to provide “a ‘strong basis in evidence . . . that remedial action [is] necessary.’ [Citation.]” (Croson, supra, 488 U.S. at p. 500.) Thus, the 2003 ordinance at issue here is in a very direct way an antidiscrimination program, in which any preference granted must be strongly justified by the tangible continuing effects of the City’s own past discrimination. To withdraw from public entities the ability to engage in this kind of racial antidiscrimination program, and to withdraw from minorities the capacity to advocate for such local antidiscrimination legislation in the future, while at the same time permitting local entities to grant preferences for other groups (without having to make a similar showing of discrimination), is precisely the kind of selective political restructuring that Hunter and Seattle sought to forbid.
*366As an additional rationale for distinguishing section 31, the majority cites two related arguments made by the Ninth Circuit. The first argument is that the Constitution does not require race-conscious measures “it barely permits.” (Coalition II, supra, 122 F.3d at p. 709; see maj. opn., ante, at pp. 327-328.) This, of course, is a red herring. No one contends the Constitution requires race- or sex-conscious affirmative action. Nor is that relevant to the inquiry, as neither the adoption of the housing ordinance in Hunter nor the busing policy in Seattle was constitutionally required. Rather, the focus of the political structure doctrine is on nonneutral structural obstacles to constitutionally permitted, not required, legislation in minorities’ interest. “What the Court did hold is that states may not place political obstacles in the way of laws that favor minorities, nor may they remove such laws wholesale from the political process. All that is constitutionally required is that minorities have the opportunity, on equal terms, to seek ‘legislation in [their] behalf’ within existing channels of government. [Citation.]” (Coalition II, at p. 715 (opn. of Norris, J., on den. rehg.).)
Second, the majority suggests that section 31 cannot violate the Constitution, as it is consistent with the “ ‘core purpose’ of the equal protection clause.” (Maj. opn., ante, at p. 332.) The Ninth Circuit put it more lyrically: “ ‘[I]t would be paradoxical to conclude that by adopting the Equal Protection Clause of the Fourteenth Amendment, the voters of the State thereby had violated it.’ (Crawford, [supra,] 458 U.S. at [p.] 535.)” (Coalition II, supra, 122 F.3d at p. 709, fn. omitted.) This is a canard. Section 31 is not coextensive with the equal protection clause. While it is true that the Constitution permits race- and sex-conscious measures only if they can survive heightened scrutiny, section 31 goes further. As the Attorney General puts it, “[section 31] closes a door to race- and gender-conscious programs that the Fourteenth Amendment leaves open.” And it does so while simultaneously leaving others free to seek preferential legislation on all other bases. This unique burden on the ability of women and racial minorities to achieve beneficial legislation in their interest is what violates the political structure doctrine, and thus the Constitution. (Hunter, supra, 393 U.S. at pp. 390-391.)34
Finally, while the majority disclaims the argument (maj. opn., ante, at p. 332, fn. 10), it bears noting that Coalition II questioned the vitality of the political structure doctrine in light of recent decisions by the high court. (Coalition II, supra, 122 F.3d at pp. 704, 705, fn. 13; see also conc. opn. of *367Corrigan, J., ante, at p. 344.)35 I do not perceive the same irreconcilability and am not convinced the Supreme Court’s more recent conventional equal protection decisions undermine Hunter and Seattle. (See Amar & Caminker, supra, 23 Hastings Const. L.Q. at pp. 1035-1039.) Whereas more recent Supreme Court cases have restricted the type of affirmative action remedies permitted under the Constitution, they do not undermine the principle that minorities may not be prevented, via nonneutral political restructuring, from seeking beneficial legislation—including constitutionally permissible affirmative action remedies.
Moreover, as noted before, our task is to faithfully apply the Supreme Court’s precedent, not to prognosticate what the court might do if presented with the opportunity to reconsider the doctrine. “If a precedent of [the high court] has direct application in a case, yet appears to rest on reasons rejected in some other line of decisions, the [lower court] should follow the case which directly controls, leaving to [the high court] the prerogative of overruling its own decisions.” (Rodriguez de Quijas v. Shearson/Am. Exp., Inc., supra, 490 U.S. at p. 484.)36
Thus, applying the Supreme Court’s decisions in Hunter and Seattle, I conclude section 31’s ban on constitutionally permissible race- and gender-conscious remedial measures violates the political structure doctrine, and I would remand the case for consideration of whether section 31 can withstand heightened scrutiny. In so concluding, I fully acknowledge the controversial nature of race- and sex-conscious measures (even when undertaken to remedy past discrimination) and the considerable passions on both sides of the issue. However, as Judge Henderson wisely commented in Coalition I, supra, 946 F.Supp. at page 1490, “this case does not call upon this Court to adjudicate whether affirmative action is right or wrong, or whether it is no longer an appropriate policy for addressing the continuing effects of past and present discrimination against racial minorities and women. Such questions, while *368they are most certainly of vital public policy interest, lie beyond the purview of this Court. . . . [f] Rather, the substantive issues raised by this action are considerably more narrow, albeit no less important: whether the particular method chosen by Proposition 209 to curtail affirmative action is unlawful because it. . . violates the rights of women and minorities to fully participate in our political system . ...” As I conclude the answer is yes, I dissent from the majority’s contrary view.
111.
The City argues it can demonstrate triable issues exist regarding whether it was constitutionally obliged to enact the 2003 ordinance, given ample evidence of pervasive intentional discrimination in the awarding of public contracts; that is, that the ordinance may be the only, or at least the most likely, means of remedying purposeful racial discrimination. As the majority explains, the trial court failed to meaningfully address this argument and remanding for a hearing would be helpful to reviewing courts. (Maj. opn., ante, at pp. 335-336.) I fully concur.
I also fully agree with the majority that the board of supervisor’s legislative findings “do not bind the court on remand” and that, where racial classifications are at issue, “ ‘simple legislative assurances of good intention cannot suffice’ [(Croson, supra, 488 U.S. at p. 500)].” (Maj. opn., ante, at p. 338, fn. 20.) When, however, as in the present case, remedial measures rest not on such “simple legislative assurances,” but on detailed factual findings, respect for constitutional separation of powers requires they be given considerable weight. (Connecticut Indemnity Co. v. Superior Court (2000) 23 Cal.4th 807, 814 [98 Cal.Rptr.2d 221, 3 P.3d 868].)
To recap, before adopting the 1989 version of the ordinance, the City considered the testimony of 42 witnesses and the written submissions from 127 minority, women, local, and other business representatives. The City subsequently held another 10 public hearings, commissioned two statistical studies, and sought additional written submissions from the public.
When the 1989 ordinance expired and the City was considering adopting a new version, it held an additional 14 public hearings, considered the live testimony of 254 witnesses, videotaped testimony of numerous other witnesses, additional statistical disparity studies, and other documentary evidence pertinent to alleged discrimination and bidding irregularities. The City also heard direct evidence concerning racial discrimination by City employees.
When the 1998 ordinance expired and the City was considering adopting the version of the ordinance challenged here, it conducted additional public *369hearings at which 134 individuals testified. It heard additional direct evidence of ongoing discrimination in the contracting process and also heard evidence that prime contractors tried to circumvent compliance with the ordinance. An additional disparity analysis was conducted, and the City’s Human Rights Commission wrote a report containing additional examples of discrimination in the process. Only after these efforts were completed did the board of supervisors make extensive legislative findings, including that the paucity of City contracts going to minority- and women-owned businesses was due to discrimination by the City and discrimination in the private sector, that the City was actively discriminating against women and minority groups in its contracting and passively participating in private sector discrimination, that the City’s contracting practices were in violation of federal law, and that the ordinance was required to remedy the discrimination against minority- and women-owned businesses.
As Justice Kennedy noted in Croson, “evidence which would support a judicial finding of intentional discrimination may suffice also to justify remedial legislative action, for it diminishes the constitutional responsibilities of the political branches to say they must wait to act until ordered to do so by a court.” (Croson, supra, 488 U.S. at p. 519 (conc. opn. of Kennedy, J.), italics added.) Questions such as whether “a race- and gender-conscious remedy is . . . the most likely . . . means of rectifying” injury from past discrimination (maj. opn., ante, at pp. 337-338) require an assessment of complex historical and socioeconomic data, and a government entity’s answer to such questions, if methodologically sound and supported by substantial evidence, should not be lightly overturned by a reviewing court.
IV.
The City also argues the ordinance is required to maintain the eligibility for funding it receives from various federal agencies, including the United States Department of Transportation. If the City is correct, the ordinance is unquestionably exempt from section 31.37 The majority rejects the City’s argument and affirms the granting of plaintiffs’ summary judgment motion. (Maj. opn., ante, at pp. 333-335.) I disagree and would remand for further proceedings.
As the majority explains (maj. opn., ante, at p. 333), title VI of the Civil Rights Act of 1964 (42 U.S.C. § 2000d) provides that “[n]o person in the United States shall, on the ground of race, color, or national origin, be excluded from participation in, be denied the benefits of, or be subjected to *370discrimination under any program or activity receiving Federal financial assistance.” Title VI also authorizes federal agencies that provide funding to issue regulations implementing its requirements. (42 U.S.C. § 2000d-l.) To that end, the Department of Transportation issued a regulation providing that, “Where prior discriminatory practice or usage tends, on the grounds of race ... to exclude individuals from participation in, to deny them the benefits of, or to subject them to discrimination under any program or activity . . . , the applicant or recipient must take affirmative action to remove or overcome the effects of the prior discriminatory practice or usage. Even in the absence of prior discriminatory practice or usage, a recipient ... is expected to take affirmative action to assure that no person is excluded from participation in or denied the benefits of the program or activity on the grounds of race . . . .” (49 C.F.R. § 21.5(b)(7) (2009), italics added.) The regulation further makes clear that race-conscious measures do not constitute prohibited discrimination “if the purpose and effect are to remove or overcome the consequences of practices or impediments which have restricted the availability of, or participation in, the program or activity . . . .” (Ibid.)
The majority concludes that the Department of Transportation regulation permits, but does not require, funding recipients to use race-conscious measures to remedy past discrimination, as “affirmative action,” as that term is meant in the regulation, encompasses both race-conscious and race-neutral measures. (Maj. opn., ante, at pp. 334-335.) Thus, the majority reasons, a recipient can remain eligible for federal funding and comply with section 31 by using race-neutral affirmative action to address racial discrimination. (Id. at p. 335.) I fully agree in most instances.
The majority, however, fails to grapple with what happens if race-neutral affirmative action is incapable of remedying the discrimination. For example, we have recognized that, “Where the state or a political subdivision has intentionally discriminated, use of a race-conscious or race-specific remedy necessarily follows as the only, or at least the most likely, means of rectifying the resulting injury. [Citations.]” (Hi-Voltage, supra, 24 Cal.4th at p. 568, italics added.) Indeed, the majority has concluded the City should have the opportunity to show whether triable issues of fact exist as to whether the Constitution requires adoption of the ordinance because, among other things, “a race- and gender-conscious remedy is necessary as the only, or at least the most likely, means of rectifying the [discrimination.]” (Maj. opn., ante, at pp. 335, 337—338.)38 If the City is successful in so demonstrating, it would be *371incongruous to conclude that the Constitution requires a race-conscious remedy (because, in part, of the inadequacy of race-neutral measures) but that the Department of Transportation’s regulations do not.39
Thus, in light of our decision to remand the matter for further consideration of the federal compulsion issue, I would also remand this claim so the superior court can consider if the City can show triable issues of fact exist as to whether it is required to take race-conscious affirmative action to maintain its funding eligibility because race-neutral measures are inadequate.
V.
Accordingly, I dissent from parts H.A. and II.B. of the majority’s opinion. I concur with part Ü.C. of the majority’s opinion and with its decision to remand the matter for further proceedings.
Unless otherwise noted, all subsequent references to a Constitution or an equal protection clause are to the federal Constitution and the federal equal protection clause.
One more fact bears emphasizing: the only race-conscious programs uniquely affected by section 31 are those that could otherwise withstand strict scrutiny; that is, those narrowly tailored to further a compelling governmental interest, such as providing a remedy for purposeful discrimination. Programs unable to survive such scrutiny were already impermissible without section 31.
As discussed below, I generally join the majority’s analysis of the City’s federal compulsion claim and concur with its decision to remand for further proceedings. I dissent from the holding as to the City’s federal funding claim.
The high court pointed out, for example, that the charter amendment did not affect “housing discrimination on sexual or political grounds, or against those with children or dogs, nor [did] it affect tenants seeking more heat or better maintenance from landlords, nor those seeking rent control, urban renewal, public housing, or new building codes.” (Hunter, supra, 393 U.S. at p. 391.)
Indeed, while the federal district court found that the initiative’s proponents had not “address[ed] ‘its appeals to the racial biases of the voters ... ’ [citation]” (Seattle, supra, 458 U.S. at p. 463), it found that they had communicated that the measure’s passage “would result in ‘no loss of school district flexibility other than in busing for desegregation purposes,’ [citation]” and they “focused almost exclusively on the wisdom of ‘forced busing’ for integration. [Citation.]” (Ibid.)
In Lee v. Nyquist, supra, 318 F.Supp. at page 712, state education officials directed the City of Buffalo, which had an appointed school board, to implement remedies to combat de facto segregation. In response, the New York Legislature enacted a statute barring state officials and appointed (but not elected) school boards from assigning students to schools on the basis of race, while leaving the school boards and state officials with student assignment authority for all other reasons. (Id. at pp. 712, 719.) The three-judge Lee court struck down the statute, applying the political structure doctrine. (Id. at pp. 718-719.)
“Racial purpose” does not refer to the drafters’ (or voters’) invidious intent. (Seattle, supra, 458 U.S. at pp. 484-486.) Rather, a racial purpose may be found when, among other things, a law is enacted because of, rather than in spite of, its effect on the racial issue. (Id. at p. 471.)
Under Washington’s Constitution, voters could propose initiatives (Seattle, supra, 458 U.S. at p. 462, fn. 4 [citing Wn. Const., art. H, § 1]) and, if a measure passed, it could not be repealed by the legislature for two years (although it could be amended within two years by a vote of two-thirds of each house of the legislature). (Seattle, supra, at p. 462, fn. 4 [citing Wn. Const., art. II, § 41].)
The majority continued, explaining that Hunter would have been virtually identical to Seattle had the charter amendment simply precluded the city council from passing any fair housing ordinance. (Seattle, supra, 458 U.S. at pp. 474-475, fn. 17.) “Surely, however, Hunter would not have come out the other way had the charter amendment made no provision for the passage of fair housing legislation, instead of subjecting such legislation to ratification by referendum.” {Ibid.)
The majority also rejected Justice Powell’s claim that Seattle was different because proponents of integrated schools remained free to use Washington’s initiative system to achieve their goals. (Seattle, supra, 458 U.S. at pp. 474-475, fn. 17.) The majority pointed out the same was true in Hunter, “[i]t surely is an excessively formal exercise, then, to argue that the procedural revisions at issue in Hunter imposed special burdens on minorities, but that the selective allocation of decisionmaking authority worked by [the ballot measure] does not erect comparable political obstacles.” {Ibid.)
The majority also emphasized that, despite the dissent’s assertions to the contrary, it was not creating a constitutional right either to local decisionmaking or to desegregative busing. (Seattle, supra, 458 U.S. at p. 480, fn. 23.) Rather, its decision was predicated upon “the comparative burden [the ballot measure] imposes on minority participation in the political *351process—that is, the racial nature of the way in which it structures the process of decisionmaking. . . . [T]he State remains free to vest all decisionmaking power in state officials, or to remove authority from local school boards in a race-neutral manner.” (Ibid.)
It also noted that, “Certainly, a state requirement that ‘desegregation or antidiscrimination laws,’ [citation], and only such laws, be passed by a unanimous vote of the legislature would be constitutionally suspect. It would be equally questionable for a community to require that laws or ordinances ‘designed to ameliorate race relations or to protect racial minorities,’ [citation], be confirmed by popular vote of the electorate as a whole, while comparable legislation is exempted from a similar procedure.” (Seattle, supra, 458 U.S. at pp. 486-487.)
The high court issued its opinions in Seattle and Crawford on the same day. Notably, four members in the Seattle majority were also in the Crawford majority (Justices Brennan, White, Blackmun (the author of Seattle), and Stevens).
Although the court’s analysis in Hunter and Seattle concerned racial minorities, nothing suggests the doctrine would not also apply to other suspect classes, such as women. (Cf. Evans v. Romer (Colo. 1993) 854 P.2d 1270, 1279-1282 [concluding the principle cannot be logically limited to the race context], affd. on other grounds sub. nom. Romer v. Evans (1996) 517 U.S. 620 [134 L.Ed.2d 855, 116 S.Ct. 1620].)
As Hunter and Seattle demonstrate, a law can restructure the political process either explicitly (e.g., the amended charter in Hunter explicitly required that future housing discrimination ordinances receive voter approval before going into effect) or implicitly (e.g., the voter initiative taking away school boards’ power to order desegregative busing in Seattle had the practical effect of requiring another ballot measure or action by the state legislature).
Indeed, the high court indicated that, had the voters in Hunter merely repealed the fair housing ordinance or had the school board in Seattle ended its race-conscious busing plan without also altering the process for obtaining future legislation, the doctrine would not have been violated. (Seattle, supra, 458 U.S. at p. 483; Hunter, supra, 393 U.S. at pp. 389-390 & fn. 5.)
The federal district court preliminarily enjoined enforcement of Proposition 209 (Coalition I, supra, 946 F.Supp. at p. 1510), but the injunction was lifted by the Ninth Circuit (Coalition for Economic Equality v. Wilson (9th Cir. 1997) 122 F.3d 692, rehg. en banc den. Aug. 27, 1997 (Coalition IT)). I cite the district court’s factual findings, which were also cited by the Ninth Circuit (id. at p. 705) and this court (Hi-Voltage, supra, 24 Cal.4th at p. 561).
As the court in Coalition I, supra, 946 F.Supp. at page 1488, explained, section 31’s ban on discrimination “simply reaffirms existing anti-discrimination protections already provided by the United States and California Constitutions, and by the 1964 Civil Rights Act. . . [and] creates no change in existing law.”
The manner in which a city contracts is generally a municipal affair. (Amaral v. Cintas Corp. No. 2 (2008) 163 Cal.App.4th 1157, 1175 [78 Cal.Rptr.3d 572].) However, the Public Contract Code requires cities and counties to use the lowest responsible bidder on public works projects (Pub. Contract Code, §§ 20160, 20162 [cities]; id., §§ 20120, 20128 [counties]) with certain enumerated exceptions (see, e.g., id., § 2002 [local agencies can give small business preference]).
Indeed, as previously noted, the substantial majority of Blacks (74 percent), Latinos (76 percent), and Asian-Americans (61 percent), and a majority of women (52 percent), voted against Proposition 209. (See, ante, at p. 356.)
The majority does not dispute that section 31 singles out a racial issue for special treatment. While the concurring opinion does so (conc. opn. of Corrigan, J., ante, at p. 340), its rationale lacks a reasoned basis. It relies on the fact that section 31 targets not only race, but gender as well, thus “broaden[ing] the application of the measure.” (Conc. opn. of Corrigan, J., ante, at p. 340.) Of course, the charter amendment in Hunter applied not only to race, but to religion as well; this fact did not prevent the charter amendment’s invalidation. (Hunter, supra, 393 U.S. at p. 387.) Moreover, as noted by Judge Norris in his dissent from the Ninth Circuit’s decision to deny en banc rehearing of the panel decision in Coalition II, combining various suspect classes into one undifferentiated group for purposes of equal protection analysis is without basis in case law or common sense. (Coalition II, supra, 122 F.3d at p. 716 (opn. of Norris, J., on denial of rehg.).)
To collect 150 percent of the threshold number, a measure’s sponsors would have to collect 1,041,531 signatures. Given a 150-day window, approximately 7,000 signatures would have to be collected on average per day.
Even in the case of a comparatively less controversial November 2008 ballot measure, like Proposition 2 (Standards for Confining Farm Animals), the statewide campaigns in favor and against the initiative spent over $19 million. (Cal. Sect, of State, Cal-Access Web site, Campaign Finance Activity <http://cal-access.ss.ca.gov/Campaign/Measures/ Detail.aspx?id=1301652> [as of Aug. 2, 2010].)
That racial minorities and women can also seek beneficial legislation on all other bases (e.g., a local, minority-owned business can seek preferences available to all locally owned businesses) does not alter this conclusion—the same was true in Hunter and Seattle.
The concurring opinion places great weight on the fact section 31 applies to public employment and education and contracting, arguing that the broad scope of section 31 distinguishes it from the laws at issue in Hunter and Seattle. (Conc. opn. of Corrigan, J., ante, at p. 340.) I disagree. As Judge Norris noted, “Neither Hunter nor Seattle—nor common sense, for that matter—supports the proposition that expanding the levels at which the State disadvantages minorities will render that action any less constitutionally suspect.” (Coalition II, supra, 122 F.3d at p. 715 (opn. of Norris, J. on den. rehg.).)
Neither the majority nor the concurrence disputes that section 31 puts unique hurdles in front of those who would seek the future enactment of race- and sex-conscious preferential legislation as opposed to those who seek preferential legislation on all other bases.
For example, Governor Pete Wilson issued an executive order on June 1, 1995, repealing race- and gender-conscious employment practices under his immediate control. (Governor’s Exec. Order No. W-124-95 (June 1, 1995).) Similarly, on July 20, 1995, the Regents of the University of California acted of their own accord to discontinue the use of race and gender preferences in contracting, employment, and admissions decisions. (Regents of U.C. (Regents), Res. No. SP-1 [admissions]; Regents, Res. No. SP-2 [contracting and employment].) While both of these pre-Proposition 209 decisions resulted in the elimination of race- and sex-conscious preferences, they simply reflected the operation of the normal political process.
For example, while the Regents subsequently decided to rescind resolutions Nos. SP-1 and SP-2 (Policy on Future Admissions, Employment, and Contracting (May 16, 2001)), the policy reversal is essentially without effect in light of section 31.
In San Francisco, for example, 70.5 percent of the electorate opposed Proposition 209. (Coalition I, supra, 946 F.Supp. at p. 1507.)
The majority relies on decisions of the Ninth and Sixth Circuit Courts of Appeals (Coalition to Defend Affirmative Action v. Granholm (6th Cir. 2006) 473 F.3d 237; Coalition II, supra, 122 F.3d 692). (Maj. opn., ante, at pp. 327, 330, 332.) As the majority acknowledges (id. at pp. 329-330), however, while decisions of lower federal courts on questions of federal law are persuasive, they do not bind us; rather, we make an independent determination of federal law. (Barrett v. Rosenthal (2006) 40 Cal.4th 33, 58 [51 Cal.Rptr.3d 55, 146 P.3d 510]; see, e.g., Steffel v. Thompson (1974) 415 U.S. 452, 482, fn. 3 [39 L.Ed.2d 505, 94 S.Ct. 1209] (cone. opn. of Rehnquist, J.).)
To the extent the majority opinion can be read to imply the Supreme Court could not have anticipated that the phrase “beneficial legislation” might be read to include race-conscious affirmative action (maj. opn., ante, at pp. 330-332), I disagree. Executive orders promulgated by Presidents Kennedy (Exec. Order No. 10925, 26 Fed.Reg. 1977 (Mar. 6, 1961)) and Johnson (Exec. Order No. 11246, 30 Fed.Reg. 12319 (Sept. 28, 1965)) required the use of affirmative action to combat discrimination. At a minimum, by the time Seattle was decided in 1982, the notion of race- and sex-conscious affirmative action was very well understood.
The majority goes astray by wrongly focusing on the difference between race-conscious preferences on the one hand and the Hunter housing ordinance and the Seattle busing policy on the other hand. (Maj. opn., ante, at pp. 330-332.) The focus should instead be on the erected legislative hurdles: section 31, the Hunter charter amendment, and the Seattle statewide initiative. All three similarly burdened minorities’ equal and meaningful access to the political process, undermining their ability to achieve favorable legislation in a nonneutral fashion.
The difference between constitutionally permissible affirmative action and the race-conscious busing addressing de facto segregation in Seattle is even less clear. Moreover, as regarding the political structure doctrine, the Supreme Court necessarily rejected any distinction between the antidiscrimination law in Hunter and the race-based remedial efforts in Seattle.
As one commentator aptly put it, “[i]f the Equal Protection Clause does not prohibit the race and gender preferences barred by Proposition 209, compliance with the Equal Protection Clause cannot constitute a justification for Proposition 209’s discriminatory invalidation of race and gender preferences.” (Spann, Proposition 209 (1997) 47 Duke L.J. 187, 255.)
Of course, while the majority states it does not question the doctrine’s “continuing validity” (maj. opn., ante, at p. 332, fn. 10), it nonetheless uses “historical context” (ibid.) to justify a cramped reading of the doctrine. (See, ante, at pp. 363-364.)
In Romer v. Evans, supra, 517 U.S. 620, the high court invalidated a state initiative prohibiting public entities from taking actions to protect gays and lesbians. The state supreme court concluded the measure was subject to strict scrutiny, relying on the political structure doctrine. (Id. at pp. 625-626.) The Supreme Court affirmed the judgment, but based its decision on a conventional equal protection analysis. (Id. at pp. 633-634.) In dissent, Justice Scalia suggested the majority “implicitly rejected]” the state supreme court’s political structure doctrine holding. (Id. at p. 640, fn. 1 (dis. opn. of Scalia, J.).) I respectfully disagree. To the contrary, Justice Kennedy’s majority opinion is fully consistent with the underpinnings of the political structure doctrine. (See id. at p. 633.) Moreover, contrary to the concurrence’s suggestion (conc. opn. of Corrigan, J., ante, at p. 343, fn. 3), the applicability of the doctrine in Romer was not so plain. Rather, the invalidated initiative in Romer explicitly targeted and burdened gays and lesbians. Thus, a conventional equal protection analysis was called for.
Section 31, subdivision (e) exempts actions “which must be taken to establish or maintain eligibility for any federal program, where ineligibility would result in a loss of federal funds to the State.”
The majority states that its resolution of this claim would be unaffected were the City to demonstrate the existence of purposeful discrimination on remand. (Maj. opn., ante, at p. 333, fn. 14.) It reasons that, if the ordinance is required by the equal protection clause, the City’s federal funding claim “has no practical significance.” (Ibid.) Even assuming the majority is correct on this point, if the ordinance is only permitted, and not required, under the *371equal protection clause, the City should have the opportunity to demonstrate race-neutral measures will not suffice to comply with its obligations under federal regulations.
Funding recipients cannot reasonably be required to employ ineffective race-neutral measures, nor, indeed, would it appear that the use of fruitless measures would satisfy entities’ obligations under title VI and agencies’ implementing regulations.