[738]Opinion
GEORGE, C. J. We granted review to address an issue of first impression: the proper interpretation of California Constitution, article XVIII, section 1 (article XVIII, section 1), which requires in its second sentence that when the Legislature proposes an amendment of the state Constitution, “[e]ach amendment shall be so prepared and submitted that it can be voted on separately.”
We conclude, as did the Court of Appeal below, and consistent with our provision’s language and history and more than a century of out-of-state decisions construing the essentially identical provisions of nearly 30 other state constitutions, that the separate-vote provision is a limitation upon legislative power to submit constitutional amendments to the voters.
We disagree, however, with the Court of Appeal below, concerning the applicable test for determining whether, in a given case, the Legislature’s submission of constitutional changes in a single measure violates article XVIII, section 1. In addressing that question, the Court of Appeal followed a minority rule that recently was reinvigorated by Armatta v. Kitzhaber (1998) 327 Ore. 250 [959 P.2d 49] (Armatta)—a decision in which the Oregon Supreme Court construed its state’s separate-vote provision as establishing a test different from and stricter than the traditional test employed by courts under a related constitutional provision also found in most state constitutions—the “single subject rule” (see Cal. Const., art. II, § 8, subd. (d); id., art. IV, § 9). Unlike the Oregon court and a few other courts that have followed Armatta under their respective state constitutions, we find no basis in the history of the California Constitution for such a conclusion, and hence we shall follow the approach that is, and has been, the majority rule for nearly 130 years: the separate-vote provision should be construed consistent with its kindred provision, the single subject provision.
So construing the separate-vote provision of article XVIII, section 1, we conclude that the Legislature’s proposed submission, in a single constitutional amendment, of two changes to the state Constitution that are not germane to a common theme, purpose, or subject, violated the constitutional separate-vote requirement. Accordingly, we affirm this aspect of the judgment rendered by the Court of Appeal, although for reasons different from those relied upon by that court.
We also address the question of remedy. The Court of Appeal, by a two-to-one vote, ordered the Secretary of State to separate the two proposed constitutional changes at issue in this matter into two measures for submission to the voters. When ruling upon this matter in the weeks preceding the November 2004 general election (and only days before the deadline for the [739]printing of ballot materials), we declined to disturb the Court of Appeal’s order, and the voters of this state subsequently adopted each separate constitutional amendment. Although we conclude that the Court of Appeal erred by ordering bifurcation, we find it unnecessary and inappropriate to invalidate either of these separately submitted and approved constitutional amendments.
I
Proposition 62, an initiative that qualified for the November 2, 2004, statewide General Election ballot, proposed a constitutional amendment to permit so-called open primaries.1 In an apparent response to that measure, both houses of the Legislature passed by a two-thirds vote Senate Constitutional Amendment No. 18 of the 2003-2004 Regular Session (Sen. Const. Amend. No. 18, Stats. 2004 (2003-2004 Reg. Sess.) res. ch. 103, hereafter Resolution 103) for submission to the voters on the November 2004 ballot. As adopted, Resolution 103 proposed, in a single measure, two changes to the state Constitution.
The first change concerned primary elections and evidently was designed to conflict with and supersede the competing initiative measure, Proposition 62. This part of Resolution 103 proposed to amend article II of the California Constitution by adding section 5, subdivision (b), which provided that a political party’s top vote getter in a primary election must be permitted to run in the ensuing general election.2 The second change set forth in Resolution 103 concerned state property and proposed to amend article III of the California Constitution by adding a new section 9 to provide a means for the state to pay bond obligations.3
[740]After Resolution 103 was designated Proposition 60 by the Secretary of State, petitioners Californians for an Open Primary and Nick Tobey (Californians for an Open Primary)—proponents of Proposition 62—filed a petition for a writ of prohibition in the Court of Appeal, seeking to bar the Secretary of State from placing Proposition 60 on the general election ballot on the ground that its submission as a single ballot proposition would violate the separate-vote provision of article XVIII, section 1. The Legislature of the State of California filed opposition. The Court of Appeal agreed with Californians for an Open Primary that Resolution 103, submitted in a single measure as Proposition 60, violated the Constitution’s separate-vote provision. In reaching that conclusion, the appellate court relied upon the Oregon Supreme Court’s construction of that state’s own separate-vote provision and unanimously endorsed a strict test focusing upon “ ‘whether, if adopted, the proposal would make two or more changes to the constitution that are substantive and are not closely related.’ ” (Quoting Armatta, supra, 959 P.2d 49, 64, italics added.) The Court of Appeal also unanimously found that the proposed submission violated this test but, by a two-to-one vote, rejected the assertion that Proposition 60 should be stricken from the ballot, instead issuing a peremptory writ of mandate directing the Secretary of State to prepare the ballot “so that section 5 of article II and section 9 of article III, as proposed in [Resolution 103], will be submitted to the voters as separate measures to be voted on separately.”
Both Californians for an Open Primary and the Legislature petitioned this court for review. The petition of Californians for an Open Primary took issue with substantial aspects of the Court of Appeal’s analysis, but agreed with the lower court’s conclusion that submission of Proposition 60 as a single measure would violate the separate-vote provision. Petitioners urged, however, that the Court of Appeal majority erred in deciding, as a remedy for the separate-vote violation, that the measure should be bifurcated and presented on the ballot as two separate measures; the proper remedy, petitioners asserted, instead was an order barring the measure from appearing on the ballot, and hence petitioners requested a stay of the Court of Appeal’s bifurcation order. The Legislature’s petition for review, by contrast, disagreed with both (1) the Court of Appeal’s separate-vote-pro vision analysis and conclusion, and (2) the Court of Appeal majority’s bifurcation remedy, and further argued that the request for a stay should be denied “and review of the [741]remedy should occur only in conjunction with review on the merits.” The Legislature argued for an “approach that truly aids this Court’s jurisdiction,” namely, that “this Court’s review proceed in an orderly manner that does not prejudge the merits, that fairly balances the interests of the parties and that protects the people’s rights in the approaching election. To that end, given the exigencies of the case, the Legislature acquiesces in the remedy ordered by the Court of Appeal as it applies to the November election and asks that this Court leave that order in place while it reviews the case.”
We unanimously granted review to address the merits of both substantive issues presented—that is, the proper interpretation of article XVHI, section 1, and the Court of Appeal’s bifurcation remedy. In light of the impending election and ballot preparation deadlines, we also ordered the Secretary of State to place Resolution 103 on the November 2004 ballot “in the manner directed by the Court of Appeal”—that is, as Propositions 60 (the primary election provision) and 60A (the state property/bonds repayment provision), and we denied the request for a stay.4 [742]At the November 2004 election, the voters rejected Proposition 62 and enacted both Propositions 60 and 60A.5
As noted above, although Californians for an Open Primary agrees with the Court of Appeal’s conclusion that the separate-vote provision of article XIII precluded the Legislature from joining the disparate provisions of Resolution 103 in a single proposed constitutional amendment, it maintains that the Court of Appeal erred in bifurcating the resolution into two separate proposed constitutional amendments and directing that the measure be placed on the ballot as two separate propositions. In the briefing filed in this court after the November 2004 election, Californians for an Open Primary argues that the Court of Appeal’s error in this regard requires invalidation of both constitutional amendments despite the voters’ approval of each measure at the November 2004 election. Because Californians for an Open Primary maintains that the appropriate remedy in this case is invalidation of the two measures enacted by the voters, the remedy issue that is presented remains alive and is not moot.
II
Article XVIII, addressing the subject of amending and revising the Constitution, is comprised of four sections. The first section—the second sentence of which we must construe in this case—provides: “The Legislature by rollcall vote entered in the journal, two-thirds of the membership of each house concurring, may propose an amendment or revision of the Constitution and in the same manner may amend or withdraw its proposal. Each amendment shall be so prepared and submitted that it can be voted on separately.” (Art. XVIII, § 1, italics added.)6 Although this provision (hereafter sometimes [743]referred to as the separate-vote provision)—or one essentially identical to it—has existed in this article since it was added to our Constitution in 1879, no California decision has defined the scope of the provision.
A
On its face, the separate-vote provision appears to limit legislative power. The Legislature, however, insists that this provision in fact has a different and quite opposite purpose and effect. It argues that instead of limiting legislative authority to package disparate proposed changes in a single measure, the separate-vote provision actually guarantees that however the Legislature deems it appropriate to combine or separate proposed changes, those changes will be submitted to the voters in the chosen manner. Specifically, the Legislature asserts that the word “amendment” in the second sentence of article XVIII, section 1 “has always been” construed to “mean[] a Senate Constitutional Amendment ... or an Assembly Constitutional Amendment, with the resulting assurance that when more than one [Senate Constitutional Amendment] or [Assembly Constitutional Amendment] appears on the same ballot each such legislative constitutional amendment will be prepared and submitted so that it ‘can be voted on separately.’ ” The Legislature concludes: “Pursuant to this . . . view of [article XVIII,] section 1, the Legislature, by two-thirds vote of each house, is free to combine disparate substantive changes within a single legislative constitutional amendment for submission to the people. . . . The ‘separate vote’ requirement guarantees that neither the Executive [that is, the Secretary of State] nor succeeding legislative majorities may interfere with these determinations.” (Italics added.) In other words, as Californians for an Open Primary observes, the Legislature views the separate-vote provision as not a limitation upon itself, but instead as a restraint upon hypothetical “renegade Secretaries of State” who might take it upon themselves either to combine separate measures that the Legislature has determined should be submitted to the voters separately, or to separate measures that the Legislature has determined should be submitted to the voters as a package.
As explained below, we do not find support for the Legislature’s view in the language of the second sentence of article XVIII, section 1, or in the provision’s history or the case law construing that provision or similar provisions in the charters of our sister states. Nor, contrary to the Legislature’s position, do we find support for its construction of the provision in past legislative constitutional amendment measures adopted by the electorate, or in the circumstance that, since 1962, the Legislature has had authority to propose not only amendments to the Constitution, but revisions as well.
[7441]
We first review the text of the provision. Its first sentence, addressing the power to propose a constitutional amendment or revision, is directed expressly to “the Legislature.” The second sentence, which we must construe in the present case (providing that “[e]ach amendment shall be so prepared and submitted that it can be voted on separately”) also appears to be directed to the Legislature. There is no indication in the language of the provision that the second sentence was directed toward an unidentified entity within the executive branch, such as the Secretary of State.
The Legislature insists nevertheless that the word “amendment” in the second sentence of article XVIII, section 1 (the separate-vote provision) means or refers to the legislative vehicle employed by the Legislature, that is, a resolution proposing a Senate Constitutional Amendment or an Assembly Constitutional Amendment. We note, however, that the same word (“amendment”) also appears in the first sentence of section 1 of article XVIII. In that context it is clear the word refers not to the legislative vehicle for proposing a change, but instead to the substantive content of such a proposal. Although it is possible that the word “amendment” might be employed in a different sense in the second sentence of article XVIII, section 1, as explained below there is no evidence that the drafters or the electorate in 1879 or thereafter ever so intended, or that they even contemplated the construction that the Legislature now places upon the provision.7
2
Mindful of the admonitions set forth by Justice Landau in his article, A Judge’s Perspective on the Use and Misuse of History in State Constitutional Interpretation (2004) 38 Val.U. L.Rev. 451 (Landau),8 we next review the [745]history of California’s separate-vote provision. Former article X of the 1849 Constitution set forth the procedures for the Legislature to propose one or more amendments to the Constitution (id., § 1), or to revise the entire Constitution (id., § 2). Former article X, section 1 allowed the Legislature to propose a constitutional amendment based upon a majority vote of both houses of two successive legislative sessions, but it placed no limitation upon the manner in which the submission was to be made to the electorate.9 Indeed, the former provision did the opposite: it made it the “duty of the Legislature to submit such proposed amendment or amendments to the people, in such manner and at such time as the Legislature shall prescribe.” (Cal. Const. of 1849, art. X, § 1, italics added.)
By the time of California’s second (and only other) Constitutional Convention, in 1878-1879, this deferential approach to legislative submission of proposed constitutional amendments had been rejected in most jurisdictions, in favor of a provision apparently first adopted by New Jersey in 1844. The New Jersey constitutional clause provided that, with regard to a proposed amendment or amendments, the legislature had a “ ‘duty ... to submit such proposed amendment or amendments ... to the people, in such manner and at such time, ... as the legislature shall prescribe’ ”—but that state’s provision also regulated the legislature’s authority to prescribe the manner of submission, by further stating: “ ‘provided, that if more than one amendment be submitted, they shall be submitted in such manner and form that the people may vote for or against each amendment separately and distinctly.'’ ” (Thorpe, The Federal and State Constitutions (1909) 2613 (Thorpe), quoting N.J. Const. of 1844, art. IX, italics in Thorpe deleted, new italics added; see also Cambria v. Soaries (2001) 169 N.J. 1 [776 A.2d 754, 761-762] (Cambria) [construing the New Jersey provision as imposing a limitation upon the state legislature].) By the late 1870’s, at least two other states had essentially identical constitutional provisions (see Thorpe, supra, at pp. 1153 [Iowa Const. of 1857, art. X, §§ 1 & 2], 4093-4094 [Wis. Const. of 1848, [746]art. XII, § 1]), and numerous others had substantially similar provisions imposing both an obligation upon the state legislature to submit its proposed constitutional amendments to the people, and specifying that such amendments were to be submitted in a manner that allowed the people to vote separately on proposed amendments. (See Thorpe, supra, at pp. 367 [Ark. Const. of 1874, art. XIX, § 22], 873 [Ga. Const. of 1877, art. XIII, § 1, par. 1], 1091 [Ind. Const. of 1851, art XVI, §§ 1 & 2], 1257 [Kan. Const. of 1859, art. 14, § 1], 1467 & 1514 [La. Const. of 1868, tit. IX, art. 147, & La. Const. of 1879, art. 256], 1823 [Md. Const. of 1867, art XIV, § 1], 2087 [Miss. Const. of 1868, art. XIII], 2385 [Neb. Const. of 1875, art. XV, § 1], 2933 [Ohio Const. of 1851, art. XVI, § 1], 3016 [Or. Const. of 1857, art XVII, §§ 1 & 2], 3148 [Pa. Const. of 1873, art. XVIII, § 1], 3304 [S.C. Const. of 1868, art. XV, §§ 1 & 2], 4063 [W.Va. Const. of 1872, art. XIV, § 2].)10
California’s drafters in 1878 to 1879 followed this trend. Three delegates offered three different proposals for a procedure to amend the Constitution, each of which rejected the deferential approach of the 1849 Constitution under which the Legislature was granted unregulated power to “prescribe” the “manner” of submission to the electorate. (Cal. Const. of 1849, art. X, § 1.) Each delegate suggested instead a separate-vote provision that was very similar to those existing at that time in the various other state constitutions cited above.11 All three suggestions were referred to the Committee on Future Amendments for consideration.
[747]In mid-December of 1878 the committee filed its first report, proposing to retain the basic structure of the then-existing 1849 Constitution’s corresponding article X by addressing in section 1 the issue of amending the Constitution by a vote of the people, and by addressing in section 2 the issue of revising the Constitution by calling a convention. Regarding section 1 and the process of amendment, the committee proposed to require a two-thirds vote of each legislative house on “any amendment or amendments,” and to retain the language, very similar to that present in the 1849 Constitution and the then-existing charters of many other states, imposing upon the Legislature the “duty ... to submit such proposed amendment or amendments to the people in such manner and at such time as may deemed expedient.” (2 Willis & Stockton, supra, at p. 800.) But the committee, consistent with all three proposals from the convention delegates and with the corresponding provisions of the then-extant charters of most other states, also included a further clause, as follows: “Should more than one amendment be submitted at the same election, they shall be so prepared and distinguished, by numbers or otherwise, that they can be voted on separately.” (Ibid.)
The committee’s proposal thereafter came up for debate before the assembled delegates in early February 1879. (3 Willis & Stockton, supra, at p. 1276.) Insofar as section 1 was concerned, the delegates debated and ultimately voted to retain the rule requiring a two-thirds vote of both houses, and they debated and ultimately revised the provision’s publication requirement (3 Willis & Stockton, supra, at pp. 1276-1277), but there was no debate or discussion concerning the separate-vote provision. (Ibid.) When the matter again came up for vote approximately two weeks later, the delegates adopted the language quoted above and referred it to the Committee on Revision and Adjustment. (Id., at pp. 1445-1446.) Approximately 10 days later, that committee proposed to designate the provision article XVIII, sections 1 and 2, and to make various technical amendments to both sections, including one to section l’s separate-vote provision itself. That amendment proposed to substitute the word “each” for “they” in the final clause, so that the provision would read as follows: “and it shall be the duty of the Legislature to submit such proposed amendment or amendments to the people in such manner, and at such time, and after such publication as may be deemed expedient. Should more than one amendment be submitted at the same election, they shall be so prepared and distinguished, by numbers or otherwise, that each can be voted on separately.” (3 Willis & Stockton, supra, at p. 1505.)
In light of the history recounted above—that is, the apparently unchallenged view among the delegates that the 1849 Constitution’s corresponding provision allowing the Legislature unfettered discretion to prescribe the “manner” of submission should be rejected in favor of an approach limiting that [748]discretion by imposing a separate-vote requirement essentially identical to that adopted by most other states in the intervening decades—it seems highly improbable that the California drafters viewed the separate-vote provision as a limitation upon some unnamed entity rather than upon the Legislature itself.12 Indeed, as the New Jersey Supreme Court recently observed, the constitutional record in that state demonstrates that the drafters plainly saw their very similar provision as a restraint upon the state legislature, and not as a limitation upon some other entity. (See Cambria, supra, 776 A.2d 754, 762 [relying upon proceedings of the New Jersey Constitutional Conventions of 1844 and 1947].) We discern no basis for concluding that the California drafters, or the electorate who subsequently adopted the Constitution in 1879, thought otherwise.
3
We next review our case law’s interpretation of the separate-vote provision of article XVIII, section 1. This court has addressed the provision only once, in Wright v. Jordan (1923) 192 Cal. 704 [221 P. 915] (Wright). As explained below, in doing so we repeatedly characterized the provision as imposing a “limitation” upon the Legislature, and not, as the Legislature now argues, a protection of a legislative prerogative and/or a limitation upon the Secretary of State.
The petitioner in Wright (the City Clerk of San Diego) sued the Secretary of State (Frank C. Jordan) to compel him to recognize and file the results of an election concerning the consolidation of the City of San Diego and the City of East San Diego. (Wright, supra, 192 Cal. 704, 706.) Secretary of State Jordan refused to take this action on the ground that the statute under which the consolidation election had proceeded was unconstitutional because, he asserted, it in turn was based upon an improperly adopted amendment to article XI, former section 8 1/2, subdivision 7 of the Constitution. That constitutional amendment (governing consolidation elections) earlier had been adopted by the electorate as a constitutional initiative under the power conferred upon the voters by constitutional amendment in 1911. The Secretary of State argued in Wright that the prior amendment to article XI, former section 8 1/2, subdivision 7, was void because it had not been adopted “in the manner required by the provisions of section 1 of article XVIII” (Wright, [749]supra, 192 Cal. at p. 711)—that is, in the manner required of legislative constitutional amendments. In rejecting the Secretary of State’s argument, this court in Wright quoted article XVIII, section l’s separate-vote provision and then noted that, with regard to the people’s authority to adopt constitutional amendments through the initiative process, “no such limitation as is embodied in the provisions of section 1 of article XVIII . . . can be found.” (Wright, supra, 192 Cal. at p. 711.) The court in Wright twice more characterized section 1 of article XVIII as setting forth a “limitation” (Wright, supra, 192 Cal. at p. 711), and thereafter acknowledged that the amendment to article XI, former section 81/2, subdivision 7 “effectuate[d] changes in several already existing articles, sections, and clauses of the constitution, and . . . these [were] not presented in the form of separate amendments.” (Wright, supra, 192 Cal. at p. 712.) Nevertheless, the court reiterated, there had been no need to comply with article XVIII section l’s separate-vote requirement, because, again, that requirement did not apply to amendments proposed by initiative. (Wright, supra, 192 Cal. at p. 712.)13
We conclude, contrary to the Legislature’s suggestion, that Wright, supra, 192 Cal. 704, far from supporting the Legislature’s interpretation of the separate-vote provision, does the opposite and supports the view that the provision imposes a limitation upon the Legislature.
4
Finally, we consider the decisions of our sister states construing their own essentially identical separate-vote provisions. Those opinions unanimously interpret their provisions as imposing a limitation upon the legislature’s power to submit constitutional changes to the voters. Each decision has found, explicitly or implicitly, that the provision is designed to limit legislative power by barring submissions that otherwise might cause voter confusion or constitute “logrolling”—that is, the practice of combining two or more unrelated provisions in one measure, thereby forcing a single take-it-or-leave-it vote on matters that properly should be voted upon separately. (E.g., Gabbert v. Chicago, R. I. & P. Ry. Co. (1902) 171 Mo. 84 [70 S.W. 891, 897] (Gabbert); Jones v. McClaughry (1915) 169 Iowa 281 [151 N.W. 210, 216]; State v. Wetz (1918) 40 N.D. 299 [168 N.W. 835, 847] (Wetz); Kerby v. Luhrs [750](1934) 44 Ariz. 208 [36 P.2d 549, 551-552] (Kerby); Keenan v. Price (1948) 68 Idaho 423 [195 P.2d 662, 676]; Moore v. Shanahan (1971) 207 Kan. 1 [486 P.2d 506, 516] (Moore); Carter v. Burson (1973) 230 Ga. 511 [198 S.E.2d 151, 156] (Carter); In re Initiative Petition No. 314 (1981) 1980 OK 174 [625 P.2d 595, 603-605] (Petition No. 314); Andrews v. Governor of Maryland (1982) 294 Md. 285 [449 A.2d 1144, 1149-1150] (Andrews); State ex rel. Clark v. State Canvassing Bd. (1995) 119 N.M. 12 [888 P.2d 458, 461]; IWP v. State Bd. of Land Comm’rs (1999) 133 Idaho 358 [982 P.2d 358, 363] (IWP); Cambria, supra, 776 A.2d 754, 764.)14 Although decisions finding a violation of a state constitution’s separate-vote provision are relatively rare in comparison with the scores of cases that have rejected such claims, at least 19 decisions from eight jurisdictions during the past 100 years (with most occurring in the past eight years) have so concluded—determining in each instance that constitutional changes packaged in a single measure should have been separately submitted to the voters.15 No decision from any jurisdiction has suggested that a separate-vote provision has the purpose or effect advanced by the Legislature in the present case. Indeed, the Legislature’s argument has not even been raised in any of the scores of cases that we have reviewed.
[7515]
The Legislature insists, nevertheless, that its view—that the separate-vote provision protects the Legislature’s right to package constitutional amendment measures as it sees fit—is supported by (i) contemporaneous and ongoing practices, and (ii) the circumstance that, since 1962, the Legislature has had authority to propose not only constitutional amendments, but also constitutional revisions. As explained below, we are not persuaded.
a
The Legislature cites approximately 30 constitutional amendment measures that it submitted to the electorate between 1892 and 2004, and argues that many if not most of those measures would have failed the strict test proposed below by the Court of Appeal. This circumstance, according to the Legislature, demonstrates the propriety of its own interpretation of the separate-vote provision.
We agree generally that long-established and adhered-to practice with regard to a constitutional provision informs a court’s interpretation of such a provision. (E.g., People v. Southern Pac. Co. (1930) 209 Cal. 578, 595 [290 P. 25] [“ ‘contemporaneous and long continued construction’ ” of a constitutional provision “ ‘by the legislature is entitled to great deference’ ”].) Indeed, early out-of-state cases so proceeded in arriving at a lenient and accommodating—rather than a narrow and strict—construction of their own separate-vote provisions. (E.g., State v. Timme (1882) 54 Wis. 318 [11 N.W. 785, 791-793] (Timme).) On the facts of the present case, however, we find the Legislature’s argument unpersuasive.
The Legislature cites nothing to suggest that, at any time prior to the commencement of this litigation, the Legislature—by rule, legal opinion, or otherwise—actually held the presently stated view of the constitutional provision. Moreover, as the Court of Appeal below observed, “we do not know that any of [the Legislature’s] examples [of prior constitutional amendment measures] violate section 1 for no cases have been brought to test them.” In any event, even assuming that, as the Legislature suggests, some of the prior legislative constitutional amendment measures that were adopted by the electorate might have failed the strict separate-vote test endorsed by the Court of Appeal below, as we shall explain post, part II.B.3, we do not endorse that strict test or anything like it. The Legislature does not argue that any prior measure cited would fail under scrutiny of the separate-vote provision as we shall construe it in this case.
In conclusion on this point—and contrary to the Legislature’s argument based upon past practices—we find it highly improbable that, despite every [752]other jurisdiction’s long-standing view of the separate-vote provision as a limitation upon the Legislature’s authority to submit proposed constitutional amendments, California’s drafters, electors, and Legislature ever did (or reasonably could) view our own provision otherwise.
b
The Legislature argues that, even if, for the first eight decades of its existence, the separate-vote provision in California (like the essentially identical provision in numerous other jurisdictions) operated as a limitation upon the Legislature, that restraint effectively and silently was abrogated when, in 1962, the electorate amended the first sentence of article XVIII, section 1—or at least when, in 1970, the electorate adopted the present version of article XVIII, section 1. In order to address the Legislature’s contention, we must in some detail review the history upon which it relies.
Like many other states in the late 1950’s and early 1960’s, the California Legislature in 1956 authorized and thereafter appointed a Citizens Legislative Advisory Commission to make recommendations for legislative improvement and reform. (See, e.g., Advisory Com., Final Rep. to Cal. Leg. and Citizens of Cal. (Mar. 1962), p. 9.) One of the major recommendations of the Advisory Commission was that the Constitution should be changed to allow for the Legislature to propose not only amendments or the calling of a constitutional convention to revise the charter, but also to permit the Legislature to propose a wholesale or partial constitutional revision without the need to call a constitutional convention. (Id., at pp. 42-44.)16 The Legislature agreed with this recommendation and in resolution chapter 222 (Assem. Const. Amend. No. 14, Stats. 1961 (1961 Reg. Sess.) res. ch. 222, pp. 5013-5014) proposed to alter the first sentence of then-existing article XVIII, section 1, by allowing the Legislature to propose not only amendments to the Constitution, but also revisions.17 The resulting measure was submitted to the electorate at the November 1962 general election as Proposition 7. The ballot materials submitted to the voters concerning that [753]matter did not suggest that the proposed amendment in any manner would change the meaning or effect of the unaltered second sentence of the section—the sentence at issue in this case, which is directed exclusively to the question of legislative proposals for amendment to the Constitution and which contains the separate-vote provision.18
Thereafter, in 1963, the Legislature appointed a Constitution Revision Commission (Revision Commission), which undertook to analyze and propose to the Legislature revisions to the entire state Constitution. (See generally Sumner, Constitution Revision by Commission in California (1972) 1 Western St.U. L.Rev. 48.) Based upon the Revision Commission’s recommendations, the Legislature in 1966 invoked its new power to propose revision of the Constitution and submitted for the electorate’s approval resolution chapter 139 (Assem. Const. Amend. No. 13, Stats. 1966, 1st Ex. Sess. 1966, res. ch. 139, pp. 960-982) (hereafter Resolution 139)—an omnibus proposed revision of articles III, IV, V, VI, VII, VIII, XIII, and XXII of the Constitution. Prior to setting forth the various proposed changes, Resolution 139 provided: “Resolved by the Assembly, the Senate concurring, That the Legislature of the State of California . . . hereby proposes to the people of the State of California that portions of the Constitution of the state be revised as follows: . . .” (Assem. Const. Amend. No. 13, Stats. 1966, 1st Ex. Sess. 1966, res. ch. 139, p. 960, second italics added.) Thereafter Resolution 139 set forth the Legislature’s proposed amendments (including numerous repeals, modifications, and additions) to the eight disparate articles mentioned above.
Consistent with the Legislature’s characterization of Resolution 139 as a proposed revision of the Constitution, the Attorney General’s ballot title for the measure commenced as follows: “Constitutional Revision. Legislative Constitutional amendment. . . .” (Ballot Pamp., Gen. Elec. (Nov. 8, 1966) analysis of Prop. 1-a, p. 1.) The Secretary of State designated this omnibus measure as Proposition 1-a (ibid.), and the electorate adopted it at the November 1966 general election.
After successfully promoting this major revision, the Revision Commission proceeded with the second phase of its work, considering changes to numerous additional articles of the Constitution, including the one at issue in the present litigation—article XVIII, as it recently had been amended in 1962. In [754]its review of section 1 of that article, the Revision Commission focused upon “[t]he only restriction” that it contained—that is, the separate-vote provision. (Rev. Com., Article XVIII, Amending and Revising the Constitution, Background Study 7 (May 1967) p. 18 (hereafter Background Study 7).) The Revision Commission repeatedly characterized that provision as a “limitation” and, addressing its scope, cited this court’s 1923 dictum in Wright, supra, 192 Cal. 704, 712 (quoted ante, fn. 13) for the proposition that the provision “might be interpreted broadly enough to permit proposal of an amendment that would effect changes in several parts of the Constitution where all changes relate to the same general subject matter.” (Background Study 7, supra, at pp. 18-19.)19 The Revision Commission then commented: “At any rate the Legislature, it seems, could easily get around the limitation by the means of classifying the proposal as a revision. Consequently the provision as a limitation on the power of the Legislature seems to be of little practical value, except as a caution.” (Id., at p. 19; see also id., at pp. 4, 25; Rev. Com., Proposed Revision of Cal. Constitution (Feb. 15, 1968) Com. on Revised Provisions, p. 109 [formally recommending to the Legislature that the separate-vote provision be deleted “as ineffective because it can be circumscribed by entitling several amendments as a revision”].)
At some point after release of the Revision Commission’s February 15, 1968 recommendations, the Legislature rejected the Revision Commission’s recommendation to delete the separate-vote provision from the second sentence of article XVIII, section 1. Why the Legislature did so is not reflected in any document of which we are aware. As explained below, however, the Legislature’s current position appears to be this: Either in the early 1960’s, or during or after early 1968, the Legislature decided (against the Revision Commission’s 1968 recommendation) to retain the separate-vote provision based upon a theory that the provision’s language usefully could be construed—in the manner presently undertaken by the Legislature—as effectuating not a limitation upon its authority to submit amendments, but instead a protection of its asserted prerogative to package disparate proposed changes in a single measure as it wishes, without any restraint or interference by the Secretary of State. In addition, the Legislature appears to suggest that the electorate, by subsequently reenacting the separate-vote provision in November 1970 to read as it does today, confirmed the Legislature’s latter (yet previously unstated) interpretation of the provision.
In other words, the Legislature suggests, even if prior to the 1962 amendment the separate-vote provision was solely a limitation upon the [755]Legislature, subsequent to that event and culminating in November 1970, the provision metamorphosed into the opposite: a protection against any undoing of legislative packaging. In support, the Legislature relies upon inferences that it draws from the history of various ballot propositions proposed to the electorate in the late 1960’s through November 1970.
As the Legislature observes, in 1968 the Legislature, following its earlier successful support of constitutional revision via 1966’s Proposition 1-a, proposed resolution chapter 184 (Assem. Const. Amend. No. 30, Stats. 1968 (1968 Reg. Sess.) res. ch. 184, pp. 3301-3316) (hereafter Resolution 184)—a phase two omnibus revision measure designed to make numerous additional changes to 13 disparate articles of the Constitution, including the article and provision here at issue, article XVIII, section 1. As noted above, in drafting the measure the Legislature, for reasons unclear, rejected the Revision Commission’s recommendation to delete the separate-vote provision, and instead proposed to retain that provision.
Prior to setting forth the various proposed changes, Resolution 184—like the earlier Resolution 139 from 1966—provided that “the Constitution of the state be revised as follows: . . .” (Assem. Const. Amend. No. 30, Stats. 1968 (1968 Reg. Sess.) res. ch. 184, p. 3302, italics added.) Thereafter Resolution 184 set forth the Legislature’s proposed revision of 13 disparate articles (II, IV, IX, X, XI, XII, XIII, XIV, XVII, XVIII, XX, XXII, and XXIV) of the state Constitution.
As had occurred under similar circumstances in 1966, and consistent with the Legislature’s characterization of Resolution 184 as a proposed revision of the Constitution, the Attorney General’s ballot title for the measure commenced as follows: “Constitutional Revision. Legislative Constitutional amendment. . . .” (Ballot Pamp., Gen. Elec. (Nov. 5, 1968) analysis of Prop. 1, p. 1.) The Secretary of State designated this omnibus measure Proposition 1. (Ibid.)
Among the numerous changes to various articles contained within that extensive measure was a proposal that the first sentence of article XVIII be amended in various ways, including (1) allowing the Legislature, in proposing either an amendment or a revision of the Constitution, to “amend or withdraw” its proposal, and (2) providing that all future amendments to or revisions of the Constitution be effective on the day after adoption by the voters. But, as noted above, contrary to the recommendation of the Revision Commission, the Legislature’s proposed amendment retained, without substantive change, the separate-vote provision found in the second sentence of section 1. (Ballot Pamp., Gen. Elec. (Nov. 5, 1968) text of Prop. 1, p. 24 [setting forth the proposed language].) Proposition 1 also proposed related [756]changes to former article IV, section 24, subdivision (a) of the Constitution (present Cal. Const., art. II, § 10, subd. (a)), in order to make that provision (which concerns the effective date of approved initiatives and referenda) consistent with the newly proposed “effective date” rule for amendments to or revisions of the Constitution. (Ballot Pamp., Gen. Elec. (Nov. 5, 1968) text of Prop. 1, p. 1 [setting forth the proposed language].) Proposition 1 failed to gain a majority vote at the 1968 election.
After that defeat, the Legislature divided and repackaged those same proposed revisions into four different measures for submission to the voters on the June 1970 special election ballot. In each of the four resulting resolution chapters—resolution chapter 331 (Assem. Const. Amend. No. 29, Stats. 1969 (1969 Reg. Sess.) res. ch. 331, pp. 4003-4008 [repackaging proposed changes to articles II, XI, XIII, and XXII]), resolution chapter 264 (Assem. Const. Amend. No. 31, Stats. 1969 (1969 Reg. Sess.) res. ch. 264, pp. 3934-3937 [repackaging proposed changes to articles XII, XIII, XIV, and XX]), resolution chapter 263 (Assem. Const. Amend. No. 30, Stats. 1969 (1969 Reg. Sess.) res. ch. 263, pp. 3933-3934 [repackaging proposed changes to articles X, XVII, and XX]), and finally, resolution chapter 340 (Assem. Const. Amend. No. 28, Stats. 1969 (1969 Reg. Sess.) res. ch. 340, pp. 4016-4019 (hereafter referred to as Resolution 340) [repackaging proposed changes to articles IV, XVIII, and XXIV]), the Legislature—as it had done earlier with respect to Resolution 139 in 1966 and Resolution 184 in 1968—characterized the measure as proposing a revision (or partial revision) of the Constitution. (See Stats. 1969 (1969 Reg. Sess.) res. chs. 263, 340, pp. 3933, 3934, 4004 & 4016.)
Through Resolution 340, the Legislature again (as it had in 1968) proposed to amend article XVIII, section 1, by (1) allowing the Legislature, after proposing a constitutional amendment or revision, to “amend or withdraw” its proposal; (2) changing the same article by establishing the “effective date” of amendments or revisions as the day following the electors’ approval; and (3) making other substantive and nonsubstantive changes to the article—but again it did not propose any substantive change to the separate-vote provision. (Assem. Const. Amend. No. 28, Stats. 1969 (1969 Reg. Sess.) res. ch. 340, p. 4017.) The Legislature also proposed in the same measure (again, as it had in 1968) changes to article IV, section 24, subdivision (a) of the Constitution (present Cal. Const., art. II, § 10, subd. (a)), establishing the same effective date as that proposed with respect to constitutional amendments and revisions for all other ballot measures (that is, initiatives and referenda).20 [757]And finally, the Legislature proposed in the same measure various changes to article XXIV—a wholly unrelated and extensive provision concerning state civil service.21
The Legislature’s four repackaged resolutions became Propositions 2, 3, 4, and 5 on the June 1970 special election ballot. The Attorney General, consistent with the Legislature’s designation of each as a revision or partial revision, prepared a title that prominently so labeled each matter. For example, with respect to Resolution 340, which became Proposition 5, the Attorney General’s title in the ballot pamphlet read: “Partial Constitutional Revision: . . . Legislative Constitutional Amendment.” (Ballot Pamp., Special Elec, (consolidated with Primary) (June 2, 1970) analysis of Prop. 5, p. 13.) Proposition 2 received a majority vote by the electorate, but Propositions 3, 4, and 5 were rejected.
Undaunted by this second setback, the Legislature, which was then in the process of preparing to submit to the voters in November 1970 numerous additional phase three proposals from the Revision Commission (see generally Rev. Com., Proposed Revision 3 (Jan. 1970-Apr. 1971) pts. 1-6), responded by repackaging the failed Proposition 5 provisions and resubmitting them yet again. This time, however, the Legislature, perhaps wary of the electorate’s apparent reluctance to adopt multisubject revisions titled as such, further divided the previously rejected provisions (and most other phase three proposals as well) into separate measures that addressed discrete subjects (or as to which multiple amendments at least were germane to a common theme, purpose, or subject). The Legislature passed resolutions regarding these measures proposing “that the Constitution of the state be amended to read as follows: . . .” (italics added)—or essentially identical language to that effect. For example, failed Proposition 5 from the June 1970 ballot was further divided into two measures. In resolution chapter 147 (Assem. Const. Amend. No. 36, Stats. 1970 (1970 Reg. Sess.) res. ch. 147, pp. 3705-3707) (hereafter Resolution 147), the Legislature resolved to amend article XXIV’s civil service provisions—and that measure eventually was submitted to the voters as Proposition 14. Most relevant here, in resolution chapter 187 (Assem. Const. Amend. No. 67, Stats. 1970 (1970 Reg. Sess.) res. ch. 187, pp. 3780-3781) (hereafter Resolution 187), the Legislature resolved to amend both (1) article XVIII by, among other things, permitting the Legislature to “amend or withdraw” any proposed amendment or revision, establishing that such measures become effective on the day after adoption, and making no substantive change to the separate-vote provision, and (2) former article IV, section 24, subdivision (a) of the Constitution (present Cal. Const., art. II, [758]§ 10, subd. (a)), relating to the effective date of adopted initiatives and referenda—and that measure eventually was submitted to the voters as Proposition 16.22
Consistent with the Legislature’s various resolutions, the ensuing titles prepared by the Attorney General for these ballot propositions tracked the Legislature’s characterizations. For example, with regard to Resolution 147 (the civil service measure), which became Proposition 14, the Attorney General’s title in the ballot pamphlet read: “State Civil Service. Legislative Constitutional Amendment. . . .” (Ballot Pamp., Gen. Elec. (Nov. 3, 1970) analysis of Prop. 14, p. 23.) And, with regard to Resolution 187 (the measure here at issue), which became Proposition 16, the Attorney General’s title appearing in the ballot pamphlet read: “Constitutional Amendments. Legislative Constitutional Amendment. . . .” (Ballot Pamp., at p. 27.)23
Under the boldfaced subheading, Detailed Analysis by the Legislative Counsel, the voters were informed that Proposition 16 would, in addition to making some substantive changes, “restate” existing provisions of the two articles—“some with and some without substantive change.” (Ballot Pamp., Gen. Elec. (Nov. 3, 1970) analysis of Prop. 16, p. 27.) The voters next were provided with a description of the “major changes” (ibid.) that would be accomplished by Proposition 16’s proposed amendments to article XVIII24 and to former article IV, section 24, subdivision (a) of the Constitution [759](present Cal. Const., art. n, § 10, subd. (a)).25 Nowhere within this extensive recitation was there any reference to the restated separate-vote provision, much less any indication that the separate-vote provision in the interim since 1962 had taken on a new meaning under which it operated—no longer as a limitation upon the Legislature’s authority to package disparate amendments in a single measure, but the opposite—a protection of the Legislature’s asserted prerogative to package disparate proposed changes in a single measure as it wishes.
As noted earlier, upon the voters’ adoption of Proposition 16 at the November 1970 general election, article XVIII (including the restated separate-vote provision found in the second sentence of its first section) was amended to read as it does today: “Each amendment shall be so prepared and submitted that it can be voted on separately.”
In light of (and in some respects in spite of) the foregoing history, the Legislature asserts that the 1962 amendment to the first sentence of article XVIII, section 1, “[d]ispels any notion that [the second sentence of] section 1 [760]is intended to prohibit multi-pronged amendments such as [Resolution 103]; a revision or partial revision by its very nature may encompass, without constraint, any combination of disparate discrete changes.” The Legislature also argues that the subsequent “1970 amendment and restatement of section 1 and the events leading up to it unequivocally demonstrate that measures such as [Resolution 103] are entirely proper either as an ‘amendment’ or as a ‘partial revision’ of the Constitution”; that after the 1962 amendment, the “need to distinguish between ‘revision’ and ‘amendment’. . . disappeared for purposes of legislative proposals” (italics added); and that “the 1962 amendment to [the first sentence of] section 1 established beyond doubt that it is up to the Legislature to decide what proposed changes to put together in a single constitutional amendment” pursuant to the separate-vote provision found in the second sentence of section 1. In its reply brief, the Legislature further asserts that its designation of a measure (in a legislative resolution announcing a constitutional amendment proposed by the Assembly or by the Senate) as either a revision or an amendment is “no longer relevant,” and that requiring or expecting the Legislature to designate a measure “unnecessarily complicates the legislative process.”
The history recounted above does not support the Legislature’s position. The Legislature apparently fails to recognize the significance of the circumstance that the amendment made to article XVIII in 1962 was addressed exclusively to the first sentence of that article’s section 1, and that the amendment left untouched the second sentence of that section, which contains the separate-vote provision. In other words, although the 1962 amendment granted the Legislature the authority to propose either revisions or amendments (and, as the article subsequently was amended in 1970, also to amend or withdraw such proposals), it did not alter the rule that applies whenever the Legislature proceeds other than by way of a revision and instead proposes an amendment.26
The history described above demonstrates that the Legislature, after the 1962 amendment expanding its authority to propose constitutional changes amounting to revisions, decided—over the contrary recommendation of the Revision Commission—to retain the separate-vote provision. There is no evidence that the Legislature at that time or thereafter considered the separate-vote provision to have metamorphosed from what the Revision Commission in 1967 aptly characterized as a “restriction” and “limitation” upon the Legislature’s authority to package disparate amendments in a single measure, into the opposite—protection for an asserted legislative prerogative to package disparate proposed changes in a single measure as it wishes. And, most importantly, [761]there is no evidence indicating that the voters who, upon the Legislature’s third attempt in November 1970, finally adopted the Legislature’s proposed changes to article XVIII, had any reason to believe that, by doing so, they were giving the separate-vote provision a meaning opposite from what had been understood previously and that also would conflict dramatically with the meaning unanimously attributed to essentially identical provisions in scores of decisions rendered in sister-state jurisdictions.
In an effort to demonstrate that viewing the second sentence of article XVIII, section 1, as a limitation imposed upon the Legislature’s authority would conflict with past practice and assertedly call into question numerous past measures in which the electorate has adopted amendments to the Constitution, the Legislature highlights the failed Propositions from the June 1970 ballot and the successful Proposition 16 from the November 1970 ballot. It argues that “if the Legislature intended to so limit its power, how could it then simultaneously propose Propositions 2, 3, 4 and 5 on the June 1970 ballot or Proposition 16 on the November 1970 Ballot? Each measure breached any such bar. . . . The Legislature’s actions make sense only if the ‘separate vote’ sentence carried a different meaning. Only the Legislature’s interpretation of section 1 fits the contemporaneous history of the 1970 amendment.”
We disagree. Propositions 2, 3, 4, and 5 on the June 1970 ballot—each of which, at the recommendation of the Revision Commission, combined in a single measure numerous substantively disparate amendments to multiple unrelated articles—possibly might have “breached” the separate-vote provision had each proposition been submitted to the voters as a single packaged “amendment,” but they were not so presented. As noted above, each measure instead was proposed by the Legislature (upon recommendation of the Revision Commission) as a revision or partial revision, and each was so titled by the Attorney General. By its terms, the second sentence of article XVIII, section 1 (containing the separate-vote provision) simply does not apply to measures properly presented to the electorate as revisions. Nothing in the presentation of Propositions 2, 3, 4, and 5 to the electorate on the June 1970 ballot supports the Legislature’s current construction of the separate-vote provision.
For different reasons we reach the same conclusion with respect to Proposition 16 on the November 1970 ballot, which resulted in the present version of article XVIII, section 1. As noted above, that measure was characterized by the Legislature in Resolution 187 as an amendment; the Attorney General titled the resulting proposition as such, and the voters adopted that proposition as such without any reason to believe that in doing so they were endorsing a new or changed meaning for the separate-vote provision. Although Proposition 16, presented as an amendment and not as a partial [762]revision, was subject to the separate-vote provision—and although, as the Legislature suggests, that measure might have violated the separate-vote provision if judged under the strict test proposed by the Court of Appeal below—we shall, for reasons explained post, part II.B.3, reject the exacting test applied by the lower court. Under the test that we shall confirm in this case—essentially the same test employed by the vast majority of our sister-state jurisdictions for approximately 125 years—1970’s Proposition 16 is not called into question under the separate-vote provision of article XVIII, section l.27
In conclusion on this point, we find no support in the language or history of the separate-vote provision, or in any of numerous decisions of our sister states construing their own essentially identical provisions, that would lead us to adopt the Legislature’s construction of article XVIII, section 1, as protecting an asserted legislative right to combine into one package disparate and unrelated changes to the Constitution. Nor can we agree with the Legislature that, in light of its ability, since 1962, to propose revisions as well as amendments, the limitation upon legislatively proposed amendments imposed by the separate-vote provision no longer is relevant or effective. The provision’s words and history, its construction by the Revision Commission, and the uniform construction of such provisions in other jurisdictions instead reveals that it was intended to be, and remains, a limitation upon the power of the Legislature to submit constitutional amendments to the voters.28
We next consider the nature and scope of that limitation, and the test to be applied in discerning whether a violation of the separate-vote provision has occurred.
[763]B
Although we reject the Legislature’s interpretation of article XVIII, section 1, and find instead that the provision restricts legislative authority to submit disparate proposed constitutional changes in a single measure, as further explained we also disagree with the position taken by Californians for an Open Primary in its briefs, and the Court of Appeal below, that we should endorse a recent trend commenced by the Oregon Supreme Court in Armatta, supra, 959 P.2d 49, and construe our separate-vote provision as requiring a test different from and stricter than the traditional test employed under the related constitutional single subject rule. For the reasons that follow, we find no basis for that position in the words of our Constitution’s separate-vote provision or in the history of our charter. We also find no rationale for concluding—as we would have to, were we to construe the provision as proposed—that the Constitution should be interpreted in a manner that would impose a restraint upon the Legislature’s power to submit constitutional amendments to the voters greater than that imposed upon the people through the initiative process.
Instead, we shall adopt the approach that is, and has been, the majority rule in our sister state jurisdictions for approximately 125 years: the separate-vote provision should be construed consistently with its kindred provision, the single subject rule. We already have rejected, in part II. 1, the Legislature’s argument that the word “amendment” in article XVIII, section 1 (“[e]ach amendment shall be so prepared and submitted that it can be voted on separately”) refers to the legislative vehicle (the resolution proposing the constitutional amendment) by which the Legislature transmits a proposed amendment to the Secretary of State for eventual submission on the ballot. We shall explain below that the word “amendment” as used in the provision refers to a substantive change or group of substantive changes that are reasonably germane to a common theme, purpose, or subject. If (as in this case) the Legislature proposes to the electorate in such a resolution that the Constitution should be amended in a manner that presents in a single measure substantive changes that are not reasonably germane to a common theme, purpose, or subject, the presentation of such a single measure to the voters as an amendment will violate the separate-vote provision found in the second sentence of article XVIII, section 1.
1
The California Constitution, like that of most states, long has contained not only the separate-vote provision at issue in this case, but also a related “single subject” provision. Indeed, California has two single subject provisions: one, which has existed in the California Constitution since 1849, requires that [764]statutes “embrace but one subject” (Cal. Const., art. IV, § 9); the other, added to the Constitution in 1948, extends that rule to initiatives proposing either statutory or constitutional changes. (Cal. Const., art. II, § 8, subd. (d) [“An initiative measure embracing more than one subject may not be submitted to the electors or have any effect”].)
Although we have not previously construed our own separate-vote provision (except in the dictum of Wright, supra, 192 Cal. 704, 712-713, discussed ante, fn. 13), we long have construed our two single subject provisions in an accommodating and lenient manner so as not to unduly restrict the Legislature’s or the people’s right to package provisions in a single bill or initiative. (E.g., Fair Political Practices Com. v. Superior Court (1979) 25 Cal.3d 33, 39 [157 Cal.Rptr. 855, 599 P.2d 46] (Fair Political Practices Commission); Perry v. Jordan (1949) 34 Cal.2d 87, 92-93 [207 P.2d 47], and cases cited [construing identically the statutory and initiative versions of the constitutional single subject provisions].) We have found the single subject rules to have been satisfied so long as challenged provisions meet the test of being reasonably germane to a common theme, purpose, or subject. (Senate of the State of Cal. v. Jones (1999) 21 Cal.4th 1142, 1157 [90 Cal.Rptr.2d 810, 988 P.2d 1089] (Jones); Legislature v. Eu (1991) 54 Cal.3d 492, 512 [286 Cal.Rptr. 283, 816 P.2d 1309] (Eu); Brosnahan v. Brown (1982) 32 Cal.3d 236, 243-253 [186 Cal.Rptr. 30, 651 P.2d 274] (Brosnahan); Harbor v. Deukmejian (1987) 43 Cal.3d 1078, 1099 [240 Cal.Rptr. 569, 742 P.2d 1290]; Perry, supra, 34 Cal.2d at pp. 92-93.)29
[765]Most of our sister states have similarly and leniently construed their own single subject provisions.30 And, significantly, for more than a century a clear majority of the nearly 30 other jurisdictions that have a separate-vote provision similar to ours have similarly construed their separate-vote provisions, upholding amendments against challenges so long as the measure’s provisions are reasonably germane to a common theme, purpose, or subject.
Numerous out-of-state decisions long have observed that single subject provisions (sometimes called a single object provision) and separate-vote provisions share the same purpose of preventing voter confusion and “logrolling”—that is, the practice of combining in one measure two or more unrelated provisions, thereby forcing a single vote on matters that properly should be voted upon separately. For example, a leading early decision of the Missouri Supreme Court, Gabbert, supra, 70 S.W. 891, explained: “The convention which required each amendment to be separately submitted also ordained that no act of the legislature should contain more than one subject, and that subject should be clearly expressed in the title. The same common purpose actuated the convention in placing these two provisions in the Constitution. ‘It was intended to kill log-rolling, and prevent unscrupulous, designing men, and interested parties, from . . . comprising subjects diverse and antagonistic in their nature, in order to combine in its support members who were in favor of a particular measure." (Id., at p. 897, some italics added.)31 Similarly, Gabbert and numerous other decisions also have recognized, explicitly or implicitly, that those two rules, despite their different [766]phrasing, are kindred provisions that should be construed consistently—that is, as each having essentially the same substantive effect. (Id., at p. 897.)32
The early separate-vote provision decisions, many of which still are leading authorities today, adopted a lenient test focusing upon the germaneness of proposed changes relating to a common subject—and those decisions rejected tests that required a strict or close relationship between changes. For example, in the first 19th-century decision to address the issue, the Wisconsin Supreme Court in Timme, supra, 11 N.W. 785, held that an amendment changing sessions of the state legislature from annual to biennial and altering the terms of office and times of election was properly submitted in a single ballot measure in order to “accomplish a single purpose.” (Id., at p. 791.) The court rejected a rule that “every proposition in the shape of an amendment to the constitution, which standing alone changes or abolishes any of its present provisions, or adds any new provision thereto, shall be so drawn that it can be submitted separately, and must be so submitted.” (Id., at p. 790.) In an oft-quoted passage, the court in Timme noted the need to avoid a construction of the provision that would be so strict as to make it “practically impossible” or highly impracticable to amend the state constitution. (Ibid.)33 After other [767]early cases followed this same course,34 a contemporaneous commentator endorsed these lenient constructions of the separate-vote provision. (Dodd, Revision and Amendment of State Constitutions (Johns Hopkins 1910) p. 181 (Dodd).) Similarly, surveying existing case law in 1915, the editors of 6 Ruling Case Law (1915) observed that the test under the various separate-vote provisions is “whether particular amendments embrace more than one subject or whether the entire amendment is germane to a single general subject.” (Id., § 22, p. 30.)
Cases from other jurisdictions continued to follow that same course,35 leading another commentator to observe in the mid-1930’s that decisions construing the separate-vote provisions of the various state constitutions focus upon “whether particular amendments embrace more than one subject or whether the entire amendment is germane to a single general subject, and the rule has been laid down that a constitutional amendment embracing several subjects, all of which are germane to the general subject of the amendment, will, under [a separate-vote] requirement, be upheld as valid, and may be submitted to the people as a general proposition.” (Annot., Proposed Constitutional Amendment (1934) 94 A.L.R. 1510, 1511.)
This lenient construction of separate-vote provisions continued in the mid-through late 20th century. Surveying the state of the law in the late 1940’s, the Louisiana Supreme Court commented that “the courts [of various states] have accepted” the proposition that the separate-vote rule is honored so long as the “ ‘several subjects’ ” of a constitutional amendment are “ ‘germane to [768]the general subject of the amendment.’ ” (State v. City of Baton Rouge (1949) 215 La. 315 [40 So.2d 477, 480-481] [upholding constitutional amendment revising local governmental boundaries and redistributing governmental powers, submitted as a single measure].) Thereafter decisions rendered from 1960 through 1982 by the high courts of Minnesota, Idaho, Missouri, Ohio, New Mexico, Georgia, and Maryland continued in the same vein.36
2
As this lenient majority rule developed, a few decisions adopted a stricter test requiring an exacting inquiry into the relatedness of the various parts of a single measure. The earliest such case, Mississippi’s decision in Powell, supra, 27 So. 927, concerned a single measure that provided, among other things, for the popular election of judges of both the state supreme court and of the trial courts. The court in Powell adopted a very strict functional relationship [769]test that essentially required separate submission of any constitutional amendment provision that possibly could stand alone.37 Applying that test, the court found a violation of the state’s separate-vote provision, because the electorate might have voted in favor of popular elections for supreme court justices but not for trial court judges, and one provision could have been enacted without the other. (Id., at pp. 930-931.)
Soon thereafter the strict approach set forth in Powell was criticized as “too narrow” in the treatise mentioned earlier (Dodd, supra, at p. 181). A few years later the Mississippi high court reconsidered its position and overruled Powell in State ex rel. Collins v. Jones, supra, 64 So. 241—embracing instead the deferential and lenient approach articulated in the “general trend” of decisions (id., at p. 252), including those set forth in the cases described ante, part II.B.1.38
Despite Mississippi’s abandonment of an exacting functional relationship test in favor of the majority rule’s lenient test, a smattering of opinions issued in Idaho in 1909, Arizona in 1934, Utah in 1962, and Kansas in 1971—all finding violations of the separate-vote provision—continued to endorse a strict functional relationship test. (McBee, supra, 100 P. 97, 103-105; Kerby, supra, 36 P.2d 549, 554-555; Lee, supra, 367 P.2d 861, 864; Moore, supra, 486 P.2d 506, 520-521.) These cases remained an essentially dormant minority position until the Supreme Court of Oregon—without citing any of them or acknowledging the majority rule cases discussed above—revived this strict interpretation of the separate-vote provision in 1998.
In Armatta, supra, 959 P.2d 49, the Oregon high court considered a challenge under its constitution’s separate-vote provision39 to an extensive “crime victims’ rights” initiative constitutional amendment that had been adopted by the voters. After first finding that the Oregon separate-vote provision applies not only to constitutional amendments submitted by the state legislature, but also to such amendments submitted through the initiative [770]process (id., at pp. 53-55),40 the court proceeded to construe the separate-vote provision. The court rejected the state’s argument (which reflected the majority rule) that the separate-vote provision should be construed consistently with, and as having essentially the same effect as, the single subject rule. The court in Armatta explained that in its view a measure might satisfy the lenient requirements of the single subject rule, and yet still fail what the court discerned to be the considerably stricter requirements of the separate-vote provision. (Id., at p. 64.) The court announced: “[T]he proper inquiry is to determine whether, if adopted, the proposal would make two or more changes to the constitution that are substantive and that are not closely related.” {Ibid., italics added.) In other words, the court in Armatta embraced, for purposes of the separate-vote clause, a version of the minority rule’s exacting inquiry into the relatedness of provisions within a single measure.
Applying its test, the court in Armatta found the various provisions of the challenged measure (for example, those altering existing constitutional provisions concerning (1) searches and seizures, (2) unanimous verdicts in murder cases, (3) the right to bail, and (4) qualification of jurors in criminal cases) “not related closely enough” (Armatta, supra, 959 P.2d 49, 67, italics added). The court concluded that the measure violated the separate-vote provision and was invalid in its entirety. (Id., at p. 68.)
Based upon subsequent decisions applying Armatta’s rule, it is clear that the Oregon Supreme Court’s inquiry concerning whether challenged provisions are sufficiently closely related is indeed a demanding one. For example, in Lehman, supra, 37 P.3d 989, that same court in 2002 found that a 1992 initiative amendment imposing term limits violated the state’s separate-vote provision, because the measure presented questions to the voters in a single proposition concerning term limits for (1) federal elective office and (2) for state elective office. The court found the two provisions, although related, were not “ ‘closely’ related” to each other, and hence invalidated the state’s decade-old term limits provision for state offices. (Id., at p. 999.) Similarly, in Swett, supra, 43 P.3d 1094, the Oregon high court in 2002 invalidated under the separate-vote requirement a 1998 election reform initiative measure, presented to the voters in a single proposition, that contained provisions (1) requiring disclosure of certain political contributions of $500 or more, and (2) requiring that initiative signature-gatherers be registered Oregon voters. Finding no “close relation” between these two changes, the court declared the [771]entire measure invalid. (Id., at p. 1100.)41 The history of this case law and related matters has led one commentator to predict that under the strict Armatta test as applied in Oregon, most proposed constitutional amendments will fail. (Hoesly, Reforming Direct Democracy: Lessons From Oregon (2005) 93 Cal. L.Rev. 1191, 1224 (Reforming Direct Democracy); see also Bentley, Armatta v. Kitzhaber: A New Test Safeguarding the Oregon Constitution From Amendment by Initiative (1999) 78 Or. L.Rev. 1139, 1154—1156 [noting the difficulty of drafting a measure that meets the Armatta test].)42 Indeed, in the few other jurisdictions that recently have embraced (explicitly or implicitly) Armatta's interpretation of the separate-vote provision under their own constitutions, the results have, with one exception, been consistent with the recent Oregon experience.43
[7723]
The Court of Appeal below concluded that “the formulation set forth in Armatta . . . comports with the constitutional text, framework, historical development, and purpose of the separate vote requirement in [article XVIII,] section 1.” Although in its briefing before this court petitioner Californians for an Open Primary supported and defended the Court of Appeal’s adoption of the test set forth in Armatta, supra, 959 P.2d 49, at oral argument counsel for petitioners retreated and professed being “agnostic” regarding the proper [773]test to be applied under the separate-vote provision. For reasons set forth below, we disagree with the Court of Appeal below, and conclude that although Armatta’s approach and its exacting test may find support in the language, history, constitutional framework, and case law of Oregon and some other jurisdictions, the same cannot be said under the California Constitution’s language, history, constitutional framework, or case law. Instead, we shall conclude that in California—as in the vast majority of states—the separate-vote provision always has had, and retains, essentially the same effect as the single subject rule, and does not impose a stricter standard requiring a showing of close relatedness (or functional relatedness) of the challenged provisions presented in a single measure.
The court in Armatta, supra, 959 P.2d 49, asserted that because the Oregon separate-vote-rule provision is worded differently from that state’s single subject provision and because the two provisions are located in different parts of the state Constitution, the court was required to “assume that they have different meanings . . . .” (Id., at p. 56.) Likewise, in the California Constitution, the two provisions (the separate-vote provision and the two versions of the single subject rule) are worded differently from each other and are located in separate parts of the Constitution. (Compare Cal. Const., art. XVIII, § 1 [the separate-vote provision] with id., art. n, § 8, subd. (d), and id., art. IV, § 9 [the initiative and legislative single subject provisions].) Although we generally would agree that differently worded phrases often carry a different meaning, we note that historically, most jurisdictions that have both a single subject provision and a separate-vote provision have not construed those differently worded provisions differently from each other— and this has held true also in jurisdictions in which the two provisions are set forth together in the same subdivision44 or even in the same sentence.45 In any event, for a number of reasons described below, we agree with the majority rule and conclude that these provisions of the California Constitution should be construed in essentially the same manner, finding that most of the reasons that led the Oregon high court to conclude otherwise in Armatta (and that have led some other courts to follow Oregon in this regard) do not apply here.
The state constitutional framework that confronted the court in Armatta is substantially different from ours in a significant way: Whereas under the California Constitution, the separate-vote provision regulates only proposed constitutional amendments submitted by the Legislature (and not those [774]proposed by the electorate through the initiative process), in Oregon (and Arizona and Montana as well—see ante, fn. 40), the separate-vote provision applies to all constitutional amendments, whether submitted by the state legislature or the electors. Based upon this particular feature of the Oregon state Constitution, the court in Armatta drew a distinction between enactment or adoption of mere legislation—to which the state’s lenient single subject provision applies—and constitutional amendments, to which the separate-vote provision applies. The court in Armatta reasoned that imposing a stricter requirement for amending the state charter “makes sense, because the act of amending the constitution is significantly different from enacting or amending legislationand it suggested that the test under the separate-vote provision should be more exacting than under the single subject test in order to “safeguard” the integrity of the “fundamental law”—that is, the state constitution. (Armatta, supra, 959 P.2d 49, 63.)
This reasoning is inapplicable in California, because, as noted above, the California separate-vote provision applies only to constitutional amendments submitted by the Legislature; the people remain free to submit amendments to the Constitution unrestrained by the separate-vote provision and are limited only by the single subject provision. (Wright, supra, 192 Cal. 704, 711-712.) Accordingly, unlike the Oregon (and Montana and Arizona) courts, we cannot draw a distinction between mere legislation and constitutional amendments or conclude that whereas the former is subject to a lenient single subject test, the latter must be subjected to a strict separate-vote inquiry.
Indeed, our case law long ago rejected a corollary of the notion that underlies the Armatta analysis, and as explained below, our past action in doing so makes it especially inappropriate to follow the Armatta approach at the present time. Justice Manuel’s dissenting opinion in Schmitz v. Younger (1978) 21 Cal.3d 90 [145 Cal.Rptr. 517, 577 P.2d 652] argued that, although our single subject provision for statutes (Cal. Const., art. IV, § 9) has been and should be leniently construed, “the special nature of the initiative process requires a narrower [that is, more exacting] construction” under the single subject provision of California Constitution, article II, section 8, subdivision (d), because, for various reasons, “the dangers presented by a multisubject proposal are much more limited in the legislative context than in the initiative context.” (Schmitz, supra, 21 Cal.3d at p. 99, fn. omitted.) Specifically, Justice Manuel proposed that when reviewing initiative measures we should impose a “functional relationship” inquiry—a test substantially similar to the “closely related” test of Armatta, supra, 959 P.2d 49. (Schmitz, supra, 21 Cal.3d at pp. 97-100.) A majority of our court in Brosnahan, supra, 32 Cal.3d 236, 248-249, however, rejected Justice Manuel’s call for a heightened single subject inquiry with respect to initiative measures.
[775]Having declined in Brosnahan to construe the state Constitution as imposing a heightened burden on the people’s right under article II, section 8, subdivision (d) to propose legislative or constitutional change by initiative, compared with the single subject requirement under article IV, section 9, for regular legislation, it would be anomalous indeed were we now to construe the Constitution’s separate-vote provision as imposing upon the Legislature the essentially same heightened burden of establishing functional relatedness that we earlier declined to impose upon the electorate. And yet that is what would happen were we to agree with the Court of Appeal below and adopt the Armatta test under the separate-vote provision of article XVIII, section 1. We would produce a scheme under which the Legislature’s authority to submit proposals for amending the state Constitution would be significantly more circumscribed than the electors’ authority to do the same through the initiative process. The electors’ right to amend the Constitution would be controlled only by the lenient single subject rule (under which proposals for amendment may be combined so long as they are reasonably germane to a common theme, purpose, or subject), while the Legislature’s authority to submit a single measure proposing amendment of the Constitution would be subject to strict review under the separate-vote provision (under which proposals for amendment may be combined only if they are, among other things, “closely” or “functionally” related to each other).
No jurisdiction of which we are aware that allows amendment of its state constitution either by legislative submission or voter initiative discriminates in this manner by setting up a higher obstacle for legislative constitutional submissions than for initiative submissions by the voters, and we cannot imagine that doing so was the intent either of the various drafters over the years or of the voters who enacted, reenacted, and amended California’s separate-vote provision. By contrast, if we follow the majority rule of our sister states and construe our separate-vote provision as requiring no more (and no less) than the constraint effectuated by the single subject rule, we shall avoid creating such an unprecedented and anomalous scheme.
Nor do we find that the history of California’s separate-vote provision supports a strict test such as the one adopted by Armatta, supra, 959 P.2d 49, or endorsed by the Court of Appeal below. In Cambria, supra, 776 A.2d 754—one of the recent decisions that has followed Armatta—the New Jersey Supreme Court discerned in that state’s 1844 constitutional debates an intent not only that its separate-vote provision would “encompass” the concept of a single subject rule (id., at p. 761), but also an intent to “requir[e] closer examination of the relationship between the parts of a proposed constitutional amendment than does the single [subject] test.” (Id., at p. 765.) By contrast, we are unaware of any evidence in the relevant debates from the 1878-1879 California constitutional convention, or in any of the provision’s subsequent [776]history, suggesting that the separate-vote provision was intended to effectuate a limitation more exacting than that provided by the state’s single subject rule.46
Nor, unlike the decisions of some of the jurisdictions that recently have endorsed a strict construction of other separate-vote provisions, does California case law support such an interpretation. For example, the Idaho Supreme Court decision in IWP, supra, 982 P.2d 358, 362-363, found support for its strict reading of that state’s separate-vote provision in its 1909 decision in McBee, supra, 100 P. 97. Likewise, the Arizona Supreme Court’s decision in Clean Elections Institute, supra, 99 P.3d 570, 573-577, found support for a strict interpretation of its separate-vote provision in its 1934 decision in Kerby, supra, 36 P.2d 549. In neither of these minority-rule jurisdictions did the recent determination to embrace a strict test reflect a break with prior case law.47 By contrast, we do not find in the sparse California case law concerning article XVIII, section 1, any indication of support for a strict test such as the one adopted by Armatta, supra, 959 P.2d 49, or endorsed by the Court of Appeal below. Instead, we find indications of the opposite. As noted [777]ante, the only case in which this court addressed the provision was our 1923 decision in Wright, supra, 192 Cal. 704—and in that matter we suggested, in dictum, that the provision should not be construed strictly, but instead leniently, to allow changes to several parts of the Constitution if the changes are reasonably germane to each other. (Id., at pp. 712-713.)
We conclude that although the strict construction of the separate-vote provision recently embraced by Oregon and some other jurisdictions may find support in the constitutional language, history, constitutional framework, and case law of those other minority-rule states, the same cannot be said with regard to article XVIII, section 1 of the California Constitution. We hold that in California, as in the vast majority of states, the separate-vote provision has essentially the same effect as the single subject rule and requires only a showing that the challenged provisions are reasonably germane to a common theme, purpose, or subject. The separate-vote provision does not impose a stricter standard requiring a showing of “close” or “functional” relatedness.
Against our determination that the separate-vote provision in essence incorporates the requirements of the single subject rule, the Legislature insists that the delegates to the 1878-1879 constitutional convention affirmatively disclosed a contrary intention—namely, that the separate-vote provision not encompass a limitation tantamount to the single subject provision, even in the latter provision’s lenient, rather than strict, sense. In support, the Legislature cites the 1878-1879 debates concerning the statutory single subject rule—a provision that already existed in article IV, section 25 of the 1849 Constitution48 and that the convention delegates eventually reaffirmed by adopting former article IV, section 24 (current art. IV, § 9))49 (See 2 Willis & Stockton, supra, at pp. 796-799 [debating the statutory single subject rule]; 3 Willis & Stockton, supra, at pp. 1269-1270 [same].) The Legislature characterizes those debates as disclosing that the delegates “vehemently expressed concerns over impediments to legislation, encouraging litigation and judicial second-guessing of legislative actions on the one hand versus [the delegates’ concerns relating to] deceit, confusion and logrolling on the other.” The Legislature contrasts these debates concerning the single subject provision with the “complete absence of similar controversy” concerning the same [778]delegates’ consideration of article XVIII, section 1, and its separate-vote provision; the Legislature concludes that this comparative silence demonstrates the delegates must have intended “no sort of single subject requirement or limitation on the Legislature’s discretion to construct proposed constitutional amendments as it wished . . .
We do not agree. First, the Legislature’s brief fails to characterize properly the substance of the 1878-1879 debate concerning the single subject provision. The cited passages disclose that the delegates did not, as a general matter, question the efficacy or propriety of a rule confining legislation to a single subject or to related subjects. Instead, the delegates appear to have accepted and generally endorsed retention of that rule from the Constitution of 1849.50 The vigorous debate to which the Legislature alludes concerned not the efficacy or propriety of the single subject rule per se, but instead a collateral issue: the severability of the parts of an act that violate the legislative single subject rule. In this regard, Delegate Freeman offered an amendment to the proposed section, adding a clause as follows: “ ‘[B]ut if any subject shall be embraced in any Act which shall not be expressed in its title, said Act shall be void only to so much thereof as shall not be so expressed ....’” (2 Willis & Stockton, supra, at p. 796.) The ensuing debate focused upon the propriety of this provision, and upon the general problem of requiring subjects of an act to be expressed in the title of the act.51 The delegates tentatively adopted a version of the proposed severability clause (id., at pp. 798-799) and then, a few weeks later, rejected repeated attempts to strike the severability clause in favor of making an entire act void if any part of it violated the single subject rule. (3 Willis & Stockton, supra, at pp. 1269-1270.) Ultimately, the provision was adopted as proposed by Delegate Freeman. (See ante, fn. 49.) Contrary to the Legislature’s suggestions, this predominating focus of the debate concerning the legislative single subject provision provides no reason to question the delegates’ understanding of article XVIII, section l’s separate-vote provision—a provision that did not contain then, as it fails to contain today, any such title requirement or severability clause.
[779]In any event, we reject the Legislature’s premise that the absence of debate concerning the scope of the limitation imposed by the separate-vote provision is revealing in this context. (See Landau, supra, 38 Val.U. L.Rev. 451, 474 [cautioning against “minting] . . . silence for historical significance by means of negative inference”].) The Legislature reasons that, had it been understood by the delegates that the separate-vote provision would impose a substantive limitation similar to that under the single subject provision, those delegates who may have opposed the legislative single subject provision also would have opposed a constitutional amendment including such a separate-vote provision. As noted above, however, the Legislature fails to demonstrate that any delegate (let alone a majority of them) objected on the merits to the continuation of the 1849 Constitution’s single subject limitation upon regular legislation. But even if some delegates had so objected, it would not follow that they also would have opposed a similar limitation by means of the separate-vote provision; instead, those delegates might well have considered it a wholly different and unobjectionable matter to impose such a similar limitation upon the considerably more momentous act of amending the state’s fundamental charter. As the Supreme Court of New Jersey observed in Cambria, supra, 776 A.2d 754: “Amendments to an organic body of law are a serious matter. The [state constitution] contains our most fundamental ideas about the type of government we want to have and how it should function, and the relationship between that government and the people. . . . The [ilegislative constitutional amendment] process is appropriately more complex than simple lawmaking . ” (id., at p. 764, italics added)—and, we would add, the substantive rules governing the scope of legislative constitutional amendments certainly should be no less demanding than the substantive limitations governing mere legislative acts.
Ill
The Legislature does not argue that Resolution 103’s two provisions—the primary elections provision (amending California Constitution, article II by adding a new section 5, subdivision (b)), and the state property/bonds repayment provision (amending California Constitution, article m by adding a new section 9)—satisfy the traditional single subject provision test that, as we confirm today, also governs the separate-vote provision of article XVIII, section 1. Nor could the Legislature so contend; patently, the two provisions are not reasonably germane to a common theme, purpose, or subject. Accordingly, we agree with the Court of Appeal’s conclusion (although for reasons different from those relied upon by that court) that the Legislature’s proposed submission to the electorate of both provisions in a single measure as proposed in Resolution 103 violated the separate-vote provision of article XVIII, section 1.
[780]IV
As demonstrated by the out-of-state cases discussed above, the normal remedy for violation of the separate-vote provision has been either (1) a preelection order barring submission of the measure to the voters in a single package, or (2) postelection invalidation of a measure that improperly was submitted to the voters in a single package. In the present case, as noted earlier, the Court of Appeal devised an alternative remedy: it rejected, by a two-to-one vote, the assertion of Californians for an Open Primary that Proposition 60 should be stricken from the ballot, and instead issued a peremptory writ of mandate directing the Secretary of State to bifurcate the two provisions and submit them to the voters separately. Also, as noted above, after we granted review in this matter, and in light of the then-impending election and ballot preparation deadlines, we rejected the request of Californians for an Open Primary for a stay, instead ordering the Secretary of State to place Resolution 103 on the November 2004 ballot “in the manner directed by the Court of Appeal” — that is, as Propositions 60 (the primary elections provision) and 60A (the state property/bonds repayment provision). Thereafter the voters at the November 2004 election, while rejecting Proposition 62, enacted both Propositions 60 and 60A.
Californians for an Open Primary observes that the resulting bifurcated provisions that were placed on the ballot proposed two constitutional amendments, neither of which, standing alone, had received the approval of two-thirds of each house of the Legislature as required by the first sentence of the first section of article XVIII. Californians for an Open Primary argues: “[Resolution 103] was the product of coalition-building in the Legislature. . . . [I]t is not at all clear that the requisite two-thirds majority of each house would have supported the separate submission to the voters of Propositions 60 and 60A. Most importantly, nobody in the Legislature ever proposed that Proposition 60 or 60A be submitted to the voters separately for their approval.” Relying upon our observation in Livermore, supra, 102 Cal. 113, 117-118, that “the power of the legislature to initiate any change in the existing organic law ... is to be strictly construed under the limitations by which it has been conferred” and “cannot be . . . enlarged beyond these terms,” Californians for an Open Primary argues that because the two amendments were not separately approved by the Legislature, they must be invalidated now, even though the voters enacted each separately in November 2004. In other words, whereas in its petition for a writ of mandate filed prior to the November 2004 election, Californians for an Open Primary originally sought to have the combined measure withheld from the ballot, it now argues that bifurcation was improper and that the appropriate postelection remedy is to invalidate both Propositions 60 and 60A.
[781]The Legislature, in its opening brief filed prior to the November 2004 election, also questions the propriety of the Court of Appeal’s bifurcation remedy, arguing that even though it had “acquiesced in this remedy in order to permit deliberative review in this court,” as a general matter, bifurcation is “highly problematic.” The Legislature asserted: “Even when done by court order, such dissection of a measure adopted by the Legislature is inappropriate absent some express statement by the Legislature that it would prefer this alternative to barring the measure from the ballot altogether.” Subsequently, after the defeat of Proposition 62 and the adoption of both Propositions 60 and 60A, the Legislature modified its position, arguing in its reply brief that assuming Resolution 103 violated the separate-vote provision, “the goal of article XVIII, section 1 was met when the ‘amendments’ were separately voted upon by the people, which happened when the voters overwhelmingly approved each measure.” Accordingly, the Legislature asserts, “[a]though court-mandated bifurcation of a legislative proposal is generally not favored, this remedy was appropriate in the unique circumstances presented here.”
We conclude that the Court of Appeal erred in bifurcating the two measures. Nothing in the language or history of article XVIII generally, or of the separate-vote provision in particular, suggests that a violation of the provision should be remedied by bifurcation of proposed amendments and the presentation of those matters to the electorate in separate measures. Nor do we discern in our case law, or in that of any other jurisdiction, any suggestion that bifurcation is an appropriate remedy in such a circumstance. Finally, we find it instructive that the analogous initiative single subject provision (Cal. Const., art. II, § 8, subd. (d)) precludes the related remedy of severance.52 (See Jones, supra, 21 Cal.4th at p. 1168 [“when an initiative measure violates the single-subject rule, severance is not an available remedy”]; see also California Trial Lawyers Assn. v. Eu (1988) 200 Cal.App.3d 351, 361-362 [245 Cal.Rptr. 916] [concluding the same].)
Indeed, allowing bifurcation of a measure that violates the separate-vote provision would permit—if not encourage—logrolling-type manipulations that in turn would frustrate one purpose of the separate-vote provision. If, for example, it were known in advance that bifurcation was a potential and permissible remedy, factions within the Legislature, none of which on its own could gamer a two-thirds vote for a particular amendment, might join forces by agreeing to present disparate proposed amendments in a single measure, knowing that a court likely would find a separate-vote violation but thereafter could order the provisions bifurcated and presented separately to [782]the electorate as discrete amendments. In this manner, legislators constituting less than two-thirds of each house could place such measures before the voters in violation of the rule set forth in the first sentence of article XVIII, section l.53 Our conclusion that bifurcation is not a remedy for violation of the separate-vote provision avoids creating such incentives or facilitating such manipulations.
V
Although we conclude that the Court of Appeal erred in directing the Secretary of State to bifurcate the two measures and place them on the ballot, we conclude that under the unusual circumstances of this case, it would be inappropriate to invalidate the two approved measures, each of which, as noted, subsequently was separately approved by the voters after this court, in the face of the then impending election, declined to stay the Court of Appeal’s bifurcation order. (Cf. Assembly v. Deukmejian (1982) 30 Cal.3d 638, 652, 669 [180 Cal.Rptr. 297, 639 P.2d 939] [refusing, under the “unusual and unique circumstances” there presented, to invalidate redistricting referendum petitions that clearly violated the Elections Code, and ordering the use of redistricting plans that had been stayed by the referendum, because doing so “minimi ze[d] the potential disruption of the electoral process” (italics omitted)].)
The potential for manipulation of the process that we described at the close of part IV, ante, manifestly did not occur here. There is no basis upon which to conclude that any legislator could have, or did, anticipate the Court of Appeal’s adoption of its novel bifurcation remedy. Indeed, as the briefing in this matter discloses, the Legislature was surprised by, and strenuously objected to, the solution imposed by that court.
In light of the absence of any prior definitive California ruling with regard to either the scope of the separate-vote provision or the remedy for its violation, and in light of the circumstance that the two proposed amendments ultimately were separately submitted to the voters and separately adopted as Propositions 60 and 60A after the electorate was afforded an opportunity to consider the arguments for and against each measure, we do not invalidate those constitutional amendments.
[783]VI
For the reasons discussed above, the judgment of the Court of Appeal is vacated and the matter is remanded to the Court of Appeal with directions to discharge the alternative writ and to deny the request that a peremptory writ issue to invalidate Propositions 60 and 60A. Each party shall bear its own costs in this proceeding.
Kennard, J., Baxter, J., Chin, J., and Corrigan, J., concurred.
Specifically, the measure proposed to allow the electorate to vote for candidates for state and federal elected offices (except for President and Vice-President) on a primary election ballot regardless of the party registration of the candidates or voters. Pursuant to the measure, the two candidates receiving the greatest number of votes, regardless of their party, then would be listed on the general election ballot. (Ballot Pamp., Gen. Elec. (Nov. 2, 2004) text of Prop. 62, pp. 83-102.)
Resolution 103 stated: “A political party that participated in a primary election for a partisan office has the right to participate in the general election for that office and shall not be denied the ability to place on the general election ballot the candidate who received, at the primary election, the highest vote among that party’s candidates.” (Sen. Const. Amend. No. 18, Stats. 2004 (2003-2004 Reg. Sess.) res. ch. 103, pt. 1st.)
Resolution 103 stated: “The proceeds from the sale of surplus state property occurring on or after the effective date of this section, and any proceeds from the previous sale of surplus state property that have not been expended or encumbered as of that date, shall be used to pay the principal and interest on bonds issued pursuant to the Economic Recovery Bond Act authorized at the March-2, 2004, statewide primary election. Once the principal and interest on those bonds are fully paid, the proceeds from the sale of surplus state property shall be deposited into the Special Fund for Economic Uncertainties, or any successor fund. For [740]purposes of this section, surplus state property does not include property purchased with revenues described in Article XIX or any other special fund moneys.” (Sen. Const. Amend. No. 18, Stats. 2004 (2003-2004 Reg. Sess.) res. ch. 103, pt. 2d.) Californians for an Open Primary asserts in its answer brief: “By accelerating the bond repayment, this second provision made [Resolution 103] more attractive to voters by making it appear to save them millions of dollars.”
We issued the following orders: “Petitions for review GRANTED, [ft] At this preliminary point in the proceedings before this court, we neither endorse nor reject the Court of Appeal’s view of the proper interpretation of the ‘separate vote’ provision of article XVm, section 1, of the California Constitution, or that court’s determination regarding the appropriate remedy for a violation of that constitutional provision. Nonetheless, in light of (1) the novelty and difficulty of the constitutional issues presented and the importance of ensuring that our decision is rendered after full and adequate briefing, oral argument and deliberation, (2) the imminence of the deadline for submitting the Voter Information Guide for the November 2004 election to the State Printer, and (3) the concerns expressed by the Secretary of State in a letter to this court filed on his behalf by the Attorney General on August 6, 2004, we direct the Secretary of State to place Senate Constitutional Amendment No. 18 of the 2003-04 Regular Session [that is, Resolution 103] on the ballot for the November 2004 election as Propositions 60 and 60A, in the manner directed by the Court of Appeal. Accordingly, the request for a stay of the Court of Appeal’s placement order, which request was included in the petition for review of Californians for an Open Primary and Nick Tobey, filed August 3, 2004, is denied.” This portion of the order was signed by Chief Justice George and Justices Kennard, Baxter, Chin, and Moreno.
The second portion of the order, signed by two justices, read as follows: “Like the majority, we neither endorse nor reject the Court of Appeal’s view of the proper interpretation of the ‘separate vote’ provision of article XVIII, section 1, of the California Constitution, or that court’s determination regarding the appropriate remedy for a violation of that constitutional provision. We also concur in the majority’s decision to grant the petitions for review. We would not, however, direct the Secretary of State to place Senate Constitutional Amendment No. 18 of the 2003-04 Regular Session [that is, Resolution 103] on the ballot for the November 2004 election as Propositions 60 and 60A, in the manner directed by the Court of Appeal.” This portion of the order was signed by Justices Werdegar and Brown.
California Secretary of State, Statement of the Vote, November 2, 2004, General Presidential Election, Statewide Measures, available at <http://www.ss.ca.gov/elections/sov/ 2004_general/formatted_ballot_measures_detail.pdf> (as of May 25, 2006).
The next three sections of article XVIII address related matters. Section 2 concerns revision by a Constitutional Convention. It provides: “The Legislature by rollcall vote entered in the journal, two-thirds of the membership of each house concurring, may submit at a general election the question whether to call a convention to revise the Constitution. If the majority vote yes on that question, within 6 months the Legislature shall provide for the convention. Delegates to a constitutional convention shall be voters elected from districts as nearly equal in population as may be practicable.” (Cal. Const., art. XVIII, § 2.) Section 3 of article XVIII of the Constitution simply confirms what already is stated in the state Constitution, article II, section 8, subdivision (b), by providing: “The electors may amend the Constitution by initiative.” Section 4 specifies, among other things, the effective date of amendments or revisions. It provides: “A proposed amendment or revision shall be submitted to the electors and if approved by a majority of votes thereon takes effect the day after the election unless the measure provides otherwise. If provisions of 2 or more measures approved at the same election conflict, those of the measure receiving the highest affirmative vote shall prevail.” (Id,., art. XVIII, § 4.)
In another respect we find the language used in the provision to be at odds with the Legislature’s interpretation. The provision’s use of the word “submitted” (“[e]ach amendment shall be so prepared and submitted that it can be voted on separately” (art. XVIII, § 1, italics added)) seems most naturally to refer to the Legislature, and not to an entity such as the Secretary of State. Even though the Legislature does not itself physically “submit” anything to the voters, section 2 of the same article states that “the Legislature . . . may submit” (id., § 2) to the voters a call for a constitutional convention—thereby demonstrating that, insofar as this article of the Constitution is concerned, the word “submit” must be understood to cover the general process of placing a legislative proposal before the voters.
Justice Landau’s article criticizes or sets forth cautions regarding the following practices, among others: (1) reliance upon the drafters’ “original intent” in order to determine what a given provision means today; (2) the problem of generalization—that is, reasoning from what is known of the drafters’ intent, to resolve a matter beyond that fairly contemplated by the drafters; (3) selective reliance upon history to explain or justify a conclusion; (4) inferring drafters’ intent from the absence of debate or discussion; (5) inferring meaning of a clause [745]borrowed from another jurisdiction from the other jurisdiction’s subsequent interpretation of that clause; and (6) failing to recognize that the understanding and perception of history is in part a function of the availability of source materials.
Section 1 of the former article provided in pertinent part: “Any amendment or amendments to this Constitution may be proposed in the Senate or Assembly; and if the same shall be agreed to by a majority of the members elected to each of the two Houses, such proposed amendment or amendments shall be entered on their journals, with the yeas and nays taken thereon, and referred to the Legislature then next to be chosen, and shall be published for three months next preceding the time of making such choice. And if in the Legislature next chosen as aforesaid, such proposed amendment or amendments shall be agreed to by a majority of all the members elected to each House, then it shall be the duty of the Legislature to submit such proposed amendment or amendments to the people, in such manner and at such time as the Legislature shall prescribe; and if the people shall approve and ratify such amendment or amendments, by a majority of the electors qualified to vote for members of the Legislature voting thereon, such amendment or amendments shall become part of the Constitution.” (Cal. Const. of 1849, art. X, § 1.)
The various other state constitutions in existence in 1878 to 1879 typically provided for amendments proposed by the legislature, but, like the original California Constitution of 1849 (art. X, § 1), did not impose a separate-vote requirement with regard to such amendments. (See Thorpe, supra, at, e.g., pp. 547 [Conn. Const. of 1828, art. XI], 1969 [Mich. Const. of 1850, art. XX, § 1, as amended eff. 1876], 2423 [Nev. Const. of 1864, art. XVI, § 1], 2672 [N.Y. Const. of 1846, art. XIII, § 1], 3663 [Tex. Const. of 1876, art. XVII, § 1], 3897-3898 [Va. Const. of 1870, art. XII].)
On October 11, 1878, Delegate Martin offered a provision permitting the Legislature, on a majority vote of both houses, to propose constitutional “alterations or amendments” that had been “published with the laws which have been passed at the same session” and further provided that “said amendments shall be submitted to the people for their approval or rejection at the next general election .... If two or more alterations or amendments shall be submitted at the same time it shall be so regulated that the electors shall vote for or against each separately.” (1 Willis & Stockton, Debates and Proceedings, Cal. Const. Convention, 1878-1879, p. 117 (Willis & Stockton).) Four days later, on October 15, 1878, Delegate White offered a competing proposal aimed at requiring a two-thirds vote of both legislative houses, but otherwise retaining the separate-vote language quoted above. (Id.., at pp. 150-151.) Yet four days later, on October 19, 1878, Delegate Blackmer offered a third proposal under which the then-existing provision (Cal. Const. of 1849, art. X, § 1) would be amended to strike the Legislature’s authority to submit amendments “ ‘in such manner’ ” as the Legislature shall prescribe, and to substitute in its place the following: “ ‘and such amendment or amendments shall be so prepared and distinguished, by numbers or otherwise, that they be voted on separately, unless it be found impracticable to so distinguish them.’ ” (1 Willis & Stockton, supra, at p. 165.)
In addition, in view of the introductory phrase included in article XVIII as adopted in 1879 (“it shall be the duty of the Legislature to submit”)—a phrase that appeared in some form in almost every other constitution of the day—it seems highly improbable that the drafters viewed the following sentence, which contained the separate-vote provision, as imposing a limitation on any entity other than the Legislature.
Thereafter, the court commented in dictum that even if the prior constitutional amendment had been proposed to the electorate not as an initiative measure but instead by the Legislature, the prior amendment may have satisfied the separate-vote requirement: “[W]e are not at all satisfied that under a liberal interpretation of section 1 of article XVEI of the constitution the legislature might not have itself proposed and presented to the people for their adoption in the form of a single amendment to the constitution the precise measure embodied in this initiative amendment, even though its effect might be to change or abrogate other portions of the constitution not embodied in the particular article thereof thus sought to be amended.” (Wright, supra, 192 Cal. 704, 712-713.)
The Legislature argues that because the early provisions of New Jersey, quoted ante, at page 745, and of Wisconsin, prefaced their separate-vote provisions with the word “provided,” those clauses are substantively different from ours, which does not include an explicit proviso. We disagree, noting that, in any event, the current clauses of most state constitutions (including present N. J. Const., art. IX, par. 5, as construed in Cambria, supra, 776 A.2d 754) are not so prefaced, and furthermore that no court construing such a clause has drawn the distinction advanced by the Legislature in the case before us.
See State v. Powell (1900) 77 Miss. 543 [27 So. 927, 931-932] (Powell); McBee v. Brady (1909) 15 Idaho 761 [100 P. 97, 103-105] (McBee); Kerby, supra, 36 P.2d 549, 554-555 (Ariz.); State ex rel. Thompson v. Zimmerman (1953) 264 Wis. 644 [60 N.W.2d 416, 420-421]; Lee v. State (1962) 13 Utah2d 15 [367 P.2d 861, 864] (Lee); Moore, supra, 486 P.2d 506, 520-521 (Kan.); Armatta, supra, 959 P.2d 49, 71 (Or.); Marshall v. State (1999) 1999 MT 33 [293 Mont. 274, 975 P.2d 325, 328-333] (Marshall); Bergdoll v. Kane (1999) 557 Pa. 72 [731 A.2d 1261, 1269-1270] (Bergdoll); Dale v. Keisling (2000) 167 Or.App. 394 [999 P.2d 1229,1232-1235]; Sager v. Keisling (2000) 167 Or.App. 405 [999 P.2d 1235, 1238-1239]; IWP, supra, 982 P.2d 358, 363 (Idaho); Pennsylvania Prison Soc. v. Commonwealth (2001) 565 Pa. 526 [776 A.2d 971, 978-982] (lead opn. by Zappala, J.) (Pennsylvania Prison Soc.); Lehman v. Bradbury (2002) 333 Or. 231 [37 P.3d 989, 994-1001] (Lehman); Swett v. Bradbury (2002) 333 Or. 597 [43 P.3d 1094, 1099-1101] (Swett); League of Oregon Cities v. State of Oregon (2002) 334 Or. 645 [56 P.3d 892, 904—911] (League of Oregon Cities); Lincoln Interagency Narcotics Team v. Kitzhaber (2003) 188 Or.App. 526 [72 P.3d 967, 971-983], rev. allowed, 84 P.3d 1080; Clean Elections Institute, Inc. v. Brewer (2004) 209 Ariz. 241 [99 P.3d 570, 573-577] (Clean Elections Institute); Meyer v. Bradbury (2006) 205 Or.App. 297 [134 P.3d 1005].
Prior case law had established that pursuant to article XVIII, section 1, as adopted in 1879, the Legislature had authority to propose only amendments to the Constitution and not revisions. (Livermore v. Waite (1894) 102 Cal. 113, 118 [36 P. 424] (Livermore); see also McFadden v. Jordan (1948) 32 Cal.2d 330, 333-334 [196 P.2d 787] [the people have no authority to propose constitutional revision by initiative].)
The ballot pamphlet supplied to the electorate included the following analysis by the Legislative Counsel, highlighting the difference between a constitutional amendment and a revision: “This measure [that is, Proposition 7] . . . would authorize the Legislature by a vote of two-thirds of the members elected to each house to propose complete or partial ‘revisions’ of the Constitution for approval or rejection by the people. Under existing provisions the Legislature can only propose ‘amendments, ’ that is[,] measures which propose changes specific and limited in nature. ‘Revisions,’ i.e., proposals which involve broad changes in all or a [753]substantial part of the Constitution, can presently be proposed only by convening a constitutional convention.” (Ballot Pamp., Gen. Elec. (Nov. 6, 1962) analysis of Prop. 7, p. 13, italics added.)
Specifically, even after amendment of article XVIII, section 1, in 1962, the second sentence of that section continued to read, as it had since 1879: “Should more amendments than one be submitted at the same election they shall be so prepared and distinguished, by numbers or otherwise, that each can be voted on separately.” (1 Stats. 1963, p. ciii.)
Similar language appeared in each of the prior background studies, beginning with the Revision Commission’s Background Study 1 (c. 1966) at page 14. (That undated study apparently was prepared prior to July 14, 1966—the dates on which the committee reviewing article XVIII took its initial actions on the matter; see Background Study 7, supra, at pages 24-26.)
(Assem. Const. Amend. No. 28, Stats. 1969 (1969 Reg. Sess.) res. ch. 340, p. 4017.)
(Assem. Const. Amend. No. 28, Stats. 1969 (1969 Reg. Sess.) res. ch. 340, pp. 4017-4019.)
As to one of the 16 resolutions passed by the Legislature submitting the phase three constitutional changes to the voters at the November 3, 1970 General Election, the Legislature apparently felt obligated to take a different course, and its resolution characterized that measure as “a revision.” Resolution chapter 189 was a proposal to repeal parts of article XX and amend other parts of that article, the constitutional provision addressing “miscellaneous subjects” (including matters ranging from making Sacramento the state capital, to prohibiting those who fight a duel from holding public office, to reinstatement of public employees who resign to join the federal or state armed forces). The Legislature described that measure as proposing to “the people of the State of California a revision of portions of the Constitution of the state by amending Sections 1, 8, 17, 17A 1/2,18, and 20 of Article XX, and by repealing Sections 1, 3.5, 4, 5, 7, 9, 12, 13, and 14 of Article XX, relating to miscellaneous subjects.” (Assem. Const. Amend. No. 65, Stats. 1970 (1970 Reg. Sess.) res. ch. 189, p. 3781, italics omitted.)
With regard to the one measure that the Legislature had characterized as a revision (see ante, fn. 22), the Attorney General’s title for Proposition 15 (formerly resolution chapter 189) read: “Partial Constitutional Revision. Legislative Constitutional Amendment. . . .” (Ballot Pamp., Gen. Elec. (Nov. 3,1970) analysis of Prop. 15, p. 25.) In addition, the Attorney General departed from the Legislature’s characterization of resolution chapter 171 (at Stats. 1970 (1970 Reg. Sess.) p. 3732) (repealing obsolete provisions relating to the administration of aged and blind aid programs) as an amendment rather than a revision, and prepared a ballot title for that matter (designated Prop. 17) as follows: “Partial Constitutional Revision. Legislative Constitutional Amendment. Repeals obsolete provisions relating to social welfare.” (Ballot Pamp., Gen. Elec. (Nov. 3, 1970) analysis of Prop. 17, 29.)
The Legislative Counsel’s Analysis first described the existing provisions, and then explained that the measure “would retain the general substance of these provisions with the [759]following major changes: [ft] (1) A new provision would be added specifically authorizing the Legislature, by a two-thirds vote of the membership of each house, to amend or withdraw a constitutional amendment or revision which the Legislature has proposed where the action is taken before the proposal has been voted on by the electorate, [ft] (2) (a) The general requirement that the Legislature provide for the constitutional convention at the session following the voters’ approval of the proposition authorizing the convention would be replaced with a requirement that the Legislature provide for the convention within six months after the voters’ approval, [ft] (b) The existing constitutional limitations on the number of elected delegates to a constitutional convention and the requirement that they have the same qualifications and be chosen in the same manner as legislators would be deleted. A requirement would be added that the delegates, each of whom must be a voter, be elected from districts as nearly equal in population as may be practicable, [ft] (c) The existing constitutional requirement that the delegates meet within three months after their election would be deleted, [ft] (3) A provision would be added that if two or more measures amending or revising the Constitution are approved by the voters at the same election and they conflict, the provisions of the measure receiving the highest affirmative vote shall prevail. Thus, no distinction would be made in the Constitution between amendments proposed by the Legislature and by initiative measures, [ft] (4) Provisions prescribing detailed procedures for submitting to the voters, revisions proposed by the constitutional convention and for certifying the results of the election, would be deleted.” (Ballot Pamp., Gen. Elec. (Nov. 3, 1970) analysis of Prop. 16, pp. 27-28, italics added.)
The Legislative Counsel’s analysis explained: “Section 24 of Article IV of the Constitution now provides that an initiative or referendum measure takes effect five days after the official declaration of vote by the Secretary of State, unless the measure provides otherwise, while . . . constitutional amendments and revisions submitted by the Legislature take effect upon approval by voters, unless the measures provide otherwise.” (Ballot Pamp., Gen. Elec. (Nov. 3, 1970) analysis of Prop. 17, p. 28, boldface omitted.) The analysis also explained that under the proposed measure, “the provision for the effective date of all ballot measures would be the same, no matter how the ballot measures originated. Each ballot measure would become effective the day after the election at which it is approved, unless the measure provides otherwise.” (Ibid.)
To reiterate, from 1879 until amended in November 1970, the second sentence of article XVin, section 1 read: “Should more amendments than one be submitted at the same election they shall be so prepared and distinguished, by numbers or otherwise, that each can be voted on separately.” As the sentence was amended in November 1970, it presently states, “Each amendment shall be so prepared and submitted that it can be voted on separately.” (Ibid.)
The Legislature, after describing the changes made by Proposition 16, asserts that “[i]n a loose sense these changes all relate to elections. They are not, however, ‘functionally related.’ ” But even assuming, as the Legislature asserts, that the various changes proposed in Proposition 16 were not “functionally related,” as we shall explain post, part E.2.c, they did not need be. Moreover, the provisions of that measure not only were related to the general subject of elections (as the Legislature concedes), they shared a more focused commonality: each change in article XVIII and in former article ÍV, section 24, subdivision (a) of the Constitution (present Cal. Const., art. E, § 10, subd. (a)) related to election procedures for amending and revising the Constitution. As we shall explain post, part II.2.c, it is sufficient, under the separate-vote requirement, that provisions be reasonably germane to a common theme, purpose, or subject—• and all parts of Proposition 16 clearly were reasonably germane to the subject of amending and revising the Constitution.
Because, as noted above, the Legislature presented Resolution 103 as an amendment and not as a “revision” or “partial revision,” we need not address the problem of defining the contours of those terms. Furthermore, as the parties observed at oral argument, the definitional issue, which is not fully briefed in this case, is a potentially difficult one.
In setting forth the “reasonably germane” test, several of our prior decisions have stated or repeated language suggesting the standard requires that each of a measure’s parts be reasonably germane to one another as well as reasonably germane to a common theme, purpose, or subject. (See, e.g., Brosnahan, supra, 32 Cal.3d 236, 245 [“ ‘an initiative measure does not violate the single-subject requirement if ... all of its parts are “reasonably germane” to each other,’ and to the general purpose or object of the initiative” (italics omitted)]; Eu, supra, 54 Cal.3d 492, 512, quoting Brosnahan; Jones, supra, 21 Cal.4th 1142, 1157, quoting Eu.) In applying the reasonably germane test, however, our decisions uniformly have considered only whether each of the parts of a measure is reasonably germane to a common theme, purpose, or subject, and have not separately or additionally required that each part also be reasonably germane to one another. (See, e.g., Eu, supra, 54 Cal.3d at pp. 512-514; Brosnahan, supra, 32 Cal.3d at pp. 245-253; Fair Political Practices Commission, supra, 25 Cal.3d 33, 38—43; see also Jones, supra, 21 Cal.4th at p. 1158 [“The single-subject rule . .. simply precludes drafters from combining, in a single [measure], provisions that are not reasonably germane to a common theme or purpose” (italics omitted)].) The governing decisions’ consistent application of the standard suggests that a measure’s separate provisions have been considered to be reasonably germane to each other within the meaning of the standard so long as all of the provisions are reasonably germane to a single common theme, purpose, or subject. For clarity and simplicity, we believe it is appropriate to describe the test, as set forth in the text above, as simply requiring that the separate provisions of a measure be reasonably germane to a common theme, purpose, or subject. (At the same time, of course, our decisions also establish that the single subject provision “obviously forbids joining disparate provisions which appear germane [765]only to topics of excessive generality such as ‘government’ or ‘public welfare.’ ” [Brosnahan, supra, 32 Cal.3d at p. 253; see, e.g., Jones, supra, 21 Cal.4th at pp. 1161-1162.].)
See, e.g., Dubois & Feeney, Lawmaking by Initiative (1998) page 138; but see Lowenstein, Initiatives and the New Single Subject Rule (2002) 1 Elec. L.J. 35 (criticizing recent decisions, some of which employ a more exacting single subject test and find violations of states’ single subject provisions and/or separate-vote provisions); Miller, Courts as Watchdogs of the Washington State Initiative Process (2001) 24 Seattle U. L.Rev. 1053, 1079-1083 (noting with approval the same recent cases).
Accord, Wetz, supra, 168 N.W. 835, 847 (North Dakota’s two provisions are designed to prevent “log-rolling and joker practices which are so familiar to students of legislation”); Carter, supra, 198 S.E.2d 151, 156 (referring to Georgia’s separate-vote and single subject rules, the court observed that “ ‘the evils [of logrolling] to be prevented by prohibiting such a practice are as apparent in the one case as in the other’ ”); Petition No. 314, supra, 1980 OK 174, 625 P.2d 595, 603-605 (quoting and relying interchangeably upon single subject and separate-vote provision cases, and finding the underlying purpose of the two provisions of the Oklahoma charter to be the prevention of logrolling); Andrews, supra, 449 A.2d 1144, 1150 (Maryland’s constitutional single subject and separate-vote provisions “must be construed together as serving the same objectives and as being directed at the same evils”). Most recently, the New Jersey Supreme Court observed that the “primary goal underlying” “both the single [subject] and the separate vote requirements found in many state constitutions ... is the prevention of ‘log-rolling.’ ” (Cambria, supra, 776 A.2d 754, 764.)
See also, e.g., State v. Mason (1891) 43 La.Ann. 590 [9 So. 776, 800-801] (noting that the single subject provision test applies to a separate-vote provision challenge); Lobaugh v. Cook (1905) 127 Iowa 181 [102 N.W. 1121, 1123] (Lobaugh) (construing state constitution’s separate-vote provision by analogy to the constitution’s single subject provision); State ex rel. Collins v. Jones (1914) 106 Miss. 522 [64 So. 241, 253-254] (construing state’s separate-vote provision by reference to cases construing the state’s statutory single subject provision, and disapproving an earlier Mississippi decision that held otherwise); Wetz, supra, 168 N.W. 835, 847, 850 (construing North Dakota’s separate-vote provision by analogy to its single subject provision and employing the same test); Mundell v. Swedlund (1937) 58 Idaho 209 [71 P.2d 434, 442] (it is “reasonable to assume” that the tests for a single-subject-provision violation and a separate-vote-provision violation “should be determined in the same manner and by the same rule and of course reasoning”); Graham v. Jones (1941) 198 LA. 507 [3 So.2d 761, 774—775] (test for violation of state’s separate-vote provision is determined by reference to the state’s single subject provision); State v. Holman (Mo. 1962) 363 S.W.2d 552, 556 (Holman) (state’s single subject provision is “further” and “controlling” with respect to the state’s separate-vote provision); Carter, supra, 198 S.E.2d 151, 156 (construing the Georgia Constitution’s separate-vote provision by analogy to the state constitution’s single subject provision and employing the same test); Andrews, supra, 449 A.2d 1144, 1150 (the Maryland Constitution’s separate-vote provision and single subject provisions “must be construed together as serving the same objectives and as being directed to the same evils”). More recently, the Missouri Supreme Court, consistent with its view in Gabbert, supra, 70 S.W. 891, observed that its state constitution’s separate-vote provision restates in different language the same constitution’s single subject provision. (Missourians to Protect the Init. Proc. v. Blunt (Mo. 1990) 799 S.W.2d 824, 830 (Blunt).) Similarly, the Idaho Supreme Court, finding a constitutional amendment impermissible under its constitution’s separate-vote provision, applied a single subject test and concluded its analysis by referring to the state’s constitutional separate-vote provision as “the ‘single subject rule’ of the Idaho Constitution.” (IWP, supra, 982 P.2d 358, 363.)
Referring to the rejected test, the court stated: “Such a construction would, we think, be so narrow as to render it practically impossible to amend the constitution; or, if not practically [767]impossible, it would compel the submission of an amendment which, although having but one object in view, might consist of considerable detail, and each separate provision, though all promotive of the same object and necessary to the perfection and practical usefulness thereof, if adopted as a whole, in such form that a defeat of one of its important matters of detail might destroy the usefulness of all the other provisions when adopted.” (Timme, supra, 11 N.W. 785, 790.)
See, e.g., Gabbert, supra, 70 S.W. 891, 897-898 (Missouri constitutional amendment abolishing the common law jury unanimity requirement in different levels of both civil courts, and imposing a two-thirds and three-quarters vote rule, respectively, properly was submitted to voters in one ballot measure because the changes were germane to “one dominant idea” and “one subject”); People v. Sours (1903) 31 Colo. 369 [74 P. 167, 178] (amendment embracing more than one subject [consolidation of specified city and county governments, and framing of home rule charters by all cities within die state] did not violate the separate-vote provision where the subjects are “germane to the general subject of the amendment” or are functionally related such that “it might not be desirable that one be adopted, and not the other”); Lobaugh, supra, 102 N.W. 1121, 1123 (amendment creating biennial statewide elections, allowing the Iowa Legislature to select the Iowa Chief Justice, and requiring all statewide and local elections to be held at the same time, addressed a common subject and properly was submitted as one measure).
See, e.g., Wetz, supra, 168 N.W. 835, 846-848; State, ex rel. v. Cook (1932) 44 OhioApp. 501 [14 OhioLawAbs. 304, 185 N.E. 212, 213-214, 38 Ohio L.Rep. 115]; accord, Curry v. Laffoon (1935) 261 Ky. 575 [88 S.W.2d 307, 308].
See Fugina v. Donovan (1960) 259 Minn. 35 [104 N.W.2d 911, 914] (upholding a constitutional amendment, submitted as a single measure, permitting the state legislature to extend legislative terms and allowing legislators to serve as notaries and seek election to other offices, because the provisions were rationally related to a single subject; court rejected a strict functional relationship test under which “propositions that may be separately submitted without being incomplete shall be submitted separately”); Penrod v. Crowley (1960) 82 Idaho 511 [356 P.2d 73, 79] (constitutional amendment revising the selection and jurisdiction of justices of the peace properly was submitted in one measure, because “both are germane to the common object and purpose”); Holman, supra, 363 S.W.2d 552, 556 (Missouri constitutional amendment authorizing both general obligation and general revenue bonds properly was submitted in one measure because, by analogy to the single subject rule, they are related to the general subject of the authorization of municipal financing); State ex rel. Roahrig v. Brown (1972) 30 OhioSt.2d 82 [282 N.E.2d 584, 586] (constitutional amendment concerning administration of the state legislature, election of governor and lieutenant governor, repeal of provisions concerning the supreme court, and barring persons convicted of specified felonies from holding public office properly was submitted in one measure, because “each of its subjects bears some reasonable relationship to a single general object or purpose”); City of Raton v. Sproule (1967) 78 N.M. 138 [429 P.2d 336, 342] (following the “majority” rule that an “amendment, which embraces several subjects or items of change”—there, provisions concerning special elections and regular elections, and qualified voters at those elections— “may be submitted to the electorate as one general proposition, if all the subjects or items of change contained in the amendment are germane to one general object or purpose”); Carter, supra, 198 S.E.2d 151, 157 (Georgia constitutional amendment consolidating and reorganizing executive branch agencies properly was submitted in one measure, because it addressed “but a single subject matter” and “[a]ll of its parts were germane to a single purpose”); Andrews, supra, 449 A.2d 1144, 1150 (Maryland constitutional separate-vote provision and single subject provision are to be construed together as common “constraints on the legislature”; Maryland constitutional amendments restructuring the state court system and making discretionary the former right of removal, properly could have been submitted in one measure consistent with the separate-vote provision, but legislature also had discretion under that provision to submit them separately and to make passage of one contingent upon the other).
The court held that “[w]hether amendments are one or many must be solved by their inherent nature,—by the consideration whether they are separate and independent each of the other, so as that each can stand alone without the other, leaving the constitutional system symmetrical, harmonious, and independent on that subject . . . .” (Powell, supra, 27 So. at p. 931.)
In reaching this conclusion, the Mississippi court also relied heavily upon its own case law construing the state’s analogous statutory single subject rule. (State ex rel. Collins v. Jones, supra, 64 So. 251, 253-255.)
The Oregon provision reads: “When two or more amendments shall be submitted ... to the voters of this state at the same election, they shall be so submitted that each amendment shall be voted on separately.” (Or. Const., art. XVII, § 1.)
In at least Montana and Arizona as well, the separate-vote provision applies not only to constitutional amendments proposed by the state legislature, but also to constitutional amendments proposed by the people through the initiative. (See Ariz. Const., art. XXI, § 1; Mont. Const., art. XIV, § 11.) In California, by contrast, the separate-vote provision applies only to legislative proposals for constitutional amendment, and does not govern initiative proposals for constitutional amendment (Cal. Const., art. XVIII, § 1; Wright, supra, 192 Cal. 704, 711-712), which of course instead are governed by the single subject rule.
Most recently, in League of Oregon Cities, supra, 56 P.3d 892, the Oregon high court invalidated under the separate-vote provision a regulatory takings reform measure, because it presented to the voters in a single proposition provisions (1) expanding property rights and (2) implicitly reducing free speech rights; once again the court found no “close relationship” between these two changes, and declared the entire measure invalid. (Id., at pp. 909-910.) We note that all of the cases in which the Oregon courts have found separate-vote provision violations have concerned initiative constitutional amendments (as contrasted with legislative constitutional amendments).
In only one post-Armatta Oregon appellate decision raising a separate-vote provision issue has a violation not been found (Hartung v. Bradbury (2001) 332 Or. 570 [33 P.3d 972])—and the holding in that case appears to be confined to itself, having been ignored in the most recent three Oregon Supreme Court decisions addressing the separate-vote issue. In Hartung, supra, petitioners challenged the Secretary of State’s reapportionment of state legislative districts, arguing that 1952 and 1986 constitutional amendments under which reapportionment had proceeded were invalid under the state constitution’s separate-vote provision. After assuming without deciding that the challenge could be raised so many “years after the adoption of those amendments” (id., at p. 976, fn. 3), the court noted that it many years earlier had rejected a separate-vote challenge to the 1952 amendment in Baum v. Newbry (1954) 200 Or. 576 [267 P.2d 220] (a decision that, following the majority rule described ante, pt. n.B.l, employed a lenient rather than strict construction of the separate-vote rule) and that the 1986 amendment “made [only] relatively modest changes” to that prior amendment (Hartung, supra, 33 P.3d at p. 976). The court then announced—without describing the amendments or engaging in any analysis under Armatta, supra, 959 P.2d 29, 64, concerning whether the amendments would “make two or more changes to the constitution that are substantive and are not closely related”—that in light of its conclusion in Baum, “the more limited 1986 amendment necessarily withstands petitioners’ constitutional challenge.” (Hartung, supra, 33 P.3d at p. 976, italics added.) As one commentator has observed, it appears that under the Armatta test (as articulated in that case and strictly applied in subsequent Oregon decisions), the amendments at issue in Hartung,would have been found to violate the state’s separate-vote provision. (Reforming Direct Democracy, supra, 93 Cal. L.Rev. 1191, 1222-1223.)
The Supreme Court of Montana, in Marshall, supra, 975 P.2d 325, found a violation of that state’s separate-vote provision with regard to an initiative constitutional amendment tax reform measure that presented to the voters in a single proposition provisions (1) amending the [772]revenue and finance article; (2) eliminating the sovereign immunity shield for a violation of the amended provisions; and (3) making an exception to limitations placed upon the Governor’s veto powers, by allowing veto of referenda proposing new or increased taxes. (Id., at pp. 330-332.)
The Supreme Court of Pennsylvania, in Pennsylvania Prison Soc., supra, 776 A.2d 971, found a violation of that state’s separate-vote provision with regard to a legislative constitutional amendment measure that presented to the voters in a single proposition provisions (1) restructuring the state board of pardons to change its composition and require decisions to be unanimous, and (2) requiring a majority rather than two-thirds vote of the state senate to confirm the governor’s appointees to that board. (Id., at pp. 981-982.) The lead opinion, however, found it unnecessary, in the “unusual circumstances" presented (id., at p. 982), to declare the measure void, because, it determined, the measure actually made no substantive change to the state senate’s confirmation process. (Id., at pp. 981-984; see also Bergdoll, supra, 731 A.2d 1261 [invalidating, under the state separate-vote provision, a legislative constitutional amendment measure that presented to the voters in a single proposition provisions (1) eliminating a “face-to-face” requirement under the state’s confrontation clause, and (2) authorizing the legislature to enact laws regarding the manner in which children may testify in criminal proceedings].)
The Supreme Court of Idaho—consistently with its earlier minority-view decision in McBee, supra, 100 P. 97 (adopting a strict rule under which, if proposed provisions can stand alone, they must be presented alone)—invalidated, under that state’s separate-vote provision, a legislative constitutional amendment that would have (1) allowed proceeds from the sale of school endowment lands to be used to acquire other lands, and (2) provided that auctions should take place regarding only sales (and not, alternatively, leases or sales) of such lands. The court concluded that these two provisions were “ ‘ “essentially unrelated” ’ ” to each other. (IWP, supra, 982 P.2d 358, 363.)
Similarly, the Supreme Court of Arizona—again, consistently with its own earlier minority-view decision in Kerby, supra, 36 P.2d 549 (adopting a strict rule under which proposed provisions may be presented in a single package only if they “should stand or fall as a whole”)—invalidated, under that state’s separate-vote provision, an initiative constitutional amendment that would have (1) eliminated public funding of statewide political campaigns, and (2) diverted all money then dedicated to the Arizona Clean Elections Commission into the state’s general fund, thus making the commission’s funding for its other responsibilities (beyond public campaign financing) dependent upon grants from the state legislature. (Clean Elections Institute, supra, 99 P.3d 570, 575-577.)
Only in New Jersey has a high court that purports to embrace the Armatta test concluded that a challenged measure meets the requirements of the separate-vote provision. In Cambria, supra, 776 A.2d 754, that state’s high court upheld, against a separate-vote provision challenge, a legislative constitutional amendment measure dedicating two new sources of revenue to the state’s Transportation Trust Fund: (1) revenue from a petroleum products tax, and (2) revenue from a general sales and use tax on new motor vehicles. The court in Cambria found these two provisions “closely related” to each other. (Id., at p. 765.)
See Andrews, supra, 449 A.2d 1144, 1150 (construing the single subject and separate-vote provisions of Md. Const., art XIV, § 1, “together as serving the same objectives and as being directed at the same evils”); Blunt, supra, 799 S.W.2d 824, 830 (construing the single subject and separate-vote provisions of Mo. Const., art. XII, § 2(b), together and interchangeably).
See Petition No. 314, supra, 625 P.2d 595, 602-606 (construing the single subject and separate-vote provisions of Okla. Const., art. XXIV, § 1 [2d par.] together and interchangeably).
Indeed, reports prepared in the 1960’s by the Revision Commission suggest the opposite and reveal that the drafters saw no reason to view the separate-vote provision as imposing a requirement stricter than imposed by the single subject rule. In Background Study 1 (mentioned ante, fn. 19), after describing the “liberal” single subject rule applicable to initiatives proposing constitutional amendments (Background Study 1, supra, at p. 14), the report commented: “It seems likely that the permitted scope of an amendment proposed either by the Legislature or by initiative would be construed to be essentially the same, i.e., limitation to a single subject. . . .” (Ibid.) Later, the same study commented: “No good reason appears why different rules should govern initiative amendments and legislative amendments, at least with respect to limitation on the scope of an amendment. . . .” (Id., at p. 15.) The study continued by suggesting that “[t]he reasons for the differences appear[] to be historical, i.e., that Article XVm, section 1 and the initiative provisions were adopted at different times and no effort was ever made to conform their various provisions. [][] ... [f] ... In summary, there appears to be no reason why these limitations should not be the same and stated in the same way, and the single subject requirement seems more meaningful than the requirement that each amendment be submitted separately.” (Id., at p. 16.) The Revision Commission’s ensuing studies, culminating in Background Study 7, were consistent. For example, in Background Study 7, after again describing the “liberal” single subject rule applicable to initiatives proposing constitutional amendments (id., at p. 19), the committee’s study repeated the observation that the “reason for the differences between initiative amendments and amendments proposed by the Legislature seems to be historical,” and that “no attempt seems to have been made to conform them” (ibid.), yet also observed that “[tjhere appears to be no good reason why the restrictions on both classes of amendments should not be the same.” (Id., at p. 20.)
The same cannot be said with regard to Montana’s jurisprudence, however. In State ex rel. Hay v. Alderson (1914) 49 Mont. 387 [142 P. 210], the state supreme court concluded, consistently with the majority rule, that “the unity of subject [implicitly] required by the [state separate-vote provision] does not essentially differ from the unity of subject required by the [state’s statutory single subject rule].” (Id., at p. 213.) In its recent decision in Marshall, supra, 975 P.2d 325, 331, the Montana court overruled Alderson and its progeny, in favor of the Armatta rule.
The original provision—which was phrased as a single “object” rather than a single “subject” limitation—read: “Every law enacted by the Legislature shall embrace but one object, and that shall be expressed in its title . . . .” (Cal. Const. of 1849, art. IV, § 25.)
As adopted in 1879, the provision read: “Every Act shall embrace but one subject, which subject shall be expressed in its title. But if any subject shall be embraced in an Act which shall not be expressed in its title, such Act shall be void only as to so much thereof as shall not be expressed in its title.” (Cal. Const., art. IV, § 24, as adopted 1879.) Presently, as modified in 1966, the provision reads: “A statute shall embrace but one subject, which shall be expressed in its title. If a statute embraces a subject not expressed in its title, only the part not expressed is void.” (Cal. Const., art. IV, § 9.)
As Delegate McCallum observed: “The committee were of the opinion, that inasmuch as this section has stood for many years, ... we might as well let it stand as it is.” (2 Willis & Stockton, supra, at p. 797; see also 3 Willis & Stockton, supra, at p. 1269 [remarks of Delegate Barbour: “[T]his section is aimed at the practice of smuggling and logrolling bills through”].)
See, for example, 2 Willis & Stockton, supra, at pages 797-798 (remarks of Delegate McFarland, opposing the proposed severability clause: “I repeat that it is impossible to express all there is in a law in its title without making the title nearly as long as the law”); id., at page 798 (remarks of Delegates Edgerton, Laine, Johnson & Hager, noting that similar severability language was found in the constitutions of Iowa, Indiana, Illinois, and Oregon); id., at page 799 (remarks of Delegate Wilson, supporting the severability clause as being consistent with “elementary” rules of appellate adjudication).
As observed earlier, the provision states: “An initiative measure embracing more than one subject may not be submitted to the electors or have any effect.” (Cal. Const., art II, § 8, subd. (d).) By contrast, the legislative single subject and title provision contains an express severance clause.
Justice Davis, dissenting on this issue in the Court of Appeal below, made the same point as follows: “In future cases, the majority’s remedy would allow a faction of those voting for a conjoined set of amendments to accomplish with stealth what could not be secured through the legislative process; namely the separate enactment of an amendment. Those members and their allies could do so by later persuading the Secretary [of State] or the judiciary to extract their favored amendment for individual consideration. [The majority’s proposed] remedy should not be invoked in the context of the fundamental organic law of the state, where the Legislature must comply strictly with the procedure for amendment.”