Taxpayers to Limit Campaign Spending v. Fair Political Practices Commission

KENNARD, J.

I dissent. When simultaneously approved electoral measures contain conflicting provisions, our state Constitution instructs us in plain language that the measures are to be examined provision by provision and reconciled to the greatest possible extent. Both parties to this litigation agree that this is what the Constitution requires. Yet the majority, in direct violation of the Constitution, invalidates in its entirety an initiative measure approved by the voters at a statewide election without pausing to determine whether it contains provisions that are reconcilable with the measure that received more votes.

Much of the majority opinion is devoted to a demonstration of the supposed analytical difficulty of applying the provision-by-provision test to the initiative measures at issue here. I recognize that the majority’s all-or-nothing approach makes the task of judicial review much easier, but resolving difficult and complex legal issues is unquestionably this court’s constitutionally assigned responsibility. We may not elevate judicial convenience over constitutional command. 1

*775 Equally misguided is the majority’s insistence that the Constitution must be interpreted to require an all-or-nothing approach because the provision-by-provision test will result in a hybrid scheme that the voters did not intend. As I will explain, the test employed by the Court of Appeal, and endorsed by both parties to this litigation, requires that any nonconflicting provisions be tested for severability. This is the established and time-tested method for determining whether the voters would have wanted part of an electoral measure to go into effect when another part has been found invalid.

The ballot for the June 1988 Primary Election included Propositions 68 and 73, two initiative measures that contained provisions imposing restrictions on the financing of political campaigns. Although both measures received a majority of affirmative votes, Proposition 73 received a greater number of affirmative votes. It is undisputed that some of the provisions of Proposition 68 conflict with provisions of Proposition 73. It is equally undisputed that some of Proposition 68’s provisions do not conflict with anything in Proposition 73. The issue here, as presented by the parties, is which of Proposition 68’s provisions became effective.

The problem of conflicts in simultaneously approved electoral measures was foreseen and dealt with in our state Constitution in article II, section 10, subdivision (b) (hereafter section 10(b)), which reads: “If provisions of 2 or more measures approved at the same election conflict, those of the measure receiving the highest affirmative vote shall prevail.” The proper interpretation of this language has a public importance that transcends the fate of the particular measure challenged here.

The first step in interpreting a constitutional provision such as section 10(b) is a close examination of its text. A constitutional provision should be construed in accordance with the natural and ordinary meaning of its words. (Mutual Life Ins. Co. v. City of Los Angeles (1990) 50 Cal.3d 402, 407 [ 267 Cal.Rptr. 589 , 787 P.2d 996 ]; ITT World Communications, Inc. v. City and County of San Francisco (1985) 37 Cal.3d 859, 865 [ 210 Cal.Rptr. 226 , 693 P.2d 811 ]; see also, Gibbons v. Ogden (1824) 22 U.S. 1, 188 [ 6 L.Ed. 23, 68 ] [framers of the federal Constitution “must be understood to have employed words in their natural sense, and to have intended what they have said.”].) Extrinsic aids should be consulted if, and only if, some legitimate uncertainty remains after consideration of the constitution’s language. (Mutual Life Ins. Co. v. City of Los Angeles, supra, at p. 407; Sandelin v. Collins (1934) 1 Cal.2d 147, 155 [ 33 P.2d 1009 , 93 A.L.R. 956 ].)

*776 The wording of section 10(b) is significant in two respects. First, the section distinguishes between “measures,” meaning an entire initiative, and “provisions of . . . measures,” meaning the separate operative parts of an initiative. 2 It speaks not of conflicting measures but of conflicting provisions of measures. Thus the unit of analysis for purposes of determining whether a conflict exists is the individual provision, not the initiative as a whole. Section 10(b) unambiguously requires that courts undertake a provision-by-provision comparison of voter-passed initiatives, not merely a comparison of their overall import.

The manner in which a conflict is to be resolved is likewise described in terms of provisions, not measures. If the provisions of two or more simultaneously enacted electoral measures conflict, section 10(b) provides that “those of the measure receiving the highest affirmative vote shall prevail.” Thus section 10(b) unambiguously prescribes that whenever a provision-by-provision comparison reveals one or more conflicts, each such conflict is to be resolved on a provision-by-provision basis. To hold that any conflict between competing ballot measures renders invalid all provisions of all measures other than the measure receiving the highest affirmative vote, as the majority does, not only disregards section 10(b)’s plain meaning, it also renders surplusage all of section 10(b)’s references to provisions of measures. (See 37 Ops.Cal.Atty.Gen. 52 , 56, fn. 11 (1961).) It is an elementary rule of construction, applicable to constitutions no less than statutes, that “every word should be given some significance, leaving no part useless or devoid of meaning.” (City and County of San Francisco v. Farrell (1982) 32 Cal.3d 47, 54 [ 184 Cal.Rptr. 713 , 648 P.2d 935 ].) The construction adopted by the majority violates this elementary rule.

The majority’s view of section 10(b)’s language is difficult to understand. The majority ultimately concludes that when there is a conflict between simultaneously enacted ballot measures, the measures are sometimes to be examined as a whole, and at other times provision by provision, depending on whether the measures address the same subject or are “competing.” (Maj. opn., ante, at p. 771, fn. 12.) I find nothing in the language of section 10(b) to support this bifurcated approach.

A second and no less significant feature of section 10(b)’s language is its use of the term “prevail.” Had the framers of section 10(b) intended that *777 only one of two or more competing ballot measures could be adopted, it would have been a simple matter to say so. An example of such language is found in a charter amendment that gave the powers of initiative and referendum to the voters of the City and County of San Francisco. 3 The charter amendment was passed by the local electorate in 1910 and approved by the Legislature in February 1911, during the same legislative session at which the predecessor of section 10(b) was drafted as part of the constitutional amendment that reserved the power of initiative for voters at state elections. The charter amendment stated, in relevant part: “When there are two or more measures proposed to secure the same general purpose, the board of election commissioners shall so declare, and shall have the ballots so printed that the voter (first) can choose between any measure or none, and (secondly) can express his preference for any one. If a majority of the votes on the first question is affirmative, then the measure receiving the highest number of votes shall become law, and the others shall fail of passage. In case two or more measures are tied for the highest vote, they shall be resubmitted at the next ensuing general election. If there is a conflict between two or more measures or between two or more charter amendments adopted at the same election, then the measure or charter amendment receiving the highest affirmative vote shall prevail. No ordinance or measure approved by the electorate under the provision of this chapter shall be subject to veto, or be amended or repealed except by vote of the electorate, unless such ordinance or measure shall otherwise provide.” (S.F. Charter, former art. XI, ch. III, § 7, italics supplied; see Stats. 1911, ch. 25, pp. 1661, 1670, 1672. See now, S.F. Charter, § 9.114.)

This charter provision established a dual system. Initiative measures could be presented to the voters as alternatives, in which case the voters would be informed that only one could be adopted and the others would “fail of passage.” If initiative measures were not so presented, however, two or more measures that conflicted could be “adopted,” that is, could become law. If this happened, the measure receiving the highest affirmative vote would “prevail.” (S.F. Charter, former art. XI, ch. III, § 7.)

To say that one of two conflicting measures shall “prevail” is not the same as saying that the other measure shall not take effect. The same term—“prevail”—is used in Government Code section 9605 to define the resolution of conflicts in statutes enacted at the same legislative session. In relevant part, Government Code section 9605 provides: “In the absence of *778 any express provision to the contrary in the statute which is enacted last, it shall be conclusively presumed that the statute which is enacted last is intended to prevail over statutes which are enacted earlier at the same session . . . .” In this closely related context, this court has interpreted “prevail” to mean “take precedence.” (In re Thierry S. (1977) 19 Cal.3d 727, 738 [ 139 Cal.Rptr. 708 , 566 P.2d 610 ].) In other words, “prevail” connotes a comparatively higher degree of rank or authority rather than a complete extirpation of rivals.

This point is illustrated by a Court of Appeal decision concerning a conflict between two Penal Code sections that had been amended at the 1943 legislative session. As amended, Penal Code section 669 provided that when a person was sentenced to terms of imprisonment for two or more crimes, and one of the terms was for life, the other term or terms were to merge and run concurrently with the life term. Penal Code section 4530, as amended, provided that the term of imprisonment for an escape would commence from the time the prisoner would otherwise have been discharged from prison. The sections conflicted only in the case of a prisoner serving a life term who was thereafter convicted of escape. Applying Government Code section 9605, the Court of Appeal concluded that Penal Code section 4530 “prevailed” over Penal Code section 669 because it was enacted later, as evidenced by its higher chapter number, and therefore a prison term for an escape would commence only upon the conclusion of the life term. (In re McManus (1954) 123 Cal.App.2d 395, 401 [ 266 P.2d 929 ].) The court nowhere suggested, however, that the amendment of Penal Code section 669 was ineffective in all those situations in which it did not conflict with the amended Penal Code section 4530, that is, whenever a term of imprisonment for a crime other than escape was imposed on a prisoner serving a life term.

The meaning of the term “prevail” becomes a matter of critical importance when the measure receiving the higher number of votes is held invalid. If the term “prevail” in section 10(b) means that the measure receiving fewer votes fails of passage, as the majority appears to hold, then the voters who approved both measures will end up with neither. But if, as I am persuaded, “prevail” means only “take precedence,” then a conflict between measures affects only the enforceability of the measure receiving fewer votes. As both measures are enacted, the measure receiving fewer votes will become fully effective and enforceable should the measure receiving the greater number of votes be held invalid. 4

*779 The preceding examination of the text of section 10(b) leads me to conclude that it compels the provision-by-provision analysis undertaken in this case by both the Fair Political Practices Commission and the Court of Appeal, in which initiative measures enacted together are harmonized whenever possible and specific provisions are invalidated only to the extent they conflict. Because there is no ambiguity in the language of section 10(b), I see no need to resort to extrinsic construction aids. If materials other than the text of section 10(b) are considered, however, the majority’s construction of section 10(b) appears all the more untenable.

As the majority observes (maj. opn., ante, at p. 747, fn. 1), language identical to section 10(b) appears in two other parts of the California Constitution: article XI, section 3, subdivision (d) (hereafter section 3(d)), which applies to conflicts in measures to adopt or amend city or county charters, and article XVIII, section 4 (hereafter section 4), which applies to conflicts in measures to amend the state Constitution. Because parallel provisions using similar or identical language should be construed to have the same meaning (People v. Caudillo (1978) 21 Cal.3d 562, 585 [ 146 Cal.Rptr. 859 , 580 P.2d 274 ]; In re Phyle (1947) 30 Cal.2d 838, 845 [ 186 P.2d 134 ]), evidence concerning the meaning of sections 3(d) and 4 should be given great weight in interpreting section 10(b).

Sections 3(d) and 4 were both adopted as part of the comprehensive 1970 constitutional revision. To ascertain their meaning, one source that may be consulted is the official reports of the California Constitution Revision Commission (see Mosk v. Superior Court (1979) 25 Cal.3d 474, 495 [ 159 Cal.Rptr. 494 , 601 P.2d 1030 ]); such reports are ’’entitled to great weight” (District Election etc. Committee v. O’Connor (1978) 78 Cal.App.3d 261, 270 [ 144 Cal.Rptr. 442 ]). Regarding section 3(d), the California Constitution Revision Commission had this to say in its official report: “Proposed Section 3(d) resolves conflicts between provisions approved at the same election by providing that the provisions of the measure with the highest affirmative vote prevail. It further provides, in effect, that only invalid provisions of a measure fail.” (Cal. Const. Revision Com., Proposed Revision (1968) p. 56, italics added.) The commission’s comment to section 4 is similar but even more explicit: “A provision is added requiring that the measure receiving the highest affirmative vote prevails in the event measures approved at the same election contain conflicting provisions. This has been the rule governing measures proposed by the Legislature, and it is now extended to initiative measures and measures proposed by constitutional *780 convention. The revision also provides that if one portion of a measure fails because a conflicting measure receives more votes, the remainder of the measure not in conflict remains in effect.'” (Id. at p. 111, italics added.) The commission’s comments provide persuasive evidence that sections 3(d) and 4, and consequently section 10(b) as well, require a court, when confronted with conflicts in the provisions of simultaneously adopted electoral measures, to give effect to those provisions not in conflict.

The majority places considerable reliance on an analysis of historical materials, but nothing in those materials alters my view regarding the meaning of section 10(b). The most obvious flaw in the majority’s analysis of the context in which the predecessor of section 10(b) was adopted is its erroneous assumption that the terms “prevail” and “control,” as applied to one of two conflicting measures, imply that the other measure is without effect. For example, the majority observes that in 1911 a section of the Political Code addressed the problem of conflicts between county ordinances proposed at the same election. Former Political Code section 4058 provided that in this event “the ordinance receiving the highest number of affirmative votes shall control.” (Stats. 1911, ch. 342, § 1, p. 578.) Without citing any authority, the majority asserts that it was “clear on the face” of former Political Code section 4058 that “only one measure would go into effect in case of a conflict.” (Maj. opn., ante, at p. 767.) The majority is simply wrong. As demonstrated above, the terms “control” and “prevail” mean only that the one measure will take precedence over the other, not that the other will be without effect.

To cite yet another example of this usage, some statutory conflicts are resolved according to the rule that a more specific statute “controls” or “prevails over” or “takes priority over” a more general statute. (See, e.g., Code Civ. Proc., § 1859 [“So a particular intent will control a general one that is inconsistent with it.”]; Estate of Kramme (1978) 20 Cal.3d 567, 576 [ 143 Cal.Rptr. 542 , 573 P.2d 1369 ] [“the specific statute controls and takes priority over a general statute encompassing the same subject”]; County of Placer v. Aetna Cas. etc. Co. (1958) 50 Cal.2d 182, 189 [ 323 P.2d 753 ] [“the later specific statute controls”]; Board of Supervisors v. Simpson (1951) 36 Cal.2d 671, 673 [ 227 P.2d 14 ] [“a particular statutory provision should prevail over a general one”].) This does not mean that the general statute ceases to have effect, but that the specific is treated as an exception to the general. (County of Placer v. Aetna Cas. etc. Co., supra, at p. 189.)

As the majority observes (maj. opn., ante, at pp. 767-768), the predecessor of section 3(d) was adopted in the following form: “In submitting any such charter or amendment separate propositions, whether alternative or *781 conflicting, or one included within the other, may be submitted at the same time to be voted on by the electors separately, and, as between those so related, if more than one receive a majority of the votes, the proposition receiving the larger number of votes shall control as to all matters in conflict.” (Cal. Const., former art. XI, § 8; Stats. 1913, ch. 90, p. 1732, italics added.) As the majority concedes, this means that each proposition receiving a majority of the votes becomes effective as to matters not in conflict. The majority cites this as evidence that when a harmonizing approach was intended, the framers of the Constitution clearly so stated. But the majority fails to explain why the rules for state elections should differ from those applying to local elections, or why, if a different meaning was originally intended, the constitutional provisions governing state and local elections now use identical language.

I would draw a lesson different from that drawn by the majority. First, there is no indication that the former provision’s repeal and reenactment as section 3(d) was intended to effect a change in meaning. On the contrary, the official report of the California Constitution Revision Commission contains this comment regarding its proposal that section 3(d) be adopted in its current form: “This proposal clarifies and extends to counties similar existing provisions which are applicable only to cities.” (Cal. Const. Revision Com., Proposed Revision (1968) p. 56.) As noted earlier, section 3(d) in its current form is identical to section 10(b). Thus the history of sections 10(b) and 3(d) compels the conclusion that both now require, and their predecessors always have required, that every effort be made to reconcile and harmonize ballot measures adopted at the same election.

The provision-by-provision analysis required by section 10(b) is consistent with the rule that statutes relating to the same subject matter must be read together and harmonized to the greatest extent possible. (Fuentes v. Workers’ Comp. Appeals Bd. (1976) 16 Cal.3d 1, 7 [ 128 Cal.Rptr. 673 , 547 P.2d 449 ].) This cardinal rule of construction applies to statutes enacted by initiative as well as to acts of the Legislature. (Estate of Gibson (1983) 139 Cal.App.3d 733, 736 [ 189 Cal.Rptr. 201 ].) It applies with particular force “when such statutes are enacted at the same time, or at the same session of the Legislature or when they become effective on the same date.” (In re Marriage of Pinto (1972) 28 Cal.App.3d 86, 89 [ 104 Cal.Rptr. 371 ].) The majority’s construction of section 10(b) abandons these long-standing principles.

The majority’s policy arguments for adopting an all-or-nothing approach are unpersuasive. The majority argues that when two or more ballot measures are “competing initiatives,” the all-or-nothing approach is necessary *782 to prevent the creation of a regulatory scheme that the electorate did not understand or intend. Yet the same objective can more accurately and more appropriately be attained by examining the provisions individually.

A ballot measure may be invalidated in its entirety if each of its provisions conflicts with the provisions of a measure receiving a higher number of affirmative votes. Also, any ballot measure can establish a conflict by express statement. Here, for example, it is undisputed that Proposition 73’s express prohibition against public financing of contests for elective office renders ineffective all of Proposition 68’s provisions dealing with public campaign financing. Presumably a broader express statement could be used to establish a conflict with each of the provisions of a competing measure. When such a “poison pill” provision is included, voters will understand that only one of the competing measures will become law and will cast their votes accordingly. But when there is no such broad exclusionary statement, and no irreconcilable conflict as to each provision, then the voters’ intent is best determined by a severability analysis.

When one portion of an initiative statute conflicts with the state or federal Constitution, or is invalid for any other reason, the offending part must be stricken from the statute but the remainder can take effect if it is sever-able. (Santa Barbara Sch. Dist. v. Superior Court (1975) 13 Cal.3d 315, 330 [ 118 Cal.Rptr. 637 , 530 P.2d 605 ]; see Calfarm Ins. Co. v. Deukmejian (1989) 48 Cal.3d 805, 821, 835 [ 258 Cal.Rptr. 161 , 771 P.2d 1247 ].) The nonconflicting provisions are presumed to be severable, and this presumption is “fortified by the express statement of a severability clause.” (In re Blaney (1947) 30 Cal.2d 643, 655 [ 184 P.2d 892 ].) The purpose of the severability analysis is to ascertain and give effect to the voters’ intent to the greatest possible extent.

To determine whether a particular provision should take effect, a three-part test is employed: (1) the provision must be mechanically and grammatically severable, (2) it must be functionally severable, and (3) it must appear that the electorate would likely have adopted the provision had it foreseen the partial invalidity of the initiative. (Santa Barbara Sch. Dist. v. Superior Court, supra, 13 Cal.3d at pp. 330-331; Calfarm Ins. Co. v. Deukmejian, supra, 48 Cal.3d at pp. 821-822.) In the case of an initiative measure enacting a complex regulatory scheme, if part of the proposed scheme is so related to the whole that it cannot operate independently, it will fail the functional severability test. And if a provision was presented to the electorate as an indivisible part of the total package, or if the electorate’s attention was not focused upon it, or if it can otherwise be concluded that the voters *783 would not have wanted it to operate independently, then it will fail the voter-intent test of severability.

When an initiative measure is invalid in part, application of the three-part severability test is the established and time-tested way to ascertain and give effect to the voters’ intent to the greatest possible extent. I see no reason why this test, routinely used when part of an initiative is held unconstitutional, should not also be used in the case of simultaneously enacted but conflicting initiative measures.

Accepting the Constitution as it is written, though not as we might prefer it to be, this court should submit the provisions of duly enacted but conflicting initiative measures to a careful and detailed scrutiny. The voters who enacted the measures are entitled to no less. Accordingly, I would examine Proposition 68 provision by provision and would uphold any provision that is both reconcilable with Proposition 73 and severable from the irreconcilable provisions.

Broussard, J., concurred.

Petitioner’s application for a rehearing was denied January 18, 1991. Arabian, J., did not participate therein. Mosk, J., Broussard, J., and Kennard, J., were of the opinion that the application should be granted.

This court recently admonished administrative agencies that when a statute requires the performance of a task, the difficulty of the task does not excuse or exempt the agency from full performance. (Laurel Heights Improvement Assn. v. Regents of University of California (1988) 47 Cal.3d 376, 399 [ 253 Cal. Rptr. 426 , 764 P.2d 278 ].) When the state Constitution requires courts to perform a task, such as a provision-by-provision analysis of a ballot *775 measure, it is equally true that the difficulty of the task is not a legitimate or acceptable excuse for nonperformance.

This distinction appears even more strikingly in the wording of the provision as originally enacted in 1911: “If any provision or provisions of two or more measures, approved by the electors at the same election, conflict, the provision or provisions of the measure receiving the highest affirmative vote shall prevail.” (Cal. Const., former art. IV, § 1; Stats. 1911, ch. 22, § 1, p. 1657.) The change from the former to the present wording, as part of the 1970 constitutional revision, was not intended to alter the provision’s meaning. (See Cal. Const. Revision Com., Proposed Revision (1966) p. 47.)

The power of direct legislation through the initiative was first exercised in California by voters of charter cities. A constitutional amendment permitting home-rule cities to amend their charters through the initiative process had been ratified in 1902. (Key & Crouch, The Initiative and the Referendum in California (1939) p. 428.)

As of this writing, the point raised in the text is of more than academic concern, since a federal district court has ruled that Proposition 73’s contribution limits are invalid as violative of the federal Constitution. (Service Employees v. Fair Political Practices (E.D.Cal. 1990) *779 747 F.Supp. 580 .) The decision is not yet final and an appeal seems likely. The validity of Proposition 73 was not challenged and is not at issue in the proceeding before this court.