OPINION

SILLS, P. J.

California has two statutes requiring that all circulators of referendum petitions be qualified to register to vote in the city whose ordinance is the object of the referendum — Elections Code section 9238, subdivision (c) and section 9209. In this case involving a challenge to signatures on a referendum petition obtained by a group seeking to have a referendum on an ordinance restricting second-story additions, the trial court:

— (a) found that section 9238, subdivision (c) (and, by implication, § 9209 as well) had indeed been violated, but

— (b) did not invalidate the petition, thus allowing the referendum election to go forward.

The anti-referendum group has appealed, arguing that by allowing the referendum to go forward, the trial court "voided" these statutes.

We affirm. It turns out that the trial court merely "voided" statutes that are unconstitutional anyway.

I. BACKGROUND

On July 26, 2006, the San Clemente City Council enacted an ordinance (Ordinance 1319) prohibiting second-story additions in a section of the city known as "Shorecliffs." Opponents of the ordinance — we will call them "the pro-referendum group" — hired a professional signature gathering firm to obtain the signatures of the minimum 10 percent of registered voters in the city necessary to subject the new ordinance to a referendum, which in this case was 3,727 signatures. The gathering firm obtained the necessary number by late August 2006 — obtaining over 3,900 signatures on the referendum petition.

However, the gathering firm employed a peculiar methodology to avoid section 9238, subdivision (c)'s requirement that circulators of a referendum petition be eligible to vote in the city. Specifically, the circulators hired by the signature gathering firm had each petition signer also sign a separate "Declaration of Circulator" portion of the petition, the idea being that each petition signer was his or her own "circulator."

In September, proponents of the anti-second-story ordinance — that is, the opponents of the referendum on the ordinance, we will call them "the anti-referendum group" — filed this action. They sought a writ of mandate directing the city clerk to certify the referendum as insufficient, or, alternatively, directing the city council not to place the referendum on the ballot.

Their main theory for their request for writ of mandate was that section 9209 (read together with § 9022) requires that circulators be residents of the cities whose ordinances are the subject of referendum petitions, and that the circulators used by the pro-referendum group here were not residents of San Clemente. Section 9209, like section 9238, requires a circulator to declare that he or she is a resident of the relevant city; section 9022 requires that circulators be qualified to vote in the state. Section 9238, subdivision (c), which is the focus of the anti-referendum group's appellate argument, is not to be found in the group's trial court petition. That fact is interesting given that in 1999, California's Attorney General concluded that section 9209 is unconstitutional in its restriction of circulators to the relevant city, though the Attorney General's opinion does not mention 9238. (See 82 Ops.Cal.Atty.Gen 250 (1999).) In short, the verified petition was predicated on a statute that had already been opined to be unconstitutional by the Attorney General.

No one in this case contends that sufficient numbers of petition signatures had been gathered by employees of the professional signature gathering company who, by coincidence, also happened to be eligible to vote in San Clemente, so that the referendum could make it onto the ballot on the strength of those signatures alone.

The case was heard in February 2007. The issue of the constitutionality of section 9238, subdivision (c) was not raised. The omission was ironic, since it was the anti-referendum group (represented then by different counsel) which, in their verified petition for a writ of mandate, first cited the United States Supreme Court opinion that must be the basis of any constitutional challenge to section 9238, subdivision (c) (or to § 9209), Buckley v. American Constitutional Law Foundation, Inc. (1999) 525 U.S. 182 [ 142 L.Ed.2d 599, 119 S.Ct. 636]. The pro-referendum group also submitted evidence that city clerks in various areas around the state have routinely allowed signers of petitions to "witness their own signatures by signing again in the `Declaration of Circulator' portion of the petition." (And indeed, in this appeal, both the city and the Orange County Registrar of Voters have filed briefs taking no position on the validity of the process used by the signature gathering company.)

The trial court took the matter under submission, and in mid-April the court ruled in a minute order. (The minute order was not a formal statement of decision, but rather as a series of formal written responses to a joint list of controverted issues.)

In his written responses, the trial judge determined that the process of deeming every signer to be his or her own circulator violated section 9238, subdivision (c). However, the trial judge also ruled that since the city clerk had determined that there were a sufficient number of qualified signatures to put the referendum on the ballot, under Truman v. Royer (1961) 189 Cal.App.2d 240 [ 11 Cal.Rptr. 159], the city clerk was required to certify the referendum. Moreover, the trial court specifically held that, "Assuming a violation of the Elections Code," disqualification of the petition was not the proper remedy. Reasoning that the "requirements" of section 9238, subdivision (c) "are for the benefit and convenience of the clerk," the court concluded that if the clerk found that the "signatures were genuine, then the petition is sufficient."

Section 9209 was not mentioned. (The statute, which had featured so prominently in the verified petition of the anti-referendum group, was not mentioned at all in the group's trial brief.) Nor were any of the statutes (specifically §§ 104 and 9022) that simply provide for circulator declarations in the first place. (Neither of those statutes were mentioned in the anti-referendum group's trial brief either.)

A judgment declaring the referendum petition valid was entered in May. The anti-referendum group promptly filed this appeal, asserting that the trial court's failure to stop the election in effect "voided" section 9238.

II. CONSIDERATION OF CONSTITUTIONALITY FOR THE FIRST TIME ON APPEAL

The anti-referendum group argues that this court should not consider at all the question of the constitutionality of section 9238, subdivision (c), because the issue was not raised by the pro-referendum group at trial. The argument is not well taken.

Appellate courts have discretion to consider the constitutionality of a statute for the first time on appeal when the question involves "`a pure question of law which is presented by undisputed facts.'" ( People v. Hines (1997) 15 Cal.4th 997, 1061 [ 64 Cal.Rptr.2d 594, 938 P.2d 388]; e.g., People v. Blanco (1992) 10 Cal.App.4th 1167, 1173 [ 13 Cal.Rptr.2d 176] [exercising discretion to consider, for the first time on appeal, constitutionality of evidentiary statute involving character trait].)

Indeed, appellate courts "typically have engaged" in the discretionary review of otherwise forfeited arguments when — albeit only when — the "claim involves an important issue of constitutional law or a substantial right." ( In re Sheena K. (2007) 40 Cal.4th 875, 887, fn. 7 [ 55 Cal.Rptr.3d 716, 153 P.3d 282] [collecting authorities in context of discussion of whether condition of probation was unconstitutional on its face]; see Hale v. Morgan (1978) 22 Cal.3d 388, 394 [ 149 Cal.Rptr. 375, 584 P.2d 512] [considering for first time on appeal constitutional validity of statute assessing penalty against landlord for depriving tenant of utility services for purpose of eviction]; e.g., Bayside Timber Co. v. Board of Supervisors (1971) 20 Cal.App.3d 1, 5-6 [ 97 Cal.Rptr. 431] [considering for first time on appeal whether forestry statute was constitutional in light of impact of timber and logging operations on state as a whole]; see also Alfaro v. Terhune (2002) 98 Cal.App.4th 492, 512 [ 120 Cal.Rptr.2d 197] ["claim that including inmates under a sentence of death within" certain legislation was "unconstitutional on its face" was "question of law"]; Thain v. City of Palo Alto (1962) 207 Cal.App.2d 173 [ 24 Cal.Rptr. 515] ["However the question of the validity and constitutionality of the ordinance, on its face, including its notice provisions, is one of law."].)

The anti-referendum group posits three areas, or questions, where consideration of the constitutionality of section 9238, subdivision (c) would "require the Court to decide the issue without crucial evidence," that is, evidence that should have been presented to the trial court, but wasn't. The three questions are: (1) "whether any of the paid signature gatherers . . . were actual residents of San Clemente"; "(2) whether the circulator declaration requirement under EC section 9238(c) imposed any significant burden on the circulators and proponents"; and (3) "whether the governmental interests justifying these statutory requirements are supported by adequate evidence."

On scrutiny, though, none of these questions present any material issues of disputed fact. First, in regard to question (1), as we will see when we examine the controlling case on the point, Buckley v. American Constitutional Law Foundation, Inc., supra, 525 U.S. 182 ( Buckley), it makes no difference as a matter of law whether any portion of the paid signature gatherers were, by coincidence, eligible to vote in San Clemente. The issue of the constitutionality of the circulator statute is independent of the particular city in which any particular signature gatherer might reside. Indeed, we may assume for sake of this case that not a single circulator in the case before us resided in San Clemente. (And, of course, to the degree that, by coincidence, some paid gatherers actually happened to be eligible to vote in San Clemente, the signatures obtained by those gatherers would have to be counted as valid under the theory of the anti-referendum group.)

Second, the same goes for question (2). As we will see when we examine the Buckley decision, the nature of the burden created by a statute such as section 9238, subdivision (c), is also independent of any evidence that might have been submitted to the trial court. Under Evidence Code section 451, subdivision (f) courts must take judicial notice of propositions "so universally known that they cannot reasonably be the subject of dispute." Under Evidence Code section 452, subdivision (g) courts may take judicial notice of "propositions that are of such common knowledge within the territorial jurisdiction of the court that they cannot reasonably be the subject of dispute." As we will show, under Buckley, there is only one proposition necessary to a determination of the constitutionality of section 9238, subdivision (c): That the number of eligible-to-vote residents of any political subdivision in the state of California is only a small fraction — easily less than half — of the number of eligible-to-vote residents of the state as a whole.

Whether that proposition is so universal that it comes within Evidence Code section 451 (must take notice) or merely within Evidence Code section 452 (may take notice) we need not decide. No one can doubt it. Moreover, precedent certainly supports taking notice of it here. California courts have regularly taken judicial notice of such demographic propositions far less obvious and universally known than the proposition that the population of even the largest city or county in California is only a small percentage of the population of the state as a whole. (E.g., People v. Posey (2004) 32 Cal.4th 193, 215, fn. 9 [ 8 Cal.Rptr.3d 551, 82 P.3d 755] [taking judicial notice that a given city was located in a given county, with one telephone area code, while two other counties had another area code]; Sanchez v. City of Modesto (2007) 145 Cal.App.4th 660 [ 51 Cal.Rptr.3d 821] [taking judicial notice of percentage of given ethnic category as a certain percentage of state population as revealed by census figures]; People v. Bhakta (2006) 135 Cal.App.4th 631, 641 [ 37 Cal.Rptr.3d 652] [taking judicial notice of fact that population of Los Angeles exceeded 750,000]; Bennett v. Livermore Unified School Dist. (1987) 193 Cal.App.3d 1012, 1015, fn. 2 [ 238 Cal.Rptr. 819] [taking judicial notice of ethnic categories set forth in Department of Education publication].)

As to the third fact-specific question — whether there might be evidence out there that might have been submitted to the trial court supporting the "governmental interests justifying" the statutory requirements of section 9238 — the point does not preclude consideration of the constitutionality of a statute on its face. In Sheena K. our high court recently contrasted an appellate claim involving some discretionary decision by a trial judge with an appellate claim involving a "facial challenge," based on lack of constitutionality. The court said a facial challenge "does not require scrutiny of individual facts and circumstances but instead requires the review of abstract and generalized legal concepts" and is thus "a task that is well suited to the role of an appellate court." ( In re Sheena K., supra, 40 Cal.4th at p. 885.)

A court can also take the precaution in regard to belatedly raised legal issues of soliciting amicus curiae briefs in order to assure that all sides of an issue are fairly presented. (E.g., Small v. Fritz Companies, Inc. (2003) 30 Cal.4th 167, 194, fn. 3 [ 132 Cal.Rptr.2d 490, 65 P.3d 1255] (cone. opn. of Baxter, J.) [noting that since court had solicited and received supplemental briefs on issue involving efficiency of capital markets, the parties "had reasonable notice and opportunity to brief and argue the issue" and the court could "resolve it in the interest of judicial efficiency"]; In re Smith (1970) 3 Cal.3d 192, 203, fn. 3 [90 Cal.Rptr. 1, 474 P.2d 969] ["When an appellate court, in the course of its normal decision-making process, discovers a crucial matter ignored by appellate counsel which should have been raised as an assignment of error, the court ordinarily will solicit supplemental briefs."]; Sugimoto v. Exportadora de Sal, S.A. de C.V. (1991) 233 Cal.App.3d 165, 167, fn. 1 [ 284 Cal.Rptr. 275] [in light of respondent's failure to file brief, appellate court asked superior court to file brief on issue].)

Thus, in addition to requesting briefing from the parties on the issue, this court, on its own motion, also invited the California Attorney General's Office to give its views on the constitutionality of section 9238, subdivision (c). (As noted, our request for briefing was explicitly without prejudice to a determination that the anti-referendum group was correct and the issue should not be decided in this appeal.)

As it turned out, the Attorney General's office respectfully declined the invitation to address the constitutionality of the statute.

No matter, though, the Attorney General's office has already provided a formal opinion on the constitutionality of the functionally identical section 9209, and found that statute's requirement that a circulator be a "voter of the city" to be unconstitutional in light of Buckley. Said the Attorney General, and the emphasis is his own: "The first question to be resolved is whether the circulator of an initiative petition must be `a voter of the city' as required under section 9209, or whether such statutory requirement is now unconstitutional in light of the United States Supreme Court's recent decision in Buckley v. American Constitutional Law Foundation [citation]. We conclude that the statutory requirement is unconstitutional under Buckley." (82 Ops.Cal.Atty.Gen. at p. 251, original italics.)

Finally, there is one more consideration in favor of exercising discretion to consider the constitutionality issue in this particular appeal: The very nature of appellant's argument is intertwined with the constitutionality of the statute appellant defends. A litigant can hardly say that a statute not otherwise enforced by the trial court must be somehow vindicated if that selfsame statute is unconstitutional. Inherent in appellant's argument is the idea that the trial court was under an obligation to assure that the Legislature had not wasted its time in writing section 9238, subdivision (c) — that there had to be some "teeth" in the statute, and those "teeth" necessarily consist of the invalidation of the petition signatures obtained in violation of the statute. That very argument necessarily implicates the issue of the legality of the statute itself, because the federal Constitution is the supreme law of the land. (E.g., Donaldson v. National Marine, Inc. (2005) 35 Cal.4th 503, 510 [ 25 Cal.Rptr.3d 584, 107 P.3d 254] ["the federal Constitution's supremacy clause makes that law `the supreme Law of the Land' and as binding on the citizens and courts as state laws . . ."].)

It would be, to put it mildly, unseemly for a state intermediate appellate court to articulate, if not invent out of whole cloth, a remedy that itself unconstitutionally burdened First Amendment rights. Federal constitutional law trumps any state policy embodied in the statute. (See Donaldson v. National Marine, Inc., supra, 35 Cal.4th at p. 510-511 [noting that when state statute confers concurrent jurisdiction over federal right of action, federal right can be enforced in state court "regardless of contrary state policy" and further state court "may not refuse to enforce a federal right in the absence of a valid excuse consistent with federal law"].) Appellate courts are not in the business of directing trial courts to violate the Constitution.

We now turn to the merits.

III. THE MERITS A. The Controlling Decisions from the United States Supreme Court 1. Meyer

Back in the mid-1980's, Colorado had a statewide ban on paying people to gather signatures for petitions and referenda. The proponents of a proposed amendment to the Colorado Constitution obtained various approvals regarding the title and summary for the measure, but soon came to the conclusion that they could not obtain the requisite number of signatures without paying signature gatherers. The proponents challenged the ban in federal courts under 42 United States Code section 1983 on the theory that the ban violated their First Amendment rights. In Meyer v. Grant (1988) 486 U.S. 414 [ 100 L.Ed.2d 425, 108 S.Ct. 1886], a unanimous United States Supreme Court agreed with them. The court struck down the ban as unconstitutional, reasoning that the ban "limit[ed] the number of voices" who could "convey" a political message, thus making it "less likely" that the supporters of a given viewpoint would be able to "place the matter on the ballot." ( Id. at p. 423.)

In striking down the legislation, the federal high court specifically rejected the idea that a state's "interest in making sure that an initiative has sufficient grass roots support to be placed on the ballot, or by its interest in protecting the integrity of the initiative process" justified the ban. There were other, less restrictive, ways of implementing those interests than a blanket ban on remuneration for signature gatherers: The "grass roots" interest can be protected by requiring a given number of signatures ( Meyer v. Grant, supra, 486 U.S. at pp. 425-426) while the "integrity" interest is protected by laws making it a felony to obtain false signatures (e.g., by way of forging, paying a signer, or misleading in regard to a petition). Both of those protections, said the court, were "adequate to the task of minimizing the risk of improper conduct in the circulation of a petition." ( Id. at pp. 426-427.) Besides which, the court noted, professional firms who submitted too many unqualifying signatures would quickly gain a reputation in the marketplace as being unreliable. ( Id. at p. 426.)

2. Buckley

One of those new regulatory measures was that petition circulators had to be registered voters residing in the State of Colorado. ( Buckley, supra, 525 U.S. at p. 188, fn. 3.) A public interest foundation oriented in favor of "direct democracy" and a group of several political activists, ranging from the chair of the state Libertarian Party to supporters of initiatives on workplace safety, joined to challenge the requirement in federal court.

As in Meyers, the restriction was found unconstitutional by the United States Supreme Court, albeit this time in what was substantively a six-to-three decision (including a clear five-vote majority opinion), as distinct from a unanimous decision. The court noted that there were about 1.9 million registered voters statewide, and at least .4 million eligible to vote but not registered. ( Buckley, supra, 525 U.S. at p. 193.) "Beyond question," said the high court majority, the "registration requirement drastically reduces the number of persons, both volunteer and paid, available to circulate petitions." ( Ibid.) Elaborating, the registration requirement "decreases the pool of potential circulators as certainly as that pool is decreased by the prohibition of payment to circulators." ( Id. at p. 194.)

The state's "dominant" interest in shrinking the pool of potential petition circulators was in making sure the circulators complied with the law. ( Buckley, supra, 525 U.S. at p. 196.) But that interest could readily be served by a lesser restriction of having a state residency, as distinct from a state registration, requirement. Residency guaranteed that circulators would be within the state's "subpoena service" powers. ( Id. at p. 197.) (The residency aspect of the Colorado's regulation had not been challenged — only the registration aspect.) The net effect of the registration requirement was to "cut[] down the number of message carriers in the ballot-access arena without impelling cause." ( Ibid.)

B. Buckley's Progeny (So Far)

In Buckley, the federal supreme court had struck down a regulation that diminished the pool of potential petition circulators on a statewide basis by as low as (if one does the arithmetic) 18 percent — and called that reduction "drastic."

The reduction in the "number of persons, both volunteer and paid, available to circulate petitions" would, of course, be much more drastic if petition circulators were limited to residents of a given political subdivision within a state. ( Buckley, supra, 525 U.S. at p. 193.) That fact would be noted in quick succession in the year 2000 by three courts, respectively in September, early November and late November: Krislov v. Rednour (7th Cir. 2000) 226 F.3d 851 (Krislov); Lerman v. Board of Elections in the City of New York (2d Cir. 2000) 232 F.3d 135 ( Lerman); and KZPZ Broadcasting v. Black Canyon Citizens (Ct.App. 2000) 199 Ariz. 30 [ 13 P.3d 772] ( KZPZ).

Krislov involved, among other things, an Illinois requirement that signature gatherers be registered voters in the relevant political subdivision. As in Meyer, the Krislov court held that the legitimate state interest in assuring a "minimum level of support" in the relevant political subdivision was adequately served by the minimum signature requirement. ( Krislov, supra, 226 F.3d at p. 863.) The court noted that a candidate who had "reasonable support in his district might be denied the possibility of being placed on the ballot simply because his supporters who are registered voters and who reside in the relevant political subdivision may not have the time or the energy to solicit signatures." ( Id. at p. 864.) And as to the state's (obviously "compelling") interest in the "integrity" of the election process, it was served by the less restrictive means, including the opportunity to challenge fraudulent signatures. ( Id. at p. 865.) The court observed that a resident circulator was no less likely to exclude invalid signatures than a nonresident. ( Ibid.)

Lerman presented only a slight variation of the issue of restricting circulators to political subdivisions in which they reside. The State of New York had a requirement that a witness to signatures on a city council candidate's petition had to be a resident of the political subdivision in which the office or petition had to be voted for ( Lerman, supra, 232 F.3d at p. 138). If one thinks about it, such a requirement is functionally the equivalent of a requirement that petition circulators be residents of the political subdivision, since it would be grossly inefficient for any campaign to have one person circulate a petition and another person witness it. The law was challenged by a "witness" (circulator) who lived in the 49th Council District who gathered signatures for a candidate in the 50th Council District, because 58 signatures she had witnessed had been ruled invalid. ( Id. at p. 139.) The Second Circuit, directly following Buckley, struck down the requirement, reasoning that the two cases could not be distinguished — certainly not on the basis that candidate petitions are distinct from ballot measure petitions. "There is no basis to conclude," said the Lerman court, "that petition circulation on behalf of a candidate involves any less interactive political speech than petition circulation on behalf of a proposed ballot initiative — the nature of the regulated activity is identical in each instance." ( Id. at p. 148.)

Turning to various state interests supposedly justifying the political subdivision restriction, the court noted that the restriction did nothing to ensure what one might call the "integrity" interest. A witness to a petition was still "answerable to a subpoena." ( Lerman, supra, 232 F.3d at p. 150.)

The court also considered the "`modicum of support'" rationale for supporting the restriction to represent an "even weaker" state interest than what one might call the "integrity" interest. Echoing Meyer, the Lerman court reasoned that the modicum of support interest is already "advanced by the requirement that candidates obtain a minimum number of signatures from district residents." ( Lerman, supra, 232 F.3d at p. 151.)

Lerman further addressed an interesting twist on the modicum of local support rationale: The theory was that allowing out-of-subdivision petition witnesses (circulators) would have the effect of "`imposing'" the costs of an election on the subdivision's residents that otherwise might not have taken place. (See Lerman, supra, 232 F.3d at p. 152.)

The imposition-of-cost-from-outside rationale — a sort of political version of trade protectionism — drew particular fire from the Lerman court as directly contrary to the First Amendment. Said the court: "[T]o the extent that the defendants mean to argue that the witness residence requirement helps to prevent non-residents from influencing politics within the district, that interest does not appear to be legitimate at all. A desire to fence out non-residents' political speech — and to prevent both residents and non-residents from associating for political purposes across district boundaries — simply cannot be reconciled with the First Amendment's purpose of ensuring `the widest possible dissemination of information from diverse and antagonistic sources.'" ( Lerman, supra, 232 F.3d at p. 152.)

Finally, in late November 2000 the Arizona Court of Appeals decided KZPZ, supra, 13 P.3d 772, which perhaps encapsulated most elegantly the application of Buckley to the idea of political subdivision restrictions on circulators. KZPZ is the precedent most like the case before us, because, like here, the case involved an attempt to use the referendum process to overturn a local land use decision and a subsequent attack by proponents of the decision to prevent the referendum from going on the ballot because of the use of out-of-subdivision petition circulators. (Except in KZPZ, the land use decision was permissive, allowing a radio station to erect seven radio towers in a city, and the pro-referendum party wanted the permissive ordinance to be the subject of a referendum. In our case, the land use ordinance is restrictive, and the pro-referendum group wants the restrictive ordinance to be the subject of a referendum.)

Arizona had a statutory requirement which could be read to require that a petition circulator had to be a resident of the relevant political subdivision where a referendum petition was to be filed, and the radio station sued to enjoin placement of the referendum on the ballot on the theory of the invalidity of the signatures "collected by the nonresident circulators." ( KZPZ, supra, 13 P.3d at pp. 774, 776.) Much of the opinion is devoted to a statutory analysis of Arizona law independent of the constitutional question (see id. at pp. 776-778), with the court noting that the anti-referendum radio station had made "a convincing argument that the literal wording of the . . . statutes, read in para materia, support[ed] the trial court's conclusion that, in order to qualify to circulate petitions for a local referendum, the circulator must be a qualified elector of the locality." ( Id. at p. 777.)

But that "literal wording" was not the end of the matter. Like this appellate court, the Arizona appellate court was not particularly enthusiastic about enforcing a potentially unconstitutional statute. As we have in this opinion, the KZPZ court recognized that "the constitutional issue is inextricably intertwined with our statutory interpretation." It thus found "discussion" of the issue "essential." ( KZPZ, supra, 13 P.3d at p. 778.)

The KZPZ court applied Buckley to the facts at hand, noting that a county residency requirement decreased the "pool of available petition circulators" from the total number of qualified electors in Arizona — about 2.3 million — to about 88,000, the actual number of qualified electors in the county. ( KZPZ, supra, 13 P.3d at p. 779.) That, said the court, was a "heavy burden on political expression regarding referendum issues in our smaller counties." ( Ibid.)

The radio station countered by arguing that the restriction of circulators to the residents of a political subdivision furthered the "compelling" interest of assuring that circulators had "ties to the community or stake in the outcome of a referendum election." ( KZPZ, supra, 13 P.3d at p. 779.) That is, it was a kind of anticarpetbagger argument, based on the idea that, as the radio station contended, the subdivision residency requirement "prevents `abuse and capture of the local referendum process by political outsiders.'" ( Ibid.)

The idea was rejected because, after all, "only county residents" could vote in the actual election. ( KZPZ, supra, 13 P.3d at p. 779.) By the same token, the imposition-of-cost rationale that Lerman had rejected was again rejected in KZPZ, because the ultimate control of whether there would be an election did not rest with the circulators of petitions, but with "the county residents who sign the petition." ( Ibid.) The net result was that the KZPZ court construed the problematic Arizona statutory scheme not to "include a local residency requirement" in order not to "run afoul of the First and Fourteenth Amendments under the principles set forth in Buckley." ( Id. at p. 780.)

C. Contrary Authority? 1. Cases Post-Buckley

We are aware of no case since Buckley and its quickly gestated progeny of Krislov, Lerman and KZPZ that has upheld a requirement of circulator residency in a given political subdivision. Restrictions to political subdivisions necessarily, after all, entail even more drastic reductions in the pool of potential circulators (defining that "pool" by the voting age residents of a state) than statewide restrictions to registered voters.

The anti-referendum group cites a number of post- Buckley cases as upholding "residency requirements for petition circulators," but fails to tell us that these cases uniformly involved statewide residency requirements in the context of statewide ballot propositions. (See Initiative Referendum Institute v. Jaeger (8th Cir. 2001) 241 F.3d 614, 615 ["The appellants challenge the requirement that all those who circulate petitions must be North Dakota residents. . . ."]; Idaho Coalition United for Bears v. Cenarrusa (D. Idaho 2001) 234 F.Supp.2d 1159, 1163 [evaluating requirement that "any person who circulates a petition for an initiative must be a resident of the state of Idaho . . ."]; Yes on Term Limits, Inc. v. Savage (W.D. Okla. 2007) 2007 WL 2670178 at p. *4 [evaluating statute requiring "the petition circulator to verify that he or she is a `qualified elector of the State of Oklahoma'"]; Kean v. Clark (S.D. Miss. 1999) 56 F.Supp.2d 719, 725 ["ultimate issue" was "whether the requirement that petition circulators for ballot initiatives must be residents of the State of Mississippi"]; see also Hart v. Secretary of State (1998) 1998 ME 189 [ 715 A.2d 165, 167] [pre- Buckley case considering challenge to state constitutional requirement "requiring circulators to be residents" of state]; cf. Maine Taxpayers Action Network v. Secretary of State (2002) 2002 ME 64 [ 795 A.2d 75, 78, fn. 6] [refusing to address challenge to "Maine's requirement that circulators be registered voters and residents of Maine"].) Whether these cases are ones in which ideologically motivated political activists crossed state lines to promote nationally significant causes by ballot propositions in those states, or merely reflect a tendency of petition circulators to be, in essence, migrant workers, is beside the point. None of these cases involve state residents barred from acting as circulators because they do not reside in a given political subdivision within their state.

Buckley, of course, never addressed the issue of a statewide residency requirement — one must remember that the shrinkage in the potential pool of circulators in that case was the product of using a statewide registration requirement as distinct from a statewide residency requirement. And in fact Buckley, in striking down a statewide registration requirement, explicitly assumed that a statewide residency requirement would "be upheld as a needful integrity-policing measure." ( Buckley, supra, 525 U.S. at p. 197.)

There is, of course, a big difference between the two requirements — a residency requirement assures that circulators are subject to the subpoena power of the state, and thus can be held accountable for their actions even if they don't register to vote. (Cf. In re Initiative Petition No. 379 (2006) 2006 OK 89 [ 155 P.3d 32, 39-40] [emphasizing need for in-state residency requirement to protect "integrity of the circulation process"].) A registration requirement merely duplicates the benefit achieved by the residency requirement (insuring that circulators remain easily subject to the state's subpoena powers), but has the unfortunate side effect of needlessly reducing the pool of potential circulators.

Thus it is safe to say that no post- Buckley case authority to date supports the constitutionality of section 9238, subdivision (c).

2. The Post-Buckley 2007 Amendment to Section 9238, Subdivision (c)

The anti-referendum group also presents a statutory variation on the constitutionality of residency (as distinct from registration) requirements, which must fail for the same reason. The group points to 2001 Senate Bill No. 904 (2001-2002 Reg. Sess.), which amended section 9238, subdivision (c) in the wake of Buckley, and to the fact that the Legislative Counsel's Office opined that, after the amendment, "the law is constitutional." (The quotes are from the anti-referendum group's briefing, not from the Legislative Counsel.) From that the anti-referendum group invokes a kind of argument from authority to the effect that since the Legislative Counsel's office has found section 9238, subdivision (c) constitutional, so should this court.

But that is not a correct characterization of what the Legislative Counsel concluded, as shown by the changes in section 9238, subdivision (c) actually wrought by 2001 Senate Bill No. 904 (2001-2002 Reg. Sess.), and recognized in the Legislative Counsel's digest of the bill.

Specifically, section 9238 was originally enacted as part of a comprehensive revision of the Elections Code in 1994. (Stats. 1994, ch. 920, p. 4690.) As enacted in 1994, section 9238, subdivision (c) required that the circulator be a voter of the city. Mere eligibility to vote wasn't good enough.

After the Buckley decision, requirements that circulators be registered to vote, as distinct from merely eligible to vote, became untenable. Section 9238 had to be amended to avoid that particular problem. The Legislative Counsel's Digest of Senate Bill No. 904 (2001-2002 Reg. Sess.) thus implicitly recognized the impetus of the bill was to broaden the pool of circulators to all those eligible to vote. That digest provided: "Existing law requires that persons who circulate state, county, municipal, or district initiative and referendum petitions be registered voters within the jurisdiction of the governmental entity to which the initiative or referendum measure would apply. [¶] This bill would alternatively allow these petition circulators to be eligible to register to vote within the jurisdiction to which the initiative or referendum measure would apply." (Stats 2001, ch. 105, § 1, italics added.) And that is precisely what Senate Bill No. 904 (2001-2002 Reg. Sess.) did: It amended section 9238, subdivision (c) to add language that being "qualified to register as a voter" of the city was sufficient. (Stats. 2001, ch. 105, § 6.)

And, of course the change in the law in the direction of expanding the pool of circulators from registered voters to persons eligible to register was constitutional — as far as it went — but that was all the Legislative Counsel concluded. The Legislative Counsel's office did not opine on the constitutionality of restricting the pool of potential petition circulators to residents of a given political subdivision. The office only addressed "the bill" at hand.

3. pre-Buckley: The Browne Decision a. What Browne Said

The city clerk rejected about 1,100 signatures gathered by the professional circulators, and the prosmoking parties, now effectively the pro-referendum party, filed a petition for a writ attacking the city's election ordinance as violative of free speech. The trial court determined that the ordinance was indeed unconstitutional, and ordered that no signatures were to be disqualified on the basis of the city's ordinance. Antismoking parties intervened and appealed from the trial court's order, prompting the Court of Appeal to issue a writ of supersedeas. Meanwhile, the Governor signed a bill prohibiting smoking in restaurants on a statewide basis, making the case moot. The appellate court, however, decided that the issue would "arise again" and exercised its discretion "to resolve issues of continuing public interest." ( Browne, supra, 27 Cal.App.4th at p. 1122.) The court ultimately determined that the ordinance was constitutional.

Structurally, the Browne court's analysis of the constitutional issue essentially consisted of the seriatim batting down of the losing side's arguments. The first argument was that the circulator restriction "imposed a substantial burden on their free speech rights." ( Browne, supra, 27 Cal.App.4th at p. 1123.) The court basically said that (a) it was the pro-referendum's party's burden of production of evidence to establish that the circulator restriction had "burdened" its free speech rights, and (b) the pro-referendum party's evidence fell "short" of meeting their burden of production. ( Ibid.) Rather, all the pro-referendum party had shown is that the professional signature gathering company had a "policy" that the local ordinance was to be followed by its signature, and that it hadn't been. That evidence didn't prove that the ordinance was a substantial burden. The dispositive phrase in that section of the opinion is "there is no support for the claim that respondents did all they could have done" ( id. at p. 1123, italics added), suggesting that the Browne court thought the signature gathering company should simply have tried harder to find circulators who were city voters. (See generally id. at pp. 1123-1124.)

Next, the court distinguished the 1988 Meyer opinion, on which the pro-referendum party had "rel[ied] heavily." ( Browne, supra, 27 Cal.App.4th at p. 1124.) The Browne court recognized that the United States Supreme Court had reasoned that the ban on paid circulators in Meyer made the pool of circulators "`necessarily smaller,'" but (counterintuitively) juxtaposed the reduction of the pool in Meyer with the absence of a reduction in the pool in the case before it. ( Ibid.) Specifically, the Browne court reasoned that it was dealing with a "decidedly different" situation because the pro-referendum party had neither argued nor shown "that people who live and vote outside the City were more willing to circulate their petitions than those who live and vote within the City." ( Ibid.) Even more specifically, the court noted that there was a pool of about 2.3 million potential circulators in the city of Los Angeles. So the court returned to the theme that the pre-referendum party simply had not tried hard enough to recruit from that pool. The court said that the pro-referendum party's "supporting documentation does not establish that they tried but were unable to find circulators among those 2,300,000 members of the City's population." ( Ibid.)

Next the Browne court turned its attention to a variation on the issue of the reduction-of-pool problem which it had touched on in distinguishing Meyer. The pro-referendum party had argued that the circulator restriction "burdens those who live outside the City and wish to circulate a petition." ( Browne, supra, 27 Cal.App.4th at p. 1124.)

We pause to note something at this point: The fact that the pro-referendum party's argument was framed in terms of a burden on outsiders makes sense when one realizes the context of the Browne case: The pro-referendum party in Browne was partly financed by tobacco companies and other large corporations. (See Browne, supra, 27 Cal.App.4th at p. 1120, fn. 1 [listing major contributors].) It is a fair inference that the impetus for the referendum came fundamentally from "outside" corporate interests, bent on holding the line against nonsmoking ordinances in any large California city.

But we should emphasize: The Browne court did not at all address the circulator restriction in terms of whether truly indigenous inside interests were also being prejudiced by the reduction in the potential circulator pool. The opinion did not consider the problem from the point of view of the local restaurant owner who was, after all, the named plaintiff in the case — it was as if the court focused only on those it considered the "real" proponents of the referendum, the tobacco companies. The Browne court thus saw the issue in terms of moneyed outsiders barging into a given political subdivision and asserting a right qua outsiders to circulate a referendum petition. (See generally Browne, supra, 27 Cal.App.4th at pp. 1124-1125.)

And, having framed the case as one of insiders versus outsiders, the Browne court found the position of the outsiders unpersuasive: The court reasoned that the outsiders were required to cite other authority than Meyer for the proposition that Los Angeles "ha[d] a duty to offer nonresidents the opportunity to circulate petitions which concern a City ordinance" ( Browne, supra, 27 Cal.App.4th at p. 1125) since Meyer didn't stand for that idea.

It was in that context — confronting the idea that the city was required to "offer" outsiders the "opportunity to circulate petitions" that the Browne court came closest to confronting the issue of whether there was a governmental interest that would be sufficiently strong enough to justify whatever burden the circulator restriction did impose — and remember that just five paragraphs previously the court had said that the burden was not "substantial." (See Browne, supra, 27 Cal.App.4th at p. 1125.) The court weighed the effect of an ordinance — which it had already said about a paragraph prior did not impose a substantial burden on free speech rights — against the interest served by the circulator restriction ordinance, which it, in passing, deemed "compelling." Having weighted the scales beforehand, it was no surprise that the court found the position of the pro-referendum party to be wanting.

Because the passage is perhaps the most substantively important one in the Browne opinion for purposes of our present analysis, we now quote it in full: "We therefore hold that respondents have not established that section 339 restricted their right of political expression. Moreover, we hold that the City has compelling interests which are protected by section 339. The electorate of this state adopted the initiative and referendum methods to have access to and control of a legislative process because it believed that the process could otherwise be dominated by special interests. ( Citizens Against Rent Control v. City of Berkeley (1980) 27 Cal.3d 819, 825 [ 167 Cal.Rptr. 84, 614 P.2d 742], reviewed on other grounds sub nom. Citizens Against Rent Control v. Berkeley (1981) 454 U.S. 290 [ 70 L.Ed.2d 492, 102 S.Ct. 434].) `[T]he initiative and referendum processes can themselves be employed by the special interest groups whose power they were designed to curb.' ( Citizens Against Rent Control v. City of Berkeley, supra, 27 Cal.3d at p. 826.) We can think of no better way to preserve the integrity of the referendum process as it relates to local ordinances than to require that only those eligible to vote for the members of the body which passed the ordinance be able to engage in the `core political speech' of attempting to gather signatures to challenge the ordinance. ( Meyer v. Grant, supra, 486 U.S. at p. 422 [ 100 L.Ed.2d at p. 435]; see Lowing v. Faull (1964) 227 Cal.App.2d 23, 29 [ 38 Cal.Rptr. 417].) Having given section 339 the strict scrutiny required, we conclude that the trial court erred in granting respondents' petition for a writ of mandate." ( Browne, supra, 27 Cal.App.4th at p. 1125, fn. omitted.)

b. Why We Respectfully Decline to Follow Browne

The Arizona Court of Appeals in KZPZ would opine that Browne was "questionable law after Buckley." ( KZPZ, supra, 13 P.3d at p. 780.) We are forced to agree — the decision is, in light of Buckley and its subsequent precedent in the context of political subdivision restrictions, untenable.

At the most basic level, Browne decides a question of federal law in a manner that cannot be reconciled with the authoritative federal decisions in the area. Buckley applies a fortiori to section 9238, subdivision (c): If a statewide reduction in the potential pool of potential petition circulators of about 18 percent was too "drastic" and therefore unconstitutional in Buckley, how much more drastic — and hence even more unconstitutional — is a reduction of the pool of potential circulators to only a tiny fraction of the state's residents. Even Los Angeles County — the largest political subdivision in California — makes up less than 30 percent of the state's population as a whole. The reduction in the pool would be 72 percent. In San Clemente, the effect of section 9238, subdivision (c) is to reduce the pool of potential circulators (from the pool of state residents eligible to vote) by more than 99 percent. Thus the anti-referendum group's arguments that section 9238 does not pose a "severe" burden on First Amendment rights, or that the ascertainment of the burden imposed is necessarily a "fact-specific issue" is, under Buckley, incorrect. Indeed, at oral argument, counsel for the anti-referendum group essentially echoed the Browne rationale, when he argued that the pro-referendum group simply should have tried harder to find qualified circulators among the residents of San Clemente.

(In that regard, the Attorney General's 1999 opinion finding § 9209 — the statute functionally identical to § 9238, subd. (c) — unconstitutional basically quoted large swaths of the Buckley opinion. It did not discuss Browne directly. It was as if the passages from Buckley self-evidently precluded any argument that § 9209 was constitutional, and Browne could be serenely ignored. (See 82 Ops.Cal.Atty.Gen., supra, at pp. 252-254.))

Next we consider the main state interest which the Browne court thought to be sufficient to justify the restriction of circulators to a given political subdivision. That interest often is described in slightly different terms in the case law. Meyers called it the "grass roots" interest, Krislov called it the "minimum support" interest, Lerman called it the "`modicum of support'" interest, and KZPZ described it as the "stake in the outcome" interest. The anti-referendum group in its supplemental briefing here refers to it as the statutory purpose to "maintain proper local control over policymaking and elections." But whatever it is called, this "grass roots" interest was authoritatively held by the Buckley court to be sufficiently served by the lesser measure of requiring a certain number of signatures from actual voters, as distinct from the more intrusive measure of reducing the pool of potential circulators who could bring those voters a given message.

Browne's invocation of higher authority also does not fare well in retrospect. As is evident from the long passage we quoted above, the main authority for Browne invoking the spectre of domination of local politics by "special interests" was Citizens Against Rent Control v. City of Berkeley (1980) 27 Cal.3d 819 [ 167 Cal.Rptr. 84, 614 P.2d 742] ( Citizens Against I).

The Brown court acknowledged that something had happened to Citizens Against I, but it wasn't accurate in characterizing that something. The Browne court's citation to the Citizens Against I opinion was that it was: "reviewed on other grounds sub nom." (some italics added) by the United States Supreme Court in Citizens Against Rent Control v. Berkeley (1981) 454 U.S. 290 [ 70 L.Ed.2d 492, 102 S.Ct. 434] ( Citizens Against II).

But Citizens Against I was not "reviewed" by Citizens Against II. It was flatout reversed. And the "grounds" on which it was reversed were not "other grounds" separable from those passages in Citizens Against I on which the Browne court relied. Rather, those passages — whose main theme we might describe as the "danger of special interests as related to the initiative and referendum process" — were part and parcel of the rationale of the Citizens Against I opinion that did not find favor in the federal high court in Citizens Against II. In Citizens Against II the United States Supreme Court reasoned that the city's interest in protection from domination by moneyed "special interests" was already adequately served by the lesser intrusive means of identification of contributors in public filings, together with the possible preclusion of anonymous contributions. ( Citizens Against II, 454 U.S. at pp. 299-300.)

In that regard, a unanimous Supreme Court in Meyer had already relegated to a footnote the argument that "it is permissible to mute the voices of those who can afford to pay petition circulators." ( Meyer, supra, 486 U.S. at p. 426, fn. 7, italics added.) That is, the federal high court did not think much of discriminating against causes supported by those who could afford to pay petition circulators as distinct from manning tables themselves. The Meyer court rejected the idea of muting the voices of such "special interests" precisely because the voters would ultimately be "`entrusted with the responsibility for judging and evaluating the relative merits of conflicting arguments.'" ( Ibid., quoting First National Bank of Boston v. Bellotti (1978) 435 U.S. 765, 790-791 [ 55 L.Ed.2d 707, 98 S.Ct. 1407].) And yet it was the wariness of moneyed special interests that animated the Browne court more than anything else.

Looking at the Browne analysis point by point also shows its inconsistency with Buckley. The Browne court operated on the assumption that the party challenging the ordinance had the evidentiary burden of establishing that it had tried very hard to find circulators in the relevant political subdivision and still couldn't find enough "local" circulators, and that it had failed to carry that burden simply by showing the shrinkage inherent in restricting circulators to that subdivision. For Buckley, by contrast, it was enough to show that the pool of circulators statewide had been reduced by around 18 percent.

Similarly, Browne's distinguishing of Meyer on the basis that the pro-petition party had failed to argue or show that the outsiders were "more willing" than locals to work as circulators fails because both Meyer and Buckley came to their conclusions based on the fact of a shrinkage of the pool of potential circulators, as distinct from the particular willingness of members of the available pool to work for a given cause.

Finally, there is another, perhaps even more basic flaw, in Browne's, reliance on the prevention-of-domination-by-outsiders rationale to support a restriction on petition circulators. Browne did not consider whether genuine "insider" and "grass roots" non-special interests might themselves need to employ professional, that is, "outsider," circulators. It takes little imagination to realize that circulator restrictions may actually hinder local grass roots petition efforts. Not every local cause is supported by people who have the free time and energy to devote to manning tables at large shopping centers for hours on end. (See Krislov, supra, 226 F.3d at p. 864 ["a candidate who has reasonable support in his district might be denied the possibility of being placed on the ballot simply because his supporters who are registered voters and who reside in the relevant political subdivision may not have the time or the energy to solicit signatures . . ."].) A restriction on outside circulators necessarily discriminates against some local causes in favor of others.

Indeed, the case before us is a perfect example of a how a circulator restriction can work against local interests as such. Here, the issue to be put to the voters is very local — a land use regulation affecting only a particular section of a particular city. The battle is, literally, between neighbors. One set of neighbors simply found it necessary to use a professional signature gathering company because, in light of the city council's action, the only way to overturn a land use decision was to have a referendum on it. No large tobacco companies are parties to this case.

D. Miscellany

We need not address directly the issue of the degree to which the anti-referendum group, in having petition signers also sign as their own witnesses in a declaration of circulator box, contravened sections 104 and 9022 of the Elections Code. Since this case comes to us after a court trial, all conflicts and inferences in the evidence are construed in favor of the judgment (in this case, to allow the referendum to go forward). In that regard, the pro-referendum group presented substantial evidence, in the form of declarations from the heads of two signature gathering companies, that city clerks around the state routinely accept petitions where signers act as their own witnesses and sign as their own circulators. (That correlates with the position we have noted that neither the City of San Clemente nor the Orange County Registrar of Voters have voiced any opinion on the methodology of having signers also attest as witnesses their own signatures in a separate declaration of circulator.)

In Assembly v. Deukmejian (1982) 30 Cal.3d 638, 651-652 [ 180 Cal.Rptr. 297, 639 P.2d 939], our high court pointed out that for a period of three years the Secretary of State's California Ballot Initiative Handbook used an incorrect phrase in a recommended sample format for initiative petitions, which meant that referendum petitions didn't have to "comply with the requirements" of a certain statute (one that required signers to affix their residence address as distinct from their address as registered to vote). As a result, neither the Secretary of State nor "the county clerks" had ever refused to accept a tendered petition on the basis of that defect. ( Id. at p. 651.) Nor had the practice ever been subjected to legal challenge prior to that case. ( Ibid.)

Thus the proponents of a referendum in that case "relied on a practice that not only had been accepted by the government entities charged with enforcing the referendum procedures but also had never been subjected to a challenge from any source." ( Assembly v. Deukmejian, supra, 30 Cal.3d at p. 651.) In light of the "unusual and unique circumstances" of the case, the Supreme Court concluded that the "failure to comply" with the certain statute would not "be deemed to render the referendum petitions invalid." ( Id. at p. 652.)

We have the same two circumstances here as were present in Assembly v. Deukmejian — reliance as a factual matter found by the trial court as "on a practice" that has been accepted by the relevant governmental entities (many city clerks' offices) and, up to this case, never subjected to challenge in litigation. But on top of those, we have two things that make this case perhaps ever rarer than Assembly v. Deukmejian: The first is that the process that was used (however it is described and, to repeat, we express no opinion on its validity or invalidity) was done in direct response to what was unconstitutional about section 9238, subdivision (c). The only reason, in this case, that the pro-referendum group had signers also witness their own signatures as "circulators" was because section 9238, subdivision (c) had unconstitutionally restricted the pool of circulators to a tiny fraction of what it constitutionally should have been.

The second is there already was relevant case law on the precise topic of circulators not being residents of the local political subdivision — Truman v. Royer, supra, 189 Cal.App.2d 240. That case buttressed the reliance by the signature gathers on the acquiescence of the clerk's offices in the "self-witnessing" procedure. As the trial court noted, in Truman the fact that the affidavits of the circulators were defective do not invalidate the referendum petition when there was no question (as also in the case before us) that the petition had in fact been signed by the required number of qualified voters.

Under such circumstances (reliance, absence of previous challenge, a practice in precise response to an unconstitutional restriction and support in existing case law), we follow the Supreme Court's lead in concluding that, even assuming, for sake of argument, that there was an arguable failure to comply with the circulator declaration statutes, any such failure, as in Assembly v. Deukmejian, should "not be deemed to render the referendum petitions invalid." ( Assembly v. Deukmejian, supra, 30 Cal.3d at p. 652; accord, Costa v. Superior Court (2006) 37 Cal.4th 986, 1028-1030 [ 39 Cal.Rptr.3d 470, 128 P.3d 675].)

IV. DISPOSITION

The judgment is affirmed. The earlier stay issued by this court is dissolved. Respondents are to recover their costs on appeal.

Rylaarsdam, J., and Moore, J., concurred.