San Bernardino County Bd. of Supervisors v. Monell ( 2023 )


Menu:
  • Filed 6/16/23 (unmodified opn. attached)
    CERTIFIED FOR PARTIAL PUBLICATION
    IN THE COURT OF APPEAL OF THE STATE OF CALIFORNIA
    FOURTH APPELLATE DISTRICT
    DIVISION TWO
    SAN BERNARDINO COUNTY BOARD
    OF SUPERVISORS,                                        E077772
    Plaintiff and Appellant;                      (Super.Ct.No. CIVSB2025319)
    v.                                                     ORDER DENYING PETITION
    FOR REHEARING AND
    LYNNA MONELL, as Clerk, etc.,                          MODIFYING OPINION
    Defendant and Respondent;                     [NO CHANGE IN JUDGMENT]
    NADIA RENNER,
    Intervener and Appellant.
    THE COURT
    The opinion filed in this matter on May 25, 2023 is modified as follows.
    1. On page 2, the last sentence in the first full paragraph:
    At the same time, the voters also elected three new supervisors.
    is deleted and replaced with:
    At the same time, a new supervisor was elected, and two more new
    supervisors, elected in March 2020, were waiting to take office.
    2. On page 5, the first sentence in the third full paragraph:
    1
    At the same time, new supervisors Joe Baca, Jr., Col. Paul Cook, and Dawn
    Rowe were elected to the Board.
    is deleted and replaced with:
    At the same time, new supervisor Joe Baca, Jr. was elected to the Board;
    two more new supervisors, Col. Paul Cook, and Dawn Rowe, had been
    elected in March 2020 and were waiting to take office.
    3. On page 9, the third and fourth sentences in the second full paragraph:
    So far, nothing has happened in that case, other than the filing of the Civil
    Case Information Statement. In the ordinary course of business, it is likely
    to be a year or so before we decide it — to say nothing of how much time
    any subsequent Supreme Court review will consume.
    are deleted and replaced with:
    Nothing much had happened in that case until after the oral argument in
    this case, when the Board requested, and we granted, calendar preference.
    Even with calendar preference, however, in the ordinary course of business,
    it is likely to be six months or more before we decide it — to say nothing of
    how much time any subsequent Supreme Court review will consume.
    4. On page 10, the second sentence in the first full paragraph:
    There is not even any pending request for a stay; before oral argument, the
    Board requested a stay, but we denied it.
    is deleted and replaced with:
    There is not even any pending motion for a stay. The Board did ask us to
    take the oral argument off calendar, pending the resolution of the Measure
    D litigation, but we have already denied that request.
    5. On page 15, in the second full paragraph, delete point (2):
    (2) they allowed incumbency “for a number of years” and did not preclude
    an incumbent from running for a different office,
    And replace it with:
    2
    (2) they allowed incumbency “for a number of years” and did not preclude
    an incumbent from running for a different state office,
    6. On pages 15-16, after the first sentence of the paragraph, the remaining sentences:
    Those limits, however, while they were not one-term limits, allowed a state
    senator to serve for no more than eight years and a state assembly member
    to serve for no more than six years. The difference between the six or eight
    years there and the four years here is not sufficient to be constitutionally
    significant — particularly when the term limits here are similarly neutral
    and nondiscriminatory and do not preclude an incumbent from holding any
    other office. Four years is ample time for a supervisor to at least attempt to
    tick off all the boxes on his or her legislative to-do list. In this respect, the
    Board, with only five members, is very different from the California
    Senate, with 40 members, or the California Assembly, with 80 members.
    In the latter bodies, seniority and a cursus honorum of committee
    memberships both play a role. By contrast, a newly elected supervisor can
    hit the ground running.4 The Board does not point to anything that a
    supervisor could accomplish in six or eight years but not in four.
    and the appurtenant footnote:
    4       Amicus curiae Howard Jarvis Taxpayers Association takes a swipe
    at the trial court for “weighing the merits of an ideal learning curve for a
    county supervisor.” We find that we cannot help but consider this subject
    in deciding whether the term limits here allow a supervisor to serve “a
    significant period in office” within the meaning of Eu or, more generally,
    unduly injure the rights of voters and candidates.
    are deleted and replaced with:
    In general, however, a restriction is not “severe” within the meaning of the
    Anderson-Burdick test if it is “‘generally applicable, even-handed,
    politically neutral, and . . . protect[s] the reliability and integrity of the
    election process. [Citation.] . . .’ [Citation.]” (Rawls v. Zamora, supra,
    107 Cal.App.4th at p. 1116.) The one-term limit here passes these criteria
    with flying colors. While Bates also threw into the mix the stringency of
    the effect of the restriction, we believe that is more appropriately
    considered when weighing the benefits of the restriction against the burden
    on voters’ and candidates’ rights, as we do in part IV.C.2, post. (See Eu,
    supra, 54 Cal.3d at p. 518.)
    7. On page 25, the last full paragraph:
    3
    Second, the term limits do allow an incumbent to serve “a significant
    period in office.” (Eu, supra, 54 Cal.3d at p. 518.) As already discussed,
    we cannot meaningfully distinguish a four-year maximum from the six-year
    maximum applicable to assemblymembers in Eu.6 (See part III.B.2, ante.)
    and the appurtenant footnote:
    6      In conformity with Government Code section 25000, subdivision
    (a), which sets a supervisor’s term at four years, the drafters of Measure K
    did not attempt to prescribe a five, six, or seven-year term.
    are deleted and replaced with:
    Second, the term limits do allow an incumbent to serve “a significant
    period in office.” (Eu, supra, 54 Cal.3d at p. 518.) The term limits in Eu,
    while they were not one-term limits, allowed a state senator to serve for no
    more than eight years and a state assembly member to serve for no more
    than six years. The difference between the six or eight years there and the
    four years here is not sufficient to be constitutionally significant.
    The Board argues that in Eu, a termed-out assemblymember could still run
    for and serve in the Senate (or vice versa), and thus could serve a total of
    fourteen years in the Legislature. Nevertheless, these are separate
    constitutional offices. Eu specifically determined that the six-year and
    eight-year terms — considered separately, and not as a total fourteen-year
    span — each constituted “a significant period in office.” (Eu, supra, 54
    Cal.3d at p. 518.)
    The Board also cites testimony that it takes a supervisor “several years” to
    learn the job, and that a one-term limit “will impact County operations in an
    extremely negative and harmful way.” Government Code section 25000,
    subdivision (a) sets a supervisor’s term at four years. Every newly elected
    supervisor faces the possibility of not being reelected. Thus, the
    Legislature has determined that four years is a reasonable time in which to
    expect a supervisor to be effective — i.e., that it is a “significant period in
    office.” If it is not enough time in San Bernardino County — if a San
    Bernardino County Supervisor does not take off the training wheels until
    his or her second term in office — that is, at best, a local inefficiency. (In
    fact, arguably it shows a need for term limits, to spur efficiency.) When
    balanced against the voters’ right to the initiative as well as the state’s
    legitimate interests in term limits (see part IV.C.2.b), it is entitled to little
    weight.
    4
    8. On pages 53-54, the third sentence in the second full paragraph:
    They could hardly have had a reasonable expectation that they could run for
    a second term, as Measure K had already been proposed and was on the
    same ballot as they were.
    is deleted and replaced with:
    As in Sheyko, they could not reasonably expect that eligibility conditions
    would not change.
    The petition for rehearing, filed June 9, 2023, is denied.
    Except for these modifications, the opinion remains unchanged. This modification does
    not effect a change in the judgment.
    CERTIFIED FOR PARTIAL PUBLICATION
    RAMIREZ
    P.J.
    I concur:
    MILLER
    J.
    I would grant the request for a rehearing.
    MENETREZ
    J.
    5
    Filed 5/25/23 (unmodified version)
    See Dissenting Opinion
    CERTIFIED FOR PARTIAL PUBLICATION
    IN THE COURT OF APPEAL OF THE STATE OF CALIFORNIA
    FOURTH APPELLATE DISTRICT
    DIVISION TWO
    SAN BERNARDINO COUNTY BOARD
    OF SUPERVISORS,                                      E077772
    Plaintiff and Appellant;                  (Super.Ct.No. CIVSB2025319)
    v.                                                   OPINION
    LYNNA MONELL, as Clerk, etc.,
    Defendant and Respondent;
    NADIA RENNER,
    Intervener and Appellant.
    APPEAL from the Superior Court of San Bernardino County. Donald R. Alvarez,
    Judge. Reversed and remanded with directions.
    The Red Brennan Group and Aaron D. Burden; Briggs Law Corporation and Cory
    J. Briggs for Intervener and Appellant.
    The Sutton Law Firm, Bradley W. Hertz, and Nicholas L. Sanders for Plaintiff and
    Appellant.
         Pursuant to California Rules of Court, rules 8.1105(b) and 8.1110, this
    opinion is certified for publication with the exceptions of part II, VI, and VIII.
    1
    Jolena E. Grider, Deputy County Counsel, for Defendant and Respondent.
    Jonathan M. Coupal, Timothy A. Bittle, and Laura E. Dougherty for Amicus
    Curiae Howard Jarvis Taxpayers Foundation.
    Jennifer B. Henning for Amicus Curiae California State Association of Counties.
    The facts are simple. The legal issues are not.
    On November 3, 2020, the voters of San Bernardino County passed Measure K. It
    amended the county charter so as to (1) limit a supervisor to a single four-year term and
    (2) limit a supervisor’s compensation to $5,000 a month. At the same time, the voters
    also elected three new supervisors.
    The trial court ruled that the one-term limit is unconstitutional. It also ruled that
    the compensation limit is constitutional, but, because Measure K is not severable, it, too,
    must be struck down. Finally, it ruled that Measure K did not apply to the new
    supervisors (although it acknowledged that the issue was moot, in light of its other
    rulings).
    Nadia Renner — the proponent of Measure K — appeals. She contends that:
    (1) The one-term limit is constitutional.
    (2) The compensation limit is severable.
    (3) Measure K applies to the new supervisors.
    (4) The trial court erred by enjoining the certification, authentication, recordation,
    and filing of Measure K.
    2
    The San Bernardino County Board of Supervisors (Board) cross-appeals. It
    contends that:
    (5) Supervisors’ compensation cannot be set by initiative, because state law
    delegates supervisors’ compensation exclusively to boards.
    (6) The compensation limit violates minimum wage laws; alternatively, if it
    effectively forces supervisors to work part-time, it impairs essential governmental
    functions.
    (7) The compensation limit improperly acts as a referendum on San Bernardino
    County Code section 13.0614 (section 13.0614), which provides for supervisors’
    compensation.
    We perceive a preliminary issue of appealability. However, we will conclude that
    the trial court’s ruling is appealable.
    We will hold that the one-term limit is constitutional. We will further hold that
    supervisors’ compensation can be set by initiative. The Board has not shown that the
    compensation limit violates minimum wage laws. The Board also has not shown that the
    compensation limit conflicts with section 13.0614; even assuming it does, the voters can
    amend or abrogate an ordinance not only by referendum, but also by initiative. Because
    both the one-term limit and the compensation limit are valid, we need not decide whether
    3
    Measure K is severable. As to whether Measure K applies to the new supervisors, we
    reach a split decision: The one-term limit applies, but the compensation limit does not.1
    Finally, assuming the trial court erred by enjoining the certification,
    authentication, recordation, and filing of Measure K, the error is forfeited, harmless,
    and/or moot.
    I
    FACTAL AND PROCEDURAL BACKGROUND
    A.      Statement of Facts.
    The following facts are taken from Renner’s statement of facts and statement of
    the case, which the Board concedes are accurate; from the declaration, and exhibits
    introduced below; and from matters of which the trial court took judicial notice.
    Renner is the proponent of Measure K, a San Bernardino County initiative.
    Measure K amends the county charter. It has four key provisions. First, it
    provides that a supervisor can serve only one four-year term. An unexpired term counts
    toward the four-year limit, but only if it has two or more years left to run when
    Measure K goes into effect. Previously, supervisors were limited to three consecutive
    four-year terms. Second, it limits the compensation of a supervisor, including all
    benefits, to $5,000 a month. Previously, supervisors’ compensation was set at the
    average of the compensation of supervisors in Riverside, Orange, and San Diego
    1      In an unpublished portion of this opinion, we will reject the Board’s
    additional contention that Measure K violates the single-subject rule.
    4
    Counties. Third, it provides: “To the extent permitted by law, the provisions of this
    Charter Amendment shall be effective on voter approval of the initiative as provided by
    California law.” Fourth, it includes a severability provision.
    On November 3, 2020, Measure K passed, with 66.84 percent of the votes.
    Another initiative on the ballot at the same time, Measure J, completely revised
    and restated the county charter. It, too, passed, but barely, with only 50.72 percent of the
    votes. Measure J would have limited supervisors to three terms (consecutive or not) and
    would have given them a salary set at 80 percent of a superior court judge’s salary and
    the same benefits as department heads. However, because Measure K got more votes, it
    supersedes Measure J to the extent that they conflict. (Elec. Code, §§ 9102, 9123; see
    Taxpayers to Limit Campaign Spending v. Fair Pol. Practices Com. (1990) 
    51 Cal.3d 744
    , 765-768.)
    At the same time, new supervisors Joe Baca, Jr., Col. Paul Cook, and Dawn Rowe
    were elected to the Board. On December 7, 2020, they were sworn in. On December 8,
    2020, the Board (including the newly elected supervisors) certified the results of the
    election.
    B.     Statement of the Case.
    Meanwhile, on December 2, 2020, the Board filed a combined complaint and writ
    petition, seeking a declaration that Measure K was invalid plus an injunction and a writ of
    mandate preventing its enforcement. The only named defendant was Lynna Monell, in
    5
    her official capacity as the Clerk of the Board. However, the trial court gave Renner
    leave to intervene.
    On January 8, 2021, the trial court granted the Board’s application for a temporary
    restraining order (TRO).2
    The case was tried to the court as a writ petition, based on trial briefs and oral
    argument, with no live testimony.
    The trial court granted the petition. It ruled that the one-term term limit was
    unconstitutional. It further ruled that the compensation limit was constitutional; however,
    it was not severable, and therefore Measure K as a whole was invalid. Finally, it ruled
    that Measure K was prospective only, and therefore it did not apply to the new
    supervisors. However, it acknowledged that this issue was moot in light of its other
    rulings.
    C.     Subsequent Statutory Changes.
    On October 4, 2021 — after the trial court ruled — Assembly Bill No. 428 (2021-
    2022 Reg. Sess.) (AB 428) was enacted (Stats. 2021, ch. 462, p. 6507), and on January 1,
    2022, it went into effect. AB 428 was enacted specifically in response to Measure K.
    (Assem. Com. on Elections, Analysis of Assem. Bill No. 428 (2021-2022 Reg. Sess.) as
    amended Mar. 18, 2021, p. 3.) It amended Government Code section 25000, subdivision
    (b) so that it now provides, as relevant here (additions italicized):
    2        A hearing on a preliminary injunction was set, continued, and then vacated.
    Thus, while the record is not entirely clear on this point, it appears that the TRO remained
    in effect until the trial court ruled on the petition.
    6
    “(1) Notwithstanding any other provision of law, the board of supervisors of any
    general law or charter county may adopt or the residents of the county may propose, by
    initiative, a proposal to limit to no fewer than two terms or repeal a limit on the number
    of terms a member of the board of supervisors may serve on the board of supervisors.
    Any proposal to limit the number of terms a member of the board of supervisors may
    serve on the board of supervisors shall apply prospectively only . . . .
    “(2) The changes made to this subdivision by the act that added this paragraph
    shall not affect any term limits that were legally in effect prior to January 1, 2022 . . . .”
    It also amended Government Code section 25300, subdivision (b) so that it now
    provides (additions italicized):
    “The board of supervisors shall prescribe the compensation of all county officers,
    including the board of supervisors, and shall provide for the number, compensation,
    tenure, appointment and conditions of employment of county employees. Except as
    otherwise required by Section 1 or 4 of Article XI of the California Constitution, such
    action may be taken by resolution of the board of supervisors as well as by ordinance.”
    II
    APPEALABILITY
    The trial court issued a “ruling” granting the mandate petition. It did not enter a
    formal judgment. It also did not expressly rule on the Board’s second (injunction) and
    third (declaratory relief) causes of action. Moreover, as far as the record shows, it did not
    actually issue either a writ or a permanent injunction.
    7
    There is a preliminary question of whether the ruling is appealable. We conclude,
    however, that it was.
    “[A]n order granting or denying a petition for an extraordinary writ [can]
    constitute[] a final judgment for purposes of an appeal, even if the order is not
    accompanied by a separate formal judgment. [Citations.]” (Public Defenders’
    Organization v. County of Riverside (2003) 
    106 Cal.App.4th 1403
    , 1409; cf. Natomas
    Unified School Dist. v. Sacramento County Bd. of Education (2022) 
    86 Cal.App.5th 1013
    , 1027.)
    “[A] ruling on a petition for writ of mandate is not appealable if other causes of
    action remain pending between the parties. [Citation.]” (Canandaigua Wine Co., Inc. v.
    County of Madera (2009) 
    177 Cal.App.4th 298
    , 302.) But “an order constitutes a final
    judgment despite other causes of action remaining if the order effectively disposes of the
    entire case. For example, an order is appealable if it resolves an allegation that is
    essential to all of the causes of action. [Citation.]” (Id. at p. 303.)
    Here, the trial court’s ruling effectively resolved the cause of action for an
    injunction, by declaring that Measure K is unconstitutional and “cannot be implemented
    . . . .” It also effectively resolved the cause of action for declaratory relief, again by
    declaring Measure K unconstitutional. The ruling “contemplated no further action, such
    as the preparation of another order or judgment [citation], and disposed of all issues
    between all parties.” (Laraway v. Pasadena Unified School Dist. (2002) 
    98 Cal.App.4th
                                             8
    579, 583.) Moreover, it is clear from the record that the parties themselves did not
    contemplate any further judicial action.
    It follows that the ruling is appealable.
    III
    THE EFFECT OF MEASURE D
    Measure D was enacted in 2022, while this appeal was pending. It would
    supersede Measure K — if Measure D is valid. A trial court upheld Measure D, but its
    validity is currently before this court in The Red Brennan Group v. Jimenez et al.,
    E080868. So far, nothing has happened in that case, other than the filing of the Civil
    Case Information Statement. In the ordinary course of business, it is likely to be a year or
    so before we decide it — to say nothing of how much time any subsequent Supreme
    Court review will consume. With all respect to our dissenting colleague, Measure D is
    irrelevant.
    The fact that Measure D has not yet been struck down, and thus is presumptively
    valid, does not make this case moot. “A case becomes moot when events ‘“render[] it
    impossible for [a] court, if it should decide the case in favor of plaintiff, to grant him any
    effect[ive] relief.”’ [Citation.]” (In re D.P. (2023) 
    14 Cal.5th 266
    , 276, italics added.) If
    we dismiss this case as moot, and if Measure D is then struck down, any relief we grant
    as to Measure K would still be effective; however, we would have no way of bringing
    this case back to life again. And even if Measure D is eventually upheld, this case still
    9
    will not be moot; there would still be a live issue regarding the supervisors’ compensation
    from 2020 through 2022.
    The only even arguable question is whether we should stay this appeal pending the
    outcome of the Measure D litigation. There is not even any pending request for a stay;
    before oral argument, the Board requested a stay, but we denied it. We see no reason
    why the appeal filed later should have precedence over the one filed earlier. Renner has
    paid her filing fee; the parties have fully briefed the appeal; prior to oral argument, we
    drafted and issued a tentative opinion; and we heard oral argument. Courts have no
    jurisdiction to decide moot cases, On the flip side, however, “[a] court is tasked with the
    duty ‘“to decide actual controversies by a judgment which can be carried into effect
    . . . .”’ [Citation.]” (In re D.P., supra, 14 Cal.5th at p. 276, italics added.) It is time to
    decide this case.
    IV
    THE ONE-TERM LIMIT
    The trial court ruled that the one-term limit violates “voters’ and incumbents’ 1st
    and 14th amendment rights.” Renner contends that it erred.
    A. The Anderson-Burdick Framework.
    “It is beyond cavil that ‘voting is of the most fundamental significance under our
    constitutional structure.’ [Citation.] It does not follow, however, that the right to vote in
    10
    any manner and the right to associate for political purposes through the ballot are
    absolute. [Citation.]” (Burdick v. Takushi (1992) 
    504 U.S. 428
    , 433 (Burdick).)
    “Election laws will invariably impose some burden upon individual voters. Each
    provision of a code, ‘whether it governs the registration and qualifications of voters, the
    selection and eligibility of candidates, or the voting process itself, inevitably affects — at
    least to some degree — the individual’s right to vote and his right to associate with others
    for political ends.’ [Citation.]” (Burdick, supra, 504 U.S. at p. 433.)
    “A court considering a challenge to a state election law must weigh ‘the character
    and magnitude of the asserted injury to the rights protected by the First and Fourteenth
    Amendments that the plaintiff seeks to vindicate’ against ‘the precise interests put
    forward by the State as justifications for the burden imposed by its rule,’ taking into
    consideration ‘the extent to which those interests make it necessary to burden the
    plaintiff’s rights.’ [Citations.]
    “Under this standard, the rigorousness of our inquiry into the propriety of a state
    election law depends upon the extent to which a challenged regulation burdens First and
    Fourteenth Amendment rights. Thus, . . . when those rights are subjected to ‘severe’
    restrictions, the regulation must be ‘narrowly drawn to advance a state interest of
    compelling importance.’ [Citation.] But when a state election law provision imposes
    only ‘reasonable, nondiscriminatory restrictions’ upon the First and Fourteenth
    11
    Amendment rights of voters, ‘the State’s important regulatory interests are generally
    sufficient to justify’ the restrictions. [Citations.]” (Burdick, supra, 504 U.S. at p. 434.)3
    “This latter, lesser scrutiny is not ‘pure rational basis review.’ [Citation.] Rather,
    ‘the court must actually “weigh” the burdens imposed on the plaintiff against “the precise
    interests put forward by the State,” and the court must take “into consideration the extent
    to which those interests make it necessary to burden the plaintiff’s rights.”’ [Citation.]
    Review under this balancing test is ‘quite deferential’ . . . . [Citation.]” (SAM Party of
    New York v. Kosinski (2d Cir. 2021) 
    987 F.3d 267
    , 274.)
    “‘Courts will uphold as “not severe” restrictions that are generally applicable,
    even-handed, politically neutral, and which protect the reliability and integrity of the
    election process. [Citation.] . . .’ [Citation.] ‘Courts will strike down state election laws
    as severe speech restrictions only when they significantly impair access to the ballot,
    stifle core political speech, or dictate electoral outcomes.’ [Citation.]” (Rawls v. Zamora
    (2003) 
    107 Cal.App.4th 1110
    , 1116.)
    3      Burdick drew on the Supreme Court’s earlier decision in Anderson v.
    Celebrezze (1983) 
    460 U.S. 780
    . Accordingly, the approach it outlined is sometimes
    called the Anderson-Burdick test or the Anderson-Burdick framework. (See, e.g.,
    Edelstein v. City and County of San Francisco (2002) 
    29 Cal.4th 164
    , 168 (Edelstein).)
    Renner and the Board both agree that the Anderson-Burdick test is controlling here.
    Recently, one federal appellate court held that the Anderson-Burdick test is
    “inapposite” to the analysis of term limits. (Kowall v. Benson (6th Cir. 2021) 
    18 F.4th 542
    , 547, cert. den. (2022) ___ U.S. ___ [
    143 S. Ct. 88
    ].) As we will discuss, however,
    controlling California Supreme Court authority requires us to apply the Anderson-
    Burdick test.
    12
    B. Whether the One-Term Limit Is “Severe.”
    The trial court ruled that the one-term limit was “severe” because it “is not merely
    about limiting the term but limiting the term to one time . . . .”
    1.     Relevant case law.
    In Legislature v. Eu (1991) 
    54 Cal.3d 492
     (Eu), our Supreme Court upheld
    Proposition 140, which imposed lifetime term limits of two four-year terms per state
    senator and three two-year terms per state assemblymember. (See Eu at pp. 502-506,
    518.)
    The court began by saying: “[T]he initiative power must be liberally construed to
    promote the democratic process. [Citation.] Indeed, it is our solemn duty to jealously
    guard the precious initiative power, and to resolve any reasonable doubts in favor of its
    exercise. [Citation.] As with statutes adopted by the Legislature, all presumptions favor
    the validity of initiative measures and mere doubts as to validity are insufficient; such
    measures must be upheld unless their unconstitutionality clearly, positively, and
    unmistakably appears. [Citation.]” (Eu, supra, 54 Cal.3d at p. 501.)
    It then held that strict scrutiny was not required because the term limits did not
    have a “serious impact on First Amendment freedoms of speech and association,” as they
    “d[id] not affect speech interests and . . . impact[ed] all political parties on an equal basis.
    [Citation.]” (Eu, supra, 54 Cal.3d at p. 515.) Rather, the term limits were entitled to
    “wide latitude,” because they “[we]re applied in an even-handed manner without
    discriminating against particular citizens or classes of citizens.” (Id. at p. 516.)
    13
    In Bates v. Jones (9th Cir. 1997) 
    131 F.3d 843
     (en banc) (Bates), cert. den. (1998)
    
    523 U.S. 1021
    , the Ninth Circuit Court of Appeals similarly upheld Proposition 140. It
    held that its impact on the plaintiffs’ claimed “right to vote for the candidate of one’s
    choice and . . . right of an incumbent to again run for his or her office” was “not severe.”
    (Id. at p. 847.) It explained that “term limits on state officeholders is [sic] a neutral
    candidacy qualification, such as age or residence, which the State certainly has the right
    to impose. [Citation.] With regard to incumbents, they may enjoy the incumbency of a
    single office for a number of years, and . . . they are not precluded from running for some
    other state office. [¶] Most important, the lifetime term limits do not constitute a
    discriminatory restriction. [They] make[] no distinction on the basis of the content of
    protected expression, party affiliation, or inherently arbitrary factors such as race,
    religion, or gender. Nor do[] [they] ‘limit [] political participation by an identifiable
    political group whose members share a particular viewpoint, associational preference, or
    economic status.’ [Citation.]” (Ibid.)
    Significantly, the Board has not cited any case striking down any term limits on
    First and Fourteenth Amendment grounds.
    2.      Application here.
    Under Eu and Bates, at least as a general rule, term limits are not so “severe” as to
    trigger strict scrutiny. The trial court distinguished those cases, however, on the ground
    that a one-term lifetime limit is “severe.” In light of the way “severe” has been
    interpreted and applied, this was error.
    14
    In Eu, the court concluded that the term limits there were not severe because they
    were nondiscriminatory — they “[we]re applied in an even-handed manner without
    discriminating against particular citizens or classes of citizens.” (Eu, supra, 54 Cal.3d at
    p. 516.) They were not based on the content of protected speech. (Id. at p. 515.) “They
    impact[ed] all political parties on an equal basis. [Citation.]” (Ibid.) A one-term limit is
    no different in these respects.
    In Bates, the court concluded that the same term limits were not severe because (1)
    they were “neutral,” (2) they allowed incumbency “for a number of years” and did not
    preclude an incumbent from running for a different office, and (3) they did not
    distinguish “on the basis of the content of protected expression, party affiliation, or
    inherently arbitrary factors such as race, religion, or gender.” (Bates, supra, 131 F.3d at
    p. 847.) Again, a one-term limit is no different.
    The only argument to the contrary is that it allows incumbency for a lesser
    “number of years” than the term limits in Eu and Bates. Those limits, however, while
    they were not one-term limits, allowed a state senator to serve for no more than eight
    years and a state assembly member to serve for no more than six years. The difference
    between the six or eight years there and the four years here is not sufficient to be
    constitutionally significant — particularly when the term limits here are similarly neutral
    and nondiscriminatory and do not preclude an incumbent from holding any other office.
    Four years is ample time for a supervisor to at least attempt to tick off all the boxes on his
    or her legislative to-do list. In this respect, the Board, with only five members, is very
    15
    different from the California Senate, with 40 members, or the California Assembly, with
    80 members. In the latter bodies, seniority and a cursus honorum of committee
    memberships both play a role. By contrast, a newly elected supervisor can hit the ground
    running.4 The Board does not point to anything that a supervisor could accomplish in six
    or eight years but not in four.
    C.     Whether the Regulatory Interests Are Sufficient to Justify the One-Term
    Limit.
    The trial court also ruled that, even assuming Measure K’s one-term limit was not
    “severe,” “the stated reason [for it] is not sufficient to justify imposing the burden [of]
    precluding an incumbent from seeking re-election.” It explained, “A one-term limit is
    not providing a supervisor sufficient time in the governing body position.”
    “Additionally, . . . the desire to ensure a candidate seeks to serve the public interest
    cannot justify then precluding a candidate or electing an incumbent he believes is serving
    the interest of the voters at least for one or two additional terms of office. And a
    reasonable remedy exists if the incumbent seeking re-election is not performing
    competently: the electorate vote for the other candidate.”
    4        Amicus curiae Howard Jarvis Taxpayers Association takes a swipe at the
    trial court for “weighing the merits of an ideal learning curve for a county supervisor.”
    We find that we cannot help but consider this subject in deciding whether the term limits
    here allow a supervisor to serve “a significant period in office” within the meaning of Eu
    or, more generally, unduly injure the rights of voters and candidates.
    16
    1.      Relevant case law.
    In Eu, the Supreme Court balanced the claimed injury to First and Fourteenth
    Amendment rights against the State’s claimed interests, as follows.
    First, it considered “the nature of the injury to the rights affected . . . .” (Eu,
    supra, 54 Cal.3d at p. 517; see id. at pp. 517-519.) It concluded that the term limits “do[]
    affect the rights of voters and candidates to a degree . . . .” (Id. at p. 519.) With respect
    to candidates, however, the impact was “mitigate[d]” by three factors: “First, the affected
    incumbent is not barred from seeking any other public office . . . . Second, the term
    limitations arise only after the incumbent already has had the opportunity to serve a
    significant period in office . . . .” (Id. at p. 518.) Third, the term limits in most instances
    did not count terms served before they went into effect. (Ibid.)
    Similarly, with respect to voters, it was unclear whether a right to vote for a
    particular candidate even existed. (Eu, supra, 54 Cal.3d at p. 519; see also id. at pp. 516,
    518.) In any event, the court accepted the argument that the term limits there “d[id] not
    truly impair the franchise, for the voters retain the basic fundamental right to cast their
    ballots for the qualified candidate of their choice.” (Id. at p. 519; see also ibid
    [“mitigating aspects[] include[e] the voters’ continued right to vote for any qualified
    candidates”].)
    Second, it considered “the interests asserted by the state as justifications for that
    injury . . . .” (Eu, 
    supra,
     54 Cal.3d at p. 517; see id. at pp. 520-522.) It observed that
    “‘[c]onstitutional restrictions circumscribing the ability of incumbents to succeed
    17
    themselves appear in over twenty state constitutions, and exist in the Twenty-second
    Amendment to the Constitution of the United States with regard to the Presidency. The
    universal authority is that restriction upon the succession of incumbents serves a rational
    public policy and that, while restrictions may deny qualified men an opportunity to serve,
    as a general rule the over-all health of the body politic is enhanced by limitations on
    continuous tenure. [Citations and fn. omitted].’ [Citations.]” (Id. at p. 520.)
    “[T]he substantial reasons for limiting the right of incumbents to succeed
    themselves . . . include ‘[t]he power of incumbent officeholders to develop networks of
    patronage and attendant capacities to deliver favorably disposed voters to the polls,’
    ‘fears of an entrenched political machine which could effectively foreclose access to the
    political process,’ and the belief that regularly disrupting those ‘machines’ ‘would
    stimulate criticism within political parties’ and ‘insure a meaningful, adversary, and
    competitive election.’ [Citation.]
    “In addition, . . . ‘ . . . a limitation upon succession of incumbents removes the
    temptation to prostitute the government to the perpetuation of a particular administration.
    [Citation.] . . . Meretricious policies which sacrifice the well-being of economic, social,
    racial, or geographic minorities are most likely where a political figure, political party, or
    political interest group can rely upon electorate inertia fostered by the hopelessness of
    encountering a seemingly invincible political machine.’ [Citation.]” (Eu, supra, 54
    Cal.3d at p. 521.)
    18
    Third, it considered “the necessity for imposing the particular burden affecting the
    plaintiff’s rights, rather than some less drastic alternatives.” (Eu, 
    supra,
     54 Cal.3d at
    p. 517; see id. at pp. 522-524.) It noted that “incumbents do indeed appear to enjoy
    considerable advantages over other candidates. [Citations.]” (Id. at p. 523.) “It is true
    . . . that [the proponents] have not offered evidence to support all of the various premises
    on which [the term limits are] based. But . . . a state need not demonstrate empirically all
    of the various evils that its regulations seek to combat . . . .’ [Citation.]” (Id. at p. 524.)
    The opponents argued that alternative measures, such as “a limitation on
    consecutive terms, together with additional restrictions on campaign contributions to
    legislators, decreased fringe and pension benefits, and additional incentives for early
    retirement, would have been sufficient to promote and accomplish the state interests
    previously discussed.” (Eu, supra, 54 Cal.3d at pp. 522-523.) The court responded that
    “[r]ealistically, only a lifetime ban could protect against various kinds of continued
    exploitation of the ‘advantages of incumbency’ captured through past terms in office.
    The remainder of [the opponents’] suggested ‘alternatives’ essentially involve narrow
    changes in the system of providing contributions or compensation for legislators, changes
    that would afford ‘career politicians’ with independent resources little incentive to
    voluntarily terminate public service.” (Id. at p. 524.)
    “[W]e conclude,” the court said, “the interests of the state in incumbency reform
    outweigh any injury to incumbent office holders and those who would vote for them.”
    19
    (Eu, supra, 54 Cal.3d at p. 524.) Indeed, the court observed, “it is unlikely we would
    reach a different result applying strict scrutiny . . . . (Id. at p. 515.)
    By contrast, Bates upheld the state’s justification for the term limits with little
    discussion. It said, “[A] lack of term limits may create ‘unfair incumbent advantages.’
    Long-term entrenched legislators may obtain excessive power which, in turn, may
    discourage other qualified candidates from running for office or may provide the
    incumbent with an unfair advantage in winning reelection.” (Bates, supra, 131 F.3d at
    p. 847.) “California voters apparently perceived lifetime term limits for elected state
    officials as a means to promote democracy by opening up the political process and
    restoring competitive elections. This was their choice to make. [Citation.]” (Ibid.)
    2.      Application here.
    We apply the three-part Anderson-Burdick framework as our Supreme Court did
    in Eu. It must be remembered that review under this framework is deferential — all the
    more so when the electoral restriction being analyzed was enacted by initiative.
    Moreover, when the restriction is “‘reasonable [and] nondiscriminatory,’ . . . ‘the State’s
    important regulatory interests are generally sufficient to justify’ the restrictions.
    [Citations.]” (Burdick, supra, 504 U.S. at p. 434.)
    20
    a.      The nature of the injury to the rights affected.
    i.     The voters’ rights.
    (a)     The nature of the right.
    Federal courts have held that voters have no right to vote for a particular
    candidate. (Citizens for Legislative Choice v. Miller (6th Cir. 1998) 
    144 F.3d 916
    , 921;
    Stiles v. Blunt (8th Cir. 1990) 
    912 F.2d 260
    , 266 [voters do not have “an absolute right to
    support a specific candidate regardless of whether he or she has satisfied reasonable
    eligibility requirements”], cert. den. (1991) 
    499 U.S. 919
    .)
    Eu acknowledged this. (Eu, 
    supra,
     54 Cal.3d at pp. 516, 518.) However, it noted
    that, at least at the time, Canaan v. Abdelnour (1985) 
    40 Cal.3d 703
    , 714-716 (Canaan)
    was to the contrary. (Eu, 
    supra,
     54 Cal.3d at p. 516.) Canaan had held that “[a] ban on
    write-in voting affects . . . the right of . . . voters to cast ballots for the candidates of their
    choice.” (Canaan, supra, at p. 715.) Eu concluded that, while term limits “affected . . .
    the voters’ right to reelect the incumbent” (Eu, 
    supra, at p. 517
    ; see also id. at p. 519), the
    existence and scope of this right were “unclear”; hence, “the legal impact of [term limits]
    on the voters remain[ed] uncertain.” (Id. at p. 519.) It treated this uncertainty as a
    “mitigating aspect.” (Ibid.)
    After Eu was decided, the Supreme Court overruled Canaan. (Edelstein, 
    supra,
    29 Cal.4th at p. 183.) While the extent to which Edelstein overruled particular portions
    of Canaan is open to debate (see ibid.), Edelstein did disavow a specific right to vote for
    21
    a particular candidate in favor of a more general right “to make free choices and to
    associate politically through the vote . . . .” (Id. at pp. 175, 177-178, 182-183.)
    (b)    The impact on the right.
    Term limits do not affect these general rights. Voters still have the right to make
    free choices and to associate politically through the vote. As in Eu, the term limits here
    “do not truly impair the franchise, for the voters retain the basic fundamental right to cast
    their ballots for the qualified candidate of their choice.” (Eu, supra, 54 Cal.3d at p. 519.)
    ii.    The candidates’ rights.
    (a)    The nature of the right.
    In her reply brief, Renner argues that candidates do not have a fundamental right
    to run for public office. She forfeited this contention by failing to raise it below. (Delta
    Stewardship Council Cases (2020) 
    48 Cal.App.5th 1014
    , 1074.) She further forfeited it
    by failing to raise it in her opening brief. (Raceway Ford Cases (2016) 
    2 Cal.5th 161
    ,
    178.)
    In any event, we reject this contention on the merits.
    The California Supreme Court has held that “[t]he right to seek public office . . .
    [is] fundamental.” (Canaan, supra, 40 Cal.3d at p. 727.) “There has been some
    suggestion in the case law that the right to be a candidate for public office may not be
    fundamental in and of itself. [Citations.] However, this court has described the right to
    hold public office as ‘valuable, fundamental and one that is subject to First Amendment
    protection . . . .’ [Citations.]” (Id. at p. 714.) Admittedly, as mentioned, Canaan was
    22
    overruled in Edelstein, but not on this particular ground. (Edelstein, supra, 29 Cal.4th at
    p. 183.) Moreover, the earlier cases on which Caanan itself relied have never been
    overruled. (Johnson v. Hamilton (1975) 
    15 Cal.3d 461
    , 468 [“the right to hold public
    office [is] valuable, fundamental and . . . subject to First Amendment protection . . . .”];
    Zeilenga v. Nelson (1971) 
    4 Cal.3d 716
    , 723 [“the right to run for public office is as
    fundamental a right as is the right to vote”].)
    More recently, in Eu, the Supreme Court said, “Two important rights are affected
    by [term limits], namely, the incumbent’s right to run for public office, and the voters’
    right to reelect the incumbent to that office.” (Eu, 
    supra,
     54 Cal.3d at p. 517, italics
    added.) Thus, it analyzed the impact on candidates (id. at pp. 517-518), and it weighed
    that impact, along with the impact on voters, against the state’s interest in the term limits
    at issue. (Id. at pp. 519-520, 522-524.)
    Federal cases are not to the contrary. According to the United States Supreme
    Court, “The right of . . . an individual to a place on a ballot is entitled to protection . . . .”
    (Lubin v. Panish (1974) 
    415 U.S. 709
    , 716; see also Moore v. Martin (8th Cir. 2017) 
    854 F.3d 1021
    , 1025 [“Ballot access restrictions implicate . . . the rights of potential
    candidates for public office . . . .”], cert. den. ___ U.S. ___ [
    138 S.Ct. 321
    ]; Phillips v.
    City of Dallas (5th Cir. 2015) 
    781 F.3d 772
    , 778-779 [“This court has been unequivocal
    in its recognition of a First Amendment interest in candidacy.”]; Davies v. Grossmont
    Union High School Dist. (9th Cir. 1991) 
    930 F.2d 1390
    , 1397 [“the individual’s right to
    seek public office is inextricably intertwined with the public’s fundamental right to
    23
    vote”]; Branch v. FCC (D.C. Cir. 1987) 
    824 F.2d 37
    , 47 [the existence of a “right to seek
    political office . . . is undeniable . . . .”], cert. den. (1988) 
    485 U.S. 959
    .)
    Admittedly, the United States Supreme Court “has not . . . attached such
    fundamental status to candidacy as to invoke a rigorous standard of review.” (Bullock v.
    Carter (1972) 
    405 U.S. 134
    , 142-143; see also Clements v. Fashing (1982) 
    457 U.S. 957
    ,
    963 [“Far from recognizing candidacy as a ‘fundamental right,’ we have held that the
    existence of barriers to a candidate’s access to the ballot ‘does not of itself compel close
    scrutiny.’”] [plur. opn. of Rehnquist, J.] (Clements).)
    However, this simply means that strict scrutiny is not required. If it were, few
    provisions for “the selection and eligibility of candidates” (Burdick, supra, 504 U.S. at
    p. 433) would escape challenge. “[T]o say that the right to candidacy is not fundamental
    is not to say that a rational basis analysis applies.” (Brazil-Breashears v. Bilandic (7th
    Cir. 1995) 
    53 F.3d 789
    , 792, cert. den 
    516 U.S. 869
    .) Rather — just like the voter’s right
    to vote — the candidate’s right to run for office is subject to the Anderson-Burdick test.
    (See, e.g., Libertarian Party of Arkansas v. Thurston (8th Cir. 2020) 
    962 F.3d 390
    , 398.)5
    5       We respectfully disagree with Kern County Employees’ Retirement Assn. v.
    Bellino (2005) 
    126 Cal.App.4th 781
     (Kern County), which held that an asserted
    infringement of the right to run for office is subject to rational-basis review. (Id. at
    p. 794.) In our view, Kern County misread Justice Rehnquist’s plurality opinion in
    Clements, 
    supra,
     
    457 U.S. 957
    . It said that strict scrutiny is not required (id. at p. 963),
    but it did not say that rational basis review applies. To the contrary, it said, “Decision in
    this area of constitutional adjudication is a matter of degree, and involves a consideration
    of the facts and circumstances behind the law, the interests the State seeks to protect by
    placing restrictions on candidacy, and the nature of the interests of those who may be
    burdened by the restrictions.” (Ibid.) Thus, Clements foreshadowed the intermediate
    [footnote continued on next page]
    24
    Eu concluded that the term limits “do[] affect the rights of . . . candidates to a
    degree . . . .” (Eu, supra, 54 Cal.3d at p. 519.) With respect to candidates, however, the
    impact was “mitigate[d]” by three factors: “First, the affected incumbent is not barred
    from seeking any other public office . . . . Second, the term limitations arise only after
    the incumbent already has had the opportunity to serve a significant period in office . . . .”
    (Eu, supra, 54 Cal.3d at p. 518.) Third, the term limits in most (though not all) instances
    did not count terms served before they went into effect. (Ibid.)
    (b)    The impact on the right.
    With respect to candidates, a one-term limit is ipso facto more burdensome than
    the two- and three-term limits in Eu. Nevertheless, the reasons stated in Eu for
    concluding that the impact was “mitigate[d]” largely also apply here.
    First, a termed-out incumbent is free to run for any other county (or state or
    federal) office.
    Second, the term limits do allow an incumbent to serve “a significant period in
    office.” (Eu, supra, 54 Cal.3d at p. 518.) As already discussed, we cannot meaningfully
    distinguish a four-year maximum from the six-year maximum applicable to
    assemblymembers in Eu.6 (See part III.B.2, ante.)
    scrutiny subsequently announced by Anderson and refined by Burdick (neither of which
    Kern County cited).
    6       In conformity with Government Code section 25000, subdivision (a), which
    sets a supervisor’s term at four years, the drafters of Measure K did not attempt to
    prescribe a five, six, or seven-year term.
    25
    Third, Measure K does not count unexpired terms, if they have less than two years
    still to run. That is not so very different from the term limits in Eu, which did not count
    unexpired terms at all (except that it did count them against “some incumbent Senators”).
    (Eu, supra, 54 Cal.3d at p. 518.) The fact that in Eu, terms begun by some incumbent
    Senators did count demonstrates that Eu did not view this factor as dispositive.
    b.     The interests asserted by the state as justifications for the
    injury.
    Eu acknowledged that the state has a legitimate interest in term limits: “‘The
    universal authority is that restriction upon the succession of incumbents serves a rational
    public policy and that, while restrictions may deny qualified men an opportunity to serve,
    as a general rule the over-all health of the body politic is enhanced by limitations on
    continuous tenure. [Citations and fn. omitted].’” (Eu, supra, 54 Cal.3d at p. 520.) It
    indicated that this interest was sufficiently compelling for the term limits there to survive
    even strict scrutiny. (Id. at p. 524.)
    Bates agreed that the state has an interest in term limits: “Long-term entrenched
    legislators may obtain excessive power which, in turn, may discourage other qualified
    candidates from running for office or may provide the incumbent with an unfair
    advantage in winning reelection.” (Bates, supra, 131 F.3d at p. 847.)
    The trial court reasoned that “the desire to ensure a candidate seeks to serve the
    public interest cannot justify then precluding a candidate or electing an incumbent he
    believes is serving the interest of the voters at least for one or two additional terms of
    26
    office. And a reasonable remedy exists if the incumbent seeking re-election is not
    performing competently: the electorate vote for the other candidate.” These arguments,
    however, apply to a two- or three-term limit just as much as to a one-term limit;
    essentially, they are arguments against having any term limits at all. As we know from
    Eu, however, term limits imposed by initiative are presumptively valid. The trial court
    was not entitled to second-guess the voters’ position on these policy issues.
    The Board argues that “Renner raises a purported government interest by merely
    citing the language of Eu and provides no evidence that this interest is relevant to San
    Bernardino County.” Eu held, however, that such evidence is not required. (Eu, 
    supra,
    54 Cal.3d at p. 523.) “[W]e [do not] require elaborate, empirical verification of the
    weightiness of the State’s asserted justifications. [Citation.]” (Timmons v. Twin Cities
    Area New Party (1997) 
    520 U.S. 351
    , 364.)
    c.     The necessity for imposing the burden.
    When the problem is personal empire-building by entrenched incumbents, the only
    solution is term limits. As Eu said, “Realistically, only a lifetime ban could protect
    against various kinds of continued exploitation of the ‘advantages of incumbency’
    captured through past terms in office.” (Eu, supra, 54 Cal.3d at p. 524.)
    The Board argues that the preexisting three-term limit was adequate and “a far less
    drastic alternative” to a one-term limit. The voters, however, were presented with the
    choice between Measure J, which would have kept the three-term limit, and Measure K’s
    one-term limit; they overwhelmingly preferred Measure K.
    27
    The ballot arguments in favor of Measure K said, among other things: “Voting
    YES for term limits and reduced salaries will finally attract representatives interested in
    public service and committed to following the will of the people.” “Obligated to
    financial backers for reelection, the Board of Supervisors has chosen to ignore voters and
    their rights.” “A single four-year term will help shut out . . . outside interests and focus
    our leaders on doing what’s best for us all.” “Measure K will ensure our elected officials
    are inspired by service to San Bernardino County residents, not an oversized paycheck or
    raising money to win their next election.”7
    It was important to the electorate, then, to ensure that supervisors could not be
    either distracted or compromised by the prospect of a future term. A three-term limit
    could never accomplish this goal. A one-term limit was the only alternative.
    D.      Conclusion.
    In sum, in the words of Burdick, the one-term limit here is a “‘reasonable,
    nondiscriminatory restriction[],’ and therefore “‘the State’s important regulatory interests
    are . . . sufficient to justify’” it. (Burdick, supra, 504 U.S. at p. 434.)
    7       As evidence of the purposes of Measure K, Renner cites the statement of
    reasons that appeared on petitions to qualify the initiative. (See Elec. Code, § 9202.)
    Such a statement, which was not made available to all voters, does not constitute
    legislative history. (Cf. Carter v. Commission on Qualifications of Judicial Appointments
    (1939) 
    14 Cal.2d 179
    , 185 [ballot pamphlet, “sent to the electors of the state,” is an
    indicator of intent].)
    28
    V
    THE COMPENSATION LIMIT
    The trial court also ruled that the compensation limit was valid and constitutional.
    In its cross-appeal, the Board contends that this was error, for three reasons.
    A.        Exclusive Delegation.
    First, the Board contends that supervisors’ compensation cannot be set by
    initiative, because state law delegates the power to set supervisors’ compensation
    exclusively to a county’s governing body.
    The trial court rejected this contention; it found “no clear indication that the
    Legislature intended the governing body to exclusively hold the right to set their salary
    within the charter and exclude the use of the initiative power to amend the charter
    associated with the governing body’s compensation.” We agree.
    1.    Analysis under the California Constitution.
    California has two kinds of counties: charter counties and general law counties.
    (Cal. Const., art. XI, § 3, subd. (a); Roe v. County of Lake (N.D. Cal. 2000) 
    107 F.Supp.2d 1146
    , 1148.) San Bernardino is a charter county. (S.B. County Charter; Stats.
    1913, ch. 33.)
    Article XI, section 3, subdivision (a) (section 3(a)) of the Constitution states,
    “County charters . . . shall supersede . . . all laws inconsistent therewith. The provisions
    of a charter are the law of the State and have the force and effect of legislative
    enactments.” (Cal. Const., art. XI, § 3, subd. (a).) Thus, section 3(a) grants a charter
    29
    county “‘home rule,’ i.e., the authority of the people to create and operate their own local
    government and define the powers of that government, within the limits set out by the
    Constitution. [Citation.]” (Dibb v. County of San Diego (1994) 
    8 Cal.4th 1200
    , 1206.)8
    Despite the seeming breadth of section 3(a), a county charter supersedes state law
    only when the county is legislating in its proper sphere. Hence, “charter provisions
    cannot control in matters of statewide concern where the state has occupied the field.”
    (Wilson v. Beville (1957) 
    47 Cal.2d 852
    , 859; accord, American Financial Services Assn.
    v. City of Oakland (2005) 
    34 Cal.4th 1239
    , 1251.)
    Article XI, section 1, subdivision (b) of the California Constitution (section 1(b))
    provides, “[E]ach governing body [of a county] shall prescribe by ordinance the
    compensation of its members . . . .” However, section 1(b) expressly applies only
    “[e]xcept as provided in subdivision (b) of Section 4 of this article . . . .” (Ibid.) And
    Article XI, section 4, subdivision (b) says that a county charter must provide for “[t]he
    8      Charter cities also have home rule. (Cal. Const., art. XI, § 5, subd. (a).)
    “But the ‘version of “home rule” afforded to a charter city is substantially more
    expansive’ than that granted to charter counties. [Citation.] While charter cities are
    granted broad authority over ‘municipal affairs,’ ‘[t]here is no corresponding grant of
    authority and autonomy over the “county affairs” of charter counties. [Citations.]’
    [Citation.]” (People v. Minor (2002) 
    96 Cal.App.4th 29
    , 40.)
    Subject to that caveat, case law dealing with charter cities also applies to charter
    counties. (See, e.g., Marquez v. City of Long Beach (2019) 
    32 Cal.App.5th 552
    , 562, fn.
    5 [“The cases addressing the home rule doctrine have applied the same analysis to the
    authority of charter counties . . . and charter cities . . . to set the compensation for their
    employees.”].)
    30
    compensation . . . of members of the governing body.” (Cal. Const., art. XI, § 4, subd.
    (b) (section 4(b)).)
    In other words, in a general law county, the “governing body” must set its
    members’ compensation. By contrast, in a charter county, the county charter may
    provide the method of setting the compensation of members of the governing body; the
    governing body has no power to set the compensation of its members, unless the county
    charter so provides. (See Brown v. Francisco (1954) 
    123 Cal.App.2d 413
    , 417 [county
    charter may fix supervisors’ compensation or may allow supervisors to do so].) Last but
    not least, the voters have the right to amend a county charter by initiative. (Cal. Const.,
    art. XI, § 3, subds. (a), (b).) It follows that the voters in a charter county may set
    supervisors’ compensation by an initiative that amends the county charter.
    The Board asserts: “[T]he California Constitution is silent regarding whether
    citizen initiatives may initially set supervisor compensation in charter counties.” “Where
    the Constitution is silent as to a charter county’s authority to engage in a legislative act,
    courts defer to the legislature . . . .” As just discussed, however, the state Constitution is
    far from silent on this point. Indeed, it could hardly be clearer.
    Jahr v. Casebeer (1999) 
    70 Cal.App.4th 1250
     (Jahr) held that, under section 1(b),
    supervisors’ compensation cannot be set by initiative. (Id. at pp. 1254-1260.) Meldrim v.
    Board of Supervisors (1976) 
    57 Cal.App.3d 341
     (Meldrim) is in accord. (Id. at pp. 343-
    344.) Jahr, however, involved Shasta County (Jahr, supra, at pp. 1252-1253), a general
    law county. (California State Association of Counties, County Structure & Powers,
    31
    available at  [as of
    May 3, 2023].) And Meldrim involved Contra Costa County, which, as the court there
    observed, is also a general law county. (Meldrim, supra, 57 Cal.App.3d at p. 343.)
    Meldrim even acknowledged that “the State Constitution . . . gives charter counties the
    right to set the salaries of their supervisors. [Citation.]” (Id. at pp. 343-344.)
    Admittedly, Jahr stated flatly that “the process through which supervisor salaries
    are established is a matter of statewide concern.” (Jahr, supra, 70 Cal.App.4th at
    p. 1259.) Because Jahr involved a general law county, however, the court was not called
    upon to decide any issue regarding a charter county. If this statement in Jahr applied to a
    charter county, it would conflict with section 4(b), which leaves it up to a charter county
    to decide how to set supervisors’ compensation, as well as with the broad home rule that
    section 3(a) confers on charter counties. It is perfectly consistent to say that supervisor
    compensation may be a statewide concern in general law counties, but it is a local
    concern in charter counties exercising their constitutional right to home rule.
    2.      Analysis under Government Code section 25300.
    The Board relies on Government Code section 25300 (section 25300) —
    particularly as it was amended, while this appeal was pending, by AB 428.
    When the trial court ruled, section 25300 provided, “The board of supervisors
    shall prescribe the compensation of all county officers . . . .” (Former § 25300, Stats.
    1974, ch. 661, § 2, p. 1523.) As the trial court recognized, however, “officers of a
    32
    county” is defined as including the members of a board of supervisors. (Gov. Code,
    § 24000, subd. (o).)
    Section 25300 now specifically provides: “The board of supervisors shall
    prescribe the compensation of all county officers, including the board of supervisors
    . . . .” (Italics added.)
    The Board argues that this constitutes an “exclusive delegation” of the power to
    set supervisors’ compensation. “[T]he local electorate’s right to initiative . . . is generally
    co-extensive with the legislative power of the local governing body. [Citation.]” (DeVita
    v. County of Napa (1995) 
    9 Cal.4th 763
    , 775.) However, “[t]he presumption in favor of
    the right of initiative is rebuttable upon a definite indication that the Legislature, as part
    of the exercise of its power to preempt all local legislation in matters of statewide
    concern, has intended to restrict that right. [Citations.]” (Id. at p. 776.)
    The Board’s argument assumes, however, that section 25300 applies to charter
    counties. It does not, for three reasons.
    First, section 4(b) says that a county charter must provide for “[t]he compensation
    . . . of members of the governing body.” Thus, if section 25300 purported to apply to a
    charter county, it would be unconstitutional. This is true even if setting supervisors’
    salaries is a statewide matter, and hence even if the county charter does not supersede
    section 25300. The Constitution itself supersedes section 25300 directly.
    Second, section 25300 does not actually purport to apply to a charter county
    (except when the charter allows the board of supervisors to set the compensation of its
    33
    own members). The second sentence of section 25300 adds, “Except as otherwise
    required by Section . . . 4 of Article XI of the California Constitution, such action may be
    taken by resolution of the board of supervisors as well as by ordinance.” And section 4
    of article XI does require otherwise; in a charter county, it requires that the county charter
    provide for supervisors’ compensation.
    Admittedly, the first sentence of section 25300 — “The board of supervisors shall
    prescribe the compensation of . . . the board of supervisors” — does not have any similar
    exclusion. Nevertheless, it is “our duty to read the elements of the statute together,
    harmonizing and giving effect to them all. [Citations.]” (Hampton v. County of San
    Diego (2015) 
    62 Cal.4th 340
    , 351.) Both the first and second sentences must be read as
    qualified by the second. Even assuming the first sentence gives a charter county board of
    supervisors the power to set their own compensation, the second takes it away by
    forbidding them to do it either by ordinance or by resolution. The principle “that courts
    should, if reasonably possible, construe a statute ‘in a manner that avoids any doubt about
    its [constitutional] validity [citations]’ [citations]” supports this interpretation. (Kleffman
    v. Vonage Holdings Corp. (2010) 
    49 Cal.4th 334
    , 346, brackets in original)
    Third, the legislative history of AB 428 demonstrates that the Legislature did not
    intend section 25300 to apply to a charter county. Admittedly, the author of the bill
    stated: ““[I]t is the Legislature’s duty to ensure that our local governments, whether
    established through general law or charter, are equipped with the tools needed to
    properly administer the heavy burdens we place upon them. . . . We must clarify existing
    34
    law to allow for County Supervisors to obtain reasonable compensation . . . .” (Assem.
    Com. on Elections, Analysis of Assem. Bill No. 428, supra, at p. 2, italics added.)
    Other voices, however, soon chimed in. A Senate committee bill analysis stated:
    “Consistent with existing law, AB 428 reiterates that members of the board of supervisors
    are county officers for which the board must set compensation. However, because the
    California Constitution says that charters must provide for the compensation of the board,
    charters that set compensation trump state law. Accordingly, while AB 428 may clarify
    matters for general law counties, it is unclear that it materially changes how
    compensation is set in charter counties that have charter provisions setting a specific
    compensation level for members of the board.” (Sen. Com. on Gov. & Fin., Analysis of
    Assem. Bill No. 428 (2021-2022 Reg. Sess.), as amended Mar. 18, 2021, p. 4, italics
    added.)
    A bill analysis by another committee concurred: “Supervisor Compensation in
    Charter Counties. AB 428 would clarify, in Government Code Section 25300, that the
    ‘county officers’ for which the board of supervisors sets compensation includes the board
    itself. This amendment is consistent, as to general law counties, with provisions of the
    California Constitution requiring each county governing body to set the compensation of
    its members. However, this clarification may not apply to all charter counties, because
    the California Constitution specifies, as an exception to the general rule of county
    governing bodies setting their own compensation, that a county charter must provide for
    the compensation of the board, and some county charters do not give the board discretion
    35
    to set its own compensation. Accordingly, while AB 428 may clarify matters for general
    law counties, it is unclear that it materially changes how compensation is set in charter
    counties that have charter provisions setting a specific compensation level for members
    of the board.” (Sen. Com. on Elections and Const. Amends., Analysis of Assem. Bill No.
    428 (2021-2022 Reg. Sess.), as amended Mar. 18, 2021, p. 6, italics added.)
    The Board notes that the Howard Jarvis Taxpayers Association (Jarvis
    Association) opposed AB 428. The Jarvis Association argued: “‘AB 428 seeks to undo
    the overwhelming approval of Measure K in San Bernardino County that amended the
    County Charter to impose a term limit of one term and reduced the total compensation for
    each member of the Board of Supervisors to $5,000 per month.” (Assem. Com. on
    Elections, Analysis of Assem. Bill No. 428, supra, at p. 4.) The Board concludes that
    “the Legislature contemplated that the passage of AB 428 would ‘undo’ Measure K.”
    Actually, in the face of the Jarvis Association’s comment (Assem. Com. on
    Elections, Analysis of Assem. Bill No. 428, supra, at p. 1), the author of AB 428
    “indicated that the intent of this bill is not to overturn Measure K.” (Id. at p. 3, italics
    added.)
    The Senate Committee on Governance and Finance observed that the effect of
    AB 428 on Measure K was not clear; it expressed concern about “thwarting the will of
    the voters.” (Sen. Com. on Gov. & Fin., Analysis of Assem. Bill No. 428, supra, at p. 4.)
    In response to this “potential legal ambiguity,” AB 428 was amended so as to provide
    expressly that it “shall not affect any term limits that were legally in effect prior to
    36
    January 1, 2022 . . . .” (Sen. Com. on Elections and Const. Amends., Analysis of Assem.
    Bill No. 428, supra, at pp. 5-6.) There was no need to amend its compensation limit,
    because, as already discussed, the Legislature knew it most likely did not apply to a
    charter county.
    Plainly, then, the Legislature did not contemplate that AB 428 would undo
    Measure K. To the contrary, it agreed with the Jarvis Association that the unamended
    version threatened Measure K, and thus it amended AB 428 so as to let Measure K stand.
    In sum, then, the author of AB 428 may have intended it to apply to charter
    counties. The Legislature as a whole, however, was made aware that section 25300 —
    both before and after AB 428 — most likely did not and could not apply to charter
    counties.
    B.     Minimum Wage Laws.
    Second, the Board contends that the compensation limit violates federal and state
    minimum wage laws.
    The trial court rejected this contention based on lack of evidence. It ruled: “Even
    though it is implied the Board may be paid less than minimum wage after deduction of all
    benefits, such contention is not demonstrated. It is based on speculation.” Again, we
    agree.
    Measure K’s compensation limit is $5,000 a month. The Board points out that the
    minimum wage of $15 an hour for 40 hours a week works out to $2,600 a month. It adds,
    “[F]ederal law requires that the County match a Medicare tax at 1.45 percent of the
    37
    supervisor’s salary, and state and local pension laws require that the County match each
    supervisor’s pension contribution at 17.54 percent of the supervisor’s salary. These
    benefits alone require a minimum of $493.74. When adding in other legally mandated
    benefits such as worker’s compensation premiums, travel reimbursement[9] and disability
    insurance, among others, . . . Measure K’s terms immediately risk requiring that the
    County violate California’s minimum wage law.”
    It must be remembered that this action is a challenge to Measure K on its face.
    “On a facial challenge, we will not invalidate a statute unless it ‘pose[s] a present total
    and fatal conflict with applicable constitutional prohibitions.’ [Citations.] . . .
    ‘[P]etitioners cannot prevail by suggesting that in some future hypothetical situation
    constitutional problems may possibly arise as to the particular application of the statute.’
    [Citation.]” (California School Boards Assn. v. State of California (2019) 
    8 Cal.5th 713
    ,
    723-724.)
    The Board’s evidence fails to establish that Measure K will necessarily violate
    superseding federal and state law. Indeed, as the Board itself says, “it is presently
    impossible to determine whether the County will be in violation of the law . . . .” It
    hypothesizes a future increase of the minimum wage to $20 an hour, and even then it says
    9       Measure K defines “compensation” as “the actual cost to the County of all
    benefits of whatever kind or nature including but not limited to salary, allowances, credit
    cards, health insurance, life insurance, leave, retirement, memberships, portable
    communications devices, and vehicle allowances.” Thus, the compensation limit does
    not apply to the reimbursement of outlays benefiting the county, including travel
    reimbursement.
    38
    only that “a supervisor’s compensation would be a minimum of $4,125 — a number
    which does not include the numerous other required benefit payments under state and
    federal law.” We cannot invalidate Measure K on such hypothetical grounds.
    In a subsidiary contention, the Board argues that, if the county could comply with
    minimum wage laws by having supervisors work part-time, that would impair essential
    government functions. Again, however, this is a purely hypothetical argument that we
    cannot reach in this appeal.
    C.    San Bernardino County Code section 13.0614.
    Third, the Board contends that the compensation limit improperly overturns San
    Bernardino County Code section 13.0614 (section 13.0614) without a referendum.
    The Board did not raise this argument below. Nevertheless, “[g]enerally, on
    appeal, a judgment or order will be affirmed if it is correct on any theory, regardless of
    the trial court’s reasons; thus, a respondent may assert a new theory to establish that an
    order was correct on that theory ‘unless doing so would unfairly prejudice appellant by
    depriving him or her of the opportunity to litigate an issue of fact.’ [Citation.]” (Bailon
    v. Superior Court (2002) 
    98 Cal.App.4th 1331
    , 1339.) The Board’s argument is purely
    legal.
    Section 13.0614 is a lengthy ordinance governing the compensation of elected
    county officials. It requires that supervisors be provided with certain benefits, including
    retirement benefits, medical, dental, and vision insurance, life insurance, long-term
    39
    disability insurance, a cell phone allowance, health club membership, an annual physical
    examination, tuition reimbursement, and membership dues.10 (§ 13.0614, subd. (c).)
    The Board asserts that “Measure K renders it impossible for supervisors to be paid
    the amounts [section 13.0614] requires.” Once again, however (see part IV.B, ante), this
    is speculation; there is no evidence of this.
    Separately and alternatively, even assuming the Board’s assertion is true,
    Measure K is part of the county charter; section 13.0614 is a mere county ordinance.
    Accordingly, to the extent that they conflict, section 13.0614 is invalid. (See City and
    County of San Francisco v. Cooper (1975) 
    13 Cal.3d 898
    , 923-924.)
    The Board implicitly recognizes this principle. It therefore argues that the voters
    cannot invalidate an ordinance by initiative rather than by referendum. It cites article XI,
    section 1, subdivision (b) of the California Constitution, which, as relevant here,
    provides: “Except as provided in subdivision (b) of Section 4 of this article, each
    governing body shall prescribe by ordinance the compensation of its members, but the
    ordinance prescribing such compensation shall be subject to referendum.” Again,
    however (see part III, ante), section 3(b) and section 4(b), taken together, give the voters
    10      It also requires that supervisors be reimbursed for certain expenses, such as
    travel and transportation. (§ 13.0614, subds. (c)(5), (c)(6).) Once again, however,
    reimbursement is not “compensation” for purposes of Measure K’s $5,000 compensation
    limit. (See fn. 9, ante.)
    Supervisors are also entitled to make voluntary contributions that can be used for
    certain purposes (medical, 401(k) plans, 457 plans, dependent care) without being subject
    to income tax. (§ 13.0614, subds. (c)(7), (c)(8), (c)(12).) As these come out of
    supervisors’ salaries, they, too, are not compensation within the meaning of Measure K.
    40
    in a charter county the power to amend a charter by initiative, so no referendum is
    needed.
    VI
    THE SINGLE-SUBJECT RULE
    The Board contends that Measure K, by combining the one-term limit and the
    compensation limit, violates the single-subject rule.
    The trial court rejected this contention. It ruled that both provisions were
    “reasonably germane to ensuring the member of the Board is about public service versus
    being paid a high salary or becoming a career politician.”
    Under the California Constitution, “[a]n 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).) “This rule applies to local as well as statewide initiatives. [Citation.]”
    (Shea Homes Limited Partnership v. County of Alameda (2003) 
    110 Cal.App.4th 1246
    ,
    1255.)
    “‘[The California Supreme Court has] upheld a variety of initiative measures in
    the face of a single-subject challenge, emphasizing that the initiative process occupies an
    important and favored status in the California constitutional scheme and that the single-
    subject requirement should not be interpreted in an unduly narrow or restrictive fashion
    that would preclude the use of the initiative process to accomplish comprehensive, broad-
    based reform in a particular area of public concern. [Citations.]’ [Citation.]
    41
    “‘[T]he single-subject provision does not require that each of the provisions of a
    measure effectively interlock in a functional relationship. [Citation.] It is enough that the
    various provisions are reasonably related to a common theme or purpose.’ [Citation.]
    Accordingly, [the Supreme Court has] upheld initiative measures ‘“which fairly disclose
    a reasonable and common sense relationship among their various components in
    furtherance of a common purpose.” [Citation.]’ [Citations.] The governing principle is
    that ‘“‘[a]n initiative measure does not violate the single-subject requirement if, despite
    its varied collateral effects, all of its parts are “reasonably germane” to each other,’ and
    to the general purpose or object of the initiative. [Citations.]”’ [Citation.] The
    ‘reasonably germane’ standard is applied ‘in an accommodating and lenient manner so as
    not to unduly restrict . . . the people’s right to package provisions in a single bill or
    initiative.’ [Citations.]” (Briggs v. Brown (2017) 
    3 Cal.5th 808
    , 828-829.)
    The Board does not attack the trial court’s ruling that both provisions of
    Measure K are reasonably germane to ensuring that Board members are devoted to public
    service. Instead, it argues only that, if Measure K is severable, then it necessarily
    concerns more than one subject. It provides no authority for this argument. The test
    under the single-subject rule (all provisions must be “reasonably germane” to each other)
    is very different from the test for severability (“‘[t]he invalid provision must be
    grammatically, functionally, and volitionally separable”). (California Redevelopment
    Assn. v. Matosantos (2011) 
    53 Cal.4th 231
    , 271.)
    42
    We need not decide this issue. The trial court ruled that Measure K is not
    severable. The Board agrees. Because we are upholding both the one-term term limit
    and the compensation limit, there is no need for us to decide whether Measure K is
    severable. And because the Board is not contending that Measure K is severable, we
    need not decide whether, if it were severable, it would violate the single-subject rule.
    VII
    APPLICATION OF MEASURE K TO THE NEW SUPERVISORS
    Renner contends that the trial court erred by ruling that Measure K did not apply
    to the newly elected supervisors. This depends, in part, on when the new supervisors
    took office and on when Measure K went into effect. Renner and the Board disagree on
    both points.
    A.      When Did the New Supervisors Take Office?
    Under both the old county charter and Measure J, the new supervisors’ terms were
    to begin on the first Monday in December, which in 2020 was December 7.11
    Accordingly, on December 7, 2020, the new supervisors were sworn in. The trial court
    ruled that, pursuant to the county charter, the new supervisors’ terms did, in fact, begin
    on December 7, 2020.
    Renner relies on article II, section 20 of the California Constitution, which
    provides: “Terms of elective offices provided for by this Constitution, other than
    11     Measure J also provided that, after 2020, supervisors’ terms would begin on
    the first Monday after the first day of January following their election.
    43
    Members of the Legislature, commence on the Monday after January 1 following
    election.” Our Supreme Court, however, has held that this governs only state elective
    offices, not local elective offices. (In re Stuart (1879) 
    53 Cal. 745
    , 748 (Stuart); accord,
    Barton v. Kalloch (1880) 
    56 Cal. 95
    , 104-105.)
    Stuart noted that “State officers, such as the Governor and the other officers who
    constitute the Executive Department of the State Government . . . are officers whose
    election is absolutely provided for by the Constitution itself.” (Stuart, supra, 53 Cal. at
    p. 748.) It then relied on former article XI, section 5, of the California Constitution,
    which allowed the Legislature to “fix the[] terms of office” of “county . . . and municipal
    officers . . . .” (Stuart, supra, at p. 748.)
    Renner argues, however, that Stuart is no longer good law. She claims that, when
    Stuart was decided, the only officers whose election was “provided for by th[e]
    Constitution” were state officers; since then, however, the Constitution has been amended
    to provide for county supervisors. Not so.
    In the original 1879 Constitution, article V, section 2 provided: “The Governor
    shall be elected by the qualified electors at the time and places of voting for members of
    the Assembly, and shall hold his office four years from and after the first Monday after
    the first day of January subsequent to his election . . . .” (Italics added.)
    By contrast, article XI, section 5 of the Constitution, on which Stuart relied, said:
    “The Legislature . . . shall provide for the election or appointment, in the several
    counties, of Boards of Supervisors . . . and shall . . . fix their terms of office.” (Italics
    44
    added.) Thus, Stuart drew a distinction between state officers like the governor, “whose
    election is absolutely provided for by the Constitution itself,” and county supervisors,
    whose election and terms were to be provided for by the Legislature. (Stuart, supra, 53
    Cal. at p. 748.)
    Similarly, the Constitution now states: “The Governor shall be elected every
    fourth year at the same time and places as members of the Assembly and hold office from
    the Monday after January 1 . . . .” (Cal. Const., art. V, § 2, italics added.) It also states:
    “The Legislature shall provide for . . . an elected governing body in each county.” (Cal.
    Const., art XI, § 1, subd. (b), italics added.) Moreover, it provides that, in a charter
    county, the charter is to provide for “[t]he . . . terms . . . of members of the governing
    body.” (Cal. Const., art XI, § 4, subd. (b).) Thus, Stuart’s reasoning is still sound — at
    least as to a charter county, the Legislature is to provide for the election of supervisors,
    and the county charter is to provide for their terms. Article II, section 20 still does not
    apply.
    We also note that, in several charter counties, supervisors do not take office on the
    Monday after January 1. Los Angeles County provides by charter for supervisors’ terms
    to begin on the first Monday in December. (L.A. County Charter, art. II, § 6.)
    Sacramento and Santa Clara Counties provide for them to begin on the first Monday in
    January. (Sac. County Charter, art. IV, § 9; Santa Clara County Charter, art. II, § 202.)
    Likewise, San Francisco provides for them to begin on January 8. (S.F. Charter,
    45
    § 2.101.) This reflects a general understanding that a charter county can choose its own
    date.
    Thus, we agree with the trial court that the new supervisors took office on
    December 7, 2020.
    B.     When Did Measure K Go into Effect?
    Measure K itself provided: “To the extent permitted by law, the provisions of this
    Charter Amendment shall be effective on voter approval of the initiative as provided by
    California law.” The introductory clause — “[t]o the extent provided by law” — seems
    to indicate that it was not intended to override any statutory and constitutional provisions
    that may apply. We need not decide this, however, because Renner does not argue that
    Measure K became effective immediately. We deem that argument forfeited.
    The parties have cited multiple statutes that they claim govern when a charter
    amendment goes into effect.
    First, Renner cites Article II, section 10, subdivision (a) of the California
    Constitution, which provides: “An initiative statute . . . approved by a majority of votes
    cast thereon takes effect on the fifth day after the Secretary of State files the statement of
    the vote for the election at which the measure is voted on . . . .”
    Case law establishes that Article II, section 10, subdivision (a) of the Constitution
    applies only to statewide initiatives; local initiatives are governed by a parallel statutory
    scheme, which includes Elections Code section 9122. (Howard Jarvis Taxpayers
    Association v. City and County of San Francisco (2021) 
    60 Cal.App.5th 227
    , 239; City
    46
    and County of San Francisco v. All Persons Interested in Matter of Proposition C (2020)
    
    51 Cal.App.5th 703
    , 709-710 & 710, fn. 2.)
    Government Code section 23723 provides: “[A] proposed [charter12] amendment
    . . . shall not take effect until accepted and filed by the Secretary of State . . . .” (Gov.
    Code, § 23723; see also id., § 23724.) The trial court ruled that this was the controlling
    provision.13
    However, Elections Code section 9102 provides: “Any proposal to . . . amend . . .
    a county charter by initiative petition . . . shall be subject to this article.” Elections Code
    section 9122 — part of the same article — then provides: “If a majority of the voters
    voting on a proposed ordinance vote in its favor, the ordinance shall become a valid and
    binding ordinance of the county. The ordinance shall be considered as adopted upon the
    date the vote is declared by the board of supervisors, and shall go into effect 10 days after
    that date.”
    There is an inescapable conflict between Elections Code sections 9102 and 9122,
    on one hand, and Government Code section 23723, on the other. They are
    “‘irreconcilable, clearly repugnant, and so inconsistent that the two cannot have
    12       While Government Code section 23723 does not use the word “charter,” it
    is part of a statutory scheme dealing exclusively with the proposal, revision, amendment,
    and repeal of county charters. (Gov. Code, §§ 23700-23732.)
    13     The trial court mistakenly cited Government Code sections 23712 and
    23713. Those sections apply to a charter proposal or revision. However, Government
    Code sections 23723 and 23724, which apply to a charter amendment (as here) or repeal,
    are otherwise essentially identical to Government Code sections 23712 and 23713. Thus,
    the error was plainly harmless.
    47
    concurrent operation.’ [Citation.]” (See Lopez v. Sony Electronics, Inc. (2018) 
    5 Cal.5th 627
    , 637, internal quotation marks omitted.)
    The Board plumps for Government Code section 23723. It cites Board of
    Supervisors v. McMahon (1990) 
    219 Cal.App.3d 286
     (McMahon), which remarked that
    the county charter amendment in that case “became effective . . . when the Secretary of
    State accepted and filed it. (See Gov. Code § 23723.)” (Id. at p. 292.) McMahon,
    however, was decided in March 1990; as we will discuss in (perhaps excruciating) detail
    momentarily, the statutory predecessors of Elections Code sections 9102 and 9122 did
    not become effective until September 1990. In any event, there was no issue in
    McMahon as to the precise date on which the charter amendment there became effective.
    “It is axiomatic that cases are not authority for propositions that are not considered.
    [Citation.]” (California Building Industry Assn. v. State Water Resources Control Bd.
    (2018) 
    4 Cal.5th 1032
    , 1043.)
    “‘If conflicting statutes cannot be reconciled, later enactments supersede earlier
    ones [citation], and more specific provisions take precedence over more general ones
    [citation].’ [Citation.] But when these two rules are in conflict, the rule that specific
    provisions take precedence over more general ones trumps the rule that later-enacted
    statutes have precedence. [Citations.]” (State Dept. of Public Health v. Superior Court
    (2015) 
    60 Cal.4th 940
    , 960-961.)
    We cannot say that either set of statutes is more specific than the other. Arguably,
    Elections Code section 9122, standing alone, is more general than Government Code
    48
    section 23723, because it applies to initiative county ordinances as well as initiative
    county charter amendments. However, Elections Code section 9102 very specifically
    makes Elections Code section 9122 applicable to initiative county charter amendments.
    We turn, then, to which set of statutes was enacted latest.
    Government Code section 23723 was enacted in 1969 (Stats. 1969, ch. 1264, § 1,
    p. 2470) and most recently amended in 1975. (Stats. 1975, ch. 238, § 10, p. 627.)
    Elections Code sections 9102 and 9122 were both enacted in 1994. (Stats. 1994,
    ch. 920, § 2, pp. 4690-5208.)
    Elections Code section 9122 replaced Elections Code former section 3716 (Stats.
    1994, ch. 920, § 2, p. 5168), enacted in 1976. (Stats. 1976, ch. 248, § 3, p. 503.) That, in
    turn, replaced Elections Code former section 3717, enacted in 1961. (Stats. 1961, ch. 23,
    § 3717, p. 632.) That replaced Elections Code former section 1617, enacted in 1939.
    (Stats. 1939, ch. 26, p. 632.) And that replaced former Political Code section 4058,
    enacted in 1911. (Stats. 1911, ch. 342, § 1, pp. 577-580.)
    The most recent three of these were all identically worded; the 1911 original had
    more verbiage, but it was identical in all relevant respects. Thus, like current Elections
    Code section 9122, all of its statutory predecessors were worded in terms of an initiative
    county “ordinance.” Standing alone, they did not purport to apply to an initiative
    amendment to a county charter.
    Meanwhile, Elections Code section 9102 replaced Elections Code former section
    3701.5, enacted in 1990 (Stats. 1990, ch. 1161, § 12, p. 4864) and effective on September
    49
    21, 1990. (Stats. 1990, ch. 1161, § 32, pp. 4857, 4870.) Again, the wording of Elections
    Code former section 3701.5 was identical to current Elections Code section 9102.
    Crucially, before 1990, there was no statute (like Elections Code section 9102)
    making an initiative county charter amendment subject to Elections Code section 9122 or
    any of its statutory predecessors.
    It follows that Elections Code sections 9102 and 9122, taken together, must be
    regarded as enacted in 1990. (Gov. Code, § 9605, subd. (a).) Thus, they superseded
    Government Code section 23723. And thus, under Elections Code sections 9102 and
    9122, Measure K went into effect 10 days after the vote was declared by the Board —
    namely, on December 18, 2020.
    C.     Is Measure K Impermissibly Retroactive If It Applies to the New
    Supervisors?
    As we have just held, the new supervisors took office on December 7, 2020 (see
    part VI.A, ante), and Measure K went into effect on December 18, 2020 (see part VI.B,
    ante). The Board contends that, if Measure K applies to the new supervisors, it is
    retroactive in violation of Government Code section 25000, subdivision (b).
    Government Code section 25000, subdivision (b), provides: “Any proposal to
    limit the number of terms a member of the board of supervisors may serve on the board
    of supervisors shall apply prospectively only.” This language was not added by AB 428;
    it was enacted in 1995. (Stats. 1995, ch. 432, § 4, pp. 3407-3408.)
    50
    Even aside from Government Code section 25000, “statutes do not operate
    retrospectively unless the Legislature plainly indicates otherwise. [Citation.]” (In re S.B.
    (2004) 
    32 Cal.4th 1287
    , 1296.) Nothing in Measure K indicates that it intended to
    operate retroactively.14 The question, then, is whether Measure K is impermissibly
    retroactive if it applies to the new supervisors.
    “‘[D]eciding when a statute operates “retroactively” is not always a simple or
    mechanical task’ [citation] and ‘comes at the end of a process of judgment concerning the
    nature and extent of the change in the law and the degree of connection between the
    operation of the new rule and a relevant past event’ [citation]. In exercising this
    judgment, ‘familiar considerations of fair notice, reasonable reliance, and settled
    expectations offer sound guidance.’ [Citation.]
    “In general, application of a law is retroactive only if it attaches new legal
    consequences to, or increases a party’s liability for, an event, transaction, or conduct that
    was completed before the law’s effective date. [Citations.] Thus, the critical question for
    determining retroactivity usually is whether the last act or event necessary to trigger
    application of the statute occurred before or after the statute’s effective date. [Citations.]
    A law is not retroactive ‘merely because some of the facts or conditions upon which its
    application depends came into existence prior to its enactment.’ [Citation.]” (People v.
    Grant (1999) 
    20 Cal.4th 150
    , 157.)
    14     For this reason, we need not decide whether, under the home rule doctrine
    (see part IV.A.1, ante), Measure K overrides Government Code section 25000.
    51
    Armstrong v. County of San Mateo (1983) 
    146 Cal.App.3d 597
     (Armstrong) dealt
    with article XIIIA, section 2, subdivision (b) of the California Constitution, which went
    into effect on July 1, 1978. (Id. at p. 604.) It allowed tax authorities to assess real
    property for the 1978-79 tax year at its assessed value in the 1975-76 tax year, plus an
    “inflation factor” of two percent per year. (Id. at pp. 604-605.)
    The appellate court held that this was not retroactive. (Armstrong, supra, 146
    Cal.App.3d at pp. 613-614.) “It is well settled that ‘[a] statute does not operate
    retroactively merely because some of the facts or conditions upon which its application
    depends came into existence before the enactment.’ [Citation.] Application of the
    inflation factor prior to the effective date of article XIIIA does not give the 1975-76 full
    cash value assessment or the pre-1978 adjustments thereof ‘an effect different from that
    which they had under previously existing law,’ such as, for example, by retroactively
    increasing taxes due in 1975-76 or any other tax year prior to the effective date of the
    article. Rather it merely utilizes facts existing prior to enactment of the article to
    determine tax rates to be applied prospectively from the effective date.” (Ibid., fn.
    omitted.)
    In Sheyko v. Saenz (2003) 
    112 Cal.App.4th 675
     (Sheyko), the Legislature passed a
    statute requiring aid applicants and recipients to be fingerprinted in accordance with the
    Statewide Fingerprint Imaging System (SFIS) as a condition of eligibility for benefits.
    (Id. at pp. 684-686.) The trial court ruled that, as applied to persons who were applicants
    52
    but not recipients when the statute went into effect, the statute was impermissibly
    retroactive. (Id. at p. 702.)
    The appellate court held that the statute was not retroactive at all. (Sheyko, supra,
    112 Cal.App.4th at pp. 702-703.) It explained: “‘Application . . . is retroactive only
    when it gives a different and potentially unfair legal effect to actions taken in reliance on
    the preenactment law.’ [Citation.]” (Id. at p. 702.) “SFIS compliance is an eligibility
    requirement [citation] and recipients lose their entitlement as soon as they lose their
    eligibility. [Citation.] It would be no different than the Legislature changing some other
    eligibility condition, such as maximum income: That new cutoff would apply to all
    recipients, regardless of when they became recipients. No recipient has a reasonable
    expectation that eligibility conditions will not change. Nobody loses past benefits for
    present noncompliance with SFIS.” (Id. at pp. 702-703.) “[E]ligibility is continually
    reevaluated to see if the present conditions of eligibility are met. This means the SFIS
    regulations do not ‘materially alter the legal significance’ of applying for benefits, nor
    give ‘a different and potentially unfair legal effect to actions taken in reliance on the
    preenactment law.’ [Citation.]” (Id. at p. 703.)
    Here, once the new supervisors were elected, they did have a vested right to
    remain in office for one term. However, they had no vested right to run for a second
    term. They could hardly have had a reasonable expectation that they could run for a
    second term, as Measure K had already been proposed and was on the same ballot as they
    53
    were. In any event, the Board does not point to any way in which they relied on any such
    expectation.
    Much as in Sheyko, the change in the law here affects the new supervisors’
    eligibility for future office. Thus, they could always lose their eligibility due to a
    statutory change. That is true even if, as here, “‘some of the facts or conditions upon
    which its application depends came into existence prior to its enactment.’ [Citation.]”
    (People v. Grant, 
    supra,
     20 Cal.4th at p. 157.) As in Armstrong, the one-term limit
    merely looks at facts existing prior to its enactment to determine eligibility, to be applied
    prospectively from the provision’s effective date. The one-term limit is analogous to a
    minimum age requirement, which looks at an event in the past — the person’s birth — to
    determine his or her present eligibility to drink, smoke, run for office, etc.
    Measure K cannot and does not kick out of office those supervisors who were
    already serving a second or third term. That would indeed be a prohibited retroactive
    application. However, it will bar them — just as it bars the new supervisors — from
    serving another term later.15
    D.      Does Measure K Violate AB 428 If It Applies to the New Supervisors?
    The Board contends that applying Measure K to the new supervisors would violate
    AB 428.
    15     We therefore do not discuss Renner’s contention that the charter provision
    that supervisors’ terms begin on the first Monday of December is unconstitutional.
    54
    Government Code section 25000, subdivision (b), as amended by AB 428,
    provides:
    “(1) . . . [T]he residents of [a] county may propose, by initiative, a proposal to
    limit to no fewer than two terms . . . a member of the board of supervisors may serve on
    the board of supervisors. . . .
    “(2) The changes made to this subdivision by [AB 428] shall not affect any term
    limits that were legally in effect prior to January 1, 2022, in any county.”
    As we held in part VI.B, ante, Measure K went into effect on December 18, 2020.
    Accordingly, the one-term limit does not violate AB 428.
    The Board argues that the Legislature intended AB 428 to overturn Measure K.
    “It is well established, however, that legislative intent should not be resorted to where a
    statute is clear on its face.” (Greb v. Diamond Internat. Corp. (2013) 
    56 Cal.4th 243
    ,
    256.) In any event, as we discussed in part IV.A.2, ante, the Legislature did not intend
    AB 428 to abrogate Measure K; indeed, it amended AB 428 and added the January 1,
    2022 start date to make sure it would not.
    55
    E.     Does Measure K Violate Provisions Protecting Supervisors’
    Compensation?
    1.     Statutory protection against reduction of compensation.
    The Board relies on Government Code section 1235,16 which provides: “The
    salary of any elected public office shall not be reduced during an election year after any
    candidate for that particular office has filed the requisite forms declaring his or her
    candidacy for that particular office.” Renner does not contend that, under the home rule
    doctrine (see part IV.A.1, ante), Measure K trumps Government Code section 1235. We
    deem her to have forfeited any such contention.
    Again, Measure K went into effect on December 18, 2020. (See part VI.B, ante.)
    That was “during an election year.” While the record does not show exactly when the
    new supervisors filed their candidacy forms, it had to be sometime before they were
    elected, on November 3, 2020. It follows that the compensation limit cannot apply to the
    new supervisors.
    2.     Constitutional protection against reduction of compensation.
    The Board also asserts that its members have constitutional protection against the
    reduction of their compensation during their terms. Somewhat unhelpfully, Renner does
    not respond to this contention.
    16      There are two section 1235s in the Government Code. We refer to the one
    enacted in 1980 (Stats. 1980, ch. 737, § 1), not the one enacted in 1994 (Stats. 1994,
    ch. 991, § 2).
    56
    Because we have just held that, under Government Code section 1235, the
    compensation provision of Measure K cannot apply to the new supervisors (see part
    VI.E.1, ante), we do not decide this question. (See Thompson v. Department of
    Corrections (2001) 
    25 Cal.4th 117
    , 129 [“we do not address constitutional questions
    unless necessary.”].)
    VIII
    THE TEMPORARY RESTRAINING ORDER
    Renner contends that, whether Measure K was valid or not, the trial court erred by
    enjoining City Clerk Monell from certifying, authenticating, and recording it.
    A county charter amendment must be “certified and authenticated by the
    chairperson and clerk of the governing body[,] . . . attested by the county elections
    official, . . . recorded in the office of the recorder of the county[,] . . . filed in the office of
    the county elections official,” and “file[d] . . . with the Secretary of State . . . .” (Gov.
    Code, § 23713, incorporated by Gov. Code, § 23724.)
    Renner does not cite any portion of the record showing that the trial court ever
    actually issued an injunction against this. She has not included the TRO itself in the
    record, and, as noted in part III, ante, the trial court never issued a permanent injunction.
    Thus, Renner has forfeited this contention. (See Cal. Rules of Court, rule 8.204(a)(1)(C);
    Brown v. El Dorado Union High School Dist. (2022) 
    76 Cal.App.5th 1003
    , 1021.)
    Separately and alternatively, any such error is harmless and/or moot. Measure K
    went into effect on December 18, 2020, regardless of when or even whether it was
    57
    certified, authenticated, recorded, or filed. (See part VI.B, ante.) Renner has not pointed
    out any other way in which Monell’s failure to perform her statutory duties was
    prejudicial.
    In any event, when we reverse a judgment, we have the power to “direct that the
    parties be returned so far as possible to the positions they occupied before the
    enforcement of . . . the judgment or order. In doing so, the reviewing court may order
    restitution on reasonable terms and conditions of all property and rights lost by the
    erroneous judgment or order, so far as such restitution is consistent with rights of third
    parties . . . .” (Code Civ. Proc., § 908.) Hence, we will direct the trial court to allow
    county officials to perform their duties, and also to order that their performance be
    deemed to relate back, nunc pro tunc, to the dates when they should have performed
    them.
    At first glance, these dates might appear speculative. Significantly, however,
    Measure J and Measure K were passed at the same time. It stands to reason that
    Measure K would have been certified, authenticated, recorded, and filed at the same time
    as Measure J (unless the Board deliberately slow-walked Measure K). Therefore, we will
    direct the trial court, on remand, to ascertain when Measure J was certified, authenticated,
    recorded, and filed and to set forth these dates in its nunc pro tunc order.
    58
    IX
    DISPOSITION
    The trial court’s order granting the petition is reversed. On remand, the trial court
    must ascertain the dates on which Measure J was certified, authenticated, recorded, and
    filed. It must then enter a new order, granting a writ and/or an injunction prohibiting the
    application of the compensation limit to the new supervisors, but otherwise denying the
    petition and entering judgment against the Board on the complaint. The order must
    provide that Monell and all other county officials may carry out their duties to certify,
    authenticate, record, and file Measure K. It must also provide that Measure K is deemed
    to have been certified, authenticated, recorded, and filed on the same dates as Measure J,
    and it must specify those dates.
    Renner is awarded costs on appeal against the Board.
    CERTIFIED FOR PARTIAL PUBLICATION
    RAMIREZ
    P. J.
    I concur:
    MILLER
    J.
    59
    [San Bernardino County Board of Supervisors v. Monell; Renner, E077772]
    MENETREZ, J., Dissenting.
    Nearly every issue in this appeal is moot, so nearly everything in the majority
    opinion is advisory. Because I would not issue an advisory opinion on moot issues, I
    respectfully dissent.
    Measure K, approved by the voters of San Bernardino County (the County) in
    2020, imposed certain term limits and compensation limits for members of the board of
    supervisors. After passage, Measure K was challenged in the superior court, which
    invalidated it as unconstitutional. This appeal was taken from that ruling.
    Measure D, approved by the County’s voters in 2022, imposed new term limits
    and compensation limits for members of the board of supervisors. Measure D supersedes
    Measure K in its entirety. Measure D has been challenged in court, but to date no part of
    it has been invalidated.
    After Measure D passed, the County requested that we stay this appeal pending
    resolution of the Measure D litigation. We invited and received supplemental briefs from
    the parties on the stay request and whether the appeal is moot. The County’s request for
    a stay was denied (over my objection). But the County renewed the request at oral
    argument, and we can also stay an appeal on our own motion.
    Because Measure D is currently in effect, no provision of Measure K is currently
    in effect. Consequently, it is presently impossible for us to grant effective relief on this
    appeal. The appeal is from an order invalidating Measure K. But no matter what we do,
    1
    Measure K cannot go back into effect unless and until Measure D is invalidated. The
    appeal is therefore moot. (Lockaway Storage v. County of Alameda (2013) 
    216 Cal.App.4th 161
    , 174–175.)
    The only issue that presently is not moot is whether Measure K’s compensation
    limits apply from 2020 to 2022, between the passage of Measure K and the passage of
    Measure D. But in the future, additional issues might cease to be moot if Measure D is
    invalidated in whole or in part. We should accordingly stay this appeal until there is a
    final judgment in the Measure D litigation. Only then will we have a final determination
    of which issues concerning Measure K are not moot. At present, only one of them is.
    I would vacate submission and stay the appeal pending final resolution of the
    Measure D litigation. I therefore respectfully dissent.
    MENETREZ
    J.
    2