Debbane v. City and County of San Francisco

Filed 9/11/26 CERTIFIED FOR PUBLICATION IN THE COURT OF APPEAL OF THE STATE OF CALIFORNIA FIRST APPELLATE DISTRICT DIVISION ONE ERIC DEBBANE et al., Plaintiffs and Respondents, v. A172067 CITY AND COUNTY OF SAN FRANCISCO et al., (San Francisco City & County Super. Ct. No. CGC-23-604600) Defendants and Appellants. In 2022, San Francisco voters approved Proposition M to “tax” owners of certain residential units in buildings with more than two units for any unit kept “vacant” for more than 182 days, which need not be consecutive, during a tax year. (S.F. City & County Sample Ballot and Voter Information Pamp., Gen. Elec. (Nov. 8, 2022) text of Prop. M, pp. 240–242 (hereafter “Pamphlet”).) The stated purpose of the measure is to “disincentivize prolonged vacancies, thereby increasing the number of housing units available for occupancy.” (Id. at p. 241.) Plaintiffs filed the instant action against the City and County of San Francisco and two of its individual officers, challenging the proposition on both statutory and constitutional grounds. They also named as real party in interest, San Francisco’s Treasurer.1 Eventually, the parties filed cross- motions for summary judgment. After full briefing and hearing, the trial 1 We collectively refer to defendants and real party as “the City.” 1 court granted plaintiffs’ motion, denied the City’s motion, and subsequently entered judgment prohibiting the City and its officers and agents from administering or enforcing Proposition M. The City maintains the trial court erred in multiple respects. We affirm. BACKGROUND Residential Vacancies in San Francisco At the request of a former supervisor, the City’s Budget and Legislative Analyst’s Office issued a report in January 2022 analyzing “the number of residential vacancies in San Francisco, reasons for vacancies, comparison to other cities, and policy options for addressing vacancies based on other jurisdictions.” (S.F. Budget & Legis. Analyst’s Office, Policy Analysis Report: Residential Vacancies in San Francisco (Jan. 31, 2022) p. 1 (“Report”).) Based on census and survey data, the Report indicated that in 2019 there were “40,458 total housing units vacant in San Francisco, or approximately ten percent of the City’s 406,399 housing units.” (Report, supra, at p. 3.) This included 8,548 units being offered for rent and/or sale, 10,444 units rented or sold but not yet occupied, 8,565 units used part time or occasionally throughout the year (e.g., “non-primary housing units” and “timeshares”), and 12,991 other vacant units (e.g., units held vacant for “personal or family reasons”). (Id., at p. 4.) The Report identified units sold but not yet occupied as “the fastest growing segment of vacant housing in San Francisco over five years through 2019.” (Ibid.) The Report also identified various “policy interventions” used by several cities to address residential vacancies, including a “vacancy tax” to “discourage extended vacancies based on discretionary choices by owners for financial or other reasons that have a social cost to the City and further 2 constrain housing supply.” (Report, supra, at pp. 27, 46.) The Report pointed out a vacancy tax on residential units would require voter approval under the California Constitution. (Id., at p. 43; see California Cannabis Coalition v. City of Upland (2017) 3 Cal.5th 924, 936 (Cannabis Coalition) [“Article XIII C [of the California Constitution] was added by Proposition 218, an initiative constitutional amendment adopted at the 1996 general election” and provides “ ‘[n]o local government may impose, extend, or increase any general tax unless and until that tax is submitted to the electorate and approved by a majority vote.’ ”].) Proposition M Proposition M was submitted to San Francisco voters at the November 2022 general election pursuant to the initiative process. (Pamphlet, supra, p. 171.) The proposition sought to add sections 2950 through 2963 as article 29A of the San Francisco Business and Tax Regulations Code.2 The proposed new article was titled the “Empty Homes Tax Ordinance” (some capitalization omitted) and the proposed “tax” was denominated the “ ‘Empty Homes Tax.’ ” (Pamphlet, p. 241, italics omitted.) The measure passed by a 54.51 percent majority vote. (Cal. State University, Sacramento Institute for Social Research & The Center for Cal. Studies: Cal. County, City, and School District Election Outcomes: Candidate and Ballot Measures 2022 Elections, table 1.1, p. 10; <https://elections.cdn.sos.ca.gov/county-city-school-district- election-results/2022/county.pdf> [as of Aug. 19, 2026].) Section 2951, subdivision (a) begins by characterizing residential vacancies as an “ongoing concern in San Francisco,” citing the Report’s 2 All further citations to “sections” are to those in article 29A of the San Francisco Business and Tax Regulations Code unless otherwise indicated. 3 finding that approximately 40,500 units were vacant in 2019. Section 2951, subdivision (e) sets forth the purpose of Proposition M—“to disincentivize prolonged vacancies, thereby increasing the number of housing units available for occupancy, while also raising funds for rent subsidies and affordable housing.” Section 2952 defines certain terms as used in Proposition M. The term “ ‘Residential Unit’ ” is broadly defined as “a house, an apartment, a mobile home, a group of rooms, or a single room that is designed as separate living quarters, other than units occupied or intended for occupancy primarily by travelers, vacationers, or other transient occupants.” (§ 2952.) The term “ ‘Vacant’ ” is defined as “unoccupied, uninhabited, or unused for more than 182 days, whether consecutive or nonconsecutive, in a tax year.” (Ibid.) Section 2953 imposes “an annual Empty Homes Tax on each person that owns a Residential Unit for keeping that Residential Unit Vacant,” (§ 2953, subd. (a)) and identifies a tiered rate based on square footage of the unit that increases over time. For the first tax year the Empty Homes Tax is in effect, the rate for each residential unit is $2,500 (with square footage less than 1,000), $3,500 (with square footage from 1,000 to 2,000), or $5,000 (with square footage greater than 2,000). (Id., subd. (b).) If the unit continues to be “Vacant” through the following tax year, the rate increases for each square footage tier to $5,000, $7,000, and $10,000, respectively. (Id., subd. (d).) And if that vacancy continues through a third tax year, the rate again increases for each tier to $10,000, $14,000, and $20,000, respectively. (Id., subd. (g).) These rates are also subject to an annual adjustment for inflation. (Id., subd. (h).) Some residential units are expressly exempted from the Empty Homes Tax. Section 2951, subdivision (c), for example, states the “[t]ax is limited to 4 buildings with more than two residential units because such buildings are more likely to include one or more units held vacant by choice and are more likely to include multiple vacancies.” Section 2952 excludes units in a “currently operational nursing home” or “residential care facility” from the definition of a “ ‘Residential Unit.’ ” And section 2953, subdivision (j) identifies various time periods that are to be “disregarded” in calculating whether a residential unit has been kept “Vacant” for more than 182 days. These include certain periods of repair, rehabilitation, construction, as well as certain leases of the unit. (§ 2952.) Section 2952 defines “ ‘Lease Period’ ” as a period during which a residential unit is leased “to one or more tenants under a bona fide lease intended for occupancy, but not including any lease or rental of that Residential Unit to anyone in the Owner’s Group or to travelers, vacationers, or other transient occupants.” The term “ ‘Owner’s Group’ ” is defined to include “any Related Person or Affiliate of the owner or any current or former co-owner,” and “ ‘Related Person’ ” is defined as “a spouse, domestic partner, child, parent, or sibling.” (§ 2952.) In other words, the “ ‘Lease Period’ ” exclusion in Proposition M does not include leases or rentals to these family members. Section 2955, subdivision (a) states: “For only so long as and to the extent that the City is prohibited from imposing the Empty Homes Tax, any person upon whom the City is prohibited under the Constitution or laws of the State of California or the Constitution or laws of the United States from imposing the Empty Homes Tax shall be exempt from the Empty Homes Tax.” And section 2963 states: “No section, clause, part, or provision of this [a]rticle 29A shall be construed as requiring the payment of any tax that 5 would be in violation of the Constitution or laws of the United States or of the Constitution or laws of the State of California.” Section 2958 directs proceeds from the Empty Homes Tax to San Francisco’s Housing Activation Fund to be used for rental subsidies for older individuals and low-income households, as well as to acquire and rehabilitate certain buildings for affordable housing. (§ 2952, subds. (a), (c)(1)(A)–(B).) San Francisco voters were provided ballot materials for the November 2022 general election that included arguments in favor of and against Proposition M, as well as a statement from the then City Controller. Proponents argued: “Prop M isn’t about taxing those who call San Francisco home. It’s about tackling the large, corporate landlords keeping units vacant, and those wealthy individuals who purchase units but don’t use them. [¶] In the first year alone, it is expected that 4,500 new units will return on the market— more than our annual goals— with no increase in taxes, no construction time, no multi-million dollar price tag, and no waiting.” (Pamphlet, supra, argument in favor of Prop. M, p. 171.) Opponents argued Proposition M was based on “overstated statistics that manipulate the perceived number of vacancies citywide,” and would sweep up “small property owners and intergenerational households, not corporate landlords.” (Id., argument against Prop. M, p. 173.) Proponents responded: “We hope no one pays this tax. We want every vacant unit filled with people who need homes.” (Id., rebuttal to argument against Prop. M, p. 173.) The then City Controller opined the proposed ordinance “could result in additional revenue to the City exceeding $20 million annually,” but “if the tax achieves its stated purpose of reducing the number of residential vacancies, it will result in lower revenue.” (Pamphlet, supra, statement on Prop. M by City Controller, p. 170.) 6 This Action Named plaintiffs Eric and Andrew Debbane, Robert Friedland, Natasa Zec, San Francisco Apartment Association, Small Property Owners of San Francisco Institute, and San Francisco Association of Realtors filed the instant action in February 2023. They asserted five causes of action in the operative first amended complaint. The first alleged Proposition M is an unconstitutional taking of property. (U.S. Const., 5th Amend.) The second alleged the proposition is preempted by the Ellis Act. (Gov. Code, § 7060 et seq.) The third and fourth causes of action alleged that, because Proposition M’s lease period exclusion does not cover leases or rentals to a spouse, domestic partner, child, parent, or sibling, it burdens a fundamental liberty interest in familial relations protected by the due process clause (U.S. Const., 14th Amend.; Cal. Const., art. I, § 7) and violates the equal protection clause. (U.S. Const., 14th Amend.; Cal. Const., art. I, § 7.) The fifth alleged the proposition violates the fundamental right to privacy under the California Constitution. (Cal. Const., art. I, § 1.) Summary Judgment Motions Plaintiffs subsequently moved for summary judgment. In support of the motion, Eric Debbane submitted a declaration stating he and his brother Andrew co-own several residential buildings in San Francisco, including a five-unit building where they live with their partners. The declaration went on to state the Debbane brothers have kept the other three units vacant for their “own personal use” and “have no desire to share the property that we own and live in with persons other than those already living on the property with us.” Robert Friedland submitted a declaration stating he owns a four-unit building in San Francisco and lives in one of the units. He has kept the other three units vacant because, as a 72-year-old retiree with “significant health 7 issues,” he “no longer wish[es] to bear the physical and mental burdens of being a landlord.” Friedland further averred Proposition M would effectively evict him from his home, as his remaining sources of income (Social Security and “some modest savings”) would be insufficient to cover the Empty Homes Tax on the three units, in addition to his other living expenses. Natasa Zec submitted a declaration stating she owns a “micro- condominium” in both San Francisco and Boston for her personal use. She splits time between the two cities, and in 2022, spent 126 days in San Francisco and more than 183 days in Boston, and wishes to maintain a similar division of time in future years. The ability to continue to travel is “of great importance” to her and owning both units is the “most cost-effective means of being able to pursue that course of action.” Zec further averred she has no interest in renting out her San Francisco unit because she has had negative experiences in the past with “extremely difficult tenants,” resulting in damage to her property. She values the privacy of her home and has “no desire to share it part-time with other people.” Proposition M would effectively evict her from her home because she cannot afford to pay the Empty Homes Tax. The City also moved for summary judgment. It maintained the named plaintiffs lack taxpayer standing under Code of Civil Procedure section 526a to challenge Proposition M, and even if they have standing, each of their five causes of action fails as a matter of law. After full briefing and hearing, the trial court granted the plaintiffs’ motion and denied the City’s motion. 8  The court first concluded plaintiffs have standing to bring the action.3 It then concluded Proposition M violates the Takings Clause of the Fifth Amendment (first cause of action), is preempted by the Ellis Act (second cause of action), violates “property-owners’ fundamental liberty interests in familial living arrangements, protected by the due process and equal protection clauses insofar as it taxes (actually, penalizes) units that are rented to family members of the owner while exempting units that are leased to strangers” (third and fourth causes of action); and “violates property- owners’ constitutional right to privacy under the California Constitution, insofar as it seeks to compel them to share the property on which they reside with others, against their will” (fifth cause of action). Judgment for plaintiffs followed, prohibiting the City and its officers, agents, employees, service providers, and all others acting by, through, or in concert therewith, from administering or enforcing Proposition M.4 DISCUSSION Standard of Review “ ‘A trial court properly grants a motion for summary judgment where “all the papers submitted show that there is no triable issue as to any material fact and that the moving party is entitled to judgment as a matter of law.” ’ ” (Hampton v. County of San Diego (2015) 62 Cal.4th 340, 347.) 3 The City does not challenge this ruling on appeal and has therefore waived the issue. (Cal. Rules of Court, rule 8.883(a)(1)(A); see Telish v. State Personnel Bd. (2015) 234 Cal.App.4th 1479, 1487, fn. 4 (Telish) [“An appellant’s failure to raise an argument in the opening brief waives the issue on appeal.”].) 4 During the pendency of this appeal, the San Francisco Board of Supervisors passed Ordinance No. 41-25, amending portions of Proposition M to suspend the Empty Homes Tax pending a final decision in this action. 9  A plaintiff moving for summary judgment bears the initial burden of “showing that there is no defense to a cause of action” and meets that burden if it “has proved each element of the cause of action entitling [it] to judgment on the cause of action.” (Code Civ. Proc., § 437c, subd. (p)(1).) If the plaintiff does so, the burden shifts to the defendant to “set forth the specific facts showing that a triable issue of material fact exists as to the cause of action or a defense thereto.” (Ibid.) Our review of a judgment entered after the grant of a motion for summary judgment is de novo. (Hampton v. County of San Diego, supra, 62 Cal.4th at p. 347.) “ ‘ “We liberally construe the evidence in support of the party opposing summary judgment and resolve any doubts concerning the evidence in favor of that party.” ’ ” (Ibid.) The appellant, however, still “has the burden of showing error, even if [it] did not bear the burden in the trial court.” (Claudio v. Regents of the University of California (2005) 134 Cal.App.4th 224, 230.) The Ellis Act Preempts the Empty Homes Tax We consider first plaintiffs’ claim that Proposition M is preempted by the Ellis Act in accordance with “ ‘the familiar principle that we should address and resolve statutory issues prior to, and if possible, instead of, constitutional questions [citation], and that “we do not reach constitutional questions unless absolutely required to do so to dispose of the matter before us.” ’ ” (Levy v. City and County of San Francisco (2025) 114 Cal.App.5th 997, 1012, quoting Facebook, Inc. v. Superior Court (Hunter) (2018) 4 Cal.5th 1245, 1275, fn. 31; see Coyne v. City and County of San Francisco (2017) 9 Cal.App.5th 1215, 1235 (Coyne) [“ ‘[c]onstitutional issues will be resolved only if absolutely necessary and not if the case can be decided on any other ground’ ”].) 10  The Ellis Act provides, with certain exceptions not relevant here, that no statute, ordinance, regulation, or administrative action shall “compel the owner of any residential real property to offer, or to continue to offer, accommodations in the property for rent or lease. . . .” (Gov. Code, § 7060, subd. (a); see generally San Francisco Apartment Assn. v. City and County of San Francisco (2016) 3 Cal.App.5th 463, 477 (San Francisco Apartment Assn.) [“ ‘ “The legislative history of the Act consistently demonstrates the purpose of the Act is to allow landlords who comply with its terms to go out of the residential rental business by evicting their tenants and withdrawing all units from the market, even if the landlords could make a fair return, the property is habitable, and the landlords lack approval for future use of the land.” ’ ”] Quoting Los Angeles Lincoln Place Investors, Ltd. v. City of Los Angeles (1997) 54 Cal.App.4th 53, 61.) The Act does not, however, “[d]iminish[] . . . any power which currently exists or which may hereafter exist in any public entity to grant or deny any entitlement to the use of real property, including, but not limited to, planning, zoning, and subdivision map approvals.” (Gov. Code, § 7060.1, subd. (b); San Francisco Apartment Assn., supra, 3 Cal.App.5th at p. 478.) Nor does it “ ‘[i]nterfere with local governmental authority over land use,’ ‘[p]reempt local . . . land use regulations, procedures, or controls that govern the demolition and redevelopment of residential property,’ or ‘[o]verride procedural protections designed to prevent abuse of the right to evict tenants.’ (Gov. Code, § 7060.7, subds. (a)–(c).) . . . ‘[I]t “completely occupies the field of substantive eviction controls over landlords who wish to withdraw” all units from the residential rental market.’ ” (San Francisco Apartment Assn., at p. 478, quoting Johnson v. City and County of San Francisco (2006) 137 Cal.App.4th 7, 14.) 11  The City maintains the trial court erred in two respects in ruling Proposition M contravenes and is preempted by this state law. Home Rule Powers Do Not Override Conflicting State Law The City first asserts Proposition M is a local tax measure which it was entitled to enact pursuant to its “home rule powers” under article XI, section 5 of the California Constitution. Article XI, section 5, subdivision (a) of the California Constitution provides: “It shall be competent in any city charter to provide that the city governed thereunder may make and enforce all ordinances and regulations in respect to municipal affairs, subject only to the restrictions and limitations provided in their several charters and in respect to other matters they shall be subject to general laws.” Home rule powers include “the power to tax for local purposes.” (City and County of San Francisco v. Regents of University of California (2019) 7 Cal.5th 536, 545 (Regents).) Assuming Proposition M is properly characterized as a tax, as the City maintains, rather than a penalty as plaintiffs assert—an issue we need not and do not decide5—we agree with the City that the measure concerns “municipal affairs” and thus is within the ambit of the City’s home rule powers. But the fact that a local law concerns local affairs does not, in and of itself, insulate that law from conflicting directives of state law. (See Regents, 5 We note, however, that Proposition M was not presented to San Francisco voters as a revenue-raising measure, but rather as a means to dissuade residential property owners from choosing not to enter, or to remain out of, the residential rental business. The proponents of the measure flatly stated, for example, that their “hope” was that “no one pays this tax”—what they “want[ed]” was “every vacant unit filled with people who need homes.” (Pamphlet, supra, rebuttal to argument against Prop. M, p. 173.) The City Controller similarly observed that as Proposition M’s stated purpose— putting tenants in vacant residential units—was achieved, any tax revenue from the measure would decline. (Id., at p. 170.) 12 supra, 7 Cal.5th at p. 550, fn. 4 [“a charter city tax—like a charter city regulation—may be preempted by a state statute in appropriate circumstances”].) In other words, concluding that a municipal enactment concerns municipal affairs is only the first step in determining whether the enactment survives a preemption challenge. As our high court explained in California Fed. Savings & Loan Assn. v. City of Los Angeles (1991) 54 Cal.3d 1 (California Fed. Savings), in determining whether a local law falling within the bounds of a charter city’s home rule powers is preempted by state law, a court must first determine whether the ordinance at issue regulates an activity that can be characterized as a “ ‘municipal affair’ ” and then whether it presents an “actual conflict” with state law. (Id. at p. 16.) If so, the court must next determine whether the state law addresses a matter of “statewide concern,” is “reasonably related” to the resolution of that concern, and is “narrowly tailored” to avoid unnecessary interference in local governance. (Id. at pp. 17, 24.) If the state law at issue meets these requirements, it prevails over the challenged local law.6 Proposition M and the Ellis Act Are in “Actual Conflict” As we have recited, Proposition M imposes a significant tax on owners of certain residential units (in buildings with more than two residential units) for any unit that is not occupied, inhabited, or used for more than 182 days in a tax year. As we have also recited, the Ellis Act provides that no 6 We note that, as used in the context of legislative enactments by home rule municipalities, the “preemption” analysis is slightly different than that in other preemption contexts, and it is perhaps more appropriately termed an inquiry into whether state law “supersedes” the local law. (AIDS Healthcare Foundation v. Bonta (2024) 101 Cal.App.5th 73, 82.) However, because relevant authority consistently uses the term “preemption” (see, post, at pp. 14–15), we do so here. 13 local ordinance shall “compel the owner of any residential real property to offer, or to continue to offer, accommodations in the property for rent or lease.” (Gov. Code, § 7060, subd. (a).) Despite the apparent tension between the directives of Proposition M and the prohibitions of the Ellis Act, the City maintains there is no “actual conflict” between the two. A conflict between a local law and state law exists if the local law “ ‘ “ ‘duplicates, contradicts, or enters an area fully occupied by general law, either expressly or by legislative implication.’ ” ’ ” (Sherwin-Williams Co. v. City of Los Angeles (1993) 4 Cal.4th 893, 897; San Francisco Apartment Assn., supra, 3 Cal.App.5th at p. 475.) “[L]ocal legislation is ‘contradictory’ to general law when it is inimical thereto.” (Sherman-Williams, at p. 898; San Francisco Apartment Assn., at p. 475.) In determining whether that is the case, California courts have applied the “prohibitive price” standard. (Coyne, supra, 9 Cal.App.5th at p. 1226 [concluding “prohibitive price standard” is the appropriate standard to determine whether San Francisco ordinances conflict with Ellis Act].) In Bullock v. City and County of San Francisco (1990) 221 Cal.App.3d 1072 (Bullock), for example, a hotel owner challenged a local ordinance conditioning issuance of a permit to convert residential hotel units on either furnishing the units or “making a substantial ‘in lieu’ payment to a fund maintained by the City.” (Id. at p. 1099.) The appellate court concluded the ordinance was preempted because it impermissibly conditioned the owners’ right to depart the business of renting residential hotel units on compliance with requirements not found in the Ellis Act. (Bullock, at pp. 1100–1101.) “The Ellis Act does not,” said the court, “permit the City to condition plaintiff’s department upon the payment of ransom.” (Bullock, at p. 1101.) Rather than recognize the right of the owner under the Ellis Act to “ ‘ “just 14 simply go out of that business,” ’ ” the ordinance “is attempting to ‘impose[] a prohibitive price on the exercise of th[at] right under the Act.’ ” (Bullock, at p. 1101.) In Reidy v. City and County of San Francisco (2004) 123 Cal.App.4th 580 (Reidy), a residential hotel owner challenged a local ordinance making it unlawful to eliminate, demolish, or change the use of residential hotel units without a permit. (Id. at p. 589.) The appellate court ruled the ordinance was preempted because it “effectively conditioned” the right of a hotel owner to go out of the rental business on compliance with requirements not found in the applicable version of the Ellis Act. (Reidy, at p. 593.) And in Coyne, supra, 9 Cal.App.5th at page 1218, property owners challenged an ordinance that increased relocation assistance payments owed to tenants under the Ellis Act. The appellate court again ruled the ordinance was preempted because the relocation payment provisions placed conditions on the right to go out of business not found in the Ellis Act. (Coyne, at p. 1227.) The ordinance was, in short, “a form of ransom” that interfered with this right and therefore imposed “a prohibitive price on the ability of landlords to exercise their rights under the Ellis Act.” (Coyne, at p. 1230.) In so ruling, the court made clear the ordinance was a “categorical infringement[]” of the owners’ Ellis Act rights and declined to consider what particular payment threshold imposed a “prohibitive price,” as there was “no set of circumstances under which we view this type of payout obligation as valid.” (Coyne, at p. 1232.) Bullock, Reidy, and Coyne collectively instruct that local ordinances like Proposition M that place conditions on a residential property owner’s right to not offer their property for rent are preempted if those conditions are not required by the Ellis Act. 15  The City does not suggest Proposition M imposes any condition that is remotely within the contemplation of the Ellis Act. Instead, it argues Bullock, Reidy, and Coyne are distinguishable because they addressed a residential property owner’s right under the Ellis Act to exit the rental market, whereas Proposition M focuses on placing residential units into the rental market, or returning such units to the rental market. To begin with, the Ellis Act, by its plain language, protects both the right not to “offer” accommodations for rent—i.e., the right not to enter the rental market—as well as the right not to “continue to offer” accommodations for rent—i.e., the right to exit the rental market. (Gov. Code, § 7060, subd. (a).) Further, there is no material difference between the “absolute right to exit the residential rental business” (San Francisco Apartment Assn., supra, 3 Cal.App.5th at p. 477, italics added) discussed in Bullock, Reidy, and Coyne, and the right not to enter or reenter the residential rental business asserted here. In fact, one could say the right at issue here is a residential property owner’s right to remain exited from the rental market. Otherwise, as the City apparently sees it, while a property owner, indeed, has the right to exit the residential rental business, they have no correlative right to remain out of that business and can be pressured back into it on pain of significant taxation. This circularity in reasoning is a patently unreasonable reading of the Ellis Act. (See San Francisco Apartment Assn., at p. 480 [explaining decade-long “ban on applying for merger approval beg[an] to run when the landlord exit[ed] the residential rental business rather than before the landlord exit[ed] the business[,] [did] not make this ban any less of a penalty triggered by the landlord’s exercise of Ellis Act rights”].) 16  The City further argues Bullock, Reidy, and Coyne are not persuasive, in any event, because Proposition M assertedly does not impose a “prohibitive price” on a residential property owner’s exercise of Ellis Act rights. This is so, says the City, because the owner still has “choices”—they can use but not reside on the property, they can reside on the property, or they can choose not to use or reside on the property and pay the Empty Homes Tax. With respect to the first supposed choice, the City posits residential property owners who would otherwise be subject to Proposition M can avoid the tax by using their residential units, not as a place they or anyone else resides, but as an office, personal gym, or private art studio. The City did not advance this argument in the trial court and has therefore forfeited the issue. (See Howitson v. Evans Hotels, LLC (2022) 81 Cal.App.5th 475, 489 [“It is well settled that the failure to raise an issue in the trial court typically forfeits on appeal any claim of error based on that issue.”].) But even if it were preserved, it is not supported by the language of Proposition M, the Report that prompted it, or the ballot materials submitted to the voters. In support of its proffered construction, the City points to Proposition M’s definition of a “ ‘Vacant’ ” residential unit as one “unoccupied, uninhabited, or unused, for more than 182 days, whether consecutive or nonconsecutive, in a tax year.” (§ 2952, italics added.) When construing a local ordinance enacted by initiative, general rules of construction apply and our primary concern is giving effect to the intent of the voters. (Cannabis Coalition, supra, 3 Cal.5th at p. 933; Lesher Communications, Inc. v. City of Walnut Creek (1990) 52 Cal.3d 531, 540–542.) We begin with the plain language of the measure, “ascribing to words their ordinary meaning, while taking account of related provisions and the structure of the relevant statutory and constitutional scheme.” (Cannabis 17 Coalition, at p. 933.) Where possible, we avoid literalism that would result in absurd or arbitrary results. (Commission on Peace Officer Standards & Training v. Superior Court (2007) 42 Cal.4th 278, 290.) As a preliminary matter, the City seems to suggest our consideration of its proffered reading of Proposition M is confined to examining its text, as plaintiffs have advanced “facial challenges” to the measure’s lawfulness. That is not the case. In considering a facial challenge to a local or state law, we are not precluded from employing any of the rules of construction that may be appropriate, including those that may be brought to bear when some ambiguity inheres in the language of the challenged measure. (See California Redevelopment Assn. v. Matosantos (2011) 53 Cal.4th 231, 257– 261 [examining full constitutional and statutory context, including ballot materials and legislative history of pertinent provisions, in addressing facial challenge to statutes curtailing redevelopment agencies].) As the City points out, Proposition M does, indeed, define “ ‘Vacant’ ” to mean “unoccupied, uninhabited, or unused.” (§ 2952.) It does not, however, provide any definition of the salient terms “unoccupied,” “uninhabited,” or “unused.” We also agree, as the City urges, that under the rules of statutory construction, we must presume these three different terms have different meanings. (See In re A.T. (2025) 110 Cal.App.5th 722, 733 [where different words “appeared in the same statute, we presumed they had different meanings”].) But even assuming these terms have different meanings, it is not at all clear what each means. And it certainly is not clear that the term “unused” permits the owner of a residential property to not reside there but nevertheless avoid paying the Empty Homes Tax if they use the property for a nonresidential purpose, which the City suggests could be office space, a 18 gym, or art studio. At best, the City has posited an arguable construction of the language and thus has identified an ambiguity as to the meaning of the term. We therefore proceed to “consider extrinsic sources, such as an initiative’s ballot materials” to resolve this supposed ambiguity. (Cannabis Coalition, supra, 3 Cal.5th at p. 934.) The Report prepared prior to the drafting of Proposition M, which is cited in the text of the measure and was cited in the ballot arguments in support of it, evidences the origin of, and the distinction between, the terms “unoccupied, uninhabited, or unused.” (Pamphlet, supra, text of Prop. M, p. 241; Id., argument in favor of Prop. M and rebuttal, p. 172.) As we have discussed, the Report grouped the 40,000-plus residential vacancies it identified from data into specific subcategories. (Report, supra, at p. 4.) One category included unoccupied units: units that had been rented or sold but were not yet occupied by the renter or owner. (Ibid.) Another category included uninhabited units: units that had been kept vacant for some other reason, including “personal or family reasons.” (Ibid.) And a third category included unused units: units “used or intended for use part time or occasional use throughout the year,” such as units “used seasonally or recreationally” like “second or non-primary housing units, and timeshares.” 7 (Ibid.) The ballot arguments utilized this same terminology. Proponents, for example, argued Proposition M would target owners with unused units: “those wealthy individuals who purchase units but don’t use them.” (Pamphlet, supra, argument in favor of Prop. M, p. 172.) They made specific 7 At oral argument, the City argued the Report is irrelevant because it was not among the materials placed before the voters. However, Measure M was a City drafted initiative that followed in the wake of the City drafted Report. Thus, the Report provides contextual background to the language of the measure. 19 mention of “corporate landlords” who were assertedly “holding units vacant, waiting to flip them for profit years down the road.” (Id., rebuttal to argument against Prop. M, p. 173.) Opponents, in turn, maintained the proposition would impact “small property owners and intergenerational households, not corporate landlords.” (Id., argument against Prop. M, p. 173.) Proponents disputed this and asserted, “We want every vacant unit filled with people who need homes.” (Id., rebuttal to argument against Prop. M, p. 173.) The Report and the ballot materials, thus, indicate the phrase “unoccupied, uninhabited, or unused” refers to residential units that are not occupied as a residence, not inhabited by a resident, or not used as a residence for more than half a year. Indeed, this is the only reading that aligns with the stated purpose of Proposition M—“to disincentivize prolonged vacancies, thereby increasing the number of housing units available for occupancy” (§ 2951, subd. (e), italics added)—as well as the name and descriptor of the tax—“an annual Empty Homes Tax on each person that owns a Residential Unit for keeping that Residential Unit Vacant.” (§ 2953, subd. (a), italics added.) In short, there is not the slightest suggestion in either the language of Proposition M—or the Report that preceded it, or the voter materials for and against it—that the Empty Homes Tax does not apply to the owner of a residential property who chooses not to reside thereon if, for example, they festoon the walls of their property with personal photographs and call it their personal photo gallery, or roll out a yoga mat on the living room floor and pronounce the property a personal workout studio. To the contrary, such a construction is squarely at odds with the intent of the proposition—to require 20 owners of residential properties, on pain of significant taxation, to put such properties to residential use and keep them in such use. Furthermore, the City’s assertion that making nonresidential use of a residential property is a choice an owner can legitimately make to avoid the Empty Homes Tax is at odds with the uses of a “Residential Unit” allowed under San Francisco’s Planning Code. San Francisco Planning Code section 102 provides that the “Principal” permitted use of a “Residential Unit” (or “Dwelling Unit”) is “Residential Use,” defined to “provide housing for San Francisco residents.” While the owner of a “Residential Unit” may also engage in “Accessory Use,” like a home office, such use is limited to a “related minor Use that is either necessary to the operation or enjoyment of a lawful Principal Use or Conditional use, or appropriate, incidental, and subordinate to any such use. . . .” (S.F. Planning Code, § 102; see S.F. Planning, General Planning Information: Accessory Uses for Dwellings (Apr. 2015), p. 2 (“Planning Information”) [identifying h