WERDEGAR, J.

Plaintiffs, the owners and operators of the San Remo Hotel in San Francisco, sought approval from the City and County of San Francisco to rent all rooms in the San Remo Hotel to tourists or other daily renters, rather than to longer term residents. Plaintiffs eventually received approval but, in the process, were required to (1) comply with zoning laws by obtaining a conditional use permit for use of their property as a tourist hotel, and (2) help replace the residential units San Francisco claimed would be lost by the conversion, pursuant to the city's Residential Hotel Unit Conversion and Demolition Ordinance (S.F.Admin.Code, ch. 41) (hereafter the HCO), which plaintiffs elected to do by paying an in lieu fee into a governmental fund for the construction of low- and moderate-income housing.

Plaintiffs challenged the conditional use permit requirement by petition for writ of administrative mandate (Code Civ. Proc, § 1094.5), pled as the first cause of action in their second amended complaint, and challenged the housing replacement requirement by four additional causes of action alleging the taking of private property without just compensation in violation of article I, section 19 of the California Constitution. [1] The trial court denied the writ petition and sustained a demurrer to the takings counts. The Court of Appeal reversed.

Based on the administrative record and the pleadings, and guided by established legal principles, we conclude the trial court properly denied the petition for writ of administrative mandate and sustained the demurrer as to the remaining causes of action. We will therefore reverse the judgment of the Court of Appeal insofar as the appellate court reversed the trial court's judgment for defendant City and County of San Francisco.

Plaintiffs are San Remo Hotel L.P., a limited partnership; its general partner, T & R Investment Corp.; and its limited partners, Thomas and Robert Field. Defendants are the City and County of San Francisco, and the various agencies and agents through which it acted (collectively the City, or San Francisco).

Because the proceedings below turned on the application and validity of two bodies of local San Francisco law, we first summarize those ordinances.

The HCO, codified in chapter 41 of the San Francisco Administrative Code, was first enacted in 1981 (S.F.Ord. No. 330-81) and was substantially revised in 1990 (S.F.Ord. No. 121-90). [2] Its purpose is to "benefit the general public by minimizing adverse impact on the housing supply and on displaced low income, elderly, and disabled persons resulting from the loss of residential hotel units through their conversion and demolition." (HCO, § 41.2.) Accompanying the ordinance are findings that the City suffers from a severe shortage of affordable rental housing; that many elderly, disabled and low-income persons reside in residential hotel units; that the number of such units had decreased by more than 6,000 between 1975 and 1979; that loss of such units had created a low-income housing "emergency" in San Francisco, making it in the public interest to regulate and provide remedies for unlawful conversion of residential hotel units; that the City had instituted a moratorium on residential hotel conversion effective November 21, 1979; and that because tourism is also essential to the City, the public interest also demands that some moderately priced tourist hotel rooms be available, especially during the summer tourist season. (HCO, § 41.3.)

Each hotel room's initial status for purposes of the HCO was determined by having the owner or operator of each hotel file an initial unit usage report stating the number of residential and tourist units in their hotel as of September 23, 1979. (HCO, § 41.6, subd. (b)(1).) The HCO defines a "Residential Unit" as a "guest room" that was occupied by a "permanent resident" on September 23, 1979, or that was designated residential under section 41.6's procedures for initial status determination. (HCO, § 41.4, subd. (q).) A "Tourist Unit" is defined as a guest room not occupied by a permanent resident on September 23, 1979, or a guest room certified as a tourist unit under section 41.6. (HCO, § 41.4, subd. (s).) A permanent resident is a person who occupies a guest room for at least 32 consecutive days. ( Id., subd. (n).)

The HCO makes it unlawful to eliminate a residential hotel unit without obtaining a conversion permit or to rent a residential unit for a term shorter than seven days. (HCO, § 41.20, subd. (a).) Violators are subject to civil penalties. ( Id., subd. (c).)

An application to convert residential units to tourist use must include, inter alia, "[a] statement regarding how one-for-one replacement of the units to be converted will be accomplished." (HCO, § 41.12, subd. (b)(9).) The applicant may satisfy the replacement requirement by constructing or bringing onto the market new residential units comparable to those converted (HCO, § 41.13, subd. (a)(1), (2)); constructing or rehabilitating certain other types of housing for low-income, disabled or elderly persons ( id., subd. (a)(3)); or paying to a public or nonprofit housing developer, or to the City's Residential Hotel Preservation Fund Account, an in lieu fee equal to the replacement site acquisition costs plus a set portion of the replacement construction costs ( id., subd. (a)(4), (5)). [3] The replacement costs are to be determined by the City's Department of Real Estate based on two independent appraisals. (HCO, § 41.3, subd. (a)(4), (5).)

The 1981 and 1990 versions of the HCO differ in their treatment of temporary tourist rentals of residential units. The 1981 ordinance allowed summer season rentals (May 1 September 30) of vacant residential units without numerical restriction, with the proviso that the room "shall immediately revert to residential use on application of a prospective permanent resident," but contained no provision for winter tourist rentals. (1981 HCO, § 41.16, subd. (a)(3)(B).) The 1990 revision additionally restricted summer tourist rentals of residential units by, among other things, limiting such rentals, absent special permission from the City's Bureau of Building Inspection, [4] to 25 percent of a hotel's residential rooms. (HCO, § 41.19, subd. (a)(3).) The revision, however, also allowed a limited number of residential rooms to be rented to tourists during the winter months as well. ( Id., subd.(c).)

In 1987, San Francisco adopted article 7 of its Planning Code, a set of zoning regulations for "neighborhood commercial districts." (S.F. Planning Code, § 701, added by S.F. Ord. No. 69-87.) The San Remo Hotel is within the North Beach neighborhood commercial district (hereafter North Beach district), created in 1987. The North Beach district "functions as a neighborhood-serving marketplace, citywide specialty shopping, and dining district, and a tourist attraction, as well as an apartment and residential hotel zone." (S.F. Planning Code, § 722.1.) While most new commercial development is permitted on the first two stories of buildings, new housing development is encouraged above the second story and "[e]xisting residential units are protected by prohibitions of upper-story conversions and limitations on demolitions." ( Ibid. )

Tourist hotels are a conditional use in the North Beach district. (S.F. Planning Code, § 722.55.) "Conditional uses are permitted in a Neighborhood Commercial District when authorized by the Planning Commission.... Conditional uses are subject to the provisions set forth in Sections 178, 179, and 316 through 316.8 of this Code." (S.F. Planning Code, § 703.2, subd. (b)(1)(B).)

A "permitted conditional use," the zoning category into which plaintiffs, before the City's zoning administrator, claimed the San Remo Hotel's tourist rental fell, is defined generally in San Francisco Planning Code sections 178 and 179. Generally, a permitted conditional use is defined, inter alia, as "[a]ny use or feature which is classified as a conditional use in the district in which it is located and which lawfully existed either on the effective date of this Code, or on the effective date of any amendment imposing new conditional use requirements upon such use or feature." (S.F. Planning Code, § 178, subd. (a)(2).) As to neighborhood commercial districts, a permitted conditional use is "[a]ny use or feature in a Neighborhood Commercial District which lawfully existed on the effective date of Ordinance No. 69-87 which is classified as a conditional use by the enactment of Ordinance No. 69-87." (S.F. Planning Code, § 179, subd. (a)(2).)

A permitted conditional use may continue "in the form in which it lawfully existed" on the effective date of the new conditional use requirement. (S.F. Planning Code, § 178, subd. (b).) But a permitted conditional use "may not be significantly altered, enlarged, or intensified, except upon approval of a new conditional use application." ( Id., subd. (c).)

"Residential conversion," defined as the change in occupancy from residential to nonresidential use (S.F. Planning Code, § 790.84), is prohibited above the first floor in the North Beach district. (S.F. Planning Code, § 722.38.) The Planning Code's definition of residential conversion, however, "shall not apply to conversions of residential hotels, as defined and regulated in [the HCO]." (S.F. Planning Code, § 790.84.)

One of plaintiffs' causes of action, the petition for writ of administrative mandate, was decided on an administrative record, while the other counts were dismissed on demurrer. On review of the latter ruling we, like the trial court, may consider only the factual allegations of the complaint and matters subject to judicial notice (Code Civ. Proc, § 430.30, subd. (a)), and not the administrative record. We therefore summarize the factual background of the writ petition separately from that of the remaining counts.

By their petition, pled as the second amended complaint's first cause of action, plaintiffs sought to overturn the City's administrative determination that, under the San Francisco Planning Code, plaintiffs were required to obtain a conditional use permit in order to use all rooms in the San Remo Hotel for tourist rentals. Plaintiffs allege that the City's zoning administrator, in a decision affirmed by the City's Board of Permit Appeals, [5] should have classified the proposed tourist use of the San Remo Hotel as a permitted conditional use under section 179, subdivision (a)(2) of the Planning Code. Plaintiffs prayed for a peremptory writ of mandate directing the City to "recognize and acknowledge" that tourist use of the San Remo Hotel "is a permitted conditional hotel use."

The administrative record shows the following:

On September 25, 1981, Jean Iribarren, then operating the San Remo Hotel under a lease from plaintiffs Thomas and Robert Field, filed the initial unit usage report required under the HCO. He reported the San Remo Hotel had 61 units in residential use both as of September 23, 1979, and seven days prior to filing the report, and zero units in tourist use on the same dates. On November 18, 1981, the Bureau of Building Inspection issued Iribarren a certificate of use reflecting the same usage numbers. According to a 1992 declaration by plaintiffs Thomas and Robert Field, Iribarren filed the "incorrect" initial unit usage report without their knowledge. They first discovered the report in 1983 when they resumed operation of the hotel. They protested the residential use classification in 1987, but were told it could not be changed because the appeal period had passed.

On September 9, 1992, the Zoning Administrator of the San Francisco Department of City Planning, responding to plaintiffs' request for a written ruling, determined that operation of the San Remo as a tourist hotel was not a permitted conditional use under section 179, subdivision (a)(2) of the San Francisco Planning Code. The ruling observes that beginning in May 1982, the San Remo Hotel was included in interim zoning districts requiring conditional use permits for operation of tourist hotels. That requirement became permanent in April 1987 with the establishment of the North Beach district. Planning Code section 790.47 defines a residential hotel as a hotel containing one or more residential units, a term defined by reference to the HCO. Because under the HCO all 62 units in the San Remo Hotel were classified as residential, its zoning classification was also as a residential hotel (a type of "group housing" under the Planning Code). [6]

The zoning administrator considered, but rejected, plaintiffs' claim that because they rented vacant rooms in the San Remo Hotel to tourists during the summer (permitted under the HCO with some conditions), and in some cases in the winter as well (permitted with special approval under the 1990 HCO), they operated a tourist hotel. According to the administrator, "[s]uch temporary authorization for tourist uses does not constitute a change of use and therefore all of the units in the Hotel remain residential units." The zoning administrator further found that, according to annual unit usage reports submitted by the San Remo Hotel to the Bureau of Building Inspection, between 25 and 57 units in the hotel were in fact occupied by residents from 1982 to 1992. Even if only these units were considered residential, and the remaining units deemed to have been in tourist use when the conditional use permit requirement took effect, the administrator ruled, plaintiffs would still be required to obtain a conditional use permit to convert the hotel to complete tourist use, because under San Francisco Planning Code section 178, subdivision (c), a permitted conditional use may not be significantly altered, enlarged or intensified without new conditional use authorization.

Hotel tax returns also indicate mixed residential and tourist use. Of the $60,942 in gross rent the San Remo Hotel earned in the quarter ending December 31, 1988, $56,676 was rent for occupancy by permanent residents (not subject to the hotel tax). For the first quarter of 1989, $33,897 of the hotel's $57,563 gross rent was from residential rentals. Even in the summer, a significant portion of the hotel's rental revenue was residential: for the quarter ending September 30, 1989, gross rent was $121,117, of which $34,469 was earned from residential rentals. According to a May 1991 decision of the Bureau of Building Inspection hearing officer approving plaintiffs' request to rent more than 25 percent of their rooms to tourists in the 1991 summer season, plaintiffs had presented evidence that in May 1991, 30 San Remo Hotel guests had occupied their rooms "for periods ranging from two months to ten years." Nine more had rented for more than a week, but less than a month.

The Board of Permit Appeals upheld the zoning administrator's determination that a conditional use permit was required to change the San Remo Hotel's use to a tourist hotel.

In their remaining causes of action, plaintiffs allege that the HCO, and the various actions of City agencies requiring plaintiffs to comply with that ordinance, constituted a taking in violation of the California Constitution. In reviewing a dismissal following the trial court's sustaining of a demurrer, we take the properly pleaded material allegations of the complaint as true. ( ABC Internat. Traders, Inc. v. Matsushita Electric Corp. (1997) 14 Cal.4th 1247, 1253 , 61 Cal.Rptr.2d 112 , 931 P.2d 290 ; Aubry v. Tri-City Hospital Dist. (1992) 2 Cal.4th 962, 966-967 , 9 Cal. Rptr.2d 92 , 831 P.2d 317 .)

Plaintiffs allege the following as to the hotel's legal status and history. "From 1916 to the present, the guest rooms in the San Remo Hotel have been rented primarily to transient and tourist guests on a commercial basis." Fewer than 20 percent of the rooms are, or were on August 5, 1987 (the effective date of the neighborhood commercial district zoning ordinance), occupied as primary residences. When plaintiffs purchased the San Remo Hotel in 1970, it was zoned C-2 commercial, with no restrictions on its use as a tourist hotel. After plaintiffs spent more than $250,000 on repairs and renovations, the City issued them a permit of occupancy for a 62 room hotel. From 1970 on, the City issued plaintiffs a hotel license and collected hotel taxes on the transient rental of the San Remo Hotel's rooms.

After the 1990 revision of the HCO, which revised the 1981 ordinance to place additional significant restrictions on tourist rental of residential rooms, plaintiffs applied under the HCO to "convert" to tourist use all the San Remo Hotel's rooms designated residential. The City's Planning Department told them they also had to apply for a conditional use permit for the conversion to be allowed under the zoning laws. The planning commission subsequently granted plaintiffs' application for a conditional use permit on three conditions: plaintiffs were to mitigate the loss of housing by complying with the HCO's housing-replacement provisions, offer current long-term residents lifetime leases, and obtain variances from floor-area ratio and parking requirements. Plaintiffs appealed the imposition of these conditions, but the City's Board of Supervisors upheld the planning commission's decision. Plaintiffs ultimately satisfied the conditions, paying a $567,000 in lieu fee assessed by the Department (formerly the Bureau) of Building Inspection, which thereafter issued a permit for full tourist use of the hotel.

The second amended complaint alleges a taking, in violation of the California Constitution, in the City Planning Commission's requirement that plaintiffs pay a housing-replacement fee as a condition of receiving a conditional use permit. Plaintiffs claim that imposition of the fee "fails to substantially advance a legitimate government interest" and that "[t]he amount of the fee imposed is not roughly proportional to the impact" of the proposed tourist use of the San Remo Hotel. They further allege that "the requirement that plaintiffs pay the City $567,000 to obtain an HCO permit to convert" constituted a taking of their property without just compensation, entitling them to a refund of the fee, and that the City's regulatory scheme as a whole and its application to the San Remo Hotel constituted a taking in violation of the California Constitution. The prayer is for damages of $567,000 plus interest from December 11, 1996 (apparently the date the fee was paid), and other damages according to proof.

The trial court sustained the City's demurrer to the second through fifth causes of action. Some causes of action were, the court held, barred by Pfeiffer v. City of La Mesa (1977) 69 Cal.App.3d 74 , 137 Cal. Rptr. 804 , in which the appellate court held that an applicant's compliance with building permit conditions barred a later action for damages resulting from imposition of the conditions. The remaining causes of action challenging the constitutionality of the HCO failed, the court concluded, because the HCO, as a legislative regulation, was not subject to the heightened scrutiny outlined in Notion v. California Coastal Comm'n (1987) 483 U.S. 825 , 107 S.Ct. 3141 , 97 L.Ed.2d 677 ( Nollan ) and Dolan v. City of Tigard (1994) 512 U.S. 374 , 114 S.Ct. 2309 , 129 L.Ed.2d 304 ( Dolan ).

Subsequently, after hearing argument directed to the writ petition, but without receiving evidence outside the administrative record, the trial court denied the requested writ of administrative mandate. Whether reviewed on the substantial evidence or independent judgment test, the court found, the Board of Permit Appeals' decision finding the San Remo Hotel was in residential use, thus requiring a conditional use permit for use as a tourist hotel, must be upheld. The trial court relied primarily on the City's 1981 issuance of a certificate of use designating all rooms in the hotel residential, concluding therefrom that "residential use of the San Remo was the only lawful use." Plaintiffs' temporary tourist rental of vacant rooms designated residential, as permitted under the HCO, the court held, did not effect a change of use under the zoning law.

Having denied the writ sought by the first cause of action and sustained a demurrer to the remaining counts without leave to amend, the trial court entered judgment for the City on plaintiffs' complaint. [7]

The Court of Appeal reversed. The City's demurrer should have been overruled, the appellate court held, because plaintiffs pled facts showing that the HCO, as applied to require payment of the $567,000 conversion fee, effected a taking under the Fifth Amendment to the United States Constitution and article I, section 19 of the California Constitution. The claim of a taking in exaction of the in lieu fee was one to which " Nollan/Dolan/Ehrlich [8] heightened or intermediate scrutiny analysis" should apply, because "the monetary sum of $567,000 exacted by the City here is a fee, exaction or payment in a `discretionary context' which presents 'an inherent and heightened risk that local government will manipulate the police power' in a manner which `avoid[s] what would otherwise be an obligation to pay just compensation.' ( Ehrlich, supra, 12 Cal.4th at p. 869, 50 Cal.Rptr.2d 242 , 911 P.2d 429 .)" The fee failed both the "essential nexus" and "rough proportionality" parts of that test, the Court of Appeal concluded, because it was based on the "fiction" of full residential use created by the 1981 survey and certificate of use. [9]

With regard to the writ petition, the Court of Appeal reversed and remanded for factual determinations. The trial court erred, the Court of Appeal held, in holding that the City's classification of all rooms in the San Remo Hotel as residential rendered any tourist use unlawful for purposes of zoning. "The City's local zoning laws applicable to hotels in the early 1980's made no distinction between the use permits for hotels based upon their `residential' or `tourist' status under the HCO. Thus, such Hotel rentals to tourists in the 1980's would have been legal, under the City's planning code then in effect. We therefore conclude that the Hotel could be an existing legal nonconforming use under the [North Beach district], notwithstanding the 1981 certificate of use." Because the trial court, believing prior tourist use would have been unlawful, had not determined its historical existence, the Court of Appeal remanded for further factual findings by the trial court or administrative agency concerning actual use of the hotel.

We granted the City's petition for review.

With the above background, we may proceed to resolve the issues raised by the parties. We address, first, the correctness of the trial court's denial of the petition for writ of administrative mandate; second, the proper level of scrutiny applicable to plaintiffs' claim that the HCO housing-replacement fee constituted a taking of their property without just compensation; and third, the merits of plaintiffs' facial and as-applied attacks on the HCO.

As already explained, San Francisco's zoning administrator, in a decision affirmed by the Board of Permit Appeals, determined that the proposed operation of the San Remo Hotel in full tourist use required plaintiffs to apply for and obtain a conditional use permit. After analysis, we conclude the trial court correctly denied the petition for writ of administrative mandate challenging this administrative decision.

Before the zoning administrator and in the trial court, plaintiffs contended that no conditional use permit was needed because tourist rental of the hotel was a permitted conditional use under the San Francisco Planning Code. A permitted conditional use in a neighborhood commercial district must have "lawfully existed" on the effective date of the 1987 neighborhood commercial district ordinance. (S.F. Planning Code, § 179, subd. (a)(2).) Section 178, subdivision (a)(2) of the same code defines the same term, in its more general application, as a use that "lawfully existed" at the time such uses became subject to a conditional use permit requirement. Although plaintiffs, echoing the Court of Appeal, focus on the "lawfulness" of tourist use under the 1981 HCO, a prior consideration is whether, and to what extent, tourist use of the San Remo Hotel "existed" as of 1987 or as of the time the San Remo first became subject to a conditional use permit requirement for tourist use, apparently by interim measures first added in 1982.

The administrative record shows that both residential and tourist rentals were significant uses of the San Remo Hotel at the relevant times. The zoning administrator cited annual unit usage reports filed by the San Remo Hotel with the Bureau (and later, the Department) of Building Inspection, which showed that during the period 1982 to 1992, between 25 and 57 units in the 62 unit hotel were occupied by residents. City hotel tax records from the last part of that period show that, even in the summer, a significant part of the San Remo's rental revenue was derived from (nontaxable) residential rentals, which constituted the majority of revenues in some autumn and winter seasons. As of May 1991, according to a hearing officer's decision allowing summer tourist rentals at the San Remo Hotel, some San Remo residents had occupied their rooms for as long as 10 years. A similar ruling in May 1989 noted that 15 to 20 of the residents then living at the San Remo had been there for six months or longer, some for "many years." Thus, the record supports the Board of Permit Appeals' finding that the San Remo Hotel was operated "at least in part" as a residential hotel in the early and mid 1980's.

In their application for HCO conversion, plaintiffs sought not to maintain the status quo but, in the words of their complaint, to "convert the San Remo Hotel's residentially designated hotel rooms [i.e., all the hotel's rooms] to tourist use." That application, according to plaintiffs, prompted the City to require a conditional use permit. Thus, as the zoning administrator understood plaintiffs' application, they sought authorization "to convert all of the units of the hotel to tourist use."

The zoning administrator correctly determined that such conversion required a conditional use permit even if some tourist use had previously lawfully existed. A permitted conditional use may continue "in the form in which it lawfully existed," but "may not be significantly altered, enlarged, or intensified, except upon approval of a new conditional use application." (S.F. Planning Code, § 178, subds. (b), (c).) Clearly a change from partial tourist use to complete tourist use would be a significant alteration or enlargement of the existing use, requiring a new conditional use permit.

We agree with plaintiffs that the superior court erred in stating, "The only lawful use of the San Remo ... was residential." The 1981 version of the HCO allowed vacant residential units to be rented on a short-term basis during the May to September tourist season. Since such rentals were also permitted under the San Remo Hotel's historical zoning (i.e., that preceding the 1982 and 1987 zoning changes), some tourist use lawfully existed prior to the 1982 and 1987 zoning restrictions. But the lawful temporary rental of vacant residential units, permitted with the further restriction that such units must immediately revert to residential use if needed (1981 HCO, § 41.16, subd. (a)(3)(B)), was not authority to use the hotel's rooms full time for tourist use, regardless of residential occupancy or demand. As the superior court observed, "This temporary tourist use is not a change of use under the Planning Code." Again, such full-time unrestricted tourist rental would be a significant alteration or enlargement of the historical lawful use, requiring new conditional use permission under San Francisco Planning Code section 178, subdivision (c). Plaintiffs do not claim that the actual tourist use of the San Remo Hotel went beyond that permitted by the HCO, nor does the record contain evidence of such illegally extensive use in the relevant period; indeed, plaintiff Robert Field stated in the trial court that plaintiffs had always complied with the HCO. Even as to those rooms that had, on occasion, been lawfully rented to tourists, therefore, the zoning administrator and Board of Permit Appeals were correct to require a conditional use permit for permanent tourist use.

Plaintiffs attribute to the City the argument that "the San Remo Hotel's zoning classification as a tourist hotel was changed by the 1981[HCO]." To rebut the City's supposed claim, plaintiffs cite a 1981 opinion letter by the San Francisco City Attorney that stated the then proposed HCO was not a zoning law and thus could be enacted without a hearing before the San Francisco Planning Commission. (S.F. City Atty, Opn. No. 81-54 (Sept. 14, 1981) pp. 7-8.) In reply, the City disavows any claim that the HCO changed the San Remo Hotel's zoning status, but insists that "the trial court correctly considered the HCO as a legal restriction on the use of the hotel." As we have explained, however, the critical issue in this case is not the lawfulness of the historical tourist use, but its extent. The HCO's restrictions on tourist use are pertinent in that they limited the hotel's actual tourist use during the 1980's. But, as plaintiffs do not claim they engaged in any tourist use beyond what the HCO permitted, and the record shows no such unauthorized use before 1987, we are not concerned here with whether tourist rentals in violation of the HCO would or would not have constituted a "lawful" use for purposes of the Planning Code provisions on permitted conditional uses. The record demonstrates that prior to 1987 the San Remo Hotel had substantial residential use and, as plaintiffs do not dispute, tourist use was restricted, in compliance with the HCO, to summer rentals of vacant units. Conversion to complete full-time tourist use would therefore be a significant expansion of the hotel's historical tourist use, requiring a conditional use permit. Nothing in the City Attorney's 1981 opinion alters our analysis or affects our conclusion. [10]

The Court of Appeal also criticized the City (i.e., the zoning administrator and the Board of Appeals) for characterizing the San Remo Hotel's historical residential use as "group housing" when, at the same time, the San Remo operated under a City-issued "hotel" permit and paid "hotel" taxes. In response, the City cites a zoning provision (S.F. Planning Code, § 209.2, subd. (a)), added in 1978, that defines the housing use category, "Group housing, boarding," in a manner that, on its face, arguably includes a residential hotel. But another part of the same section, also dating from 1978, defines the use "Hotel, inn or hostel," in a manner that includes the San Remo Hotel's historic use if the San Remo's rooms were, during the relevant period, used "primarily for the accommodation of transient overnight guests" ( id., subds. (d), (e), italics added), a question not clearly answered by the administrative record excerpt in the appellate record. Nor does the 1978 law appear to preclude a mixed-use hotel from having both classifications.

Fortunately, we need not determine the correct zoning categorization of the San Remo Hotel's pre-1987 use in order to decide this case. The question before us is whether plaintiffs' proposed full tourist use of the hotel qualifies as a permitted conditional use under the current zoning laws. As we have seen, such use does qualify to the extent it lawfully existed when the current laws' restrictions came into effect, but does not qualify to the extent plaintiffs propose to significantly alter or expand it. This is true regardless of whether the San Remo was historically classified as a hotel, as group housing, or both.

In classifying uses for purposes of neighborhood commercial district zoning, we are directed to consider separately each use in a multiple-use structure. (S.F. Planning Code, § 703.2, subd. (b)(1).) That the San Remo might legitimately have been classifiable as a "Hotel, inn or hostel" under the 1978 zoning law ( id., § 209.2, subd. (e)) or a "Hotel, tourist" under the 1987 neighborhood commercial district law ( id., § 790.46) as well as a "hotel, residential" ( id., § 790.47) is therefore not determinative. (Cf. Tenderloin Housing Clinic, Inc. v. Astoria Hotel, Inc. (2000) 83 Cal.App.4th 139, 144, fn. 2 , 98 Cal.Rptr.2d 924 [hotel with 13 rooms designated for tourist use and 79 for residential use was both a tourist hotel and a residential hotel under the Planning Code].) The historical extent of tourist use itself determines the extent to which the San Remo Hotel can, under current law, be put to that use without conditional use permits.

Both minority opinions misapprehend the factual context in which the conditional use permit was issued. The concurring and dissenting opinion, which would have us remand the case for factual findings (cone. & dis. opn, post, 117 Cal.Rptr.2d at pp. 301-302, 41 P.3d at pp. 114-115), proceeds on the incorrect assumption that conditional use permits are issued on a room-by-room basis, and that a determination of the precise number of rooms in tourist use when the conditional-use permit requirement came into effect would materially affect the permit requirement ( id. at pp. 301-303, 41 P.3d at pp. 114-115). Contrary to suggestions in the concurring and dissenting opinion, the number of rooms for which a new tourist use was proposed is of no import as to whether a conditional use permit was required in the first instance, so long as some expansion of tourist use was proposed. Nor did the conditional use permit specify the number of rooms subject to one-to-one replacement under the HCO, calculate the in lieu replacement fee to be assessed, or impose any other condition dependent on the number of rooms previously in tourist use. Because the record shows some residential use at all relevant times, and because plaintiffs' tourist use of the hotel was, as required under the HCO, temporary and subject to preemption by residential demand, a conditional use permit was required regardless of the exact number of rooms being rented to tourists at any time. Hence, no basis exists for "grandfather[ing]" tourist use in either the entire hotel or individual rooms as a permitted conditional use, as the concurring and dissenting opinion argues should have been done. (Cone. & dis. opn, post, 117 Cal. Rptr.2d at pp. 301-303, 41 P.3d at pp. 114-116.)

The dissent argues at length that the City Planning Commission acted in violation of the state Ellis Act (Gov.Code, §§ 7060-7060.7), which allows the withdrawal of residential accommodations from the market. (Dis. opn, post, 117 Cal. Rptr.2d at pp. 310-311, 41 P.3d at pp. 121-122.) This argument founders on the stubborn fact that, so far as the record or briefing here shows, plaintiffs never took the measures necessary to invoke their statutory rights under the act. (See Gov. Code, § 7060.4 [permitting local governments to establish notice requirements for withdrawal of accommodations]; S.F. Admin. Code, § 37.9A(f) [establishing such notice requirements].) Hence, we have no occasion here to discuss the preemptive effect of the Ellis Act addressed in Bullock v. City and County of San Francisco (1990) 221 Cal.App.3d 1072 , 271 Cal.Rptr. 44 , where the plaintiff hotel owner had "in no uncertain terms and in accordance with the [Ellis Act] procedures established by the City, advised the City of his intent to depart the business of renting residential hotel units." ( Id. at p. 1100, 271 Cal.Rptr. 44 .) Nor, contrary to the dissent's suggestion, can the Ellis Act be read as occupying the fields of local real property regulation, zoning, or residential hotel regulation. (See Gov.Code, § 7060.1, subds. (b), (c).)

Similarly, the dissent's claim that "the planning commission chose to require HCO compliance and thereby used the leverage it gained ... to exact a $567,000 fee" (dis. opn, post, 117 Cal.Rptr.2d at p. 312 , 41 P.3d at p. 123 ) misrepresents the facts in the appellate record. In their application for a conditional use permit, plaintiffswho had already applied for a conversion permit under the HCO (see ante, 117 Cal.Rptr.2d at pp. 278-279, 41 P.3d at p. 95 )assured the planning commission that they would comply with the HCO. The commission incorporated that assurance as a condition of the permit, but did not itself assess any fee.

Because, as the administrative record demonstrates, tourist use of the San Remo Hotel before enactment of the conditional use requirements neither encompassed all the hotel's units nor occurred full time without regard to residential occupancy and demand, plaintiffs' proposal to convert to full-time tourist use constitutes an expansion of the hotel's historical use requiring conditional use authorization. Consequently, the trial court correctly denied the petition for writ of administrative mandate, and the Court of Appeal erred in reversing this aspect of the trial court's judgment.

The takings clause of the California Constitution (art. I, § 19) provides: "Private property may be taken or damaged for public use only when just compensation, ascertained by a jury unless waived, has first been paid to, or into court for, the owner." The federal takings clause (U.S. Const., 5th Amend.) provides: "nor shall private property be taken for public use without just compensation."

By virtue of including "damage[ ]" to property as well as its "taking]," the California clause "protects a somewhat broader range of property values" than does the corresponding federal provision. ( Hensler v. City of Glendale (1994) 8 Cal.4th 1, 9, fn. 4 , 32 Cal.Rptr.2d 244 , 876 P.2d 1043 ; accord, Varjabedian v. City of Madera (1977) 20 Cal.3d 285, 298 , 142 Cal.Rptr. 429 , 572 P.2d 43 ; see Bacich v. Board of Control (1943) 23 Cal.2d 343, 350 , 144 P.2d 818 ; Reardon v. San Francisco (1885) 66 Cal. 492, 501 , 6 P. 317 .) But aside from that difference, not pertinent here, we appear to have construed the clauses congruently. (See, e.g., Santa Monica Beach, Ltd. v. Superior Court (1999) 19 Cal.4th 952, 957, 962-975 , 81 Cal. Rptr.2d 93 , 968 P.2d 993 ( Santa Monica Beach ) [takings challenge to rent control regulation under both clauses considered without separate discussion of the state clause]; Hensler v. City of Glendale, supra, at p. 9, fn. 4, 32 Cal.Rptr.2d 244 , 876 P.2d 1043 [conclusion that U.S. Const., 5th Amend, was not violated "applies equally" to Cal. Const, art. I, § 19].) Despite plaintiffs' having sought relief in this court only for a violation of article I, section 19 of the California Constitution, therefore, we will analyze their takings claim under the relevant decisions of both this court and the United States Supreme Court.

"In determining whether a government regulation of property works a taking of property under the Fifth Amendment to the United States Constitution, the United States Supreme Court has generally eschewed any `set formula' for determining whether a taking has occurred, preferring to engage in `"essentially ad hoc, factual inquiries"' ( Lucas v. South Carolina Coastal Council (1992) 505 U.S. 1003, 1015 [ 112 S.Ct. 2886, 2893 , 120 L.Ed.2d 798 ]), which focus in large part on the economic impact of the regulation (see Penn Central Transp. Co. v. New York City (1978) 438 U.S. 104, 124 [ 98 S.Ct. 2646, 2659 , 57 L.Ed.2d 631 ]; Keystone Bituminous Coal Assoc. v. DeBenedictis (1987) 480 U.S. 470, 493-501 [ 107 S.Ct. 1232, 1246-1250 , 94 L.Ed.2d 472 ]).... Other than this ad hoc inquiry, the court has held categorically that property is taken when a government regulation `compel[s] [a] property owner to suffer physical "invasion" of his property' or `denies all economically beneficial or productive use of land.' ( Lucas, supra, 505 U.S. at pp. 1015-1016, 112 S.Ct. at p. 2893 .) The court has also stated that `the Fifth Amendment is violated when a landuse regulation "does not substantially advance legitimate state interests."' ( Lucas, supra, 505 U.S. at p. 1016 , 112 S.Ct. at p. 2894 .)" ( Santa Monica Beach, supra, 19 Cal.4th at p. 964 , 81 Cal.Rptr.2d 93 , 968 P.2d 993 .)

As in Santa Monica Beach , it is the last-mentioned prong of the high court's takings analysis that is at issue here. In particular, the parties debate whether a heightened level of means-ends scrutiny, the force and application of which has been developed in Nollan, supra, 483 U.S. 825 , 107 S.Ct. 3141 , 97 L.Ed.2d 677 , Dolan, supra, 512 U.S. 374 , 114 S.Ct. 2309 , 129 L.Ed.2d 304 , Ehrlich, supra, 12 Cal.4th 854 , 50 Cal.Rptr.2d 242 , 911 P.2d 429 , and Santa Monica Beach, supra, 19 Cal.4th 952 , 81 Cal.Rptr.2d 93 , 968 P.2d 993 , applies to review of the conversion fee the City required plaintiffs to pay under the HCO.

In Nollan, a California agency conditioned its approval for the plaintiffs to rebuild a beachfront house on their dedication of a public easement providing lateral access across their portion of the beach. Although the easement would constitute a physical invasion of property, the high court recognized it could nonetheless be demanded as a condition of the development permit, if the permit could otherwise have been denied and if the easement condition "serve[d] the same legitimate police-power purpose as a refusal to issue the permit." ( Nollan, supra, 483 U.S. at p. 836 , 107 S.Ct. 3141 .) "The evident constitutional propriety disappears, however, if the condition substituted for the prohibition utterly fails to further the end advanced as the justification for the prohibition." ( Id, at p. 837, 107 S.Ct. 3141 .) Without this "essential nexus," between the permit condition and the development ban, "the building restriction is not a valid regulation of land use but `an out-and-out plan of extortion.'" ( Ibid. )

Because the conditional exaction in Nollan failed to meet "even the most untailored standards" ( Nollan, supra, 483 U.S. at p. 838 , 107 S.Ct. 3141 ), the court did not need to elucidate with any precision the required "`fit'" between the exaction and its purposes. But the court cautioned that, when the circumstances created a potential for the government to extort property by withholding otherwise unrelated permits, judicial scrutiny would be searching: "[O]ur cases describe the condition for abridgment of property rights through the police power as a 'substantial advancing]' of a legitimate state interest. We are inclined to be particularly careful about the adjective where the actual conveyance of property is made a condition to the lifting of a land-use restriction, since in that context there is heightened risk that the purpose is avoidance of the compensation requirement, rather than the stated police-power objective." ( Id. at p. 841, 107 S.Ct. 3141 .) "Thus in Nollan, the rule that the government's physical occupation of private property is a per se taking is transformed, in the context of a development application, into a rule of heightened scrutiny to ensure that a required development dedication is not a mere pretext to obtain or otherwise physically invade property without just compensation." ( Ehrlich, supra, 12 Cal.4th 854, 890 , 50 Cal. Rptr.2d 242 , 911 P.2d 429 (cone. opn. of Mosk, J.).)

Dolan, like Nollan, involved a government agency's conditioning a development permit on dedication of a portion of the applicant's real property. In Dolan , the high court addressed the question it had reserved in Nollan, the "required degree of connection between the exactions and the projected impact of the proposed development." ( Dolan, supra, 512 U.S. at p. 386, 114 S.Ct. 2309 .) The court concluded that a "`rough proportionality'" standard "best encapsulates what we hold to be the requirement of the Fifth Amendment. No precise mathematical calculation is required, but the city must make some sort of individualized determination that the required dedication is related both in nature and extent to the impact of the proposed development." ( Id. at p. 391, 114 S.Ct. 2309 .)

The Dolan court also briefly addressed the scope of applicability of the heightened scrutiny and shifted burden of persuasion outlined in that decision and in Nollan: "Justice Stevens' dissent takes us to task for placing the burden on the city to justify the required dedication. He is correct in arguing that in evaluating most generally applicable zoning regulations, the burden properly rests on the party challenging the regulation to prove that it constitutes an arbitrary regulation of property rights. [Citation.] Here, by contrast, the city made an adjudicative decision to condition petitioner's application for a building permit on an individual parcel. In this situation, the burden properly rests on the city." ( Dolan, supra, 512 U.S. at p. 391, fn. 8 , 114 S.Ct. 2309 .) Most land use regulations "involve[ ] essentially legislative determinations classifying ... areas of the city, whereas here the city made an adjudicative decision to condition petitioner's application for a building permit on an individual parcel." ( Id. at p. 385, 114 S.Ct. 2309 .)

In Ehrlich , this court addressed the question of whether the heightened scrutiny outlined in Nollan and Dolan applied to a monetary exaction. The defendant city in Ehrlich had conditioned permits for the development of a condominium complex on the site of a former private tennis club on the owner's payment of a $280,000 fee to be used for city recreational facilities. Though the members of this court disagreed on various parts of the analysis, we unanimously held that this ad hoc monetary exaction was subject to Nollan/Dolan scrutiny. ( Ehrlich, supra, 12 Cal.4th at pp. 874-881, 50 Cal.Rptr.2d 242 , 911 P.2d 429 (plur. opn. of Arabian, J.); id. at pp. 899-901, 50 Cal.Rptr.2d 242 , 911 P.2d 429 (cone. opn. of Mosk, J.); id. at p. 907, 50 Cal.Rptr.2d 242 , 911 P.2d 429 (cone. & dis. opn. of Kennard, J.); id. at p. 912, 50 Cal.Rptr.2d 242 , 911 P.2d 429 (cone. & dis. opn. of Werdegar, J.).) In such cases, the exaction must be more than "theoretically" or "plausibly" related to the ends that would be served by permit denial; Nollan and Dolan require "a factually sustainable proportionality between the effects of a proposed land use and a given exaction." ( Ehrlich, supra, at p. 880 , 50 Cal.Rptr.2d 242 , 911 P.2d 429 (plur. opn. of Arabian, J.).)

In holding the fee at issue subject to Nollan/Dolan, we emphasized that because the city had exercised its discretionary powers in imposing and calculating the recreational impact fee, rather than doing so pursuant to a legislative mandate or formula, imposition of the fee bore much the same potential for illegitimate leveraging of private property as did the real property exactions in Nollan and Dolan . Thus, the plurality concluded that heightened scrutiny was appropriate "[w]hen such exactions are imposed-as in this case-neither generally nor ministerially, but on an individual and discretionary basis." ( Ehrlich, supra, 12 Cal.4th at p. 876 , 50 Cal.Rptr.2d 242 , 911 P.2d 429 (plur. opn. of Arabian, J.).) The plurality further distinguished "generally applicable development fee[s] or assessments]," as to which "the courts have deferred to legislative and political processes," from "special, discretionary permit conditions" like the one at issue in Ehrlich. (Id. at p. 881, 50 Cal.Rptr.2d 242 , 911 P.2d 429 .) Justice Mosk, concurring, explained that although "general governmental fees" are "judged under a standard of scrutiny closer to the rational basis review of the equal protection clause than the heightened scrutiny of Nollan and Dolan " ( id. at p. 897, 50 Cal. Rptr.2d 242 , 911 P.2d 429 (cone. opn. of Mosk, J.)), "when a municipality singles out a property developer for a development fee not imposed on others, a somewhat heightened scrutiny of that fee is required to ensure that the developer is not being subject to arbitrary treatment for extortionate motives" ( id. at p. 900, 50 Cal.Rptr.2d 242 , 911 P.2d 429 ). Finally, Justice Kennard agreed that "[b]ecause the $280,000 recreational mitigation fee was imposed on Ehrlich's development application individually, and not pursuant to an ordinance or rule of general applicability, the constitutionality of this fee is evaluated using the Nollan Dolan `essential nexus' and `rough proportionality' analysis." ( Id. at p. 907 (cone. & dis. opn. of Kennard, J.).)

A majority in Ehrlich further agreed that to the extent a development mitigation fee is not subject to heightened scrutiny under Nollan and Dolan , there must nonetheless be a "reasonable relationship" between the fee and the deleterious impacts for mitigation of which the fee is collected. ( Ehrlich, supra, 12 Cal.4th at pp. 865, 867, 50 Cal.Rptr.2d 242 , 911 P.2d 429 (plur. opn. of Arabian, J.); id. at p. 897, 50 Cal.Rptr.2d 242 , 911 P.2d 429 (cone. opn. of Mosk, J.).)

In Santa Monica Beach supra, 19 Cal.4th 952 , 81 Cal.Rptr.2d 93 , 968 P.2d 993 , considering a challenge to a municipal rent control ordinance, we reviewed and synthesized the prior decisions as follows. "From the above, it can be inferred that the `substantially advance' standard in the takings context is applied differently depending on the type of government action under consideration. As Nollan and Dolan both attest, government requirements that property owners dedicate land as a condition of receiving a development permit will receive the highest scrutinya type of intermediate scrutiny in which a government's dedication requirements will pass constitutional muster as long as the government `make[s] some sort of individualized determination that the required dedication is related both in nature and extent to the impact of the proposed development.' ( Dolan, supra, 512 U.S. at p. 391 , 114 S.Ct. at pp. 2319-2320, fn. omitted.) ... The most deferential review of land use decisions appears to be for those that pertain to `essentially legislative determinations' that do not require any physical conveyance of property." ( Santa Monica Beach, supra, 19 Cal.4th at p. 966 , 81 Cal.Rptr.2d 93 , 968 P.2d 993 , quoting Dolan, supra, 512 U.S. at p. 385 , 114 S.Ct. 2309 .) `

"We recognized these different levels of takings scrutiny in [ Ehrlich, supra, 12 Cal.4th 854 ]. We rejected the claim that the Nollan and Dolan standards do not apply to development fees imposed on an individualized basis as a condition for development. ... But a different standard of scrutiny would apply to development fees that are generally applicable through legislative action `because the heightened risk of the "extortionate" use of the police power to exact unconstitutional conditions is not present.' ( Id. at p. 876, 50 Cal.Rptr.2d 242 , 911 P.2d 429 ; see also id. at p. 897, 50 Cal.Rptr.2d 242 , 911 P.2d 429 (cone. opn. of Mosk, J.); id. at p. 903, 50 Cal.Rptr.2d 242 , 911 P.2d 429 (cone, and dis. opn. of Kennard, J.).) Thus, individualized development fees warrant a type of review akin to the conditional conveyances at issue in Nollan and Dolan," while generally applicable development fees warrant a more deferential type of review. ( Santa Monica Beach, supra, 19 Cal.4th at pp. 966-967, 81 Cal.Rptr.2d 93 , 968 P.2d 993 .)

The Court of Appeal held that housing replacement fees assessed under the HCO were subject to Nollan/Dolan/Ehrlich review because they were exacted discretionarily and applied only to a relatively small number of property owners rather than to "every other property in the City." Plaintiffs defend that analysis, while the City argues for the more deferential constitutional scrutiny applicable to land use regulations made generally applicable by legislative enactment to a class of property owners.

We agree with the City. Contrary to the Court of Appeal's assertion, and unlike Ehrlich , the HCO does not provide City staff or administrative bodies with any discretion as to the imposition or size of a housing replacement fee. Under the HCO, the responsible city agency, the Department (formerly the Bureau) of Building Inspection, "shall ... deny" an application to convert residential units to tourist use if the housing replacement requirement is not satisfied and "shall issue" the permit if the ordinance's requirements, including that for housing replacement, are met. (HCO, §§ 41.14, 41.15.) The applicant chooses how to satisfy the replacement requirement, whether by constracting or bringing onto the market new units; by sponsoring such construction by a public or nonprofit private housing developer; or by paying, in lieu of such construction, a fee to a designated City housing fund. ( Id., § 41.13.) If the applicant chooses the in lieu fee, its amount is determined according to a set formula based on replacement cost, which in turn is determined by a different City agency, the City's Department of Real Estate, through two independent appraisals. ( Ibid. ) Thus, no meaningful government discretion enters into either the imposition or the calculation of the in lieu fee. [11]

Nor did the City single out plaintiffs for payment of a housing replacement fee. The HCO is generally applicable legislation in that it applies, without discretion or discrimination, to every residential hotel in the city. All proposals to convert residential to tourist use are subject to the same ordinance. In suggesting that an ordinance, to be considered generally applicable, must apply to "every other property in the City," the Court of Appeal invoked an impossible standard, one that would be met by almost no rationally drawn land use regulation. The HCO applies to all property in the class logically subject to its strictures, that is, to all residential hotel units; no more can rationally be demanded of local land use legislation in order to qualify for deferential review. (We do not speak of a legislative "class" artificially tailored to encompass only a single property; no such claim has been or could be made as to the HCO.) [12]

In these respects a housing replacement fee assessed under the HCO stands in sharp contrast to the recreational facilities replacement fee we found subject to heightened scrutiny in Ehrlich . In that case, the city relied on no specific legislative mandate to impose the fee condition and no legislatively set formula to calculate its size. The condition was imposed ad hoc, entirely at the discretion of the city council and staff. ( Ehrlich, supra, 12 Cal.4th at p. 862 , 50 Cal.Rptr.2d 242 , 911 P.2d 429 .) So far as the court's opinions reveal, the plaintiffs property development proposal was the only one upon which such a fee condition had been imposed. We concluded that applying Nollan/Dolan review to such a "special, discretionary permit condition[ ]" ( Ehrlich, supra, at p. 881 , 50 Cal.Rptr.2d 242 , 911 P.2d 429 (plur. opn. of Arabian, J.)) was necessary "to ensure that the developer is not being subject to arbitrary treatment for extortionate motives" ( id. at p. 900, 50 Cal. Rptr.2d 242 , 911 P.2d 429 (cone. opn. of Mosk, J.)). At the same time, we distinguished cases such as the present one, involving a "generally applicable development fee or assessment" ( id. at p. 881, 50 Cal.Rptr.2d 242 , 911 P.2d 429 (plur. opn. of Arabian, J.)) imposed not "individually" but "pursuant to an ordinance or rule of general applicability" ( id. at p. 907, 50 Cal.Rptr.2d 242 , 911 P.2d 429 (cone. & dis. opn. of Kennard, J.)).

The "sine qua non" for application of Nollan/Dolan scrutiny is thus the "discretionary deployment of the police power" in "the imposition of land-use conditions in individual cases." ( Ehrlich, supra, 12 Cal.4th at p. 869 , 50 Cal.Rptr.2d 242 , 911 P.2d 429 (plur. opn. of Arabian, J.).) Only "individualized development fees wan-ant a type of review akin to the conditional conveyances at issue in Nollan and Dolan." ( Santa Monica Beach, supra, 19 Cal.4th at pp. 966-967, 81 Cal.Rptr.2d 93 , 968 P.2d 993 ; see also Landgate, Inc. v. California Coastal Com. (1998) 17 Cal.4th 1006, 1022 , 73 Cal.Rptr.2d 841 , 953 P.2d 1188 ( Landgate ) [heightened scrutiny applies to "development fees imposed on a property owner on an individual and discretionary basis"].)

Under our precedents, therefore, housing replacement fees assessed under the HCO are not subject to Nollan/Dolan/Ehrlich scrutiny.

Plaintiffs argue that a legislative scheme of monetary exactions (i.e., a schedule of development mitigation fees) nevertheless should be subject to the same heightened scrutiny as the ad hoc fees we considered in Ehrlich , because of the danger a local legislative body will use such purported mitigation feesunrelated to the impacts of developmentsimply to fill its coffers. Thus, plaintiffs hypothesize that absent careful constitutional scrutiny a city could "put zoning up for sale" by, for example, "prohibit[ing] all development except for one-story single-family homes, but offering] a second story permit for $20,000, an apartment building permit for $10,000 per unit, a commercial building permit for $50,000 per floor, and so forth." [13]

We decline plaintiffs' invitation to extend heightened takings scrutiny to all development fees, adhering instead to the distinction we drew in Ehrlich, supra, 12 Cal.4th 854 , 50 Cal.Rptr.2d 242 , 911 P.2d 429 , Landgate, supra, 17 Cal.4th 1006 , 73 Cal.Rptr.2d 841 , 953 P.2d 1188 , and Santa Monica Beach, supra, 19 Cal.4th 952 , 81 Cal.Rptr.2d 93 , 968 P.2d 993 , between ad hoc exactions and legislatively mandated, formulaic mitigation fees. While legislatively mandated fees do present some danger of improper leveraging, such generally applicable legislation is subject to the ordinary restraints of the democratic political process. A city council that charged extortionate fees for all property development, unjustifiable by mitigation needs, would likely face widespread and well-financed opposition at the next election. Ad hoc individual monetary exactions deserve special judicial scrutiny mainly because, affecting fewer citizens and evading systematic assessment, they are more likely to escape such political controls.

Nor are plaintiffs correct that, without Nollan/Dolan/Ehrlich scrutiny, legislatively imposed development mitigation fees are subject to no meaningful means-ends review. As a matter of both statutory and constitutional law, such fees must bear a reasonable relationship, in both intended use and amount, to the deleterious public impact of the development. (Gov.Code, § 66001; Ehrlich, supra, 12 Cal.4th at pp. 865, 867, 50 Cal.Rptr.2d 242 , 911 P.2d 429 (plur. opn. of Arabian, J.); id. at p. 897, 50 Cal.Rptr.2d 242 , 911 P.2d 429 (cone. opn. of Mosk, J.); Associated Home Builders etc., Inc. v. City of Walnut Creek (1971) 4 Cal.3d 633, 640 , 94 Cal.Rptr. 630 , 484 P.2d 606 .) Plaintiffs' hypothetical city could only "put [its] zoning up for sale" in the manner imagined if the "prices" charged, and the intended use of the proceeds, bore a reasonable relationship to the impacts of the various development intensity levels on public resources and interests. While the relationship between means and ends need not be so close or so thoroughly established for legislatively imposed fees as for ad hoc fees subject to Ehrlich, the arbitrary and extortionate use of purported mitigation fees, even where legislatively mandated, will not pass constitutional muster.

Finally, we should not lose sight of the constitutional background. "To put the matter simply, the taking of money is different, under the Fifth Amendment, from the taking of real or personal property. The imposition of various monetary exactionstaxes, special assessments, and user feeshas been accorded substantial judicial deference." ( Ehrlich, supra, 12 Cal.4th at p. 892 , 50 Cal.Rptr.2d 242 , 911 P.2d 429 (cone. opn. of Mosk, J.).) "There is no question that the takings clause is specially protective of property against physical occupation or invasion.... It is also true ... that government generally has greater leeway with respect to noninvasive forms of land-use regulation, where the courts have for the most part given greater deference to its power to impose broadly applicable fees, whether in the form of taxes, assessments, user or development fees." ( Id. at pp. 875-876, 50 Cal. Rptr.2d 242 , 911 P.2d 429 (plur. opn. of Arabian, J.).)

Nollan and Dolan involved the government's exaction of an interest in specific real property, not simply the payment of a sum of money from any source available; they have generally been limited to that context. (See, e.g., Monterey v. Del Monte Dunes at Monterey, Ltd. (1999) 526 U.S. 687, 703 , 119 S.Ct. 1624 , 143 L.Ed.2d 882 [ Dolan "inapposite" to permit denial]; Clajon Production Corp. v. Petera (10th Cir.1995) 70 F.3d 1566, 1578 [heightened scrutiny limited to exaction of real property]; Commercial Builders v. Sacramento (9th Cir.1991) 941 F.2d 872, 875 [ Nollan inapplicable to housing mitigation fee]; cf. United States v. Sperry Corp. (1989) 493 U.S. 52, 62 , 110 S.Ct. 387 , 107 L.Ed.2d 290, fn. 9 ["It is artificial to view deductions of a percentage of a monetary award as physical appropriations of property. Unlike real or personal property, money is fungible"].) In Ehrlich, we extended Nollan and Dolan slightly, recognizing an exception to the general rule of deference on distribution of monetary burdens, because the ad hoc, discretionary fee imposed in that case bore special potential for government abuse. We continue to believe heightened scrutiny should be limited to such fees. (Accord, Krupp v. Breckenridge Sanitation Dist. (Colo.2001) 19 P.3d 687, 698 [to the extent Nollan/Dolan review applies to purely monetary fees, it is limited to "exactions stemming from adjudications particular to the landowner and parcel"].) Extending Nollan and Dolan generally to all government fees affecting property value or development would open to searching judicial scrutiny the wisdom of myriad government economic regulations, a task the courts have been loath to undertake pursuant to either the takings or due process clause. (See, e.g., Dolan, supra, 512 U.S. at p. 384 , 114 S.Ct. 2309 [reiterating "the authority of state and local governments to engage in land use planning" even when such regulation diminishes individual property values]; Penn Central Transp. Co. v. New York City, supra, 438 U.S. at p. 133 , 98 S.Ct. 2646 [that landmarks law burdens have more severe impact on some landowners than others does not render its application a taking: "Legislation designed to promote the general welfare commonly burdens some more than others"]; Usery v. Turner Elkhorn Mining Co. (1976) 428 U.S. 1, 19 , 96 S.Ct. 2882 , 49 L.Ed.2d 752 [wisdom of particular cost-spreading scheme "not a question of constitutional dimension"].)

Plaintiffs attack the housing replacement provisions of the HCO both on their face and as applied to the San Remo Hotel. In the discussion that follows, we address only the substantive contentions made in this court by the plaintiffs. [14]

Challenging the ordinance on its face, plaintiffs assert there is no connection between the housing replacement fees assessed and the housing lost by conversion to tourist use. We conclude, to the contrary, that the housing replacement fees bear a reasonable relationship to loss of housing. Under the ordinance, the amount of the in lieu fee is based on the number of rooms being converted from residential to tourist designation; the number of rooms designated residential is, in turn, based on the self-reported use as of September 23, 1979, shortly before a City moratorium on residential hotel conversion first came into force. (HCO, § 41.3, subd. (g).) On its face, the use of a defined historical measurement point is reasonably related to the HCO's housing preservation goals (see HCO, §§ 41.2, 41.3), and the use of individualized self-reported survey results, with inspection by City staff if needed, and opportunities for appeal by the hotel owner or challenge by other interested parties (HCO, § 41.6), is a facially reasonable means of determining initial status. Plaintiffs fail to demonstrate from the face of the ordinance that fees assessed under the HCO bear no reasonable relationship to housing loss in the generality or great majority of cases, the minimum showing we have required for a facial challenge to the constitutionality of a statute. (See Easier v. Lockyer (2000) 23 Cal.4th 472, 502 , 97 Cal.Rptr.2d 334 , 2 P.3d 581 ; California Teachers Assn. v. State of California (1999) 20 Cal.4th 327, 345, 347 , 84 Cal.Rptr.2d 425 , 975 P.2d 622 ; id. at pp. 358-359, 84 Cal.Rptr.2d 425 , 975 P.2d 622 (dis. opn. of Werdegar, J.).) [15]

Plaintiffs also assert the City has admitted, in the HCO itself, that the in lieu fees assessed are intended to raise money rather than mitigate the loss of housing. They cite to HCO section 41.3, subdivision (m), a legislative finding in the 1990 ordinance explaining why the in lieu fees were raised from 40 percent of replacement cost to 80 percent. That finding states that the 40 percent figure was found inadequate because of lower than expected contributions by government sources. "Federal, state and local funds were incorrectly assumed at that time to be available and sufficient to make up the shortfall between the 40 percent in lieu fee and actual replacement costs. For example, in 1979 the federal government was spending 32 billion dollars on housing and is spending only 7 billion dollars in 1989." (HCO, § 41.3, subd. (m).) Contrary to plaintiffs' suggestion, this finding does not tend to show an impermissible revenue-raising purpose for the in lieu fees, but only the legitimate purpose of more fully funding the replacement of housing lost through conversion.

We note, as well, that the structure of the HCO's housing replacement provisions rebuts plaintiffs' claim that they are intended merely to raise general revenue. No hotel owner is required to pay a fee to the City as a condition of conversion. Rather, to comply with the replacement provisions and receive a conversion permit, an owner may construct comparable housing units for rent; bring units onto the market from any building not subject to the HCO; construct or rehabilitate, even at less than a one-to-one ratio, apartment units for elderly, disabled or low-income renters, or transitional or emergency housing; or contribute to a private nonprofit housing developer for construction of comparable units. (HCO, § 41.13, subd. (a)(1)(3), (5).) Even when the hotel owner chooses to pay a fee in lieu of such replacement, the fee is not paid to the City's general fund but to a separate residential hotel preservation account. (HCO, § 41.13, subd. (a)(4).) The HCO was clearly not designed as a means of raising general revenue.

In their last facial claim, plaintiffs assert that the HCO does not preserve available housing because "[t]iny hotel rooms without baths and without kitchens are not housing." We disagree. While a single room without a private bath and kitchen may not be an ideal form of housing, such units accommodate many whose only other options might be sleeping in public spaces or in a City shelter. Plaintiffs do not dispute that San Francisco has long suffered from a shortage of affordable housing or that residential hotel units serve many who cannot afford security and rent deposits for an apartment. (See HCO, § 41.3, subds. (a)-(f).) Maintaining the availability of residential hotel rooms is a reasonable means of serving one segment of San Franciscans' housing needs. [16]

Our dissenting colleague argues that the HCO constitutes a facial taking because, in the well-known phrase of Justice Holmes, it affords insufficient "`reciprocity of advantage'" to owners of the hotels affected. (Dis. opn., post, 117 Cal.Rptr.2d at p. 316 , 41 P.3d at p. 126 , quoting Penna. Coal Co. v. Mahon (1922) 260 U.S. 393, 415 , 43 S.Ct. 158 , 67 L.Ed. 322 ( Penna. Coal Co.) . ) The dissent would apparently approve an economic regulation affecting property only if each property owner restricted by the regulation were guaranteed, at the same time, a proportionate benefit from the same regulation. Whether Holmes's conception of the justifiable regulation of property was as narrow as the dissent's is unclear, but, in any case, such a restrictive view has generally not controlled the development of takings law.

In Penna. Coal Co. itself, Justice Brandeis observed that in many cases where the high court had approved the exercise of police powers to regulate the use of property, the burdened property owner had received no reciprocal benefit from the regulation "unless it be the advantage of living and doing business in a civilized community." ( Penna. Coal Co., supra, 260 U.S. at p. 422 , 43 S.Ct. 158 (dis. opn. of Brandeis, J.).) In the many difficult cases that have followed, it has generally been the Brandeis view that has prevailed: "Under our system of government, one of the State's primary ways of preserving the public weal is restricting the uses individuals can make of their property. While each of us is burdened somewhat by such restrictions, we, in turn, benefit greatly from the restrictions that are placed on others. [Footnote 21:] The Takings Clause has never been read to require the States or the courts to calculate whether a specific individual has suffered burdens under this generic rule in excess of the benefits received. Not every individual gets a full dollar return in benefits for the taxes he or she pays; yet, no one suggests that an individual has a right to compensation for the difference between taxes paid and the dollar value of benefits received." ( Keystone Bituminous Coal Assn. v. DeBenedictis, supra, 480 U.S. at p. 491 & fn. 21, 107 S.Ct. 1232 , italics added; see also Penn Central Transp. Co. v. New York City, supra, 438 U.S. at p. 133 , 98 S.Ct. 2646 ( Penn Central ) ["that the Landmarks Law has a more severe impact on some landowners than on others ... does not mean that the law effects a 'taking'"]; Agins v. Tiburon (1980) 447 U.S. 255, 262 , 100 S.Ct. 2138 , 65 L.Ed.2d 106 [restrictive zoning ordinances "benefit the [property owners] as well as the public by serving the city's interest in assuring careful and orderly development of residential property with provision for openspace areas"].) Thus, the necessary reciprocity of advantage lies not in a precise balance of burdens and benefits accruing to property from a single law, or in an exact equality of burdens among all property owners, but in the interlocking system of benefits, economic and noneconomic, that all the participants in a democratic society may expect to receive, each also being called upon from time to time to sacrifice some advantage, economic or noneconomic, for the common good.

The federal and state takings clauses, to be sure, place a limit, imprecise as it may be, on the regulatory burdens an individual property owner may be made to bear for public purposes. The breadth or narrowness of the class burdened by the regulation, the extent to which a regulation defeats the owner's reasonable investment-backed expectations, and the extent to which the affected property is also benefitted by the regulation are certainly pertinent to whether a regulation works a taking. ( Agins v. Tiburon, supra, 447 U.S. at p. 262 , 100 S.Ct. 2138 ; Penn Central, supra, 438 U. S. at pp. 124, 132, 98 S.Ct. 2646 .) But the HCO neither targets an arbitrary small group of property owners, nor deprives all the burdened properties of so much of their value, without any corresponding benefit, as to constitute a taking on its face. As discussed earlier, the HCO affects all the approximately 500 residential hotels in San Francisco, comparable to the "over 400" New York City landmarks the United States Supreme Court relied upon in holding that landmark laws could not be characterized as "discriminatory, or 'reverse spot,' zoning." ( Penn Central, supra, at p. 132 , 98 S.Ct. 2646 .) Also like the landmarks law upheld in Penn Central, the HCO allows the property owner to continue the property's preordinance use unhindered; like the landmarks law, therefore, the HCO "does not interfere with what must be regarded as [the property owner's] primary expectation concerning the use of the parcel." ( Id. at p. 136., 98 S.Ct. 2646 ) Finally, the chief purpose of the HCO, ensuring affordable and available housing for those San Franciscans who would otherwise be without it, carries benefits for all the City's property owners, including those operating tourist hotels. (See id. at pp. 134-135, 98 S.Ct. 2646 [landmarks law benefits all New Yorkers].) We cannot agree with the dissent that a law applying on equal terms to all properties in a sizeable class defined by use, designed to benefit the City as a whole, and merely prohibiting a change of use from residential to commercial unless the owner mitigates the detrimental impact of that change, constitutes a facial taking of property.

The above may help to explain why the dissent's hypothetical concerning governmental appropriation of an automobile is inapposite. A law arbitrarily selecting a private automobile owner to dedicate his or her car to public use or pay for the government to buy another one would, as the dissent suggests (dis. opn, post, 117 Cal.Rptr.2d at p. 309 , 41 P.3d at p. 121 ) clearly require compensation. Less clear, but more like the present case, would be a law requiring all common carriers to take certain mitigation measures before converting from passenger to freight service. A burden placed broadly and nondiscriminatorily on changes in property's use is not the equivalent of an arbitrary decision to hold an individual's property for ransom. As elsewhere in takings law, the answers are found not in absolute rules for all cases, but by the particularized weighing of public and private interests. ( Agins v. Tiburon, supra, 447 U.S. at pp. 260-261, 100 S.Ct. 2138 .)

Finally, the dissent insists that owners of residential hotels cannot be required to continue the use of their property as low-income housing, or to mitigate the impact of ending that use, because they "did not cause poverty in San Francisco." (Dis. opn, post, 117 Cal.Rptr.2d at p. 309 , 41 P.3d at p. 120 .) But, of course, the owners of undeveloped property in Agins v. Tiburon, supra, 447 U.S. 255 , 100 S.Ct. 2138 , 65 L.Ed.2d 106 , restricted by zoning in how intensely they could develop the property, had not (yet) caused "the ill effects of urbanization" ( id. at p. 261, 100 S.Ct. 2138 ) the zoning law was designed to protect against, and the owners of New York City's Grand Central Terminal had not (yet) caused the loss of historic structures that motivated that city's landmarks law ( Perm Central, supra, 438 U.S. at p. 107 , 98 S.Ct. 2646 ). Here, as in those cases, it is the detrimental effects of a change in the use of property that motivates the regulation. A use not in itself noxious or harmful, such as the operation of a tourist hotel, may nonetheless call for mitigation when the change of property to that use results in the loss of an existing use of public importance. (See id. at p. 134, fn. 30, 98 S.Ct. 2646 ["Nor ... can it be asserted that the destruction or fundamental alteration of a historic landmark is not harmful"].)

If, as Justice Holmes warned, the Constitution "does not enact Mr. Herbert Spencer's Social Statics" ( Lochner v. New York (1905) 198 U.S. 45, 75 , 25 S.Ct. 539 , 49 L.Ed. 937 (dis. opn. of Holmes, J.)), it just as surely does not enact the late Robert Nozick's "Minimal State." (See Nozick, Anarchy, State and Utopia (1974) pp. ix, 171-172, 272-274.) However strongly and sincerely the dissenting justice may believe that government should regulate property only through rules that the affected owners would agree indirectly enhance the value of their properties (dis. opn, post, 117 Cal.Rptr.2d at p. 316 , 41 P.3d at p. 126 ), nothing in the law of takings would justify an appointed judiciary in imposing that, or any other, personal theory of political economy on the people of a democratic state.

Turning to the as-applied challenge, plaintiffs argue "[t]he $567,000 fee imposed by the Hotel Ordinance has no connection at all to the Field Brothers' tourist use of the San Remo Hotel" because (quoting the Court of Appeal) the permit simply "`allow[ed] an existing use to continue.'" As explained above, however, the mitigation fee was based on the number of units designated residential that were proposed for conversion, and the residential designation of the San Remo Hotel's rooms was reasonably based on the hotel management's own report of the rooms' use on the HCO's initial status date of September 23, 1979. [17] Plaintiffs' operative complaint, moreover, contains no allegations specifically relating to the San Remo Hotel's use as of the initial status date. The only use allegation covering that date is a general assertion that the hotel has been "primarily" used by "transient and tourist" guests since 1916. Nowhere do plaintiffs allege that the San Remo Hotel was, in 1979 or at any time, entirely in tourist use, as would be required to support their claim that the housing replacement fee has "no connection at all" to the hotel's historical use. (As discussed earlier, the administrative record indicates that plaintiffs could not truthfully so allege, since it shows mixed tourist and residential use throughout the 1980's.) The complaint, therefore, fails to state a cause of action on the ground that the amount of the fee paid by plaintiffs bore no reasonable relationship to the impacts of their proposed conversion to tourist use.

Plaintiffs further argue that, because the conditional use permit granted them by the City Planning Commission contained a condition requiring them to offer lifetime leases to existing residential tenants, no housing was lost by conversion to tourist use. Plaintiffs' conclusion, however, does not follow from their premise. The HCO seeks to preserve the supply of affordable housing units, not merely to extend the tenancy of any individual resident. (HCO, § 41.2.) Rooms designated residential under the ordinance that were vacant or temporarily rented to tourists at the time of conversion were nonetheless housing units that would be lost in the conversion, since after conversion they would no longer be held available, by law, for residential tenants. The same is true of rooms occupied by residential tenants who declined the offered leases. Even as to any rooms for which lifetime leases were accepted, the residential designation to be lost by conversion would have preserved the residential availability of those units after the lessees moved or passed away. Accordingly, even if no current resident were required to move, the City could reasonably base the in lieu fee on the number of units designated and reserved for residential use that would be made unavailable by the plaintiffs' proposed conversion of all of the hotel's rooms to tourist use.

Finally, plaintiffs, citing Ehrlich, supra, 12 Cal.4th at page 883 , 50 Cal. Rptr.2d 242 , 911 P.2d 429 , contend that the size of a development mitigation fee may not constitutionally be based on the loss of the property's prior use. In Ehrlich, of course, the court examined the recreational facilities fee under the "rough proportionality" standard ( id. at p. 882, 50 Cal. Rptr.2d 242 , 911 P.2d 429 ), a type of scrutiny inapplicable here. Perhaps more significantly, the fee in Ehrlich was imposed as a condition of a requested change in zoning for the subject property; there was no existing recreational use, the club having already been closed as uneconomical and its facilities demolished. ( Id. at pp. 861-862, 50 Cal.Rptr.2d 242 , 911 P.2d 429 .) The particular recreational facilities previously existing on the property having been permissibly demolished, the city could not use the value of their loss to impose a mitigation fee for the change in zoning sought by the property owner, although the majority held the fee could be based on the costs of planning and rezoning other properties for the needed recreational use. ( Id. at pp. 883-884, 50 Cal.Rptr.2d 242 , 911 P.2d 429 (plur. opn. of Arabian, J.); id. at p. 902, 50 Cal.Rptr.2d 242 , 911 P.2d 429 (cone. opn. of Mosk, J.).) In the present case, the housing that was to be lost by conversion of rooms from residential to tourist use had not been abandoned or demolished; nor had plaintiffs invoked their statutory right (Gov.Code, § 7060) to withdraw residential accommodations from the market. Plaintiffs sought not merely a change in the zoning affecting the site of the San Remo Hotel, but permission to change the use of existing residential facilities on the property. A mitigation fee measured by the resulting loss of housing units was thus reasonably related to the impacts of plaintiffs' proposed change in use.

The judgment of the Court of Appeal is reversed insofar as it reversed the superior court's judgment for defendant on plaintiffs' complaint. In all other respects the judgment of the Court of Appeal is affirmed.

WE CONCUR: GEORGE, C.J., KENNARD, and MORENO, JJ.

CONCURRING AND DISSENTING OPINION BY BAXTER, J.

I concur in the majority opinion insofar as it finds the Court of Appeal erred in assessing plaintiffs' as-applied challenge to the City and County of San Francisco's (City's) housing replacement fee under a standard of heightened scrutiny. I also agree with much of what the majority says about the test that does apply to this exaction. My objections are that in the remainder of its analysis, the majority fails to decide the issues on which we granted review and strains to reach questions that are not fairly included in the review petition.

I ask different questions and, not surprisingly, come up with different answers. Unlike the majority, I would affirm the Court of Appeal to the extent that it remanded for further proceedings concerning the petition for writ of administrative mandate and would remand for further proceedings under the correct takings standard.

The majority frames the first question as whether the City properly required plaintiffs to obtain a conditional use permit for full tourist use of the hotel. The problem with this formulation is that it treats the issue of the conditional use permit as an all-or-none proposition. This is a mistake. The critical issue is not whether plaintiffs needed to obtain a conditional use permit if they wished to rent all of the hotel's rooms to touristsas the majority notes, there is evidence in the record to support the finding that some rooms had historically been rented to long-term residents (maj. opn, ante, 117 Cal.Rptr.2d at pp. 281, 285, 41 P.3d at pp. 97, 100)but how many rooms required a permit before plaintiffs could rent them to tourists. To answer that question, it is necessary first to determine how many tourist rooms are grandfathered as a permitted conditional use.

As the City points out, the process of converting residential hotel rooms to tourist hotel rooms requires two separate permits: a conversion permit from the City Department of Building Inspection under the Residential Hotel Unit Conversion and Demolition Ordinance (HCO), and a conditional use permit from the City Planning Commission under the City Planning Code. Plaintiffs do not independently challenge here the requirement of obtaining a conversion permit from the Department of Building Inspection, merely the need to obtain a conditional use permit from the Planning Commission.

The parties agree that no duty exists to obtain a conditional use permit to continue a property use that qualifies as a permitted conditional use. [1] A permitted conditional use is one that "existed lawfully at the time a new zoning prohibition or restriction came into force" and "is conditionally permitted by the new zoning law." (Maj. opn., ante, 117 Cal.Rptr.2d at p. 283, fn. 10 , 41 P.3d at p. 98, fn. 10 .) Thus, a permitted conditional use in the North Beach neighborhood commercial district must have "lawfully existed" on the effective date of the 1987 ordinance. ( Id. at pp. 280-281, 41 P.3d at p. 97 , citing S.F. Planning Code, § 179, subd. (a)(2).) It was on this ground that plaintiffs disputed the need for a conditional use permit. (See maj. opn., ante, 117 Cal.Rptr.2d at p. 277 , 41 P.3d at p. 94 .) [2]

The City asserts that no tourist rental could have "lawfully" existed in the district because of the legal restrictions on tourist rental imposed by the HCO, which was enacted in 1981 and amended in 1990. The City, however, has been unable to explain why the HCO, which is part of the San Francisco Administrative Code, should have been considered by the City Planning Commission in determining whether tourist rental "lawfully existed" under section 179, subdivision (a)(2) of the San Francisco Planning Code, where the zoning ordinances are found. Indeed, the City's statements throughout these proceedings preclude any attempt to cut and paste the HCO into the Planning Code.

As the City repeatedly asserts, the HCO is "separate" from the City Planning Code and, in particular, did not change the zoning classification of the San Remo Hotel under the City Planning Code. The City likewise concedes that "[n]either the Planning Commission nor the Zoning Administrator has responsibility for administering the HCO." The HCO is administered instead by the City Department of Building Inspection.

These concessions are consistent with an opinion letter by the city attorney issued contemporaneously with the enactment of the HCO in 1981. In that letter, the city attorney stated that the HCO was not a zoning law, that the City Planning Code "makes no distinction between the use of a hotel devoted to permanent residents or to tourists," that the adoption of the HCO "does not render an existing structure or use a non-conforming structure or use," and that whether a tourist use may continue as a permitted conditional use is governed by the applicable sections of the Planning Code. (S.F. City Atty., Opn. No. 81-54 (Sept. 14, 1981) pp. 7-8; see maj. opn, ante, 117 Cal.Rptr.2d at p. 282 , 41 P.3d at p. 98 .)

The city attorney's opinion letter, together with the City's concessions, undercut the City's efforts here to claim that the HCO was a critical component of the determination whether tourist use lawfully existed as a permitted conditional use under the City Planning Code. Indeed, the zoning administrator was apparently so uncomfortable with the City's belated change of heart that he declined to accept the City's newly minted position and found instead that "[i]t has been the Department of City Planning's (`Department') administrative practice to classify residential hotels designated under the Hotel Ordinance as residential uses under the Planning Code." (Italics added.) But "administrative practice" is merely a convenience; it does not aid in determining whether a prior use was lawful or unlawful. The body of law relevant to that determinationi.e., the Planning Codedid not distinguish between tourist hotels and residential hotels and therefore did not prohibit tourist rentals at the San Remo. [3]

The superior court, on this same record, impliedly rejected the zoning administrator's restrained construction and found instead that "[s]ince the issuance of the Certificate of Use in 1981, residential use of the San Remo Hotel was the only lawful use." The majority holdsand I agree that the superior court erred. (Maj. opn, ante, 117 Cal.Rptr.2d at p. 282 , 41 P.3d at p. 98 .) But the majority declines to issue the administrative writ on the separate ground that plaintiffs' proposal to rent all of the rooms to tourists is "a significant alteration or enlargement of the historical lawful use" and therefore required a new conditional use permit. (Maj. opn., ante, 117 Cal.Rptr.2d at p. 282 , 41 P.3d at p. 98 .) I agree that a conditional use permit is required to the extent plaintiffs wish to alter or enlarge the historical lawful use. I disagree strongly, however, with the majority's unstated and unproven assumption that this encompasses all of the San Remo's rooms. The majority has simply terminated its analysis prematurely.

Nothing in the record supports the majority's implied finding that none of the rooms could be grandfathered as a permitted conditional tourist use. Indeed, neither the zoning administrator nor the superior court bothered to determine how many rooms might have been used historically as tourist rooms because of their erroneous belief that the HCO barred any tourist use from qualifying as a lawful permitted conditional use. (See maj. opn., ante, 117 Cal.Rptr.2d at p. 280 , 41 P.3d at p. 96 .) Similarly, the Board of Permit Appeals, in affirming the zoning administrator, apparently found it sufficient that the San Remo was a residential hotel "at least in part."

The majority's reliance on the unproven assumption that no rooms could be grandfathered as a permitted conditional use is all the more puzzling given that the opinion elsewhere is quite cognizant of the prospect that some rooms in the same building may be subject to differing classifications. As the majority explains, we must "consider separately each use in a multiple-use structure," such as the San Remo, in classifying uses "for purposes of neighborhood commercial district zoning." (Maj. opn., ante, 117 Cal.Rptr.2d at p. 283 , 41 P.3d at p. 101 , citing S.F. Planning Code, § 703.2, subd. (b)(1).) [4] The majority also recognizes that "[t]he administrative record shows that both residential and tourist rentals were significant uses of the San Remo Hotel at the relevant times." (Maj. opn., ante, 117 Cal.Rptr.2d at p. 281 , 41 P.3d at p. 97 .) Accordingly (and as the majority acknowledges), tourist use at the San Remo would qualify as a permitted conditional useand is grandfathered under the Planning Code"to the extent it lawfully existed when the current laws' restrictions came into effect, but does not qualify to the extent plaintiffs propose to significantly alter or expand it. " ( Id. at p. 283, 41 P.3d at p. 99 , italics added.)

The majority thus appears to recognize not only that the critical question here is the number of tourist rooms that should be grandfathered as a permitted conditional use, but also that the current record is inadequate to ascertain that number. For example, the zoning administrator described the annual unit usage reports from 1982 to 1992 as showing residents occupying between 25 and 57 of the designated residential units. "Even if only the 25-57 units actually shown to be occupied by residents were designated as residential units," the zoning administrator explained, "[plaintiffs] would still be required to procure a conditional use permit to convert them under the terms of Planning Code section 178(c)." (Italics added.) Yet, as the majority recognizes, the number of eligible rooms "during the relevant period" is "not clearly answered" by the zoning administrator's aggregation of usage reports over a 10 year period. (Maj. opn., ante, 117 Cal.Rptr.2d at p. 283 , 41 P.3d at p. 101 .) Plaintiffs, for their part, offered evidence that no more than 10 to 20 percent of the hotel's room had been rented to long-term residents, which would fix the number of rooms to be converted well below the number suggested by the zoning administrator. In their complaint, they allege that no more than nine rooms should have been deemed residential.

As interesting as this dispute may be, it is beyond cavil that the current record is insufficient to resolve how many tourist rooms might be grandfathered as a permitted conditional use. Under the circumstances, a remand is appropriate to enable factual findings to be made regarding the actual use of the hotel during the relevant period. [5] On remand, the City would also be free to argue that plaintiffs had discontinued or abandoned any permitted conditional tourist use they are able to establish. "`[Abandonment of a nonconforming use ordinarily depends upon a concurrence of two factors: (1) An intention to abandon; and (2) an overt act, or failure to act, which carries the implication the owner does not claim or retain any interest in the right to the nonconforming use (8A McQuillin, [Municipal Corporations (3d ed.1994)], § 25.192; 1 Anderson, American Law of Zoning, § 6.58). Mere cessation of use does not of itself amount to abandonment although the duration of nonuse may be a factor in determining whether the nonconforming use has been abandoned (101 C.J.S. Zoning § 199).'" ( Hansen Brothers Enterprises, Inc. v. Board of Supervisors (1996) 12 Cal.4th 533, 569 , 48 Cal.Rptr.2d 778 , 907 P.2d 1324 ; see S.F. Planning Code, § 178, subd. (d) [permitted conditional uses deemed abandoned if discontinued for three years]; S.F. Admin. Code, § 41.19(a)(1) [a tourist unit may be rented to a permanent resident without changing the legal status of the unit]; cf. Tenderloin Housing Clinic, Inc. v. Astoria Hotel (2000) 83 Cal.App.4th 139, 144-145 , 98 Cal.Rptr.2d 924 .) The current record, however, renders it impossible to determine which rooms, if any, could be rented to tourists as a permitted conditional use and which rooms, if any, have abandoned or discontinued that permitted conditional use. How the majority, which is plainly aware of the significance of the hotel's historical use and the inadequacy of the record on that point, can do anything other than order a remand is a mystery.

In sum, the answer to the question framed by the majority"Did San Francisco Properly Require Plaintiffs to Obtain a Conditional Use Permit for Full Tourist Use of the Hotel?"is a conditional yes. The need to obtain a permit exists only for those rooms for which plaintiffs cannot establish tourist rental as a permitted conditional use, which is precisely the issue left unresolved by the majority. In my view, the writ petition should be granted in part and the matter remanded to permit the superior court or the appropriate administrative agency to take evidence concerning actual tourist use and to resolve any claims that tourist use was abandoned or discontinued. The majority's abrupt termination of the litigation grants the City a windfall housing replacement fee for each room that further proceedings would have revealed to be grandfathered as a permitted conditional tourist use.

I agree with the majority that in-lieu fees assessed under the HCO are not subject to the "rough proportionality" test articulated in Dolan v. City of Tigard (1994) 512 U.S. 374 , 114 S.Ct. 2309 , 129 L.Ed.2d 304 ( Dolan ). Dolan envisioned that some "land use regulations" would not be subject to the "rough proportionality" test. ( Id. at p. 385, 114 S.Ct. 2309 .) To identify which land use regulations would be subject to the more stringent test, the court relied on two "relevant" distinctions: between "an adjudicative decision to condition the [owner's] application for a building permit on an individual parcel" and "essentially legislative determinations classifying entire areas of the city," and between "a requirement that [the owner] deed portions of the property to the city" and "a limitation on the use [the owner] might make of her own parcel." ( Ibid., italics added.)

In Ehrlich v. City of Culver City (1996) 12 Cal.4th 854 , 50 Cal.Rptr.2d 242 , 911 P.2d 429 ( Ehrlich ), this court unanimously concluded that the "rough proportionality" test applied when "special, discretionary permit conditions on development by individual property owners" ( id. at p. 881, 50 Cal.Rptr.2d 242 , 911 P.2d 429 (plur. opn. of Arabian, J.); id. at p. 912, 50 Cal.Rptr.2d 242 , 911 P.2d 429 (cone. & dis. opn. of Werdegar, J.)) were " adjudicatively imposed" ( id. at p. 906, 50 Cal.Rptr.2d 242 , 911 P.2d 429 (cone. & dis. opn. of Kennard, J.); id. at p. 891, 50 Cal.Rptr.2d 242 , 911 P.2d 429 (cone. opn. of Mosk, J.)). Here, as the majority notes, the in-lieu fee is a product of "generally applicable legislation" (maj. opn., ante, 117 Cal.Rptr.2d at p. 289 , 41 P.3d at p. 104 ) and its calculation is subject to "no meaningful government discretion." ( Id. at p. 289, 41 P.3d at p. 104 .) Under those circumstances, the in-lieu fee here must be viewed as one of those land use regulations that is not subject to the "rough proportionality" test.

This much is sufficient to answer "no" to the main question the City presented in its review petition: "Where a legislatively adopted impact fee applies equally to 500 residential hotels and 18,000 residential hotel units, is it nevertheless a `particularized' exaction subject to `heightened scrutiny' because it does not apply to every property in the city?" Although the majority might have stopped the takings analysis at that point, it seems prudent to me to discuss, as the majority does, the legal standard that does apply to the in-lieu fee here, namely that "[a] land use regulation does not effect a taking if it `substantially advances legitimate state interests'...." ( Dolan, supra, 512 U.S. at p. 385 , 114 S.Ct. 2309 .)

Admittedly, this test is not easy of application. The Supreme Court acknowledges that it has not provided "a thorough explanation of the nature or applicability of the requirement that a regulation substantially advance legitimate public interests" ( Monterey v. Del Monte Dunes at Monterey, Ltd. (1999) 526 U.S. 687, 704 , 119 S.Ct. 1624 , 143 L.Ed.2d 882 ( Del Monte Dunes )) other than to distinguish this requirement from that used by the high court in due process and equal protection claims. (See ibid., citing Nollan v. California Coastal Comm'n (1987) 483 U.S. 825, 834-835 , 107 S.Ct. 3141 , 97 L.Ed.2d 677, fn. 3 ( Nollan ).) As Nollan stated in that footnote, "our opinions do not establish that these standards are the same as those applied to due process or equal protection claims." ( Nollan, supra, 483 U. S. at pp. 834-835, fn. 3, 107 S.Ct. 3141 .) Rather, takings jurisprudence has diverged from due process and equal protection both in "verbal formulations" and in application. ( Ibid. ) "[T]here is no reason to believe (and the language of our cases gives some reason to disbelieve) that so long as the regulation of property is at issue the standards for takings challenges, due process challenges, and equal protection challenges are identical." ( Ibid.; see generally Santa Monica Beach, Ltd. v. Superior Court, (1999) 19 Cal.4th 952, 1018-1021 , 81 Cal.Rptr.2d 93 , 968 P.2d 993 (dis. opn. of Chin, J.) ( Santa Monica Beach ).)

Thus, to the extent that the majority suggests that the property owner must prove the land use regulation is arbitrary to prevail on a takings claim as opposed to a due process claim (see maj. opn., ante, 117 Cal.Rptr.2d at p. 294, fn. 16 , 41 P.3d at p. 108, fn. 16 , quoting Dolan, supra, 512 U.S. at p. 391, fn. 8 , 114 S.Ct. 2309 ), it is not faithful to the distinction the high court has drawn between those two legal doctrines. [6] For the most part, however, the majority takes pains to distinguish the requirement under the takings clause that a land use regulation "substantially advance[] legitimate state interests" from the requirement of due process and equal protection that a regulation be "not arbitrary." To pass scrutiny under the takings clause, the majority says that in-lieu fees "must bear a reasonable relationship, in both intended use and amount, to the deleterious public impact of the development," although "the relationship between means and ends need not be so close or so thoroughly established for legislatively imposed fees as for ad hoc fees," which are subject to the "rough proportionality" test. (Maj. opn., ante, 117 Cal.Rptr.2d at p. 291 , 41 P.3d at p. 106 .) Although this formulation makes plain that something more is required than mere rational-basis review, its meaning is still opaque. The defect, I submit, is that the majority's test is too much defined by what it is not, rather than by what it is. For the courts who will be called upon to apply this standard, we must be more illuminating.

The majority's formulation correctly describes the subjects of the inquiryi.e., the governmental regulation and the public impact of the developmentas well as the intensity of the relationship between themi.e., a reasonable relationship. What is missing is a description of the nature of the relationship between the public impact of the development and the governmental regulation. On this point, I find helpful the concurring and dissenting opinion in Pennell v. San Jose (1988) 485 U.S. 1, 15-24 , 108 S.Ct. 849 , 99 L.Ed.2d 1 (cone. & dis. opn. of Scalia, J.). Pennell involved the kind of land use regulation that, like the in-lieu fees here, is not subject to the "rough proportionality" test. (See Santa Monica Beach, supra, 19 Cal.4th at p. 968 , 81 Cal.Rptr.2d 93 , 968 P.2d 993 .) Justice Scalia, joined by Justice O'Connor, observed that "[t]raditional land use regulation (short of that which totally destroys the economic value of property) does not violate this [takings clause] principle because there is a cause-and-effect relationship between the property use restricted by the regulation and the social evil that the regulation seeks to remedy. Since the owner's use of the property is (or, but for the regulation, would be) the source of the social problem, it cannot be said that he has been singled out unfairly." ( Pennell v. San Jose, supra, 485 U.S. at p. 20 , 108 S.Ct. 849 , italics added.)

From Justice Scalia's separate opinion, it is apparent that the missing element in the majority's formulation is the causal connection between the property use restricted by the regulation and the social evil that the regulation seeks to remedy. Thus, I would reformulate the standard as follows: The in-lieu fee does not violate the takings clause so long as (1) there is a cause-and-effect relationship between the owner's desired use of the property and the social evil that the fee seeks to remedy, and (2) the fee is reasonably related in both intended use and amount to that social evil. This two-part standard best implements the high court's broadly stated requirement that the fee substantially advance legitimate state interests.

The majority, after announcing the correct standard, strays beyond the questions presented and applies the standard to some of the claims in plaintiffs' complaint and then, without any additional discussion, dismisses the remaining claims by affirming the trial court's judgment sustaining the demurrer. I would refrain from reaching the merits of these selected takings claims and, without analysis or even an acknowledgement of doing so, from summarily denying the others. Well-settled principles of appellate review counsel us to remand the matter to the lower courts to apply the correct legal standard in the first instance.

The most immediate reason for remanding to the lower courts, of course, is that we already need to remand for further factual findings to determine the number of rooms that require further permitting before they may be offered to tourists. When that is completed, plaintiffs will need to choose anew whether and, if so, how to satisfy the housing replacement requirement, "whether by constructing or bringing onto the market new units; by sponsoring such construction by a public or nonprofit private housing developer; or by paying, in-lieu of such construction, a fee to a designated City housing fund." (Maj. opn., ante, 117 Cal.Rptr.2d at p. 289 , 41 P.3d at p. 104 .) Should plaintiffs choose again to pay the in-lieu fee, the appropriate City agency will need to recalculate it, taking into account the correct number of units and their replacement cost. (See ibid. ) Only if plaintiffs then challenge the fee imposed will we be presented with an actual case or controversy that resembles the one the majority addresses here.

Even if we were not already remanding for further factual development involving the petition for writ of administrative mandate, a remand to permit the Court of Appeal to apply the correct legal standard to plaintiffs' claims would be the prudent course. The majority finds, and I agree, that the Court of Appeal erred in analyzing plaintiffs' as-applied takings challenge under a heightened scrutiny standard. (Maj. opn., ante, 117 Cal.Rptr.2d at pp. 289, 290, 41 P.3d at pp. 103-104, 105.) It is our practice, where a lower court has applied an incorrect legal standard, to remand for application of the correct standard, even when a remand is not required for other reasons. (E.g., Yamaha Corp. of America v. State Bd. of Equalization (1998) 19 Cal.4th 1, 15 , 78 Cal.Rptr.2d 1 , 960 P.2d 1031 [where the Court of Appeal applied an erroneous standard, "regard for the structure of appellate decisionmaking suggests that the case should be returned to the Court of Appeal"]; id. at p. 25, 78 Cal.Rptr.2d 1 , 960 P.2d 1031 (cone. opn. of Mosk, J, joined by George, C.J. and Werdegar, J.) ["It is therefore appropriate to remand to the Court of Appeal for reconsideration in light of the proper standard of review"]; see also People v. Cox (2000) 23 Cal.4th 665 , 677-678 & fn. 7, 97 Cal.Rptr.2d 647 , 2 P.3d 1189 ; Ramirez v. Yosemite Water Co. (1999) 20 Cal.4th 785, 803 , 85 Cal.Rptr.2d 844 , 978 P.2d 2 ; People v. Breverman (1998) 19 Cal.4th 142, 164, 178-179 , 77 Cal.Rptr.2d 870 , 960 P.2d 1094 ; People v. Cahill (1993) 5 Cal.4th 478, 510 , 20 Cal.Rptr.2d 582 , 853 P.2d 1037 .) There is no reason to deviate from our practice in this case. Indeed, the justifications for adhering to our traditional practice are numerous and compelling.

First, no hardship exists to warrant our extraordinary intervention to bring an abrupt close to the litigation. Plaintiffs have done everything the City has ordered them to do: they have secured the permits to rent to tourists and have paid the $567,000 housing replacement fee. The only issue in the litigation, as conceded by the City at oral argument, is whether plaintiffs will get some or all of their money back-and the majority's approach plainly does not aid plaintiffs. Moreover, the majority should not be under any illusion that sustaining the demurrer will bring an end to a decade of litigation over that issue. Plaintiffs have reserved their federal claims and, if rebuffed here, will resume their federal litigation, which is now subject to Pullman abstention. (See The San Remo Hotel v. City and County of San Francisco (9th Cir.1998) 145 F.3d 1095 , 1106 & fn. 7.)

Second, remand would give us, a reviewing court, the benefit of a reasoned decision applying the correct standard to the merits of plaintiffs' claims. The Court of Appeal declined to address plaintiffs' facial challenge on the grounds that plaintiffs "failed to seek leave of court to replead such causes of action" and that "the present state of the pleadings is insufficient to allow us to fully assess the ultimate legal validity of the facial constitutionality of the HCO." The trial court did purport to reach the merits, but relied solely on Terminal Plaza Corp. v. City and County of San Francisco, supra, 177 Cal.App.3d 892 , 223 Cal.Rptr. 379 and Bullock v. City and County of San Francisco (1990) 221 Cal. App.3d 1072 , 271 Cal.Rptr. 44 . Yet, neither Terminal Plaza nor Bullock addresses the "substantially advance" branch of takings analysis. As for the as-applied challenges, the Court of Appeal applied the wrong standard, and the trial court failed to reach the merits of these claims at all, choosing instead to reject the claims categorically on the authority of Pfeiffer v. City of La Mesa (1977) 69 Cal.App.3d 74 , 137 Cal.Rptr. 804 . The Court of Appeal held Pfeiffer inapplicable for several reasons, none of which the City challenges here. (See maj. opn., ante, 117 Cal. Rptr.2d at p. 280, fn. 9 , 41 P.3d at p. 96, fn. 9 .) In sum, only one category of claims has even been addressed on the merits by a lower courtthe as-applied claims by the Court of Appealand that court applied what we have now determined to be the wrong standard. I do not view these circumstances as crying out for us to deviate from our practice of remanding to permit the lower courts to apply the correct standard in the first instance.

Third, remand for application of the correct standard would be consistent with the way the parties have framed the issues and the relief the City has sought here. The questions presented in the City's petition for review do not invite us to resolve the merits of plaintiffs' claims. Indeed, when asked at oral argument what this court should do if it adopts (as it has) a standard not urged by either party, [7] the City's counsel replied, "I assume the court would remand for a consideration by the trial court of what standard of review the court would order to be applied."

Fourth, remand is appropriate to permit the lower courts to address the claims articulated in the complaint that are not discussed by the majority. The majority purports to affirm the trial court's judgment sustaining the demurrer to the entire complaint, yet limits its discussion only to a few facial and as-applied challenges to the HCO. Our task on demurrer is to determine whether "the complaint states a cause of action under any theory, regardless of the title under which the factual basis for relief is stated." ( Quelimane Co. v. Stewart Title Guaranty Co. (1998) 19 Cal.4th 26, 38 , 77 Cal.Rptr.2d 709 , 960 P.2d 513 .) Viewed in this light, it is apparent that this complaint articulates a number of potential takings and other claims, none of which has yet been addressed by the majority or rejected by the Court of Appeal, including (1) that the lifetime leases effect a physical taking (see Yee v. Escondido (1992) 503 U.S. 519, 528 , 112 S.Ct. 1522 , 118 L.Ed.2d 153 ["A different case would be presented were the statute, on its face or as applied, to compel a landowner over objection to rent his property or to refrain in perpetuity from terminating a tenancy"]) [8] ; (2) that the HCO and fee deprived plaintiffs of their reasonable investment-backed expectations; (3) that the HCO denies plaintiffs all economically viable use of the property, inasmuch as a surplus of vacant residential hotel rooms already exists in San Francisco [9] ; and (4) that state law preempted the HCO. [10]

I can think of no reason for the majority's failure to address these claims, other than the fact that none of them can even remotely be shoehorned into the issues presented by the City's petition. But the City's decision to limit issues for review can hardly be deemed a license for us to dismiss plaintiffs' entire complaint without comment, especially where viable issues remain. To affirm the demurrer here would punish plaintiffs for complying with rule 29.3(c) of the California Rules of Court, which tells the parties that "[u]nless otherwise ordered, briefs on the merits shall be confined to those issues, and issues fairly included in them." The lesson here, if there is one, is that litigants in this court should be careful to brief against any conceivable contingency that could jeopardize any favorable ruling below, even if the arguments fall well outside the questions presented. I confess I do not think this is a good idea.

I would therefore affirm the judgment of the Court of Appeal to the extent that it ordered a remand for further proceedings relating to the petition for writ of administrative mandate and affirmed the imposition of nominal penalties under the City's cross-complaint. I would reverse the Court of Appeal to the extent that it applied a heightened-scrutiny standard to plaintiffs' taking claims and would then remand the cause to enable the Court of Appeal to apply the correct legal standard. To the extent the majority prevents plaintiffs from demonstrating their entitlement to writ relief, prematurely analyzes plaintiffs' as-applied takings claims, and summarily disposes of plaintiffs' other claims without any analysis whatsoever, I respectfully dissent.

I CONCUR: CHIN, J.

DISSENTING OPINION BY BROWN, J.

Americans are a diverse group of hard-working, confident, and creative people molded into a nation not by common ethnic identity, cultural legacy, or history; rather, Americans have been united by a dreama dream of freedom, a vision of how free people might live. The dream has a history. The idea that property ownership is the essential prerequisite of liberty has long been "a fundamental tenet of Anglo American constitutional thought." (Ely, The Guardian of Every Other Right (1998) p. 43.) "Indeed, the framers saw property ownership as a buffer protecting individuals from government coercion. Arbitrary redistribution of property destroyed liberty, and thus the framers hoped to restrain attacks on property rights." ( Ibid. ) "Property must be secured, or liberty cannot exist" (Adams, A Balanced Government (1790) in Discourses on Davila (1805), reprinted in 6 Works of John Adams (1851 ed.) p. 280), because property and liberty are, upon examination, one and the same thing.

Private property is in essence a cluster of rights inuring to the benefit of the owner, freely exchangeable in accordance with the terms of private agreements, and recognized and protected by common consent. In the case of real property, this cluster of rights includes the right to exclude persons from certain physical space. In the case of intellectual property, it may include the right to employ a valuable method or process to the exclusion of others. In other words, private property represents zones of individual sovereignty regions of autonomy within which we make our own choices.

But private property, already an endangered species in California, is now entirely extinct in San Francisco. The City and County of San Francisco has implemented a neo-feudal regime where the nominal owner of property must use that property according to the preferences of the majorities that prevail in the political process or, worse, the political powerbrokers who often control the government independently of majoritarian preferences. Thus, "the lamb [has been] committed to the custody of the wolf." (6 The Works of John Adams, supra, at p. 280.) San Francisco has redefined the American dream. Where once government was closely constrained to increase the freedom of individuals, now property ownership is closely constrained to increase the power of government. Where once government was a necessary evil because it protected private property, now private property is a necessary evil because it funds government programs.

The City and County of San Francisco (the City), like other cities, seeks to provide affordable housing to its low-income residents. The most egalitarian way to achieve this goal would be to distribute the cost of subsidies as broadly as possible, but the forces attacking private property in Californiathough claiming the moral high groundhave proved themselves anything but egalitarian in their approach. In 1981, the City enacted the Residential Hotel Unit Conversion and Demolition Ordinance (S.F.Admin.Code, ch. 41) (the HCO; all citations to HCO are to chapter 41 of the San. Francisco Administrative Code), the details of which are summarized in the majority opinion, ante, 117 Cal. Rptr.2d at pp. 274 through 275, 41 P.3d at pp. 91 through 92. The HCO places the burden of providing low-income housing disproportionately on a relatively small group of hotel owners. These hotel owners certainly did not cause poverty in San Francisco; indeed, for a long time they voluntarily helped relieve the problem by leasing some or all of their rooms on a long-term basis to low-income residents. But as the economy of the City shifted, this residential use of their hotel rooms became increasingly unprofitable, and hotel owners began to abandon the residential rental business. It was then that the City, facing constitutional constraints on taxation and other sources of revenue, began to see the hotel owners as the most convenientif not the most equitable-off-budget solution to its housing problems. If the City were devising a tax that would subsidize low-cost housing, I strongly doubt it would limit its tax to the owners of a few hundred residence hotels, but in the often surreal world of political expedience, these ill-fated business people were ordered to use their property for the benefit of the poor, thereby greatly depressing the market value of that property.

The express purpose of the HCO was to preserve the City's stock of low-income residential housing by requiring hotel owners to continue leasing their rooms as residences, or to replace those residential units if they chose to convert the rooms to tourist use. (HCO, §§ 41.2, 41.3.) Obviously, the HCO is facially unconstitutional. If a person took my car and asked a ransom for its return, he or she would be guilty of theft. But what if the City, seeking to provide transportation to the poor, orders me to operate an informal carpool, or if I prefer, to buy the City a replacement car? When presented with a similar hypothetical at oral argument, the San Francisco City Attorney declared such a rule a mere "regulation of use." I disagree. The essence of private property is the right to use that property as one sees fit and for one's own advantage. The police power permits the government to regulate that use so as to promote health, safety, and the general welfare, but it does not permit the government to achieve its social agenda by ordering a political minority to dedicate its property to the benefit of a group the government wishes to favor. As I explain in more detail in part II. below, such a regulation amounts, in practical effect, to a transfer of title and requires the government to pay its way.

But constitutional issues aside, the City had another problem with its HCO. In 1985, the state Legislature enacted the Ellis Act, which unequivocally guarantees the right of property owners to abandon the residential rental business. Government Code section 7060, subdivision (a), provides: "No public entity ... shall, by statute, ordinance, or regulation, or by administrative action implementing any statute, ordinance or regulation, compel the owner of any residential real property to offer, or to continue to offer, accommodations in the property for rent or lease." If this clear language left any doubt about the continuing viability of the HCO, that doubt was finally resolved against the City in Bullock v. City and County of San Francisco (1990) 221 Cal.App.3d 1072, 1102 , 271 Cal.Rptr. 44 ( Bullock ), which held that the Ellis Act preempts the HCO. "We conclude," said the court in Bullock , "that [the HCO] is preempted by the Ellis Act and is therefore invalid to the extent it is applied to prevent plaintiff from going out of the residential hotel business. " ( Bullock, at p. 1102, 271 Cal.Rptr. 44 , italics added.) In other words, state law expressly permits property owners to do what the HCO bars them from doing: it permits them to stop leasing to residents.

The City, however, was unwilling so easily to concede defeat. The Ellis Act affirms the continuing power of public entities "to grant or deny ... zoning ... approvals." (Gov.Code, § 7060.1, subd. (b).) This provision seems reasonable on its face: property owners are free to stop leasing to residents, but they still must obey zoning laws regulating the new use to which they intend to put their property. (See also Gov.Code, § 7060.7.) Conveniently, the City amended its zoning laws shortly after the Ellis Act became law. As related in the majority opinion (maj. opn., ante, 117 Cal.Rptr.2d at p. 276 , 41 P.3d at p. 93 ), one effect of this change was to require the owners of the San Remo Hotel to obtain a conditional use permit before "intensif[ying]" any use of their property as a tourist hotel. (S.F. Planning Code, § 178, subd. (c).) The parties dispute the extent to which the rooms in the San Remo Hotel were ever, in fact, being used as residences, but as the majority points out (maj. opn., ante, 117 Cal.Rptr.2d at pp. 281-285, 41 P.3d at pp. 97-100), at least some of those rooms were historically used as residences, and to the extent the owners sought to convert those rooms to tourist use, the change required a permit under the new zoning ordinance. In short, the City put in place a zoning mechanism by which it could try to achieve the goals of its preempted HCO, if it chose to do so.

But the City had to proceed cautiously. Bullock made clear that the City could not use zoning laws pretextually to bar landlords from exercising their rights under the Ellis Act. As the court stated, "[nothing in the Ellis Act gives any landlord invoking its protection the unilateral power to effect what amounts to a rezoning of his property.... [¶] The City is ... not precluded from seeking to have enjoined a violation of [zoning] ordinances ..., so long as this claim is not used as a pretext for halting [ the property owner's ] departure from the residential hotel business. " ( Bullock, supra, 221 Cal.App.3d at p. 1104 , 271 Cal.Rptr. 44 , italics added.) Implicitly conceding that the Ellis Act is in direct conflict with the HCO, the majority argues (maj. opn., ante, 117 Cal.Rptr.2d at pp. 284-285, 41 P.3d at p. 100 ) the Ellis Act has no effect here because the record does not establish that plaintiffs complied with its notice requirements. (See Gov.Code, § 7060.4.) But the power of local governments to implement the Ellis Act by requiring notice in no way negates the fact that state law has preempted the City from using zoning as a way to discourage property owners from exiting the residential rental business. Government Code section 7060.4 permits local governments to enact notice requirements, but it does not permit those governments to enact zoning restrictions that abrogate the very protections the Ellis Act affords. Therefore, whether plaintiffs followed the notice requirements of the Ellis Act is beside the point; the state has occupied this area of law, and the City may not enact and enforce contrary laws. Nevertheless, after the City amended its zoning laws, the preempted HCO was poised to become the ghost that drove zoning decisions in San Francisco.

Here, for example, the owners of the San Remo Hotel sought to convert their hotel to full tourist use. The City's Planning Department told them thatconsistent with the recent amendments to the zoning lawsthey would need a conditional use permit. Whether to issue a conditional use permit is an adjudicative decision that is exercised at the discretion of the planning commission (S.F. Planning Code, §§ 303, 316, 316.8), and the planning commission exercised its discretion here by conditioning the San Remo Hotel permit on, among other things, compliance with the preempted HCO. The owners of the San Remo Hotel met this condition by paying, in protest, a $567,000 "in lieu fee" (representing a portion of the appraised cost of building replacement housing) and brought this action alleging, among other things, a taking without just compensation in violation of the state Constitution.

This constellation of facts makes this case indistinguishable from Ehrlich v. City of Culver City (1996) 12 Cal.4th 854 , 50 Cal.Rptr.2d 242 , 911 P.2d 429 ( Ehrlich ). Ehrlich addressed the same problem that the United States Supreme Court recognized in Nollan v. California Coastal Comm'n (1987) 483 U.S. 825 , 107 S.Ct. 3141 , 97 L.Ed.2d 677 ( Nollan ) and Dolan v. City of Tigard (1994) 512 U.S. 374 , 114 S.Ct. 2309 , 129 L.Ed.2d 304 ( Dolan ). When a government agency has discretionary authority to permit or prohibit a new use of property, the risk arises that governmental greed will consciously or unconsciously distort the decisionmaking process, causing the agency to exact a condition from the property owner that has nothing to do with mitigating the effects of the proposed new use. This sort of regulatory leveraging, taken to its extreme, might cause a government to impose "stringent land-use regulation which [it] then waives to accomplish other purposes" ( Nollan, at p. 837, fn. 5, 107 S.Ct. 3141 ), and this concern warrants a more searching inquiry by a reviewing court than might otherwise be appropriate.

In Nollan, for example, a California agency conditioned a permit to develop beachfront property on dedication of a public easement. The easement, which permitted the public to cross the property to gain access to the ocean, "utterly fail[ed] to further the end" of mitigating the impact of the proposed development on ocean views, and therefore lacked the "essential nexus" that the federal Constitution required and amounted to "`extortion.'" ( Nollan, supra, 483 U.S. at p. 837 , 107 S.Ct. 3141 .) Similarly, in Dolan the City of Tigard conditioned a permit to greatly expand a retail sales complex on dedication of a strip of property as a pathway for pedestrians and bicycles. The high court held that a permit condition must be "`rough[ly] proportional[]'" "in nature and extent" to mitigating "the impact of the proposed development" ( Dolan, supra, 512 U.S. at p. 391 , 114 S.Ct. 2309 , fn. omitted), and the City of Tigard had failed "to quantify its findings" that the pathway would mitigate increases in traffic. ( Id. at p. 395, 114 S.Ct. 2309 .) In Ehrlich, we unanimously extended the principle of Nollan and Dolan to a case in which a city demanded a monetary fee from a developer rather than a dedication of an interest in real property, noting that the same risk of governmental abuse was present. ( Ehrlich, supra, 12 Cal.4th at p. 876, 50 Cal. Rptr.2d 242 , 911 P.2d 429 (plur. opn. of Arabian, J.); id. at pp. 899-901, 50 Cal. Rptr.2d 242 , 911 P.2d 429 (cone. opn. of Mosk, J.); id. at p. 907, 50 Cal.Rptr.2d 242 , 911 P.2d 429 (conc. & dis. opn. of Kennard, J.); id. at p. 912, 50 Cal.Rptr.2d 242 , 911 P.2d 429 (conc. & dis. opn. of Werdegar, J.).) The majority reaffirms the holding of Ehrlich today. (Maj. opn., ante, 117 Cal.Rptr.2d at p. 287 , 41 P.3d at pp. 102-103.)

Here, as in Ehrlich, a public agency (the City's Planning Commission) has made a discretionary adjudicative decision with respect to a specific permit application. Contrary to the argument of the majority (maj. opn., ante, 117 Cal.Rptr.2d at pp. 289-290, 41 P.3d at pp. 103-105), this case does not involve a legislatively imposed fee for which the risks of abuse are arguably held in check by the political process. (Maj. opn., ante, 117 Cal.Rptr.2d at p. 291 , 41 P.3d at p. 105 .) Rather, in an adjudicative proceeding concerning a single property owner's permit request, the planning commission chose to require HCO compliance and thereby used the leverage it gained by regulating commercial uses of property to exact a $567,000 fee. On these facts, the fee must be roughly proportional in nature and extent to mitigating the impact of the proposed new use of the property. ( Ehrlich, supra, 12 Cal.4th at p. 876, 50 Cal.Rptr.2d 242 , 911 P.2d 429 (plur. opn. of Arabian, J.); id. at pp. 899-901, 50 Cal.Rptr.2d 242 , 911 P.2d 429 (conc. opn. of Mosk, J.); id. at p. 907, 50 Cal.Rptr.2d 242 , 911 P.2d 429 (conc. & dis. opn. of Kennard, J.); id. at p. 912, 50 Cal.Rptr.2d 242 , 911 P.2d 429 (conc. & dis. opn. of Werdegar, J.).) In short, because we are dealing with a discretionary decision of the planning commission rather than direct enforcement of the HCO, the government's decision here, like the decision at issue in Ehrlich, was adjudicative.

The majority tries to sidestep this gaping hole in its argument with a footnote asserting that under applicable zoning laws "conversion of residential rooms to commercial use is unconditionally prohibited above the first floor in the North Beach district except insofar as permitted by the HCO (S.F. Planning Code, §§ 722.38, 790.84)." (Maj. opn., ante, 117 Cal.Rptr.2d at p. 289, fn. 11 , 41 P.3d at p. 104, fn. 11 .) The majority concludes: "The planning commission, therefore, had no discretion to permit such change in use absent HCO compliance." ( Ibid. ) In other words, the majority argues the planning commission's decision here was legislative, not adjudicative, thereby permitting the majority to exploit the exception it reads into Ehrlich for legislatively created permit fees.

First, the majority simply misreads the City's Planning Code. San Francisco Planning Code section 722.38 prohibits conversion from residential to nonresidential use above the first floor, but section 790.84 of that code does not limit the exception from this prohibition to cases where the property owner complies with the HCO. Rather, section 790.84 creates an exception for conversions that are "defined and regulated in [the HCO]." The San Francisco Planning Code therefore remains neutral with respect to HCO compliance. It simply provides that its flat prohibition on conversions does not apply when the HCO applies, thereby giving space within which the HCO can operate. Nothing in the Planning Code requires the planning commission to enforce the HCO, nor does the Planning Code state that conversions are prohibited except "as permitted by the HCO." (Maj. opn., ante, 117 Cal.Rptr.2d at p. 289, fn. 11 , 41 P.3d at p. 104, fn. 11 .) In other words, the Planning Code leaves a regulatory gap with respect to conversions regulated in the HCO, and the commission has, in its discretion, chosen to fill that gap by requiring HCO compliance. Its decision was not compelled by any legislative rule, and therefore it is no different than the decision at issue in Ehrlich. At the very least, the legislative rule was ambiguous, and the planning commission's interpretation of the rule was in that sense adjudicative.

Second, the majority's exception for legislatively created permit fees is mere sophism, particularly where the legislation affects a relatively powerless group and therefore the restraints inherent in the political process can hardly be said to have worked. (Cf. United States v. Carolene Products Co. (1938) 304 U.S. 144, 152, fn. 4 , 58 S.Ct. 778 , 82 L.Ed. 1234 .) If the agency in Nollan had passed a rule requiring all beachfront property owners to dedicate an easement as a condition of developing their properties, those easements would have no better mitigated the effects of development (and they would have been no less objectionable) than the easement that the agency exacted adjudicatively. Of course, when the government may prohibit a certain use of property entirely, it may opt instead to place conditions on that use ( Nollan, supra, 483 U.S. at p. 836 , 107 S.Ct. 3141 ), but the conditions must be related in some real way to the justification advanced for a complete prohibition. ( Id. at p. 837, 107 S.Ct. 3141 .) Otherwise, the government has transformed the police power into an efficient way to raise money by regulating political minorities and then selling exemptions from the regulatory scheme, without any real intent to advance the scheme's purported purpose. It becomes "as if California law forbade shouting fire in a crowded theater, but granted dispensations to those willing to contribute $100 to the state treasury." ( Ibid. ) The government, in effect, says: We have the power; therefore, pay us to leave you alone. By any measure, that is extortion. Moreover, it turns the takings clause on its head. Instead of the government having to pay compensation to property owners, the government now wants property owners to compensate it to get back the fair value of property the government took away through regulation.

A public agency can just as easily extort unfair fees legislatively from a class of property owners as it can adjudicatively from a single property owner. The nature of the wrong is not different or less abusive to its victims, but the scope of the wrong is multiplied many times over. Therefore, I believe Ehrlich should apply whenever the risk is great that greed for public revenues has driven public regulatory policy. In other words, where a legislative scheme imposes a burdensome fee on a small class of property owners as a condition to buying relief from a regulation, I believe careful judicial scrutiny is appropriate, including finding a close link between the fee and the purpose of the regulation. In light of the majority's decision, however, we can be sure that agencies will now act legislatively, rather than adjudicatively, and thereby insulate their actions from close judicial scrutiny.

In addition, the HCO is structured in such a way that the same sort of discretionary, case-specific decisionmaking that triggered our holding in Ehrlich also takes place under the HCO. The fee that a hotel owner may pay under the HCO so as to buy the right to lease residential rooms to tourists is a fixed portion of the estimated cost of replacing the lost residential housing. (HCO, § 41.13, subd. (a)(4), (5).) Though this cost estimate must be based on two independent appraisals ( ibid. ), it nevertheless permits a discretionary element to enter into the process, with the result that the City and the property owner inevitably end up in a back-and-forth negotiation over the amount of the fee. Obviously, the City has the upper hand in this negotiation-because it is holding the conversion permit hostage-and therefore the City is free to squeeze as large a fee as possible from the property owner. That dynamic brings into play all the concerns that justified our holding in Ehrlich, and therefore the same careful scrutiny ought to apply.

Finally, and most importantly, the majority fails to appreciate that, if the San Francisco Planning Code somehow requires across-the-board compliance with the HCO, then the Ellis Act preempts that requirement, just as it preempts the HCO itself. The zoning exception to the Ellis Act permits the City to place certain limits on how a property owner may use property, but as Bullock held, the City may not use that zoning power to erect barriers to a property owner's decision to exit the residential rental business, even if the effect is to reduce the residential housing stock of the City. ( Bullock, supra, 221 Cal.App.3d at p. 1104 , 271 Cal.Rptr. 44 .) In other words, the planning commission, acting under the zoning exception to the Ellis Act, might be able to block the owners of the San Remo Hotel from renting more rooms to tourists, assuming its purpose is to limit the number of tourist hotel rooms in the area, but the commission may not do so to protect the City's stock of low-cost residential housing, because that would be inconsistent with the state law right of property owners to stop leasing to residents. Nevertheless, here the majority argues (and the City concedes) that the preservation of low-cost housing was the planning commission's purpose when it required HCO compliance. (See, e.g., maj. opn., ante, 117 Cal.Rptr.2d at pp. 292-298, 41 P.3d at pp. 106-111.) That purpose was simply impermissible under the Ellis Act, and even if it were somehow permissible, the majority acknowledges that at least some rooms at the San Remo Hotel were historically used as tourist rooms (maj. opn., ante, 117 Cal.Rptr.2d at pp. 281-285, 41 P.3d at pp. 97-100,) and therefore it cannot, in any case, justify the planning commission's decision to charge a conversion fee for all rooms. (Cf. id. at pp. 296-297, 41 P.3d at p. 110 .)

Of course, despite the Ellis Act problem, the majority and the City have no choice but to assert that the planning commission's purpose was the preservation of residential housing, because if the planning commission had some other purpose, it could notunder any standard of review (rational basis, reasonable relationship, or rough proportionality)have attached the condition that the owners of the San Remo Hotel take steps to preserve low-cost residential housing. Just as in Nollan, where the easement providing access to the beach was not related in any way to mitigating the obstruction of an ocean view, similarly here the preservation of low-cost housing is not related in any way to mitigating the impact of more tourist hotel rooms in the neighborhood of the San Remo Hotel. In other words, by requiring HCO compliance as a condition for renting more rooms to tourists, the planning commission has revealed its true colors: it does not care about the number of hotel rooms in the area; what it really cares about is preventing property owners from exiting the residential rental business. But this the City simply cannot do under the Ellis Actif not for the fact that the majority has chosen to turn a blind eye.

Our takings jurisprudenceboth state and federal-has become so labyrinthine and compartmentalized that attempts to find just the right standard for the case often entirely miss the underlying point of the exercise. We speak of ad hoc inquiries, relevant factors, per se takings, and means-end relationships. We chip away at the problem with separate lines of cases addressing distinct issues such as development permits and price controls. And all these efforts, valid as far as they go, leave us still groping for a basic conceptual approach that takes seriously the constitutional prohibition against uncompensated takings of private property. Thus, like the Wizard of Oz, we mystify our audience with the look and feel of great erudition, while concealing the humble reality that we have yet to solve the problem in a satisfactory way.

But, here, we need not consider all these arcane standards and fragmentary theories. These are analytical tools relevant to tough casescases in which it is unclear whether property has been taken. This is not a tough case. Here, property unquestionably has been taken. No matter the analysis, the facts of this case come down to one thingthe City and County of San Francisco has expropriated the property and resources of a few hundred hotel owners in order to ameliorate-off budget and out of sight of the taxpayerits housing shortage. In short, this ordinance is not a matter of efficiently organizing the uses of private property for the common advantage; instead, it is expressly designed to shift wealth from one group to another by the raw exercise of political power, and as such, it is a per se taking requiring compensation.

The majority rejects the legal theories on which the property owners have proceeded, but it fails to confront the more basic issue that prefigures all others: the City has replaced taxation and the provision of public services with a regulation that orders certain people to use their private property to do the government's work. If the relevant case law is sparse, it is only because no public agency has ever been so bold.

"Private property may be taken or damaged for public use only when just compensation ... has first been paid to ... the owner." (Cal. Const., art. I, § 19.) In a simple world where "property" is understood to refer to tangible property and "tak[ing]" is understood to be a formal transfer of title, this constitutional injunction is relatively easy to apply. But such a narrow application of the takings clause would trivialize the right. Restriction of any one of the several rights that constitute private property in effect takes that property. For example, a property owner cares little about the formalities of title possession when the property in question has been rendered nearly valueless by regulations prohibiting its most productive uses. In that case, the regulations have deprived the owner of so many of the rights that originally constituted the property that the property has, in effect, ceased to exist, or it has become a mere empty shell. Furthermore, in a complex and increasingly serviceand informationbased economy, the constitutional protection of intangible property is just as critical as the protection of physical property was to an agrarian and manufacturing economy.

Nevertheless, in countless ways, government takes propertyin the sense of regulating its usewithout always having to compensate the owner. In Penna. Coal Co. v. Mahon (1922) 260 U.S. 393, 413 , 43 S.Ct. 158 , 67 L.Ed. 322 ( Penna. Coal Co. ), Justice Holmes noted that "[government hardly could go on" if it had to pay its way every time a regulation restricted the use of property. The law has long recognized, for example, that government might, in the exercise of the police power, act to proscribe a nuisance, and in so doing it need not pay compensation. (See, e.g., Civ. Code, § 3479; Code Civ. Proc, § 731.) Holmes spoke of "an average reciprocity of advantage" whereby a property regulation ultimately works for the enrichment of all, though it imposes specific limitations on the use of certain property. ( Penna. Coal Co., at p. 415, 43 S.Ct. 158.)

For example, business owners on a popular shopping street might generally agree that their properties would be more attractive, and hence more valuable, if all the businesses used small, attractive signs rather than huge, garish billboards. Nevertheless, without regulation, competitive forces will inevitably cause business signs to become ever larger and more visually intrusive. No business owner wants to be the only one on a shopping street to have a small sign, and transaction costs often prevent owners from coming together to negotiate an agreement that would work to their common advantage. In that case, a regulation that has the immediate effect of reducing property value by restricting sign size, has the indirect effect of enhancing that value for all affected businesses. (See generally Epstein, Takings: Private Property and the Power of Eminent Domain (1985) pp. 195-215.)

A similar justification can be articulated in the case of the regulations proscribing nuisance. For example, if the law prohibits a property owner from operating a slaughterhouse in a residential neighborhood, the immediate effect might be to lower incrementally the value of the regulated property, but the indirect effect is to place that property in a neighborhood rendered more desirable for residential use, thereby enhancing its value. Again the same rationale can be extended to justify zoning and planning regulations that constrain the freedom of developers but enhance all property values over the long term by making cities more attractive and efficient. Finally, when a regulatory scheme creates new value for a property owner by establishing an artificial monopolyas, for example, might be true in the case of a utilitythe government can regulate the profit the property owner gains from its governmentcreated preferential status. All these examples point to the same fundamental principle: property owners, given a choice, will prefer to own property in a community having appropriate and mutually beneficial regulations, because such property has greater value by reason of the regulation. Accordingly, regulatory authority is not inherently confiscatory in all cases.

But the corollary of this ruleone I think is implicit in the takings clause of the state Constitutionis that a regulation is a taking if, rather than promoting "an average reciprocity of advantage" ( Penna. Coal Co., supra, 260 U.S. at p. 415 ), it is merely designed to benefit one class of citizens at the expense of another; that is, if it simply shifts wealth by a raw act of government power. The government, in that case, has deprived the property owner of a right associated with his property, shifting that right to another party, but it has in no sense compensated the owner by enhancing, in some real way, the value of the rights the owner has retained.

In short, it might be perfectly legitimate for the City to help the low-income residents of San Francisco, but it may not do so at the expense of some small class of persons simply by legislating a transfer of property rights. Of course, providing assistance to low-income residents of the community incrementally benefits all members of the community both by removing the blight of homelessness and by representing a general moral good, but here the burden of this common benefit falls so disproportionately on 500 business owners in a city of 776,700 residents that careful judicial scrutiny is warranted. Where the impact is so disproportionate, we cannot say that we have "an acceptable level of assurance that over time the burdens associated with collectively determined improvements will have been distributed `evenly' enough so that everyone will be a net gainer." (Michelman, Property, Utility, and Fairness: Comments on the Ethical Foundations of "Just Compensation" Law (1967) 80 Harv. L.Rev. 1165, 1225, original italics.) Moreover, it simply stretches the police power too far to suggest that the City is somehow regulating the use of property for the common advantage when it redistributes wealth by ordering a political minority to dedicate its property to the benefit of another group. The police power can no more be used in this way than it could be used to order a rich man to give a beggar a dime. Here, the primary beneficiary of the regulation is not the common advantage but the low-income individuals who obtain the inexpensive housing. Laudable as that goal might be, the takings clause precludes the government from achieving the goal by police power regulation.

Nor can the HCO be justified under the theory that the City is merely requiring property owners to continue the existing use of their property. (See, e.g., Penn Central Transp. Co. v. New York City (1978) 438 U.S. 104, 125 , 98 S.Ct. 2646 , 57 L.Ed.2d 631 .) Such a rule would punish a property holder for using property in a way that proved popular. Moreover, it fails to recognize the effect of shifting economic conditions and therefore locks property into unproductive uses. But most important, such a rule represents a dedication of property rights to the public, and in all fairness the public should have to pay for these rights, just as a private party would. For these reasons, I would not extend the holding of Penn Central beyond its unique factual context, including the fact that the regulatory authority in that case might have permitted a more moderate development of the property than the one the property owner sought and it gave the property owner transferable development rights as compensation. ( Id. at pp. 136-137, 98 S.Ct. 2646 .)

Here, the City has essentially said to 500 unlucky hotel owners: We lack the public funds to fill the need for affordable housing in San Francisco, so you should solve the problem for us by using your hotels to house poor people. The City might as well have ordered the owners of small grocery stores to give away food at cost. The federal takings clause "bar[s] Government from forcing some people alone to bear public burdens which, in all fairness and justice, should be borne by the public as a whole." ( Armstrong v. United States (1960) 364 U.S. 40, 49 , 80 S.Ct. 1563 , 4 L.Ed.2d 1554 .) I believe the same principle underlies our state takings clause. Accordingly, I would find the HCO facially unconstitutional.

I agree with part I of the concurring and dissenting opinion of Justice Baxter to the effect that the San Francisco Planning Code does not require a conditional use permit for tourist use of hotel rooms that were traditionally dedicated to tourist use, and therefore with respect to those rooms, a conversion fee was inappropriate under any takings standard. In addition to the points Justice Baxter makes, I would find the HCO preempted by the Ellis Act and facially unconstitutional, and therefore that the plaintiffs were free to disregard the HCO, and the HCO cannot render tourist use of their hotel illegal. Furthermore, because the HCO was in effect void, plaintiffs' admission under the HCO concerning residential use of their hotel is not binding on them. Moreover, the admission was based on a 1981 version of the HCO, which permitted tourist use of the rooms during the all-important summer season (1981 S.F. Admin. Code, § 41.16(a)(3)(B)), and therefore the City cannot use that admission against plaintiffs now that it seeks to restrict plaintiffs under a different ordinance from tourist use during any season. Therefore, I would affirm the Court of Appeal's decision with respect to the writ petition.

I would also affirm the Court of Appeal's conclusion that the Ehrlich "`rough proportionality'" standard ( Ehrlich, supra, 12 Cal.4th at p. 876, 50 Cal.Rptr.2d 242 , 911 P.2d 429 ) applies to the San Francisco Planning Commission's decision to require HCO compliance. This decision was adjudicative both in fact and form, and even if it were not, it would nevertheless be extortionate in nature and therefore suspect. I would also affirm the Court of Appeal's conclusion that, under the Ehrlich standard, the City has failed to show a sufficiently close link between its fee, which will subsidize low-cost residential housing, and the regulatory purpose of limiting the number of hotel rooms in the neighborhood (that being the only purpose that the planning commission could have had without impinging on plaintiffs' rights under the Ellis Act).

Once again a majority of this court has proved that "`If enough people get together and act in concert, they can take something and not pay for it.'" ( Landgate, Inc. v. California Coastal Com. (1998) 17 Cal.4th 1006, 1035, fn. 1 , 73 Cal.Rptr.2d 841 , 953 P.2d 1188 (dis. opn. of Brown, J.), quoting O'Rourke, Parliament of Whores (1991) p. 232.) But theft is still theft. Theft is theft even when the government approves of the thievery. Turning a democracy into a kleptocracy does not enhance the stature of the thieves; it only diminishes the legitimacy of the government. Like Justice Rehnquist, I "see no reason why [constitutional protections or property rights] should be relegated to the status of a poor relation." ( Dolan, supra, 512 U.S. at p. 392 , 114 S.Ct. 2309 .) The right to express one's individuality and essential human dignity through the free use of property is just as important as the right to do so through speech, the press, or the free exercise of religion. Nevertheless, the property right is nowin California, at leasta hollow one. I dissent and hope the plaintiffs find a more receptive forum in the federal courts.

[1] Plaintiffs sought no relief in state court for violation of the Fifth Amendment to the United States Constitution. They explicitly reserved their federal causes of action. As their petition for writ of mandate, as well, rests solely on state law, no federal question has been presented or decided in this case.

[2] Unless otherwise specified, all references to the HCO are to the 1990 version of the ordinance.

[3] The 1990 revision raised the portion of construction costs to be paid by the applicant from 40 to 80 percent. (1981 HCO, § 41.10, subd. (a)(4); 1990 HCO, § 41.13, subd. (a)(4), (5).) Plaintiffs, however, apparently applied during a "window" period (1990 HCO, § 41.13, subd. (d)) qualifying them for the 40 percent rate.

[4] The Bureau of Building Inspection is now known as the Department of Building Inspection. For clarity, we shall generally refer to the Department as the Bureau, the name operative at the time of most of the events related herein.

[5] The Board of Permit Appeals is now known as the Board of Appeals. For clarity, we shall continue to use the earlier name as it was the name operative at the time of most of the events related herein.

[6] The administrative record does not explain the variance between the 61 rooms initially certified by the Bureau of Building Inspection in 1981 and the 62 rooms referred to by the zoning administrator in 1992. The parties appear to agree the hotel now contains 62 rooms.

[7] Judgment was also entered for the City on its cross-complaint for penalties under the HCO, but as no issue regarding the cross-complaint is before us on review, that action need not be described further.

[8] Nollan, supra, 483 U.S. 825 , 107 S.Ct. 3141 , 97 L.Ed.2d 677 ; Dolan, supra, 512 U.S. 374 , 114 S.Ct. 2309 , 129 L.Ed.2d 304 ; Ehrlich v. City of Culver City (1996) 12 Cal.4th 854 , 50 Cal.Rptr.2d 242 , 911 P.2d 429 ( Ehrlich ).

[9] The Court of Appeal held Pfeiffer v. City of La Mesa, supra, 69 Cal.App.3d 74 , 137 Cal. Rptr. 804 , upon which the trial court partially relied, inapposite for several reasons, including that it predated enactment of the Mitigation Fee Act (Gov.Code, § 66000 et seq.), which allows a developer to pay a mitigation fee under protest and subsequently litigate its validity. As the City did not challenge this holding in its petition for review or its brief on the merits in this court, its correctness is not before us.

[10] The parties and the Court of Appeal address, at points, the question whether tourist use of the San Remo Hotel qualifies as a "lawful nonconforming use." Like a permitted conditional use, a lawful nonconforming use is one that existed lawfully at the time a new zoning prohibition or restriction came into force, the difference being that a permitted conditional use is conditionally permitted by the new zoning law, while a nonconforming use is prohibited by that law. (See S.F. Planning Code, §§ 178, subd. (a)(2), 180, subd. (a)(1); Hansen Brothers Enterprises, Inc. v. Board of Supervisors (1996) 12 Cal.4th 533, 540, fn. 1 , 48 Cal.Rptr.2d 778 , 907 P.2d 1324 .) Since tourist hotels are a conditionally permitted, rather than a prohibited, use in the North Beach district (S.F. Planning Code, § 722.55), a tourist use, if it had lawfully existed before the neighborhood commercial district restrictions became effective, would be classified as a permitted conditional use rather than a lawful nonconforming use. But regardless of terminology, the same result obtains in this case, since the rule against expansion or alteration of an existing use applies to nonconforming uses as well as to permitted conditional uses. (S.F. Planning Code, §§ 178, subd. (c), 181, subd. (a); Hansen Brothers Enterprises, supra, at p. 552, 48 Cal.Rptr.2d 778 , 907 P.2d 1324 .)

[11] That the planning commission allegedly required compliance with the HCO as a condition of the conditional use permit does not alter our conclusion as to City discretion. While issuance of a conditional use permit is generally discretionary (see S.F. Planning Code, §§ 303, 316, 316.8), conversion of residential rooms to commercial use is unconditionally prohibited above the first floor in the North Beach district except insofar as permitted by the HCO (S.F. Planning Code, §§ 722.38, 790.84). The planning commission, therefore, had no discretion to permit such change in use absent HCO compliance.

[12] According to the City, the ordinance applies to more than 500 properties containing (as stated in the 1990 HCO) more than 18,000 guest rooms. (HCO, § 41.3, subd. (d).) Plaintiffs accept these numbers but nonetheless characterize the City as imposing housing preservation costs on only "a few" property owners. The Court of Appeal, similarly, alluded to "a small group" of property owners as bearing the HCO's costs. Whether or not 500 or more property owners are properly deemed "a few" or a "small group," however, the critical fact remains that the HCO is generally and nondiscriminatorily applicable within a class of properties reasonably defined according to the purpose of the ordinance. (See Penn Central Transp. Co. v. New York City, supra, 438 U.S. at p. 132 , 98 S.Ct. 2646 [owners of landmarked properties not arbitrarily singled out to bear costs; law was a comprehensive plan to preserve historic structures, applying to over 400 landmarks and 31 historic districts throughout the city].)

[13] Alternatively, plaintiffs suggest that "[i]f this Court believes that some degree of scrutiny less than Nollan and Dolan is appropriate for legislative exactions," we could articulate a lesser standard by shifting the burden of proof to the property owner while maintaining the substantive Nollan/Dolan test. This case having been decided on demurrer, the burden of proof is not at issue; we assume the facts as pled in the second amended complaint. We therefore decline to address burden of proof issues here.

[14] The concurring and dissenting opinion, in criticizing us first for addressing these issues at all (cone. & dis. opn., post, 117 Cal.Rptr.2d at p. 305 , 41 P.3d at p. 117 ) and then for failing to address issues that have not been raised in this court ( id. at pp. 307-308, 41 P.3d at p. 119 ), seemingly ignores the choices plaintiffs have made in refining their claims as they climbed the appellate ladder. In particular, had plaintiffs wished to resurrect theories asserted in their pleading, but not raised in the City's petition for review, they could have done so by answer to the petition. (See Cal. Rules of Court, rule 28(e)(5).)

[15] In support of their claim of a lack of "nexus" between the City's housing goals and the in lieu fees assessed under the HCO, plaintiffs rely on Seawall Associates v. City of New York (1989) 74 N.Y.2d 92 , 544 N.Y.S.2d 542 , 542 N.E.2d 1059 , in which an ordinance similar in some ways to the HCO (but differing in some respects as well) was found to work a facial taking of hotel owners' property. But in finding that the ordinance did not substantially advance the city's goal of alleviating homelessness, the New York court explicitly exercised the heightened scrutiny described in Nollan, which, of course, we have concluded does not apply to the HCO's in lieu fees. ( Seawall, supra, at pp. 1068-1069 .) The Seawall court, moreover, applied Nollan to burdensome land use restrictions generally, not only to exactions imposed as conditions of permit approvals. (See Seawall, supra, at p. 1068 [discussing the ordinance's "ban on converting, destroying and warehousing [single-room occupancy] units"].) To that extent, Seawall was impliedly overruled by Monterey v. Del Monte Dunes at Monterey, Ltd., supra, 526 U.S. at pages 702-703, 119 S.Ct. 1624 , in which the high court held heightened scrutiny was "inapposite" to permit denials and other land use restrictions not involving exactions. The New York Court of Appeals acknowledged the overruling in Bonnie Briar Syndicate v. Mamaroneck (1999) 94 N.Y.2d 96 , 699 N.Y.S.2d 721 , 721 N.E.2d 971, 975 .

[16] We note as well that plaintiffs' challenge in this respect goes not to the housing replacement fee, or to any other exaction made as a condition of permit approval, but to the City's underlying reasons for restricting residential hotel conversion. A challenger to the justification for such a legislatively imposed, generally applicable restriction on changes in real property use "bears the burden of proving that the regulation `constitutes an arbitrary regulation of property rights.'" ( Santa Monica Beach, supra, 19 Cal.4th at p. 966, 81 Cal.Rptr.2d 93 , 968 P.2d 993 , quoting Dolan, supra, 512 U.S. at p. 391, fn. 8 , 114 S.Ct. 2309 .) Other than to assert that single-room rentals cannot be considered housing, a view we reject, plaintiffs make no attempt at such a showing.

[17] Plaintiffs assert they were unaware of the 1981 survey that established the hotel's initial status, and did not know of the hotel's classification until 1983. No facts alleged in their complaint, however, would show that they were prevented by City action from learning of and participating in the survey and certification process. Under the HCO, the City issued certificate of use, from which plaintiffs could have appealed, was required to be posted in the hotel lobby when issued in 1981. (HCO, § 41.6(d).) Even when they took possession in 1983, moreover, plaintiffs made no effort to correct the allegedly incorrect designation, waiting until 1987 even to draw it to the City's attention by letter. In any event, plaintiffs do not contend the HCO's survey and classification procedures deprived them of procedural due process, and we express no opinion in that regard.

[1] It follows that no conversion permit would be needed to continue a permitted conditional use, either. Yet, under the result endorsed by the majority, property owners who wish to convert only one residential room to tourist use must also convert all remaining rooms in the buildingeven rooms grandfathered as a permitted conditional tourist useand satisfy the one-for-one replacement requirement for all the rooms. Remarkably, the majority makes no effort to justify such a condition under the takings test the court has adopted.

[2] The majority purports to find it significant that the conditional use permit did not "specify the number of rooms subject to one-to-one replacement under the HCO, calculate the in-lieu replacement fee to be assessed, or impose any other condition dependent on the number of rooms previously in tourist use." (Maj. opn., ante, 117 Cal.Rptr.2d at p. 284 , 41 P.3d at p. 99 .) Plaintiffs' challenge, however, is not to the contents of the permit, but to the administrative decision that preceded issuance of the permit, in which the City refused to grandfather any prior tourist use of the hotel as a permitted conditional use. (See id. at p. 277, 41 P.3d at p. 94 .) Although plaintiffs thereafter applied for a conditional use permit for all their hotel rooms, they did so, as the City Planning Commission noted, "under protest" and without waiving their claim that the majority of the rooms should have been grandfathered as a permitted conditional tourist use.

Were plaintiffs on remand to establish the truth of their allegation that at least 53 of their 62 rooms should be grandfathered as a permitted conditional tourist use, it is certainly plausible that they would not undertake the permitting process for the remaining handful of rooms. Without a remand, however, the majority has approved a procedure in which plaintiffs, no matter the number of rooms sought to be converted, are obligated to pay $567,000. It is difficult to imagine how such a scheme satisfies even the most lax of constitutional tests.

[3] The majority's resolution of this hotly disputed issue is, to put it charitably, unclear. The majority begins by declining to wade into the fray, insisting that it is unnecessary to resolve whether the City Planning Commission properly considered the HCO to be a legal restriction on tourist use of the hotel since "the critical issue in this case is not the lawfulness of the historical tourist use, but its extent." (Maj. opn., ante, 117 Cal.Rptr.2d at p. 282 , 41 P.3d at p. 98 .) Later on, however, the majority appears to leave the sidelines and adopt the City's newly minted position, rejecting plaintiffs' effort to have prior tourist use grandfathered as a permitted conditional use "because plaintiffs' tourist use of the hotel was, as required under the HCO, temporary and subject to preemption by residential demand." (Maj. opn., ante, 117 Cal.Rptr.2d at p. 284 , 41 P.3d at p. 99 .) If the latter statement represents the majority's true views, the majority ought in all fairness explain why the city attorney, the zoning administrator, and the Court of Appeal (see Terminal Plaza Corp. v. City and County of San Francisco (1986) 177 Cal.App.3d 892, 902 , 223 Cal.Rptr. 379 [HCO "does not regulate land use in the same manner as zoning laws"]) all are wrong about the role of the HCO in determining whether a permitted conditional use has been established.

[4] I do not assume, as the majority suggests, "that conditional use permits are issued on a room-by-room basis." (Maj. opn., ante, 117 Cal.Rptr.2d at p. 284 , 41 P.3d at p. 99 .) I do, however, agree with the quotation in the text from the majority that, in classifying uses for zoning purposes, we must consider separately each use in a multiple-use structure. The need for a conditional use permit, in other words, is determined on a room-by-room basisas the majority acknowledges. (See maj. opn., ante, 117 Cal.Rptr.2d at p. 282 , 41 P.3d at p. 98 [affirming the need for a conditional use permit "[e]ven as to those rooms that had, on occasion, been lawfully rented to tourists...."].)

[5] It is unclear whether the critical date is in 1987, when the neighborhood commercial district ordinance took effect, or in 1982, when interim measures first took effect (and before plaintiffs resumed operation of the hotel), or some date in between. (See maj. opn., ante, 117 Cal.Rptr.2d at p. 281 , 41 P.3d at p. 97 .) Because the selection of the appropriate effective date may depend in part on the evidence adduced by plaintiffs demonstrating an existing tourist use, it would be premature to resolve the issue here. Inasmuch as the parties agree that the relevant effective date lies somewhere between 1982 and 1987, the purpose of the majority's discussion of usage rates between 1988 and 1991 (see maj. opn., ante, 117 Cal.Rptr.2d at pp. 278, 281, 41 P.3d at pp. 94-95, 97) is lost on me.

[6] Justice Chin cogently explained the meaning of Dolan's footnote 8 in his dissenting opinion in Santa Monica Beach, supra, 19 Cal.4th at pages 1020 through 1021, 81 Cal. Rptr.2d 93 , 968 P.2d 993 .

[7] The City misperceives the applicable standard to be "akin to the rational basis test."

[8] The Court of Appeal found that plaintiffs had properly alleged a physical taking in the complaint. The City disagrees, and urges us to reverse the Court of Appeal. On this, as with the remaining causes of action in the complaint, the majority is silent.

[9] The federal court deemed this claim unripe because state remedies for inverse condemnation on this claim were available. ( San Remo Hotel v. City and County of San Francisco, supra, 145 F.3d at pp. 1101-1102.) The majority's decision to affirm the judgment sustaining the demurrer without even mentioning this claim casts grave doubt on the correctness of the federal court's understanding of state remedies-and thus will have repercussions on ripeness far beyond this case.

[10] Since the Court of Appeal, having ordered a remand to permit plaintiffs to prove up a legal nonconforming use, found it unnecessary to reach the preemption claim, it is not surprising that plaintiffs did not use their answer to the review petition to seek our review of this issueor of the other unrelated issues that had not yet been resolved in the litigation. What is surprising is the majority's rush to have this court, in the first instance, opine on a number of constitutional issues in this difficult area of law despite the existence of an unaddressed, nonconstitutional basis for decision. The majority's eagerness to discard cherished views of judicial restraint (see, e.g., People v. Hernandez (1998) 19 Cal.4th 835, 843 , 80 Cal.Rptr.2d 754 , 968 P.2d 465 (cone, opn. of Werdegar, J.); People v. Bennett (1998) 17 Cal.4th 373, 393 , 70 Cal.Rptr.2d 850 , 949 P.2d 947 (conc. opn. of Werdegar, J.)) merely to facilitate an abrupt and unsolicited termination of this litigation in the City's favor is puzzling.