The opinion
HANING, J.
Although I fully concur in Justice Peterson's opinion, I write to emphasize the fundamental principles upon which my concurrence rests. First, as Justice Peterson states, it is undisputed that the sole authority in Berkeley to set and adjust rents, to ensure a fair return on investment in compliance with its constitutional mandate (Fisher v. City of Berkeley (1984) 37 Cal.3d 644 [ 209 Cal.Rptr. 682 , 693 P.2d 261 ]), and to promulgate regulations to accomplish those ends resides within the rent stabilization board (the Board). Second, since neither the state or federal Constitution nor any federal or state statute prohibits or preempts the regulations under attack (id., at p. 680), any limitation on the Board's discretionary power must come from the rent control ordinance itself. I do not find any language in the ordinance that specifically prohibits the subject regulations. In fact, there is nothing in the ordinance that assists or guides the Board in establishing or determining a fair return on investment except in regulation 1264, which was promulgated subsequent to the litigation herein and consequently not at issue in this case.
As Justice King notes in his dissent, the ordinance is designed to protect tenants against “unwarranted” rent increases and “arbitrary, discriminatory, or retaliatory evictions.” There are no evictions at issue in this action; none of the parties are tenants claiming unlawful or retaliatory eviction, or landlords seeking to evict tenants. As for “unwarranted” rent increases, the dissent fails to give adequate consideration to the fact that the Board is also constitutionally required to ensure a fair return on investment. The Board spent a substantial amount of time studying this problem, retained various outside economists and experts to assist it, conducted public hearings before promulgating its regulations, and the record contains ample evidentiary support for the Board’s actions.
The opponents of the Board’s regulations attempt to portray all Berkeley landlords as rapacious developers draining the life blood from the poor and the underprivileged. This simply is not the case. Based on our long history with this ongoing litigation, we know that many tenants are middle-class professionals on their way up the economic ladder, and are climbing a little faster than their contemporaries in other communities because of their ability to take advantage of Berkeley’s lower rents. Many landlords are elderly, retired and minority property owners renting out small units in the attempt to *989 make a living that their Social Security incomes cannot alone provide. The Board’s obligation in setting reasonable rents and ensuring a fair return on investment is to balance the interests of the entire community. It has done so in a fashion to avoid the undue delay found unconstitutional in Birkenfeld v. City of Berkeley (1976) 17 Cal.3d 129, 169 [ 130 Cal.Rptr. 465 , 550 P.2d 1001 ]. No party to this proceeding has suggested that individual cases of hardship or overreaching cannot or will not be addressed under the regulations. Consequently, I see no “unwarranted” rent increases, nor any other action by the Board that constitutes an abuse of its discretion under the ordinance.
The litigation over this ordinance commenced sometime shortly after an amendment to the Berkeley Charter in 1972—nearly a quarter of a century ago. Whatever the intent of the drafters, it was not that the courts establish the rents for each rental unit in the City of Berkeley—that task is clearly entrusted to the discretion of the Board. (See Aguirre v. Lee (1993) 20 Cal.App.4th 1646, 1652-1653 [ 25 Cal.Rptr.2d 367 ].) However, if every act of fine tuning the regulations to adjust to changing economic conditions prompts another legal challenge, this litigation will never end. In making this observation I am not advocating an affirmance of the Board’s action for the mere sake of expediency or economy. I simply believe that the Board’s action is constitutional, within its discretion and that interference by the courts is not warranted. (Ibid.)
KING, J.
Although I concur in parts of the majority opinion, I respectfully but most vigorously dissent from other parts. The author of the lead opinion obviously does not believe, despite our Supreme Court’s interpretation of clear limiting language in the very ordinance before us, that its purpose is to control rents.
It may well be that the City of Berkeley’s ordinance should be amended to take into account cases decided since its enactment and current rent control thinking, but that is for the citizens of Berkeley to decide, not this court. Our duty is limited to preventing the Berkeley Rent Stabilization Board (the Board) from exceeding the narrow authority granted by the ordinance. The lead opinion fails to carry out that duty, holding in effect that the Board may adopt regulations which disregard its primary statutory obligation to control rents. It consistently ignores the fact that, as interpreted by the Supreme Court, the ordinance limits annual general rent increases to "amounts reflecting the universal effect of inflation on certain specified expenses and on profits, and authorizes no other rent increases unless a landlord, by individual petition, demonstrates that he or she is not receiving a fair return on investment.
*990 I concur in the majority’s holding that the Board has properly granted a one-time increase in rents by indexing for inflation which has occurred since the adoption of the ordinance in 1979, but my concurrence is only to the extent that this is applied to the portion of net operating income (NOI) not used for debt service. I also agree, but with the same limitation, that future annual general adjustments (AGA’s) of rent should take into account the effects of inflation on NOI. This is required by Fisher v. City of Berkeley (1984) 37 Cal.3d 644 [ 209 Cal.Rptr. 682 , 693 P.2d 261 ], and by the decision I authored for this court in Searle v. City of Berkeley Rent Stabilization Board (Cal.App.) (Searle II). However, I would hold the Board acted in excess of the authority it is granted by the ordinance in also applying this one-time and annual inflation index to that portion of NOI used for debt service, because landlords with fixed-rate mortgages pay the same monthly debt service throughout the life of the mortgage, regardless of inflation. Additionally, for those Berkeley landlords with no debt service at all, the increase the majority upholds is also a complete windfall unrelated to fair return on investment.
I concur with the majority’s approval of the Board’s use of 1980 rents as the base rents, although there really is no reason why the actual rents charged at the time rent control was imposed in 1979 could not have been used. I also reluctantly concur with the majority’s approval of granting the entire one-time rent increase for inflation since 1979 to all landlords, including those who purchased their property after that time, and who thus get a rent increase reflecting prepurchase inflation.
I wholeheartedly dissent from the majority’s conclusion that the Board possesses the authority to adopt regulations which grant automatic increases in rent (1) to the one-half of all Berkeley landlords whose 1979 rents were below the median rent for comparable units, and (2) to those landlords whose 1979 rents were under 75 percent of the Department of Housing and Urban Development (HUD) rent levels (which the Board automatically increases to the 75 percent level). Both these regulations, on their face, require a landlord to petition for an individual rent adjustment (IRA) which, under section 12 of the ordinance, may be granted only if it is necessary in order to provide the landlord with a fair return on investment. Yet these two regulations, the so-called “comparable rents” and “historical low rents” regulations, provide for automatic rent increases upon the mere filing of an IRA petition, thereby eliminating the ordinance’s compulsory showing that such an increase is necessary to provide the landlord with a fair return.
*991 Therefore, I would hold the trial court erred in upholding regulations 1100 and 1113 to the extent they inflation-index the debt service portion of NOI, and regulations 1262(C)(3) (comparable rents) and 1280 (historically low rents). My dissent arises from a fundamental disagreement with the lead opinion’s general approach as well as with its reading of both the ordinance and the relevant case law.
I.
The issue before us is not, as the lead opinion repeatedly asserts, whether the Board abused its discretion in enacting the challenged regulations, but whether under an ordinance which limits the Board’s power to increase rents it exceeded its authority by enacting them. In reviewing the legislative acts of an administrative agency the court asks “first, did the agency act within the scope of its delegated authority.” (California Hotel & Motel Assn. v. Industrial Welfare Com. (1979) 25 Cal.3d 200, 212 [ 157 Cal.Rptr. 840 , 599 P.2d 31 ].) “In order for a regulation to be valid, it must be (1) consistent with and not in conflict with the enabling statute and (2) reasonably necessary to effectuate the purpose of the statute. ... [^D In other words, the rulemaking authority of an agency is circumscribed by the substantive provisions of the law governing the agency.” (Physicians & Surgeons Laboratories, Inc. v. Department of Health Services (1992) 6 Cal.App.4th 968, 982 [ 8 Cal.Rptr.2d 565 ], citations omitted.) “Administrative regulations that alter or amend the statute or enlarge or impair its scope are void and courts not only may, but it is their obligation to strike down such regulations.” (Morris v. Williams (1967) 67 Cal.2d 733, 748 [ 63 Cal.Rptr. 689 , 433 P.2d 697 ].)
Our first duty, therefore, is to determine whether the Board exercised its quasi-legislative authority within the bounds of the ordinance. That issue “comes within our respectful but nondeferential standard of review.” (Henning v. Division of Occupational Saf. & Health (1990) 219 Cal.App.3d 747, 758 [ 268 Cal.Rptr. 476 ].) “While the construction of a statute by officials charged with its administration, including their interpretation of the authority invested in them to implement and carry out its provisions, is entitled to great weight, nevertheless . . . final responsibility for the interpretation of the law rests with the courts.’ ...” (Morris v. Williams, supra, 67 Cal.2d at p. 748 , citation omitted.)
In this instance, the meaning of the enabling legislation is not in dispute, although the relevant provisions are set out in somewhat truncated form in the lead opinion. Section 3 of the ordinance states in its entirety, “The *992 purposes of this Chapter are to regulate residential rent increases in the City of Berkeley and to protect tenants from unwarranted rent increases and arbitrary, discriminatory, or retaliatory evictions, in order to help maintain the diversity of the Berkeley community and to ensure compliance with legal obligations relating to the rental of housing. This legislation is designed to address the City of Berkeley’s housing crisis, preserve the public peace, health and safety, and advance the housing policies of the City with regard to low and fixed income persons, minoritie s, students, handicapped, and the aged." (Italics added.)
Section 6, subdivision (f), enumerates the powers and duties of the Board, which include adjusting base rent ceilings “in accordance with Sections 11 and 12.” As summarized in Fisher v. City of Berkeley, supra, 37 Cal.3d at page 653 , 1 “Section 10 establishes base rent ceilings that landlords may not exceed except as permitted by the Board under sections 11 and 12. Section 11 provides for annual general adjustment (AGA) of rent ceilings to cover increases or decreases relating to utilities and taxes. In making such general adjustment, the Board is given authority to adopt a general formula based on available data relating to such expenses. If a landlord is not satisfied with this general increase, he may petition the Board for an individual rental adjustment (IRA) under section 12. In ruling on this petition the Board must consider many nonexclusive factors, including a landlord’s individual costs, but in no event may it deny a rent increase needed to allow a landlord a ‘fair return on investment.’ ” (Fn. omitted.)
The lead opinion not only ignores Fisher’s apt description of the ordinance’s limitation on the Board’s authority under section 12 to adopt regulations assuring landlords a fair return on investment in response to individual petitions, but also completely rewrites section 11 to give the Board power it is not granted by the ordinance. This ordinance severely restricts the power granted to the Board to increase rents. Unless the limitations on the Board’s power violate the Constitution, this court must enforce them. The majority refuses to do so.
*993 II.
As the lead opinion notes, the Supreme Court had invalidated an earlier Berkeley ordinance which contained no mechanism for adjusting rent ceilings without confiscatory delay. Specifically, the previous ordinance did not allow the Board to “order general rental adjustments for all or any class of rental units based on generally applicable factors such as property taxes.” (Birkenfeld v. City of Berkeley, supra, 17 Cal.3d at p. 171 , fn. omitted.) “However, the exact mechanism found wanting in 1976 is present in the ordinance before us now in section 11—a comprehensive scheme that provides for annual across the board adjustment based on ‘cost’ factors.” (Fisher v. City of Berkeley, supra, 37 Cal.3d at pp. 687-688, fns. omitted.) “This is, in essence, a variation on the so-called ‘maintenance of net operating income,’ or ‘cost passthrough’ approach.” (Id. at p. 679, fn. 32.)
Maintenance of net operating income is a commonly used formula for implementing the fair return on investment standard approved in Fisher . (See 37 Cal.3d at pp. 679-686.) Net operating income is defined as gross income less “property taxes, reasonable maintenance and operating expenses, and the amortized cost of capital improvements (pursuant to Section 1267).” (Reg. 1263.) 2 Debt service, as the lead opinion notes, is not one of the enumerated operating expenses; it is presumed the landlord pays the mortgage out of net operating income, the remainder of which, perforce, is the landlord’s “profit.”
For the first time anywhere, the lead opinion challenges the use of long-established rent control terminology. NOI minus debt service is not merely “City’s definition of ‘profit’ ” (lead opn., ante, p. 971), but the definition that all parties in the interminable litigation over Berkeley rent control, including all parties to this appeal and seven amici curiae, have used for at least 15 years. It comports with the dictionary definition, “the excess of returns over expenditure” (Webster’s New Collegiate Diet. (9th ed. 1984) p. 939) as well as with the plain meaning; Profit is what remains after payment of all expenses of the property, to be used as the landlord wishes, usually for personal goods and services. It is because inflation raises the cost of these goods and services, that profits must be inflation-indexed in order to avoid confiscation, as discussed in Searle II.
*994 It must also be remembered that originally, “maintenance” 3 of NOI did not mean, as the lead opinion appears to believe, maintenance of the real value of NOI through annual increases to compensate for inflation. On the contrary, it meant exactly what it says: the exact dollar amount of base year net operating income indefinitely maintained. In response to the Birkenfeld court’s rejection of Berkeley’s attempt to freeze rents indefinitely, section 11 of the new ordinance “provides for annual rent adjustment, but precludes the Board from granting such citywide rent adjustment except to offset certain increases in general costs.” (Fisher v. City of Berkeley, supra, 37 Cal.3d at p. 679, fn. 32 , italics added.) Debt service was not included among those costs that could be expected to rise annually for every landlord at about the same rate, but was considered a part of NOI which was to remain fixed. 4 The lead opinion not only ignores, but contravenes, the specific statutory limitations on the power of the Board.
The Fisher court warned that NOI could not remain fixed forever either. Because of the effect of inflation, although the ordinance might “properly restrict landlords’ profits on their rental investments, it may not indefinitely freeze the dollar amount of those profits without eventually causing confiscatory results.” ( 37 Cal.3d at p. 683 , italics in original, citing Cotati Alliance for Better Housing v. City of Cotati (1983) 148 Cal.App.3d 280, 293 [ 195 Cal.Rptr. 825 ].) As the court explained, “if the fixed amount of a landlord’s profit remains the same year after year his return will in time diminish in real value: it is obvious that a $1,000 ‘profit’ in 1990 will have a much lower value than the same dollar amount of profit in 1980.” (Fisher v. City of Berkeley, supra, 37 Cal.3d at p. 683 .) Thus Fisher required the annual general adjustment to include consideration of not just the enumerated expenses, but also the effect of inflation on profits, in order to assure continued fair return on investment.
It is not, as the lead opinion suggests, through oversight or imprecision that the discussion in Fisher is framed entirely in terms of the effect of inflation on profits, rather than NOI. (37 Cal.3d at pp. 682-684; see also Palos Verdes Shores Mobile Estates, Ltd. v. City of Los Angeles (1983) 142 Cal.App.3d 362, 368-372 [ 190 Cal.Rptr. 866 ], upholding a “maintenance of profit” or “fair operating income” approach.) That is the way the plaintiffs presented the issue because they and everyone else involved understood that “profits” meant the portion of NOI remaining after payment of debt service. *995 It also makes sense, for while a given number of dollars would not buy in 1984 the same goods and services as it did in 1979, in most cases it would still pay the mortgage. The price of goods and services on which profits may be spent rises as a result of inflation; payments on fixed-rate mortgages do not.
It is true, as the lead opinion points out, that in Searle II we discussed adjustment of NOI, not profits, for inflation. That is because we were reviewing a regulation which purported to index NOI, not profits, for inflation. Neither party suggested, nor did we consider, that partial inflation-indexing might be justified by the distinction between the profit and debt service portions of NOI. We held in Searle II that on the record before us an inflation index of 40 percent of the consumer price index (CPI) was arbitrary precisely because the Board had offered no rational basis for its choice. This does not mean, as the lead opinion suggests, that there could never be a rational basis for that choice. For example, the City of Santa Monica indexes NOI at 40 percent of the annual CPI increase, based on studies indicating that 60 percent or more of its owners’ base year NOI went to long-term fixed-rate debt service. Thus, in Santa Monica, 40 percent indexing of base year NOI is tantamount to 100 percent indexing of base year profits.
We also held in Searle II that there was no rational basis for relegating adjustments for the universal impact of inflation on all landlords to the individual petition system. What reason exists then, the lead opinion asks, for treating the portion of NOI which goes to pay debt service differently from “all other components of NOI,” i.e., profits? The answer is obvious: They are not similarly affected by inflation.
For the vast majority of landlords paying off long-term fixed-rate mortgages, 5 inflation has no impact whatever on the purchasing power of that portion of NOI which goes toward debt service. It is true, as the lead opinion observes, that debt service and financing costs are not fixed for every landlord, 6 but it does not follow that they are subject to “the universal impact of inflation” (Searle II, supra, A044761), that is, that all landlords’ debt *996 service is uniformly affected by inflation. Some landlords choose to finance their properties through other than long-term fixed-rate mortgages, e.g., adjustable-rate mortgages. Some may have refinanced or bought rental property during periods of high interest rates, while others purchased or refinanced during mortgage rates’ recent descent to the lowest levels in two decades, and of course, some landlords have no mortgage at all. These are very personal decisions, involving various levels of risk, just as when one is financing a home. They are based on innumerable very personal factors, psychological as well as economic; each decision, each choice, is an integral part of an entirely individual financial plan. Some landlords, for example, may refinance rental property in order to obtain funds for investment elsewhere. Even if this involves a variable-rate mortgage, there is no reason a tenant should have to finance the increased costs of a landlord’s personal investments. Even if, as the lead opinion suggests, mortgage rates “generally move in tandem with inflation rates” (lead opn., ante, p. 975) (which has certainly not been the case in recent years), these individual choices cannot trigger a constitutional entitlement to annual general rent increases—available as well to landlords with no debt service at all because they have paid off their mortgages or received rental property through gift or inheritance— based on fluctuations in the consumer price index.
The lead opinion places great emphasis on the fact that the Board acted on the basis of “expert” evidence of “outside economists” (lead opn., ante, pp. 956, 961, 963). The author of the lead opinion does not name these experts, but presumably refers to the consulting firm of Hamilton, Rabinowitz & Altschuler (HR&A). Whatever the nature and extent of its experience “in the design of rent control programs” (lead opn., ante, p. 961), HR&A expressly eschewed any legal expertise, telling the Board it was not a law firm and “does not purport to offer professional legal opinions or assessments.” Thus the Board’s consultants, by their own admission and contrary to the lead opinion’s assertion, could not and did not render advice on what was necessary to provide a “fair return” or avoid “confiscation.”
The experts did testify to the following more or less self-evident mathematical facts: given a scheme in which rent is composed of operating expenses plus net operating income, and net operating income is composed of debt service plus profits, if you inflation-index operating expenses and profits but not debt service, the result will be (1) the ratios among the *997 various components will change, and (2) net operating income and therefore rents will not keep pace with inflation. This in turn slows down the appreciation rate of rental properties. The Fisher court recognized and approved this “antispeculation” effect. (Fisher v. City of Berkeley, supra, 37 Cal.3d at pp. 691-692.) Neither rents, nor NOI, nor property values actually “shrink,” as the lead opinion suggests, but they do increase more slowly than they would without rent control. This is one of the reasons for adopting rent control.
If, as a matter of policy, the Board wants rents to keep up with inflation, and property to appreciate quickly, the expert testimony indubitably provides a rational basis for 100 percent inflation-indexing, including debt service. It is not arbitrary, capricious or otherwise prohibited by the Constitution. But neither is it required by the Constitution or permitted by the ordinance. Fisher established the constitutionality of Berkeley’s maintenance of net operating income system for guaranteeing landlords a fair return on investment, except insofar as it indefinitely froze the dollar amount of profits in the face of constant inflation. Fisher and its progeny do not require maintenance of the original ratio among, or “real dollar” value of, all the elements in the formula. If they did, confiscation would have begun with the very first AGA, which adjusted only that portion of rent revenues dedicated to operating expenses. The Constitution does not guarantee a certain ratio among the elements in a rent control formula, nor rents that keep pace with inflation, nor a certain rate of property appreciation. It guarantees a fair return on investment. As the Board admitted in its reply brief, even a decline in property value does not necessarily deny a landlord a fair return. 7
We have come, full circle, back to the intersection of the ordinance and the Constitution. To recapitulate: the Board is bound by its ordinance. The ordinance grants power to make annual general rent adjustments in section 11. Section 11 allows AGA’s only to cover increases in specified costs. The Constitution requires that landlords receive a fair return on their investment. The Fisher court said freezing profits would eventually be confiscatory, but section 12’s fair return guarantee provided a sufficient adjustment mechanism to save the ordinance’s facial constitutionality. In Searle II, we said section 12’s individual petition process was not sufficient to adjust rents for the universal and uniform impact of inflation, because it is both irrational and likely to lead to confiscatory delay. 8 Thus, as the lead opinion notes, the AGA power expressly granted to the Board in section 11 has been “expanded by the constitutional requirement” (lead opn., ante, p. 962) of inflation *998 indexing to guarantee a fair return. This is true, however, only to the extent that the individual petition process in section 12 is constitutionally insufficient. As we found in Searle II, the IRA process is constitutionally insufficient when a universal threat to fair return, here inflation, affects all landlords every year to the same extent. As has been shown, this can be said of the profit but not of the debt service portion of NOI, as held by the lead opinion.
Thus, the Board exceeded its authority under the ordinance in enacting regulations 1100 and 1113 to the extent they provide for across-the-board inflation indexing of that portion of net operating income which, on average, is dedicated to debt service. To the extent the Board interpreted its ordinance otherwise, its interpretation is “clearly erroneous or unauthorized” (Anderson v. San Francisco Rent Stabilization & Arbitration Bd. (1987) 192 Cal.App.3d 1336, 1343 [ 237 Cal.Rptr. 894 ]; Cole v. City of Oakland Residential Rent Arbitration Bd. (1992) 3 Cal.App.4th 693, 698 [ 4 Cal.Rptr.2d 593 ]).
Neither of the above two cases addresses, contrary to the lead opinion’s suggestion, “analogous debt service issues” (Lead opn., ante, p. 973). For starters, the lead opinion does not assert, nor could it, that either the San Francisco or the Oakland rent control law is as stringent as Berkeley’s. Oakland’s ordinance, for example, “allows landlords to impose a rental increase where there is an increase in debt service” (Cole v. City of Oakland Residential Rent Arbitration Bd., supra, 3 Cal.App.4th at p. 695 ), while Berkeley’s does not (Ord. § 12, subds. (d) & (e); Fisher v. City of Berkeley, supra, 37 Cal.3d at pp. 691-692). At cited page in Anderson , the court held the San Francisco ordinance did not authorize exclusion of preexisting debt service “from the term ‘monthly loan payments’ ([San Francisco Admin. Code] 32.73, subd. (a)).” {Anderson v. San Francisco Rent Stabilization & Arbitration Bd., supra, 192 Cal.App.3d at p. 1347 .) The quoted phrase (lead opn., ante, p. 973), inexplicably replaced by an elipsis in the lead opinion, undermines any applicability Anderson might have to this case. The “sole *999 issue” presented in Cole was “the propriety of the Board’s method of computing gross income under [Oakland City Council Ordinance No. 9980] section 10.14.1” (Cole v. City of Oakland Residential Rent Arbitration Bd., supra, 3 Cal.App.4th at p. 697 ), the source of the “formula for allowing permissible rent increases based on increased debt service costs” (id. at p. 696). Neither of these cases is remotely on point.
The dictum in Kavanau v. Santa Monica Rent Control Bd. (1993) 19 Cal.App.4th 730, 733 , footnote 3 [ 23 Cal.Rptr.2d 724 ], with which the lead opinion agrees (“at a minimum,” debt service is a factor relevant to fair return determination) (lead opn., ante, p. 977) has, of course, no precedential value, especially as it appears to conflict with the Fisher court’s express approval of Berkeley’s two “ ‘antispeculation’ clauses” which explicitly exclude increased debt service as a factor in individual fair return analysis in most cases. (Fisher v. City of Berkeley, supra, 37 Cal.3d at pp. 691-692.)
III.
The lead opinion also concludes, virtually without analysis, that the Board did not abuse its discretion in enacting the comparable rent and historically low rent regulations, 9 largely “for the reason stated in the preceding section.” (Lead opn., ante, p. 983.) But these regulations have an entirely different source, and are based on an entirely different rationale from the inflation-indexing regulations. They provide for individual rather than general rent adjustments, and are designed to implement section 12’s guarantee of a fair return on investment, although as will be seen, as written they grant automatic rent increases without considering whether the landlord is already receiving a fair return.
It will be recalled that Berkeley’s maintenance of NOI system is based on the net operating income produced by rents in the base year, usually 1979. (Reg. 1264.) Since base year rents were set in an unregulated market, base year NOI is presumed to have provided landlords with a fair return on their investment in the base year. Regulation 1262, as originally enacted, permitted individual adjustment of base year NOI upon an administrative determination that, contrary to the express presumption, base year NOI had not provided a fair return on a particular landlord’s investment in the base year. Factors that might be considered in making such a determination included “the fact that the rent was exceptionally high or low, due to a family or *1000 special relationship between the landlord and tenant, or other factors.” (Italics added.)
As amended in 1987, the regulation permitted individual adjustment of base year NOI upon an administrative determination that it was “exceptionally high or low” for either of two reasons, one of which was the “special relationship” factor reworded to cover any situation in which “the rent was not established in an arms-length transaction or was otherwise not set under market conditions.” Effective April 1991, regulation 1262 mandates (rather than merely permitting) individual adjustment of base year NOI under the specified circumstances, to which a third has been added: “base year rent was below prevailing market rents for comparable units in the base year.” (Reg. 1262(C)(3).) Thus, in an unprecedented interpretation of a rent control ordinance, the Board permits half of all base year rents to be increased automatically to the median. This increase has no direct relationship to fair return on investment, the rationale the lead opinion uses to uphold it. 10
The lead opinion asserts the Board enacted regulation 1262(C)(3) “to comply with Birkenfeld and its progeny,” (lead opn., ante, p. 957) that it was “guided by the decision” (lead opn., ante, p. 961) in Vega v. City of West Hollywood (1990) 223 Cal.App.3d 1342 [ 273 Cal.Rptr. 243 ], and that it concluded such an adjustment mechanism was mandated by those decisions. If so, the Board was mistaken. Contrary to the lead opinion’s repeated but unexamined conclusion, neither Birkenfeld nor Vega “required rent boards to allow upward rent adjustments in these circumstances.” (Lead opn., ante, pp. 966, 982-983.)
In Vega , the rent control ordinance (like Berkeley’s) contained an express presumption that net operating income produced by property during the base year provided the landlord a fair return. The presumption could be rebutted by a finding that, “The rent on the base date was disproportionately low due to the fact that it was not established in an arms-length transaction or other peculiar circumstances.” (Vega v. City of West Hollywood, supra, 223 *1001 Cal.App.3d at p. 1345 , fh. 1.) It was uncontested in Vega that “peculiar circumstances” had caused the individual landlord’s base rents to be “disproportionately low.” The issue was how to determine the amount of the base rent adjustment to which she was therefore entitled. The court ordered her base date rents set in a manner consistent with evidence of rents for comparable units. Since it was undisputed that the landlord was eligible for a base rent adjustment, the Vega court did not even address the issue raised by Berkeley’s new comparable rent regulation. 11
In Birkenfeld v. City of Berkeley, supra, 17 Cal.3d at page 169 , the court held a base rent adjustment mechanism was constitutionally necessary to provide for “situations in which the base rent cannot reasonably be deemed to reflect general market conditions.” As one factor that might have prevented base rent from reflecting general market conditions, the court mentioned “the existence of a special relationship between landlord and tenant resulting in an undercharging of rent.” {Id. at p. 168.) This is the source of the “arms-length transaction” provisions in the ordinance at issue in Vega and in Berkeley’s regulation 1262, both of which focus on the manner in which base rents were set rather than on the outcome.
The challenged “comparable rents” provision is not mandated by Birkenfeld because it requires no showing that rents were set in such a manner that they cannot be deemed to have reflected general market conditions. Neither Birkenfeld nor Vega suggests, even in dicta, that all base rents which fall below “prevailing” rents for comparable units were necessarily set under other than market conditions and should automatically be increased. The most that can be said is that they were on the lower end of the range established by an unregulated market, which the ordinance presumes provided a fair return on investment.
My analysis is bolstered by the recent case of Apartment Assn, of Greater L.A. v. Santa Monica Rent Control Bd. (1994) 24 Cal.App.4th 1730,1736 [ 30 Cal.Rptr.2d 228 ] (review den. Aug. 11, 1994) in which the court entitled the first section of its discussion, “There Is No General Constitutional Entitlement to Base Date Rents Adjusted to Market Levels.” The court explained, “Though both [Birkenfeld and Vega] stated that the base date rents should reflect general market conditions, in both cases the landlord’s entitlement to an increase in the base rent depended on the existence of circumstances that prevented the base rent from reflecting market conditions. . . . Neither case *1002 held that all landlords have the right to have their base rent adjusted merely by showing that the base rent was below market.7 Respondents’ position that ‘Birkenfeld and Vega establish a constitutional standard of general application to all historically low base rent properties without exception’ is not supported by the opinions in those cases, and we hold that there is no general entitlement to an increase in base date rents predicated on market conditions.” (Id. at pp. 1736-1737.) In footnote 7, the court observed, “Indeed, the Vega opinion begins, ‘The issue in this case is how, under a particular municipal rent control ordinance and peculiar circumstances, rental rates must be determined.’ ( 223 Cal.App.3d at p. 1344 , italics added.)”
Since Vega does not compel regulation 1262(C)(3), does the ordinance permit the Board to adopt it? This is the dispositive question, the one which the lead opinion once more fails to ask or answer. The ordinance contains no express provision at all for base rent adjustments. Section 12, which governs IRA’s, provides in subdivision (c), “In making individual adjustments of the rent ceiling, the Board or the hearing examiner shall consider the purposes of this Ordinance and shall specifically consider all relevant factors, including (but not limited to): ... (8) The landlord’s rate of return on investment.” Subdivision (c) concludes, “It is the intent of this Ordinance that individual upward adjustments in the rent ceilings on units be made only when the landlord demonstrates that such adjustments are necessary to provide the landlord with a fair return on investment.” Thus, the ordinance clearly denies the Board the authority to enact a regulation automatically raising all base year rents below the median to the median, unless necessary to provide a fair return on investment.
Although the original express “fair return” language does not occur in later versions of regulation 1262, it remains logically implicit. Otherwise, the entire regulation would be prohibited by the governing section of the enabling legislation. 12 Read to conform with the ordinance, regulation 1262 sets out the circumstances which will establish a presumption that base year NOI did not provide a fair return on investment, which can be rebutted by contrary evidence. A regulation automatically increasing rents merely because they are below the median is unauthorized by the ordinance unless it is shown to be necessary to provide a fair return on investment. Since there is no reason to believe that every unit renting for an amount in the lower half of the range established by an unregulated market failed to provide a fair return in 1979, the ordinance does not permit (C)(3) in its current form.
*1003 To summarize, neither Birkenfeld nor Vega requires the challenged comparable rents regulation. The Constitution requires a fair return on investment. The ordinance presumes base year NOI provided a fair return on investment. Regulation 1262 enumerates certain unusual circumstances in which it may be presumed that base year net operating income did not provide a fair return on investment. Such a presumption does not reasonably arise from the mere fact that base year rent was “below prevailing market rents for comparable units.” Therefore, the ordinance does not permit subdivision (C)(3).
A brand new “historically low rent” regulation, regulation 1280, provides for units with 1979 rents below 75 percent of HUD rents for Alameda County in 1979, 13 automatic increases up to that 75 percent level adjusted to reflect all AGA’s granted since 1979. The administrative record suggests regulation 1280 was intended to be an alternative “quick ‘n’ easy” version of the comparable rent regulation just discussed. In essence, regulation 1280 allows some landlords with low base year rents to escape such burdens as gathering and presenting data on comparable rents for comparable units in the base year, and making a “fair return on investment” showing (in the absence, as we have noted, of appropriate guidelines). In return for this procedural shortcut, these landlords accept a preestablished limit—75 percent of the HUD figures—on their rent increases. Such an increase may or may not result in a fair return on investment since the regulation provides an increase without considering whether the landlord already receives a fair return on investment.
The lead opinion upholds regulation 1280 on the same bases as it upheld regulation 1262(C)(3). Under the analysis we applied to that regulation, regulation 1280 as enacted must also fail. It is not required by Birkenfeld or Vega , and exceeds the bounds of the ordinance to the extent that it circumvents the individual petition process 14 and grants adjustments of base year NOI without regard to fair return.
*1004 On the other hand, it would not be unreasonable for the Board to create a rebuttable presumption that rents falling below 75 percent of HUD rents for Alameda County in 1979 failed to provide a fair return on investment. Thus, regulation 1280 might validly be recast as a replacement for the comparable rent regulation 1262(C)(3). That is, the criterion embodied in regulation 1280, subdivision (b) (base year rents below 75 percent of HUD rents) which now establishes an automatic right to rent increases in specific amounts might replace the arbitrary and unreasonable criterion (base year rents below prevailing market rents for comparable units in the base year) as the third basis for an individual adjustment of base year NOI, subject to fair return analysis.
IV.
I cannot agree with my colleagues that the regulations providing for inflation-indexing of the debt service portion of NOI or automatically increasing rents which were below the median for comparable units or were “historically low rents” are either required by prior case law or within the Board’s discretion. On the contrary, I conclude the case law does not require nor does the ordinance permit the Board to adopt such regulations. Moreover, I believe the source of our disagreement is our very different understanding of the “logical foundations of . . . rent control . . . itself.” (Lead opn., ante, pp. 968-969.)
The lead opinion obviously believes “. . . the proper purpose of rent control, as limited by constitutional standards, is to allow local rent boards to set rent levels at fair and reasonable levels—which protect the value of property owners’ investments in their property from confiscation by maintaining their net operating income against the continual and progressive ravages of inflation.” (Lead opn., ante, p. 974.) But this statement turns rent control on its head! The purpose of rent control is to control rents, to keep them down, to protect tenants from landlords who may be greedy. This is especially so in Berkeley, where the purpose provision of the rent control ordinance is essentially a statement of tenants’ rights, as recognized by Fisher . The constitutional standard (fair return on investment) which limits rent control (even in Berkeley) is what protects landlords, ensuring that rent control does not go so far in protecting tenants that landlords’ property is confiscated.
Perhaps we differ as well in our understanding of the “respective roles of administrative agencies and the courts.” (Lead opn., ante, p. 976.) The lead opinion stresses the deference which courts generally accord to administrative agencies, but ignores the deference we owe to an ordinance enacted by *1005 initiative of the people, “ ‘ “one of the most precious rights of our democratic process.” ’ ” (Amador Valley Joint Union High Sch. Dist. v. State Bd. of Equalization (1978) 22 Cal.3d 208, 248 [ 149 Cal.Rptr. 239 , 583 P.2d 1281 ], citation and internal quotation marks omitted). In this case, we are presented with an administrative agency which, for historical reasons, may well be irremediably at odds with its enabling legislation. Not so long ago, really, but in a different time, the people of Berkeley adopted an ordinance which was meant to and did impose strict limitations on rent increases. Much more recently, they elected what is sometimes referred to as a “moderate” rent board. Does this mean Berkeley has changed its collective mind about rent control? Maybe so, maybe not; it is not for us to decide. Until the ordinance is repealed or amended, our job is to ascertain what it says and what authority it gives to the Board, not what we think perhaps Berkeley voters or the Board, might now wish it said.
The lead opinion concludes, “The net effect of our decision will be to return to rent boards the task of interpreting and shaping the regulatory details of the rent control process . . . .” (Lead opn., ante, p. 987.) I agree, but based upon the lead opinion they will now be able to do so without regard to the limitations placed upon their power by their enabling legislation.
The lead opinion apparently believes that by electing this Board, the citizens of Berkeley have signaled their disenchantment with stringent rent control, so we ought to help them out by reading into their ordinance a grant of discretion broad enough to permit the enactment of what we consider “reasonable” regulations. It may be that courts have already intruded too far into the legislative process where rent control is concerned. But there is a difference between judicial construction aimed at preserving the constitutionality of an ordinance passed by the people, and judicial construction which essentially undermines its very foundation. To paraphrase the court in Abramson v. City of West Hollywood (1992) 7 Cal.App.4th 1121, 1128 [ 9 Cal.Rptr.2d 507 ], no matter how great the deference or how strong the policy, I am unwilling to read into the ordinance a grant of authority or discretion which does not exist. To do so, in my view, is to engage in unwarranted judicial activism.
The petition of appellant City of Berkeley for review by the Supreme Court was denied December 1, 1994. Mosk, J., and Kennard, J., were of the opinion that the petition should be granted.
Fisher is not only the lead case on the very ordinance under which the challenged regulations were enacted, but also the first Supreme Court case to establish an inflation-indexing requirement of constitutional dimension. Fisher should be our guide in this appeal. It is therefore curious that the majority gives this landmark decision short shrift compared to Birkenfeld v. City of Berkeley (1976) 17 Cal.3d 129 [ 130 Cal.Rptr. 465 , 550 P.2d 1001 ], which concerned neither this ordinance nor the issue of inflation-indexing. Perhaps this is because the lead opinion is completely inconsistent with the Supreme’s Court’s decision in Fisher .
Since section 12, subdivision (c)(3), of the ordinance, as implemented by regulation 1267, provides for rent increases by individual petition to pass through to tenants the cost of capital improvements necessary to maintain habitability, the author of the lead opinion worries needlessly that less than full inflation-indexing through the AGA process will discourage landlords from completing needed repairs, or result in deteriorating rental housing stock.
“Maintenance” means “the act of maintaining.” To “maintain” is “to keep in an existing state.” (Webster’s, op. cit. supra, at p. 718.)
The lead opinion’s “note that debt service adjustments are not forbidden by section 11 of the ordinance, and that the other considerations specifically listed [there] are not defined so as to exclude the power of a board to adjust them” (lead opn., ante, p. 970) is thus incorrect.
The record does not disclose the percentage of mortgages on Berkeley residential rental properties that are fixed as opposed to adjustable rate. We can assume the vast majority are fixed rates and thus for the entire life of the 20- or 30-year mortgage, the monthly payment remains the same although paid with inflated dollars. For example, if the mortgage payment for the first month is $1,000, every payment for the life of the mortgage is $1,000, not an amount increased by inflation. Of course, as each payment is made, an increased portion of it is applied to principal, increasing the landlord’s investment in the property.
It is not true, on the other hand, that, in the lead opinion’s words, “City argues that. . . all property owners’ debt service costs are necessarily fixed in actual dollars.” (Lead opn., ante, p. 975.) That would be as foolish as suggesting, as the lead opinion, again incorrectly, *996 says the City has, that “. . . the falling real value of NOI will not adversely affect landlords.” (Lead opn., ante, p. 975.) To the extent it is successful in controlling rents, all rent control adversely affects landlords to some degree. Fisher recognized and approved this effect as long as the ordinance, as in section 12, has a mechanism for a landlord to petition to assure a fair return on investment.
Thus, it is not entirely accurate to say, as the lead opinion does, that in Searle II we ordered the Board to take action “in order to preserve the value of property.” (Lead opn., ante, p. 960.)
We also said in Searle II that Berkeley’s fair return regulation (then Reg. 1275) remained constitutionally inadequate in that it failed to furnish specific guidelines enabling landlords to *998 understand the burden they must bear to obtain relief and the benefits to be expected if they succeed. “It does not say,” we noted, “what fair return on a landlord’s investment is, nor provide a formula for determining it.” {Searle II, supra, A044761.) Regulation 1264 now states that fair return will “ordinarily” be measured by maintenance of inflation-indexed base year NOI. But it still does not say how fair return will be measured extraordinarily, that is, in those cases where a landlord contends his or her inflation-indexed base year NOI (even adjusted as the regulations allow) does not provide a fair return on investment. It is the Board’s responsibility to determine what constitutes a fair return on investment. Courts should only become involved if the exercise of that discretion is abused by a return falling outside broad legal parameters. In other words, it is the responsibility of the Board to determine what constitutes a fair return on investment so that landlords can ascertain whether to file for an IRA.
The lead opinion refers to these regulations as applying to “very low rents” (lead opn., ante, p. 984), “abnormally low rents” (lead opn., ante, p. 982), rents “far below the rents of comparable units” (lead opn., ante, p. 957), and rents “in a hardship category.” (lead opn., ante, p. 966.) As will be seen, both regulations define the covered units rather specifically; neither uses any of the rhetorical but imprecise formulations adopted by the lead opinion, nor does the record before us demonstrate these characterizations are correct.
Regulation 1262(C)(3), does not define the term “prevailing.” According to the dictionary it means: most frequent, generally current, common. (Webster’s, op. cit. supra, p. 932.) Section 10 of the ordinance provides that “where no rent was in effect on May 31, 1980, or during the six months preceding that date, the base rent ceiling shall be a good faith estimate of the median rent in effect for comparable units in the City of Berkeley on May 31, 1980.” (Italics added.) Thus, it is reasonable to suppose, as at least one Board hearing officer has done, that regulation 1262(C)(3), permits upward adjustment of all base year rents below the median, affecting half of all rent-controlled units. This reading is supported by the addition in the most recent revision of regulation 1262, effective February 5, 1993, of a provision that adjustments under (C)(3) may not be granted to a unit whose base rent was set according to the quoted portion of section 10 of the ordinance.
The lead opinion consistently mischaracterizes certain statements in Vega (e.g., 223 Cal.App.3d at pp. 1349, 1351, 1352) as “holdings.” (Lead opn., ante, pp. 961, 983.) In context, these broad generalities mean at most that “comparable rents” are more appropriate criteria than fair return on individual investment for adjusting base rents not established in arm’s-length transactions.
The Board seems to have recognized as much when, in the current version of regulation 1262 (amended since this litigation began), it added, “Adjustments of current lawful rent ceilings under this subdivision (C) may only be awarded in conjunction with an NOI analysis under regulation 1264,” that is, the “fair return” regulation.
Neither the record on appeal nor the parties’ briefs document the exact source or nature of the HUD figures employed in regulation 1280, which are described therein as “fair market rents.” The lead opinion calls them rent levels “certified as reasonable by the federal government” (lead opn., ante, p. 984). Unspecified “information before the Board” allegedly indicated that Berkeley units rented at market rates in 1979 had rents at or above 75 percent of the HUD figures.
The author in the lead opinion states the Board has power to grant rent adjustments “through the IRA process.” (Lead opn., ante, p. 984.) But the very point of regulation 1280 is that it allows certain landlords to obtain rent increases merely by applying for them, without going through the IRA process. The Board’s suggestion that minority landlords need this “simplified procedure” because they are somehow incapable of complying with regulations 1262’s procedural requirements is profoundly troubling. I agree with the lead opinion that the Board should respond to concerns about its bureaucratic maze and delays in hearings (no *1004 evidence of which is in the record before us), but it demeans minority landlords to have a simplified process only for them.