OPINION

HUFFMAN, Acting P.J.

This appeal by the Building Industry Association of San Diego, Inc., a California nonprofit corporation (BIA), presents questions as to the validity of a residential growth control initiative, chapter 32A of the City of Oceanside Municipal Code, commonly known as Proposition A (Prop. A), adopted by the voters of the respondent City of Oceanside (the City or Oceanside). Actions for declaratory and other relief by BIA and Del Oro Hills (Del Oro), a partnership, the plaintiff in a related action (judgment in which was also appealed and is pending before this appellate panel), were consolidated below, with Del Oro adopting the BIA arguments as to the invalidity of Prop. A, and additionally asserting its own damages claims in further trial proceedings. The City obtained judgment in its favor in a three-phase court trial on BIA's complaint as to the validity of Prop. A; summary judgment was then rendered for the City against Del Oro. By stipulation, appeals of the judgments against BIA and Del Oro were consolidated for hearing by this court, although we have vacated the consolidation order for the sole purpose of issuing separate opinions in the two matters. ( Del Oro Hills v. Oceanside (D017139, app. pending).)

We have concluded that the BIA judgment must be reversed because the trial court erred in its application of the doctrine of law of the case based upon this court's prior opinion in writ proceedings arising out of the trial court's denial of BIA's motion for summary adjudication on the issues of whether Prop. A conflicted with the City's general plan and with state planning and zoning law (Gov. Code, § 65000 et seq.). ( Building Industry Assn. v. Superior Court (1989) 211 Cal.App.3d 277 [ 259 Cal.Rptr. 325] [hereafter Building Industry or prior opinion].) Moreover, under the standards set by Lesher Communications, Inc. v. City of Walnut Creek (1990) 52 Cal.3d 531 [ 277 Cal.Rptr. 1, 802 P.2d 317] ( Lesher), Prop. A is facially in conflict with the City's general plan and with state planning and zoning law. In light of this finding of conflict, we need not reach the issue presented as to the appropriate burden of proof on the City under Evidence Code section 669.5 (whether the City had to show that Prop. A bore a real and substantial relationship to the public welfare of the citizens of Oceanside, or rather that its provisions were "necessary" to protect the public welfare).

FACTUAL AND PROCEDURAL BACKGROUND

1. Adoption of Prop. A

"Ch. 32A, adopted by the Oceanside electorate in April 1987, declares one of its purposes is 'to augment the policies of the City as recorded in the General Plan and City ordinances relating to the regulation of residential development,' and '[i]n order to accomplish this purpose, the City must be able to control the rate, distribution, quality and economic level of proposed development on a year to year basis.' Ch. 32A adopts a 'Residential Development Control System' (RDCS) which, with what may be significant exceptions, adopts a maximum number of dwelling units to be constructed each year, called annual allotments. The allotments are 1,000 for 1987 and 800 for each year thereafter until December 31, 1999, with power granted to the City Council to modify the annual allotment by an amount no greater than 10 percent more or less for any given year and a requirement the annual allotment for a next succeeding year be adjusted higher or lower in order to redress any excess or deficit in the preceding year. Excepted from the RDCS are the following: '(a) Projects of not more than four residential dwellings, limited to only one such project per developer per calendar year.

"'(b) Fourplexes or less numbered multiple dwellings on a single existing lot.

"'(c) Single family residential units on a single existing lot.

"'(d) Rehabilitation or remodeling of an existing dwelling, or conversion of apartments to condominiums, so long as no additional dwelling units are created.

"'(e) Units within the legally designated redevelopment project area.

"'(f) Those specific Units which are formally dedicated for occupancy by low income persons or senior citizens pursuant to the provisions of applicable federal, state, or local laws or programs provided these types of units are spread equitably throughout the city and not concentrated in one neighborhood. For the purposes of this section, a project is funded or subsidized pursuant to applicable federal, state or local laws or programs if it receives a loan, grant or continuing financial subsidy for the purpose of developing low-income or senior citizen housing units. This section does not exempt low income or senior citizen projects built with density bonuses or other development considerations under any program.

"'(g) Single family dwelling unit projects with lots an average of which are 10,000 square feet or better, which can achieve a minimum of 70% or better, of the maximum awardable points using the Residential Development Evaluation System are exempt.'

"Ch. 32A, in addition, provides for application, evaluation and award of development allotments which must be granted before a building permit may be issued." ( Building Industry, supra, 211 Cal.App.3d at pp. 285-286.)

Under Prop. A, projects are reviewed by a "Residential Development Evaluation Board" (the Board) made up of members of the City's planning commission, who evaluate proposed projects for their impact upon public facilities and services (the "A" criteria) and site and architectural quality (the "B" criteria). A project which does not receive a score of 51 percent on the "A" criteria and 70 percent on the "B" criteria is eliminated from consideration for an annual allocation. The Board's recommendations are forwarded to the city council, which makes the annual allocations.

2. Oceanside's Land-use Regulation

At the time Prop. A was adopted, there were a number of existing land-use regulations in place in the City. Since 1982, the City had had an interim growth management element (IGME) as an element of its general plan, requiring projects with negative fiscal impacts to be approved only by a supermajority of the planning commission and city council, and only if there were offsetting benefits. Since 1979, the City's general plan had a public facilities and management element (PFME), which stated as its objective "[t]o influence the timing of development and to direct it to those locations within the City that avoid or minimize any adverse fiscal, economic, social or environmental impacts." The PFME divided the City into four areas and gave priority to development in those areas where adequate services were available. The PFME stated a policy of avoiding "direct controls on the number or location of new housing units built. . . ." Neither the IGME nor the PFME gave the City the power to regulate the rate, timing or sequencing of development.

The PFME used the method of planning for growth by encouraging assessment districts and impact fees. The PFME was adopted by the City in light of a compound annual growth rate of 6.3 percent since 1970. An environmental impact report (EIR) for the PFME states that development was expected to continue in the range of 900 to 2,000 housing units per year through 1995, a range of 2.9 to 7 percent compound annual growth through 1995. The EIR for the PFME points out that that measure rejected the use of a point system, such as would later be used in Prop. A, with the explanation that "[p]oint systems usually are flawed by the potential for a high score for one service to offset a low score for another" and on the theory that "the problem of adding apples and oranges and then considering only the total is difficult to avoid." The EIR also rejected a "periodic comparative evaluation" system, such as Prop. A would later adopt.

In 1986, the year before Prop. A was adopted, the city council adopted a new land-use element (LUE) of the general plan, setting forth the city's long-term goals, objectives, and policies for development. The LUE calls for but does not establish a timing mechanism to regulate residential development. At the time the LUE was adopted, the City recognized the PFME had not solved all the City's growth-related problems. However, no official City action had been taken to invalidate the PFME.

3. The City's Growth Status as of 1987

In the 1970's, the City's growth rate had reached 8.94 percent per year. While the City's population grew by 89.4 percent from 1970 to 1980, the County's population increased by 37.1 percent. The growth rate as of 1987 was 5 percent annual population increase, equivalent to 60 percent growth if sustained for a decade. The City's expert witness, Dr. Myers, a professor of urban and regional planning, testified that a 1 percent growth rate is normal, and 2 percent per year or 25 percent per decade is a healthy growth rate. In his opinion, at 40 percent growth per decade, it is difficult to "keep up" with a city's growing population. "[W]ith rapid growth the needs escalate and you are always playing catch up."

Conflicting evidence was presented about the City's ability to accommodate growth. The trial court heard evidence that there were severe deficiencies in facilities and services, such as the road system, fire, paramedic and police facilities and response times, schools, libraries, parks and recreation facilities. There were drainage and sewer problems. However, several City officials and the City planner testified that the City's infrastructures were in reasonably good shape.

4. Regional Supply of Low-income Housing

Pursuant to state law mandate, the San Diego Association of Governments (SANDAG) determines the region's share of the state's growth and allocates a regional share for a five-year period to each city and incorporated areas of the county. Dwelling units thus allocated to the City are referred to in this opinion as the "regional share." Evidence was presented at trial as to the City's regional share for the 1985 through 1991 housing element period, by income category, as follows:

" Very Low "Total [Income] Low Moderate All Others " 0-50% 50-80% 80-120% Over 120% "9,107 2,049 1,576 1,757 3,726"

For the 1991 through 1996 housing period, the figures were as follows:

" Very Low "Total [Income] Low Moderate All Other " 0-50% 50-80% 80-120% Over 120% "7,463 1,716 1,269 1,567 2,911"

5. BIA Files Its Action

Shortly after Prop. A was adopted, BIA and 10 developers sued the City under several theories contending Prop. A was invalid on its face and seeking declaratory and injunctive relief. After the trial court denied cross-motions for summary judgment and summary adjudication, its order was upheld by this court in our prior opinion. We concluded that invalidity of a growth control ordinance "can be established only by determining facts bearing on whether the enactment truly conflicts with state law and its purposes." ( Building Industry, supra, 211 Cal.App.3d at p. 290.) We stated that whether the regional housing needs as established by SANDAG will be met by the City was a question of material fact that awaited proof. ( Id. at p. 293.) We explained that some factual determinations by the trial court would of necessity have to be projections of the reasonable probability of accommodating regional housing requirements in future housing periods. ( Id. at p. 294.) We were unable to establish the validity of any of BIA's various asserted conflicts between Prop. A and state law "without reference to established facts." ( Ibid.) We further found that there was present no clear invalidity of Prop. A due to inconsistency with the City's general plan, on the basis that "factual determinations need to be made before the question of inconsistency can be resolved." ( Id. at p. 297.)

In accordance with the prior opinion, the trial court scheduled trial of the BIA lawsuit, and consolidated Del Oro issues, in three phases. Phase I addressed the issues raised by Evidence Code section 669.5, subdivision (a), as to whether the City had successfully rebutted the statutory presumption that Prop. A was "presumed to have an impact on the supply of residential units available in an area which includes territory outside the jurisdiction of the City. . . ." (Evid. Code, § 669.5, subd. (a).) In phase I, the trial court made findings that the City had met its regional share of all new residential housing units for the period between 1986 and 1991, and that it would meet the same regional share requirement for the next housing element period, 1991 through 1996. As to the next housing element period, 1996 through 2000, the court found it probable that the City would also meet its requirements for that period.

However, the court went on to find that when Prop. A took full effect in 1988, affordable housing in the City took a dramatic decline. There was a drastic reduction in the supply of more affordable housing (apartments, condominiums and townhouses) as compared to detached residential homes after the implementation of Prop. A. The court noted that from April 1987 to March 1990 only 415 total building permits were issued under Prop. A's exception subdivision (f) for low-income/senior housing, and concluded that the Prop. A exception subdivision (f) for low-income persons or senior citizens was not generating enough new housing units to approach the required SANDAG figures. The City had not shown that any other exception to Prop. A was generating low-income housing (except for fourplex units), and that the dwelling units generated by the annual allotment would not be available to low-income families due to the point ranking system encouraging unit size, reduced density and other expensive features. Declining to consider future mitigation measures to be proposed by the City, on the grounds that the court must make its decision on the present impact of Prop. A, the trial court found that the City had not rebutted the presumption under Evidence Code section 669.5, subdivision (a) that Prop. A would have an impact on the supply of affordable residential units in the region. The City challenged the trial court's ruling on phase I by writ application to this court, which was denied by order of November 27, 1990. ( Oceanside v. Superior Court (Nov. 27, 1990) (D013199 [nonpub. opn.].)

6. Phase Two Trial and Findings

Under Evidence Code section 669.5, subdivision (b), the City had the burden of proof that Prop. A was "necessary for the protection of the public health, safety, or welfare of the population of the city. . . ." ( Ibid.) The trial court rejected BIA's argument that Evidence Code section 669.5, subdivision (b) established a very strict test for the validity of such building permit cap ordinances, holding that instead the section merely recapitulated the standards set forth in Associated Home Builders etc., Inc. v. City of Livermore (1976) 18 Cal.3d 582, 607-609 [ 135 Cal.Rptr. 41, 557 P.2d 473, 92 A.L.R.3d 1038] ( Livermore). That is, the trial court agreed with the City's interpretation of "necessary" in Evidence Code section 669.5, subdivision (b) as "reasonably related to" the protection of the public health, safety or welfare of the affected population. Thus, the court set forth as the appropriate "balancing test" for phase II:

"The City has the burden of proof by the preponderance of the evidence that (1) there are real and substantial 'problems relating to its police power to protect public health, safety and welfare of its population' (either present or reasonably anticipated), and (2) that the ordinance or initiative in question is reasonably related to and designed to remedy the identified public health, safety and welfare problems.

"The Court must then balance the above two factors against the quantified adverse effects which the ordinance will have upon the need for public housing including affordable housing in the region." (Fn. omitted.)

The court further noted that the more impact Prop. A was proven to have upon public housing, including affordability, the more substantial the police power justification must be to balance the competing interests.

In applying this balancing test, the trial court analyzed the evidence presented of the implementation thus far of Prop. A (1987-1990) and also the evidence forecasting its effect into the future (through 2000), as a means of judging the validity of Prop. A as of its adoption date, April 1987. The court noted that Prop. A was to be used to time development with relation to adequate services, over its duration. Thus, the trial court clearly ruled that evidence of the 1987-1990 implementation of Prop. A was circumstantially relevant to the determination of whether Prop. A was reasonably designed to accomplish its objectives. It explained in the statement of decision that it was required to balance the City's problems and the relationship of the ordinance to the problems versus the "quantified adverse effects which the ordinance [would] have upon the need for public housing including affordable housing. . . ." The overall balancing test had to be applied as of the adoption date, 1987. The issue presented was Prop. A's impact as presently written, not as potentially amended. In summary, the trial court had to decide whether Prop. A was valid as of 1987, when its future effects were taken into account, i.e., as shown by the implementation and forecasting evidence that was presented to show how Prop. A worked out in practice.

In making its specific findings, the trial court first calculated that based on the 443 housing unit allocations issued in the first three and one-half years of Prop. A's effective period, in the categories of very low and low income and fourplexes, there would be a deficiency of approximately 2,350 units in those categories for the current housing period (1991 through 1996). The court found there would be an additional similar deficiency in moderate income units. The court then found that Prop. A's impact during the 1991 through 1996 housing period represented at least 33 percent of the City's regional share requirements (which represented a total of 7,463 units). The court then estimated that, based upon a family of four persons, approximately ten thousand persons might be deprived of anticipated housing needs in the region in that current housing period, and a similar number would likely be denied housing over the next housing period.

Moving on to an analysis of the competing interests or justification for Prop. A, the court summarized the evidence given of the City's current infrastructure problems, and concluded that the City was undergoing an "unhealthy" 5 percent growth rate, although the court acknowledged that this rate was down from 8.94 percent in the 1970's. Although the court acknowledged that BIA had successfully called into question the City's position that there were glaring infrastructure deficiencies at the time Prop. A was adopted, the court concluded that, based on the unhealthy growth rate, the existing infrastructure deficiencies became more meaningful and subject to more concern relating to the public welfare, such that the "unhealthy growth rate" as of 1987 created real and substantial problems which the City could reasonably anticipate to occur in the near future.

The court then analyzed the state of the existing regulations at the time that Prop. A was adopted, noting that the City had growth management tools in the tentative subdivision map act and CEQA procedures, the IGME, the PFME, and the LUE and a managed growth ordinance. The court then concluded that the City had the authority to implement a stronger ordinance or initiative, such as Prop. A. The court then found that Prop. A had a reasonable relation to the problems it was designed to address, based on its ranking system and "A" and "B" list criteria, and it was too speculative to say the City would disregard the objectives of the initiative even though the number of units permitted under Prop. A had actually been higher than the number permitted before it went into effect.

To complete the balancing test of the competing interests affecting Prop. A, the trial court reiterated that about 20,000 people had, in its estimation, probably been deprived of affordable housing during the term of Prop. A, but that approximately 40,000 people would probably benefit from Prop. A, insofar as it promoted quality of life considerations. This 40,000 figure was derived from a family of 4, based on a projected 10-year allocation of 8,000 housing units (32,000 persons) during the duration of Prop. A, with an additional 7,800 persons benefiting from the 10,000-square-foot-lot exception. The trial court thus concluded that although the denial of housing to the 20,000 projected population should in some ways outweigh the disadvantages of a reduced quality of life if Prop. A were not in effect, nevertheless some persons denied housing would be able to live elsewhere, and the benefits to the 40,000 people who would obtain an enhanced quality of life under Prop. A should outweigh the deficiencies to the 20,000 persons kept out of the housing market. The trial court thus found that Prop. A represented a reasonable accommodation of the competing interests, and ruled in favor of the City in phase II.

7. Phase III Trial and Findings

Phase III of the trial relied on evidence from phases I and II, and presented the issues of whether Prop. A conflicted with either the PFME element of the City's general plan, or three specified Government Code sections: section 65008, subdivision (c) (prohibiting discrimination against low-income housing), section 65913.1 (establishing appropriate standards for zoning to contribute to the economic feasibility of producing the lowest possible cost housing), or section 65915 (the density bonus law intended to encourage low-income housing). The trial court took the matter under submission and ruled in favor of the City on all of the issues, explaining that although BIA had made a persuasive showing that Prop. A conflicts with the density bonus provisions of section 65915, the court was bound by the prior decision in the case, ruling that state law did not preempt Prop. A and there was no facial conflict between Prop. A and section 65915 at the time Prop. A was adopted. The court further ruled that even in light of BIA's showing of conflict between Prop. A and the density bonus law, the low-income/senior exception to Prop. A (subd. (f)) should not be invalidated because section 69515 was not triggered until a developer proposed a housing development, by submitting a preliminary proposal to which the City was required to respond by notifying the developer of the procedures it planned to use to comply with the density bonus law (i.e., by waiving or modifying development and zoning standards which would otherwise inhibit the utilization of the density bonus on specific sites). Other findings were made as to sections 65008, subdivision (c), and 65913.1 (to be explained in pt. II.D., post).

Judgment was entered accordingly and further proceedings as to Del Oro were severed for trial. (See Del Oro Hills v. Oceanside (D017139, app. pending).) BIA timely appealed the judgment.

DISCUSSION

BIA presents as its threshold issue on appeal whether the trial court properly interpreted Evidence Code section 669.5, subdivision (b), as requiring the City to show only that Prop. A was reasonably related to the public health, safety, and welfare, as opposed to showing it was "necessary" to promote those objectives. The City goes back even further in its respondent's brief and argues that the trial should have ended with phase I, and that the trial court should have found that the City adequately rebutted the presumption of Evidence Code section 669.5, subdivision (a), that Prop. A had an impact on the regional supply of residential units. The City responds to BIA's assertion that the wrong standard was used by claiming that in any case, whatever standard or whatever manner of balancing the trial court used, the evidence supports the judgment.

Having studied these issues, we decline to base our decision on any of them. Rather, the dispositive issue presented is the issue of whether Prop. A impermissibly conflicts with either the City's general plan, i.e., its PFME element, or state law provisions litigated at trial (§§ 65008, subd. (a), 65913.1, or 65915). When we apply the doctrine of law of the case and the rules announced in Lesher, supra, 52 Cal.3d 531, on the facts found and established by the trial court, we conclude that Prop. A was invalid when passed because it is inconsistent with policies established by the general plan, and it is further invalid because it conflicted with state law as of the time of its adoption. ( Id. at pp. 540-541, 544-545, 547.) Because the trial court should have ruled that Prop. A was facially in conflict with the PFME and with state law, as expressed in the cited Government Code housing statutes, and because the evidence showed that, as applied, Prop. A remained in conflict with those provisions, the issue of the proper standard to apply to assess Prop. A becomes moot in light of the measure's invalidity ab initio.

We begin our discussion of the conflict issue with an analysis of the law of the case effect of our prior opinion, and then discuss the tension between the policies promoted by the PFME and state law, on the one hand, and Prop. A, on the other.

I Law of the Case

(1a) In its statement of decision on phase III of the BIA trial, the trial court outlined its view of the effect of the prior opinion upon its ruling on the issues presented regarding the alleged conflict of Prop. A with the City's general plan and/or the three particular Government Code sections relied on by BIA (§§ 65008, 65913.1, 65915). The court first cited People v. Rath Packing Co. (1974) 44 Cal.App.3d 56, 66-67 [ 118 Cal.Rptr. 438], for the rule that issues of law that were before and decided by the Court of Appeal are law of the case and binding on the trial court. The court then listed the following five issues of law which it ruled precluded the parties from relitigating the claims that Prop. A is facially inconsistent with the general plan's PFME or section 65008, 65913.1, or 65915:

"1. The Court of Appeal rejected Plaintiffs' claims (among others) that there is a facial inconsistency between Prop. A and (1) the PFME of the City's general plan, or (2) Government Code § 65008, or (3) Government Code § 65913.1, or (4) Government Code § 65915. ([ Building Industry v. Superior Court, supra,] 211 Cal.App.3d 277, 294, 295 and 297.) The Court considered a 'checklist of twelve (12) asserted conflicts which BIA presents,' and held 'the arguments are unconvincing and lack merit.' The Court then specifically examined some of the code sections in the checklist ( id. at [p.] 294) and stated 'due to a similar lack of merit in BIA's other arguments of conflict, we deem it unnecessary to address them seriatim.' ( Id. at [p.] 295.)

"2. The Court of Appeal stated the rules that Prop. A must be presumed valid and comes before the Court with every intendment in its favor. 'If the validity be "fairly debatable, the legislative judgment must be allowed to control".' Plaintiffs bear the burden of overcoming this presumption and the burden of proof on the alleged conflicts issues. (See id. at [p.] 293.)

"3. The Court of Appeal held '[p]artial inconsistency with portions of a general plan or state law will not alone suffice to render a numerical growth control ordinance invalid.' ( Id. at [p.] 290.)

"4. 'Substantial compliance with State law is the test applicable to [Prop. A]. The question of substantial compliance is one of law. . . .' ( Ibid.)

"5. 'No state preemption precluding enactment of [Prop. A] is present in the field.' ( Id. at [p.] 295.)"

(2) Since at least 1930, the "modern view" has been that the doctrine of the law of the case should not be adhered to when its application results in a manifestly unjust decision. ( Standard Oil Co. v. Johnson (1942) 56 Cal.App.2d 411, 415 [ 132 P.2d 910], citing United Dredging Co. v. Industrial Acc. Com. (1930) 208 Cal. 705, 712 [ 284 P. 922].) "'[ W] here the controlling rules of law have been altered or clarified in the interval between the first and second appeal and adherence to the previous decision would result in defeating a just cause, it has been held that the court will not hesitate to reconsider its prior determination. [Citations.]'" ( Standard Oil Co. v. Johnson, supra, at pp. 415-416, original italics.)

(1b) In light of this rule, we must consider the extent to which the principles of law applied in our prior opinion have been disapproved or modified by the Supreme Court's discussion of that opinion in Lesher, supra, 52 Cal.3d at pages 545 through 547. In Lesher, the Supreme Court decided that an initiative measure limiting municipal growth was invalid at adoption because it conflicted, contrary to state law, with the city's existing general plan. The Supreme Court disagreed with two points in the analysis in our prior opinion: First, the Supreme Court disapproved any suggestion that a municipal zoning ordinance which was inconsistent with a city's general plan could be subject to a compliance decree rather than a finding of invalidity. (Although no question of a remedy for an invalid ordinance was reached by the trial court here, the point is important on the general validity analysis.) Another statement by the Supreme Court controls our review of several of the trial court's rulings: The Supreme Court expressly disapproved the prior opinion to the extent that it suggests other than that "[a] city may not adopt ordinances and regulations which conflict with the state Planning and Zoning Law. [Citations.]" ( Lesher, supra, 52 Cal.3d at p. 547.) The Supreme Court's statement was based on the following reasoning: "Conforming a general plan to an inconsistent growth-control ordinance might also be inconsistent with the legislative policy that each city and each county provide in the general plan for its appropriate share of the regional need for housing. (§ 65302.8.) Under that section, amendment of a general plan to limit the number of housing units to be built annually must be accompanied by findings that justify reduction of housing opportunities in the region." ( Lesher, supra, 52 Cal.3d at p. 546.)

These statements in Lesher, supra, 52 Cal.3d 531 undermine the validity of the prior opinion's statements that "partial inconsistency with portions of a general plan or state law will not alone suffice to render a numerical growth control ordinance invalid" ( Building Industry, supra, 211 Cal.App.3d at p. 290), and also suggest that this court erred in stating "substantial compliance with state law is the test applicable to Prop. A." Further, it seems our statement that "no state preemption precluding enactment of [Prop. A] is present in the field" ( id. at p. 295) was too broad, as it is now clear that growth control ordinances may not conflict with state planning and zoning laws.

(3), (1c) Next, as to the trial court's conclusion that our prior opinion precluded any relitigation of the claims that Prop. A is facially inconsistent with the PFME or the specified Government Code sections, another aspect of the doctrine of the law of the case must be considered: "[L]aw of the case consists in the propositions of law actually decided and applicable to the facts in judgment. It only applies when, upon a subsequent trial, the issues and facts found remain substantially the same, and has no application where the facts alleged and found are materially different from those considered on a former appeal. [Citation.] The doctrine is applied only to the principles of law laid down by the court on appeal as applicable to a retrial of fact. [Citation.] Insofar as the former decision related to the effect of the evidence or the findings on the former trial there can be no application of the doctrine until the facts have been elicited on a retrial. The doctrine not only does not apply to new and additional evidence, it does not apply when explanation of previous evidence appears in the later trial. [Citation.]" ( Weightman v. Hadley (1956) 138 Cal.App.2d 831, 841 [ 292 P.2d 909].)

With this rule in mind, we return to the prior opinion to determine what it actually decided. The prior opinion consisted of a denial of petitions for writ of mandate which sought to compel the superior court to grant motions for summary judgment or adjudication brought by BIA and Del Oro, which would have invalidated Prop. A. This court upheld the trial court's determination that triable issues of fact remained before the question could be resolved whether there is facial inconsistency between Prop. A and the City's general plan or state law. We explained that the matter of conflict between a growth control ordinance and state law or a general plan "cannot be determined without reference to facts relating to a local entity's compliance with its obligation to meet its share of regional housing needs." ( Building Industry, supra, 211 Cal.App.3d at p. 284.) In discussing the various arguments made by BIA and Del Oro, this court consistently emphasized that factual determinations remained to be made before it could be decided whether the ordinance accomplished a reasonable accommodation of competing interests. ( Id. at p. 293.)

Even when we considered BIA's 12 asserted conflicts with state law and/or the general plan, we were unable to determine the merit of the arguments "without reference to established facts." ( Building Industry, supra, 211 Cal.App.3d at p. 294.) Although we went on to demolish one of the arguments specifically, it was not one of the same arguments raised in this proceeding. ( Id. at pp. 294-295, discussing the findings requirement of § 65589.5.) Similarly, although we stated at page 297 of the prior opinion, "It is readily apparent in the case before us that there is present no clear invalidity of [Prop. A] due to inconsistency with the general plan," we went on to state that "factual determinations need to be made before the question of inconsistency can be resolved." ( Building Industry, supra, 211 Cal.App.3d at p. 297.) Thus, we are unable to agree with the trial court that this court expressly or finally rejected BIA's claims that there was facial inconsistency between the three asserted Government Code sections, section 65008, section 65913.1, or section 65915, or the PFME of the City's general plan. Instead, those conclusions were preliminary on that record, which we said lacked essential factual determinations before any such conflict ruling could be made.

II Conflict of Prop. A With the General Plan or State Law A Introduction

(4a), (5a) The above law of the case analysis does not end our inquiry. Our task now is to decide whether on the facts as now established by the trial court, Prop. A meets the standards set out by Lesher, supra, 52 Cal.3d 531 regarding conflict of a zoning or planning ordinance with either a city's general plan or the state Planning and Zoning Law. (§ 65000 et seq.) (6), (4b) First, to reiterate the statements in Lesher concerning conflict of a zoning ordinance with a general plan, such an ordinance is "invalid at the time it is passed. [Citations.] The court does not invalidate the ordinance. It does no more than determine the existence of the conflict. It is the preemptive effect of the controlling state statute, the Planning and Zoning Law, which invalidates the ordinance." ( Lesher, supra, 52 Cal. 3d at p. 544.) The Planning and Zoning Law, in this context, requires adoption and maintenance of a city's general plan. (§ 65300 et seq.; Lesher, supra, at pp. 538-539.) Under section 65860, county or city zoning ordinances must be consistent with the entity's general plan, such that "[t]he various land uses authorized by the ordinance are compatible with the objectives, policies, general land uses, and programs specified in such a plan." (§ 65860, subd. (a)(ii).) "Only the general plan in effect at the time the ordinance is adopted is relevant in determining inconsistency." ( Lesher, supra, at p. 545.)

(5b) Moreover, "[a] city may not adopt ordinances and regulations which conflict with the state Planning and Zoning Law. [Citations.]" ( Lesher, supra, 52 Cal.3d at p. 547.) The specific statutory arguments BIA raises concerning sections 65008, 65913.1, and 65915 fall into this category, and the issue is essentially preemption. Article XI, section 7 of the California Constitution allows counties and cities to make and enforce within their limits "all local, police, sanitary, and other ordinances and regulations not in conflict with general laws." In our prior opinion, we recognized that local growth control legislation is not generally preempted, impliedly or expressly, by statute. ( Building Industry, supra, 211 Cal.App.3d at pp. 295-296.) However, due to the numerous factual issues then remaining unresolved, we could not and did not rule upon the specific statutory arguments raised here concerning the effect of Prop. A, as implemented and as forecasted.

B Factual Analysis

(4c), (5c) As explained in the prior opinion, the crux of the issue was whether in light of facts relating to the local entity's compliance with its obligation to meet its share of regional housing needs, there was conflict between a numerical growth control ordinance and state law or a general plan. ( Building Industry, supra, 211 Cal.App.3d at p. 284.) In the trial in this matter, the court made certain findings, which we now summarize, regarding the City's compliance with its obligation to meet its share of regional housing needs. In phase I, the trial court made a finding that for the 1986-1991 housing period, the City had met its regional share of new residential housing units (considering all income categories). For both the 1991-1996 period and the 1996-2000 period, it was projected to do so also. The court emphasized that for the 1986-1991 housing period, the City had met and exceeded its regional share for lower/moderate income categories. However, since some of those units were attributable to pre-Prop. A permits, the court opined that it was too simplistic to say that resolves all the issues; thus, the court looked to postimplementation housing periods. The court reasoned that when Prop. A took full effect in 1988, affordable housing declined, so the next housing period is relevant.

Thus, considering the 1991-1996 period (as required by the prior opinion, which allowed projections, and to allow for pre-Prop. A permits), the court noted that only 443 units were approved from Prop. A's adoption in April 1987 through March 1990 in the low-income, senior citizen and fourplex categories; the court noted that those are the only low-income categories and the City could not place all the blame on developers for avoiding low-income housing. Moreover, the City's other housing programs had not filled the breach. Because the court found Prop. A will result in insufficient low-income housing for its regional share for the 1991-1996 housing period, it did not deem the meeting of the 1986-1991 requirements to be dispositive.

In phase II, the trial court was to balance the City's problems and the relationship of Prop. A to those problems, against the quantified adverse effect Prop. A would have on affordable housing, applying that test as of the 1987 adoption date. The Livermore criteria ( supra, 18 Cal.3d at pp. 607-609) were to be used: (a) estimating the probable effect/duration of restriction; (b) identifying the competing interests; (c) determining whether the ordinance is a reasonable accommodation of competing interests. The trial court used the evidence of the implementation of Prop. A from 1987-1990, and also evidence that forecast its effect into the future (through 2000) in order to judge the validity of Prop. A as of its adoption date. This approach was consistent with the theory of our prior opinion concerning the conflict issue.

In the statement of decision, the trial court reemphasized that although in Phase I it had found the City accommodated its regional share of low/moderate income housing for the 1986-1991 period, and the all income total regional share housing requirements will be met through 2000 despite Prop. A, the City could not show the 1991-1996 low/moderate income regional share units would be built. The court estimated a deficiency of 2,350 low income, senior citizen and fourplex category units for the "current housing period" (meaning 1991-1996.) Prop. A was also found to adversely impact availability of moderate income units for 1991-1996, although the evidence was not clear as to how much. In light of the probable effect and duration of the restriction, the court found it should give less weight to compliance with total regional share than to compliance with the all income categories share. The court then engaged in its numerical balancing effort (there was a 33 percent impact in 1991-1996 on the City's regional share of very low/low/moderate income categories; likely the same proportion in 1996-2000) and its analysis of the competing interests or justification for the ordinance, which included the previous four years' implementation evidence as to the timing requirement against the adverse effect on low income housing, all over the duration of Prop. A.

The trial court's overall conclusion on the degree of the City's compliance with its obligations to provide low and moderate income regional share units was that its 1986-1991 compliance was not enough, as the court had to go on to the 1991-1996 period to make a complete analysis, and the City could not show it would meet the regional share for low/moderate income units for that period. In the statement of decision for phase I, the trial court noted that the prior opinion required and allowed it to make projections of the reasonable probability of accommodating regional housing requirements to the year 2000, as part of the ultimate determination of the facts.

Based on the above, the City is not justified in arguing that the phase I finding that all areas of housing were accommodated (for 1986-1991) means that the phase II finding is unsupported (that Prop. A prevented the City from meeting housing needs); the phase I positive finding deals with 1986-1991, but the phase II negative finding applies to 1991-1996. Nor is the City justified in contending, as it did at oral argument, that the phase II findings about the 443 exceptions granted are on a different issue than the issue of whether the overall housing goals were met. As discussed above, the trial court had ample basis to use both the implementation and forecasting evidence to judge the validity of the ordinance as of 1987, its adoption date.

In phase III, the trial court concluded that even under the low income exemption from Prop. A, developers were discouraged from proposing density bonus units by the operation of Prop. A. (§ 65915.) Even though the City later interpreted the language of Prop. A excluding density bonus projects to lessen that impact, the court noted that the market rate units in any proposed density bonus project would still have to await building permit allocations, which would reduce or eliminate the incentive of the density bonus. The court stated that the recent amendments to the density bonus law had not alleviated that impact of Prop. A, and the new density bonus law as amended sought to create even stronger incentives for the housing needs of low income households. The court went on to find that it was premature to speculate on whether the City might comply with the density bonus law, when no developer proposal was yet before the City planning authorities. The court further stated it was bound by the prior opinion's discussion that state law did not preempt Prop. A.

Although we have determined it is not necessary to address the issues raised concerning the application of the Evidence Code section 669.5, subdivision (b) balancing test, it is relevant here to note that the City, in its response to the brief filed by amici curiae, Pacific Legal Foundation and People for Affordable Housing, has agreed that the trial court went too far in speculating how Prop. A might be implemented and in engaging in a numbers test, i.e., the estimation that while some 20,000 persons would be deprived of affordable housing under Prop. A, some 40,000 persons would be benefited by its promotion of a better quality of life. Under the circumstances, we should not place too much weight upon the trial court's application of the balancing test, and instead focus upon the various factual findings that it made for purposes of analyzing any conflict between Prop. A and the general plan or the particular statutes involved.

C The General Plan

(4d) At the time Prop. A was adopted, the general plan had in effect the PFME, which contained a statement of a number of major policies to be served by the plan. In particular, the PFME stated a policy to "[a]void direct controls on the number or location of new housing units to be built, but provide financial incentives for projects where services can be provided to new and old residents at least cost." As previously explained, we do not interpret our prior opinion as making any final determination on the actual record that there was any clear invalidity of Prop. A due to inconsistency with the general plan. (See Building Industry, supra, 211 Cal.App.3d at p. 297.) Now that factual determinations have been made regarding the extent of the City's compliance with its obligation to address regional housing needs, the trial court had the power to reach the merits of that issue. After first reasoning that the law of the case required a finding of no conflict, the trial court ruled that BIA and Del Oro had not shown any conflict of Prop. A with the PFME, as there was no substantial evidence of such conflict. The court reasoned that the PFME had been insufficient to manage the unhealthy growth situation existing at the time Prop. A was adopted. To reach that conclusion, the court discounted the criticisms in the PFME's EIR of a point system to evaluate growth, reasoning that the EIR was not a policymaking document, but only informational in nature.

In light of the clear statement in Lesher, supra, 52 Cal.3d at page 544, that a zoning ordinance that conflicts with a general plan is invalid at the time it is passed, and in light of the factual findings that Prop. A adversely affected the availability of low-income housing, the only possible finding is that Prop. A does conflict with the City's general plan, specifically the PFME, regarding the imposition of direct controls on the number or location of new housing units to be built. This restriction on land uses is not compatible with the then-existing general plan. (§ 65860, subd. (a)(ii); Lesher, supra, 52 Cal.3d at p. 545.)

Although the trial court reasoned that the City had the power to adopt a more restrictive type of growth control than that represented by the PFME, it could not do so without validly amending the existing general plan. According to the reasoning of Lesher, supra, 52 Cal.3d at pages 540-544, general plans cannot be impliedly amended and a zoning ordinance cannot be deemed a general plan amendment without a clear indication that the voters intended to accomplish such an objective when passing the initiative. Moreover, Lesher clearly indicates that no compliance decree would be appropriate to bring the conflicting zoning ordinance into compliance with the general plan, as a court may not rescue a zoning ordinance that is invalid ab initio. ( Lesher, supra, 52 Cal.3d at p. 545.) Since the record in this case contains an adequate basis to determine the validity of Prop. A both on its face and as applied up through the time of trial, there is no need for a remand for further proceedings on this issue. Finally, the housing element of Oceanside's general plan states as a policy:

"ADEQUATE PROVISION OF HOUSING

"Adequate provision for the housing needs of all economic segments of the community is an issue of the highest priority in Oceanside to meet the low income household assistance goals and to protect, encourage and, where feasible, provide low and moderate income housing opportunities within the intent of State policy to address local needs. . . ."

Prop. A does not promote this policy and accordingly must be deemed inconsistent with it. Although we are sympathetic to the goals stated in Prop. A of promoting the quality of life for the citizens of the City, we are bound by Lesher to find the ordinance invalid under those standards.

D Conflict With Statutory Provisions

(5d) BIA and Del Oro next assert that Prop. A conflicts with three particular sections in the state Planning and Zoning Law. The text of the sections on which BIA and Del Oro rely is in relevant part, in the version in effect at the adoption of Prop. A in 1987, as follows:

Section 65008, subdivision (c):

"No city, county, or city and county shall, in the enactment or administration of ordinances pursuant to this title, prohibit or discriminate against a residential development or emergency shelter because the development or shelter is intended for occupancy by persons and families of low and moderate income, as defined in Section 50093 of the Health and Safety Code, or persons and families of middle income."

Section 65913.1:

"[A] city . . . shall designate and zone sufficient vacant land for residential use with appropriate standards . . . to meet housing needs as identified in the general plan . . . 'appropriate standards' shall mean densities and requirements . . . which contribute significantly to the economic feasibility of producing housing at the lowest possible cost given economic and environmental factors, the public health and safety, and the need to facilitate the development of housing for persons and families of lowand moderate income. . . ."

Section 65915:

"(a) When a developer of housing agrees to construct at least (1) 25 percent of the total units of a housing development for persons and families of low or moderate income, . . . or (2) 10 percent of the total units . . . for lower income households, . . . a city, . . . shall either (1) grant a density bonus or (2) provide other incentives of equivalent financial value.

". . . . . . . . . . . . . . . . . . . . . . . . . . . .

"(c) . . . 'density bonus' means a density increase of at least 25 percent over the otherwise maximum allowable residential density under the applicable zoning ordinance and land use element of the general plan. . . ."

In its statement of decision for phase III of the trial, the trial court interpreted the prior opinion as deciding that there was no facial conflict between Prop. A and any of these sections. As discussed above, we disagree with that characterization of the prior opinion due to the factual determinations which remained for trial at the time the opinion was written. The trial court had another basis for each of its rulings as well: as to section 65008, the court stated Prop. A was neutral as to any particular development project, regardless of whether it was designed for low income or other future residents. As to section 65913.1, the trial court stated that Prop. A did not affect, change, or prohibit rezoning, and that because the evidence showed the housing element goals had been met for low and moderate income housing in 1986 to 1991, and because there was a current surplus of dense housing and available land zoned for small attached units, as well as a low income exemption in Prop. A., there was no substantial conflict between Prop. A and section 65913.1. As to section 65915, the density bonus law, the trial court ruled that even though it found Prop. A discouraged the development of low income housing units which would be eligible for a density bonus, and even though there was some showing that Prop. A conflicted with density bonus provisions, the trial court did not have to invalidate the low-income exemption to Prop. A because by its amended terms section 65915 was not triggered until a developer actually proposed a housing development, thus allowing it to submit a preliminary proposal to which the City must respond with its proposed action to provide the density bonus.

Again, Lesher, supra, 52 Cal.3d 531, provides strong guidance for this Court in analyzing the conflicts issue. The Supreme Court disapproved the prior opinion to the extent that it suggests that a city may adopt ordinances and regulations which conflict with the state Planning and Zoning Law. ( Lesher, supra, 52 Cal.3d at p. 547.) The Supreme Court's action undercuts the statement in the prior opinion that "[a]pparent partial inconsistency with portions of a general plan or state law will not alone suffice to render a numerical growth control ordinance invalid." ( Building Industry, supra, 211 Cal.App.3d at p. 290.) It also suggests that "substantial compliance" with state law is too lenient a test to apply to zoning and land-use control ordinances. ( Id. at p. 291.) We should thus look at the question of conflict between Prop. A and these provisions in light of the facts established by the trial court on the issue of the City's compliance with its obligation to meet its share of regional housing needs. ( Id. at p. 284.)

The trial court made key findings that since Prop. A took effect in 1988, affordable housing in the City had dramatically declined. Only 415 total low income and senior citizen units had been excepted from the effect of Prop. A in the period between 1987 and 1990, and they were all in the senior category (plus an additional 28 units in the similar fourplex category). Based on those figures, the trial court made an estimate that some 20,000 persons would be denied affordable housing based on the effect of Prop. A. The trial court expressly found that the exception allowing higher priced units on 10,000-square-foot lots was not justified by an asserted imbalance in the current housing stock, such that higher end units were needed in the City. As explained above, the trial court's overall conclusion on the degree of the City's compliance with its obligations to provide low and moderate income regional share units was that its 1986-1991 compliance was not enough, as the court had to go on to the 1991-1996 period to make a complete analysis, and the City could not show it would meet the regional share for low/moderate income units for that period.

In phase III, the court further ruled that the evidence showed developers were discouraged from proposing housing that qualified for density bonuses, due to the exclusion of density bonus units from the low-income and senior exception. This exclusion, it held, conflicted with the state density bonus law. However, it declined to invalidate the ordinance on that ground, using the theory that section 65915 had not been shown to have been triggered by a developer's specific proposal of a density bonus project.

The three cited Government Code sections, taken together, clearly show an important state policy to promote the construction of low income housing and to remove impediments to the same. Prop. A is such an impediment, and cannot survive such a conflict. Although the City has made efforts to implement a policy to mitigate the harshness of the exemption of density bonus projects from the exception in Prop. A for low-income or senior citizen housing, those efforts do not change the text of the initiative measure. We also find the trial court's hypertechnical reasoning that the density bonus law had not been triggered as of the time of trial to be unpersuasive in light of the clear facial conflict between Prop. A and the density bonus law, and the very low number of low-income or senior citizen allocations which were made between 1987 and 1990.

Similarly, when the trial court made its finding that Prop. A did not offend against the provisions of section 65008, subdivision (c), prohibiting discrimination against low income housing, because Prop. A was neutral as to the intended occupants of any development, it disregarded its own factual findings that affordable housing had taken a dramatic decline since the effective date of Prop. A. Prop. A acts to favor the development of larger units on larger lots, with more design and other amenities, through the exemption for 10,000-square-foot lots. The City admitted as much, while asserting that such an exemption was needed in order to remedy an imbalance in the current housing supply. The trial court was not persuaded by that argument in phase I, and we are not persuaded by it at all.

Finally, with respect to the asserted conflict between Prop. A and section 65913.1, the trial court ruled that Prop. A did not affect, change, or prohibit rezoning, even though the prior opinion declined to disturb the trial court's ruling that Prop. A was a zoning ordinance for some purposes. Lesher characterized a growth control ordinance regulating land use as "resembling" a zoning ordinance, and treated it like one. ( Lesher, supra, 52 Cal. 3d at p. 541.) We fail to see how Prop. A, as a zoning ordinance, can be said to comply with standards contributing to the economic feasibility of producing the lowest possible cost housing, in light of the limited exceptions to the allocation system of Prop. A for low-income or senior housing. Since the trial court found the City was unlikely to meet its low-income housing goals through the year 2000, as a result of Prop. A, a facial conflict with section 65913.1 and Prop. A is apparent.

E Conclusion

Further proceedings to develop additional factual background on the conflict issue are unnecessary. (7) Where there is a lack of substantial evidentiary dispute about the facts underlying a question of statutory interpretation, the proper interpretation of statutory language is a question of law which an appellate court may review de novo, independent of the trial court's ruling or rationale. ( Los Angeles County Safety Police Assn. v. County of Los Angeles (1987) 192 Cal.App.3d 1378, 1384 [ 237 Cal.Rptr. 920].) (5e) The facial conflict of Prop. A with these three sections stating state housing policy may be determined as a matter of law. (See § 65580, subd. (a), making the legislative finding as to housing elements that "[t]he availability of housing is of vital statewide importance, and the early attainment of decent housing and a suitable living environment for every California family is a priority of the highest order"; also, the Legislature recognized that cooperation between government and the private sector would be necessary to accommodate the housing needs of Californians of all economic levels. (§ 65580, subd. (b).)) Prop. A must be considered invalid as of the date of its adoption.

We find support for our conclusion that Prop. A is in conflict with state housing policy as expressed in the state Planning and Zoning Law (sections as cited) in Livermore, supra, 18 Cal.3d at page 602. There, the Supreme Court distinguished the growth control ordinance which it was declining to strike down by way of strict judicial scrutiny (of whether the municipality had a compelling interest in the subject of the ordinance) from those exclusionary land use ordinances "which ban or limit less expensive forms of housing while permitting expensive single family residences on large lots." ( Ibid.) The Supreme Court explained "[t]he Livermore ordinance is not made from this mold; it impartially bans all residential construction, expensive or inexpensive." ( Ibid.) The City's Prop. A does not have the same across-the-board effect, and Livermore does not permit that we uphold it. We accordingly reverse the judgment with directions to enter a different judgment in accordance with the principles of this opinion.

DISPOSITION

The judgment is reversed with directions to enter judgment in favor of BIA on the grounds that Prop. A impermissibly conflicts with the City's general plan and with Government Code sections 65008, 65913.1, and 65915. Each party to bear its own costs.

Froehlich, J., and Nares, J., concurred.

A petition for a rehearing was denied August 16, 1994, and respondent's petition for review by the Supreme Court was denied October 13, 1994. Lucas, C.J., was of the opinion that the petition should be granted.