In Re Petition for Substantive Certification Filed by the Township of Warren

622 A.2d 1257, 132 N.J. 1, 1993 N.J. LEXIS 57

• Court: Supreme Court of New Jersey
• Decided: April 1, 1993
• Status: Published

Opinions

The opinion of the Court was delivered by

STEIN, J.

These appeals concern the validity of a regulation adopted by the Council on Affordable Housing (COAH) authorizing municiparities that seek substantive certification of regional fair-share plans for low- and moderate-income housing to make available an occupancy preference for fifty percent of such housing to income-eligible households that reside or work in the municipality.

Pursuant to N.J.S.A. 52:27D-314 of the Fair Housing Act (L. 1985, c. 222), COAH granted substantive certification to the housing-element and fair-share plans filed by the Townships of Denville, Hillsborough, Holmdel, and Warren, and the Boroughs of Bloomingdale and Roseland. The Public Advocate of New Jersey (Public Advocate) appealed all of COAH’s orders, and the Appellate Division affirmed the grant of substantive certification to each of the six municipalities. Three of the Appellate Division decisions are reported: In re Petition for Substantive Certification Filed by the Township of Warren, 247 N.J.Super. 146, 588 A.2d 1227 (1991); In re Petition for Substantive Certification Filed by the Township of Denville, 247 N.J.Super. 186, 588. A.2d 1248 (1991); In re Petition for Substantive Certification Filed by the Borough of Roseland, 247 N.J.Super. 203, 588 A.2d 1256 (1991). In each of the six appeals, Judge Shebell dissented from that portion of the court’s opinion upholding the validity of the fifty-percent-occupancy preference, which had been incorporated in each of the municipality’s fair-share plans in accordance with COAH’s authorizing regulation, N.J.A.C. 5:92-15.1. In all six cases the Public Advocate appeals to this Court as of right, Rule 2:2-1, as do the Morris County Fair Housing Council arid the Morris County branch of the NAACP but only in respect of In re Township of Denville, supra, 247 N.J.Super. 186, 588 A.2d 1248.

I

The six municipal parties to this appeal had been defendants in actions instituted in the Law Division, prior to adoption of the Pair Housing Act, N.J.S.A. 52:27D-301 to -329, in which their zoning ordinances were challenged as violative of the Mount Laurel doctrine for their failure to provide a reasonable opportunity for the construction of housing affordable to lower-income households. See Southern Burlington County NAACP v. Township of Mount Laurel, 67 N.J. 151, 336 A.2d 713 (Mount Laurel I), appeal dismissed and cert, denied, 423 U.S. 808, 96 S.Ct. 18, 46 L.Ed.2d 28 (1975), and 92 N.J. 158, 456 A.2d 390 (1983) (Mount Laurel II). After adoption of the Fair Housing Act, the cases were transferred to COAH pursuant to N.J.S.A. 52:27D-316, and the municipalities filed housing elements and fair-share housing plans with COAH. See N.J.S.A. 52:27D-310, 311. The municipalities’ requests to transfer the litigation to COAH from the Law Division were treated by COAH as equivalent to petitions for substantive certification. See Hills Dev. Co. v. Township of Bernards, 103 N.J. 1, 38 n. 10, 510 A.2d 621 (1986). The Public Advocate and other parties filed objections to the municipal plans.¹ With respect to each of the six municipalities, the Public Advocate objected to the fifty-percent-occupancy preference for residents and workers, emphasizing that the percentage of minority residents in each of the municipalities was substantially smaller than the percentage of minority residents in their related housing regions. In support of that assertion, the Public Advocate presented to COAH the following percentage comparisons of minority (African-American and Hispanic) households for each of the municipalities and their respective fair-share regions, based on 1980 census data:

[[Image here]]

The Public Advocate contended that the occupancy preference constituted a racially-discriminatory standard that perpetuated exclusionary zoning in violation of the New Jersey Constitution, article 1, par. 5; the Fair Housing Act; the federal Fair Housing Act, Title VIII of the Civil Rights Act of 1968, 42 U.S.C.A. §§ 3601-3631 (federal Fair Housing Act or Title VIII); and the Law Against Discrimination (LAD), N.J.S.A. 10:5-1 to -42.

A mediation and review process ensued, see N.J.S.A. 52:27D-315, in the course of which some of the technical objections to the various fair-housing plans were resolved. The mediators uniformly declined to address the challenges to the occupancy preferences, viewing the Public Advocate’s position as tantamount to an assertion that COAH’s regulation authorizing the occupancy preference was invalid. On the basis that issues that had been submitted to mediation remained unresolved, the Public Advocate requested that the occupancy-preference issue be transferred to the Office of Administrative Law as a contested case. See N.J.S.A. 52:27D-315(c). COAH rejected the Public Advocate’s request for an adjudicative hearing, and proceeded to adopt resolutions granting substantive certification to the housing elements and fair-share plans of each of the municipalities. In the course of its deliberations concerning the Holmdel petition, COAH issued an opinion that summarily addressed the various challenges to the validity of the occupancy preference:

To the extent that [the Public Advocate’s] concerns [about the occupancy-preference] arise out of alleged racial disparities between municipalities, such actions are beyond the Council’s mandate as set forth in the Pair Housing Act, and the Council concludes that it is thus an inappropriate forum in which to seek to redress such grievances.

The Appellate Division comprehensively addressed the validity of the occupancy preference only in its Warren Township opinion, supra, 247 N.J.Super. 146, 588 A.2d 1227. The court categorized the Public Advocate’s contentions as constituting a facial challenge to the validity of N.J.A. C. 5:92-15.1, the occupancy-preference regulation. Id. at 157, 588 A.2d 1227. That regulation provides:

For all low and moderate income housing units provided in inclusionary developments, municipalities shall establish occupancy such that initially, no more than 50 percent of the units are made available to income eligible households that reside in the municipality or work in the municipality and reside elsewhere.

The Appellate Division acknowledged that COAH had refused the requested referral of the occupancy preference’s validity to the Office of Administrative Law, but observed that no adjudicative hearing was required because no material factual data were in dispute. Id. at 159-60, 588 A.2d 1227. The court upheld the validity of the occupancy preference as an appropriate exercise of COAH’s rulemaking power under the Fair Housing Act. Id. at 170-74, 588 A.2d 1227. The court rejected the contention that the occupancy preference violated either the federal or state constitutions based on the absence of any allegation that the occupancy preference was intentionally discriminatory. Id. at 175, 588 A.2d 1227. The Appellate Division also concluded that even if the occupancy preference had a disparate racial impact, it would not violate federal or state statutes because it furthered legitimate governmental interests. Id. at 176-77, 588 A.2d 1227. The dissenting member concluded that the occupancy preference violated the constitutionally-protected right to travel, id. at 183-86, 588 A.2d 1227, also noting that the preference lacked statutory authority, id. at 183, 588 A.2d 1227, and “disserv[ed] the purposes of our Mount Laurel holdings.” Id. at 186, 588 A.2d 1227.

In its Warren Township opinion, the Appellate Division also addressed and sustained the validity of Regional Contribution Agreements (RCAs), expressly authorized by the Fair Housing Act, N.J.S.A. 52:27D-312a, permitting two municipalities to contract for the transfer by one to the other of up to fifty percent of the transferor’s regional fair share of low- and moderate-income housing. The court concluded that such agreements violated neither the Mount Laurel doctrine nor constitutional and statutory prohibitions against racial discrimination. Id. at 162-70, 588 A.2d 1227. The court also rejected the Public Advocate’s contention that COAH’s regulations authorizing municipal fair-share plans violate the Mount Laurel doctrine because of their failure to provide housing units affordable to households earning less than forty percent of the region’s median income. Id. at 179-83, 588 A.2d 1227. The Public Advocate filed petitions for certification in all six cases seeking review of the Appellate Division’s determinations concerning the validity of the RCAs and the validity of COAH’s regulations that omit any requirement for housing affordable to households earning less than forty percent of the region’s median income. We denied the petitions for certification, 127 N.J. 557, 606 A.2d 369 (1992).

II

A. The Mount Laurel Doctrine and the Fair Housing Act.

On an intuitive level, the arguments offered to support the occupancy preference appear to be as plausible as those presented against it. Proponents contend that municipalities that satisfy their regional fair-share obligation to provide affordable housing should be permitted to make available some of that housing to eligible residents and workers in that municipality, thereby addressing the needs of households with an existing connection to the community and increasing local support for the governing body’s decision to seek COAH certification for a fair-housing plan that satisfies the municipality’s regional fair share of low- and moderate-income housing. Opponents assert that the exclusionary aspects of the occupancy preference could not conceivably be compatible with the statutory scheme of the Fair Housing Act and the methodology of its implementing regulations adopted, in part, to ameliorate the effects of decades of exclusionary zoning. Although intuition may enlighten our analysis of the validity of the occupancy preference, we gain deeper and more reliable insight from the origins and evolution of the Mount Laurel doctrine and the passage of the Fair Housing Act.

In Mount Laurel I, Justice Hall pointedly framed the legal issue before the Court:

[W]hether a developing municipality like Mount Laurel may validly, ■ by a system of land use regulation, make it physically and economically impossible to provide low and moderate income housing in the municipality for the various categories of persons who need and want it and thereby * * * exclude such people from living within its confines because of the limited extent of their income and resources.

[67 N.J. at 173, 336 A.2d 713.]

The Court observed that “the effect of Mount Laurel’s land use regulation has been to prevent various categories of persons from living in the township because of the limited extent of their income and resources.” Id. at 159, 336 A.2d 713. The Court noted that plaintiffs, representing poor minorities not currently residing in Mount Laurel, were not the only category of citizens excluded because of restrictive zoning:

We have reference to young and elderly couples, single persons and large, growing families not in the poverty class, but who still cannot afford the only kinds of housing realistically permitted in most places — relatively high-priced, single-family detached dwellings on sizeable lots and, in some municipalities, expensive apartments. [Ibid.]

The Court also acknowledged and accepted the representation of counsel for Mount Laurel that the Township was not motivated by “any desire or intent to exclude prospective residents on the obviously illegal basis of race, origin, or believed social incompatibility.” Ibid. As the trial court had noted, paraphrasing the then-Mayor of the Township, “when a discussion arose, as to low income housing * * * it was the intention of the township committee to take care of the people of Mount Laurel Township but not make any area of Mount Laurel a home for the county.” Southern Burlington County NAACP v. Township of Mount Laurel, 119 N.J. Super. 164, 169, 290 A.2d 465 (Law Div.1972). This Court rejected the view that Mount Laurel could discharge its zoning responsibility by providing housing for its own poor only, without addressing the housing needs of low- and moderate- income families within the region who might wish to reside there:

[T]he universal and constant need for such housing is so important and of such broad public interest that the general welfare which developing municipalities like Mount Laurel must consider extends beyond their boundaries and cannot be parochially confined to the claimed good of the particular municipality. It has to follow that, broadly speaking, the presumptive obligation arises for each such municipality affirmatively to plan and provide, by its land use regulations, the reasonable opportunity for an appropriate variety and choice of housing, including, of course, low and moderate cost housing, to meet the needs, desires and resources of all categories of people who may desire to live within its boundaries.

[67 N.J. at 179, 336 A.2d 713.]

[A] developing municipality’s obligation to afford the opportunity for decent and adequate low and moderate income housing extends at least to “ * * * the municipality’s fair share of the present and prospective regional need therefor.”

[Id. at 188, 336 A.2d 713.]

Eight years later in Mount Laurel II, this Court reaffirmed the constitutional underpinning of the Mount Laurel doctrine, and emphasized that the lawful exercise of the zoning power compels municipalities to address not only their own needs but regional needs as well:

When the exercise of that power by a municipality affects something as fundamental as housing, the general welfare includes more than the welfare of that municipality and its citizens: it also includes the general welfare — in this case the housing needs — of those residing outside of the municipality but within the region that contributes to the housing demand within the municipality. Municipal land use regulations that conflict with the general welfare thus defined abuse the police power and are unconstitutional. In particular, those regulations that do not provide the requisite opportunity for a fair share of the region’s need for low and moderate income housing conflict with the general welfare and violate the state constitutional requirements of substantive due process and equal protection.

[92 N.J. at 208-09, 456 A.2d 390.]

In Mount Laurel II, we also considered and rejected the adequacy of the zoning-ordinance amendment that had been adopted by Mount Laurel Township in response to our 1975 decision. We noted that the Township had calculated its indigenous need, based on the number of deteriorated or dilapidated housing units in the Township, to be 103 units, and had calculated its fair share of the prospective regional housing need to be 515 units. The zoning-ordinance amendments adopted by the Township would have permitted construction of 131 units of low- and moderate-income housing, only twenty-eight units more than the number required to meet the Township’s indigenous needs. Id. at 299-300, 456 A.2d 390. We concluded that the Township’s revised ordinance “fails completely to comply with the mandate of Mount Laurel I,” id. at 302, 456 A.2d 390, that even if the entire 131 units of low- and moderate-income housing were to be built, that number “would fall far short of Mount Laurel’s fair share of the prospective regional low income housing need * * *.” Id. at 302-03, 456 A.2d 390.

In the Fair Housing Act, the Legislature expressly acknowledged the constitutional obligation of every growth-area municipality “to provide through its land use regulations a realistic opportunity for a fair share of its region’s present and prospective needs for housing for low and moderate income families,” N.J.S.A. 52:27D-302a, and declared the statutory scheme of the Act to be one that “comprehends a low and moderate income housing planning and financing mechanism in accordance with regional considerations and sound planning concepts which satisfies the constitutional obligation enunciated by the Supreme Court.” N.J.S.A. 52:27D-303. The clear and recurring theme of the Fair Housing Act is its recognition and implementation of the requirement that municipalities must provide through their zoning ordinance a realistic opportunity to satisfy their fair share of their region’s present and prospective need for low- and moderate-income housing. N.J.S.A. 52:27D-302a, d, e; -311a, -314a, b. Although the Fair Housing Act recognizes the relevance of sound planning concepts, N.J.S.A. 52:27D-303, and acknowledges the importance of encouraging housing construction in urban areas, specifically authorizing Regional Contribution Agreements that permit municipalities to transfer up to fifty percent of their fair share obligation to another municipality “in accordance with sound, comprehensive regional planning,” N.J.S.A. 52:27D-312a, c, the Legislature nevertheless determined that “the provision of housing in urban areas must be balanced with the need to provide housing throughout the State for the free mobility of citizens.” N.J.S.A. 52:27D-302g.

B. Calculation of Municipal Fair Share of Present and Prospective Need.

The Fair Housing Act itself is silent with respect to how and to what extent a municipality may address needs relating to its own residents. The Act does not expressly authorize municipalities to establish occupancy preferences for their own residents and workers in respect of low- and moderate-income housing constructed to meet its fair share of the region's needs. The Public Advocate stresses that under the regulations implementing the Fair Housing Act, COAH's basic methodology for calculating each municipality’s fair share of its region’s need for low- and moderate-income housing cannot be reconciled with an occupancy preference for residents and workers.

The Public Advocate emphasizes that the data from which those municipalities’ fair shares were derived are primarily regional rather than local and, as a result, the occupancy preference provides low- and moderate-income housing on a preferential basis to households that did not comprise the data base from which municipal fair share was calculated. That contention invites a searching examination and explanation of COAH’s methodology for calculating regional need and municipal fair share.

COAH’s methodology is substantially similar to that used by Judge Serpentelli in AMG Realty Co. v. Township of Warren, 207 N.J.Super. 388, 398-410, 504 A.2d 692 (Law Div.1984). Regional need is composed of a combination of present need and prospective need. COAH uses a four-factor formula to allocate the fair share of prospective regional need among municipalities in the region and applies three of those four factors in its fair-share allocation of non-indigenous present need among those same municipalities.

1. Present Need.

A municipality’s present need for low- and moderate-income housing (present need) is the sum of its indigenous need and its reallocated present need. N.J.A.C. 5:92-5.5. Indigenous need in a municipality is either the actual number of deficient housing units occupied by low- and moderate-income households or, in municipalities in which the proportion of deteriorated low- and moderate-income housing units exceeds the regional percentage, a reduced number that bears the same proportion to the total number of housing units in the municipality as the number of deteriorated low- and moderate-income housing units in the region bears to the total housing units in the region. N.J.A.C. 5:92-5.2; N.J.A.C. 5:92 — Appendix A at 92-41 to 43. In those municipalities whose deficient housing units, as a percentage of all housing units, exceeds the regional average, the excess deficient-housing units are accumulated in a regional pool and reallocated to all municipalities in the region’s growth area, except for certain “Urban Aid Cities” designated in the Technical Appendix to COAH’s regulations. N.J.A.C. 5:92-5.4; 5:92 — Appendix A at 92-43 to 44 and 92-56, A municipality’s allocated fair share of the regional pool of such excess deficient low- and moderate-income housing units constitutes that municipality’s reallocated present need. N.J.A.C. 5:92-5.4; 5:92— Appendix A at 92-43 to 44.

We note that the term indigenous need, as calculated by COAH, is somewhat of a misnomer in that it includes only housing units that require upgrading but does not include households in the municipality that demonstrate a present need for affordable housing. Such households may reside in standard housing units, paying a disproportionate share of their income for housing, or because of poverty may be forced to share housing with their extended families. In AMG Realty Co., supra, Judge Serpentelli also excluded that category of financially-needy households from the calculation of Warren Township’s present need, observing that inclusion of those households would increase dramatically the municipality’s obligation to provide affordable housing sufficient to address its indigenous need, and noting the difficulty of calculating accurately the number of households that should be included in the category of financial need. 207 N.J.Super. at 388, 504 A.2d 692. Notwithstanding the methodology adopted by the Law Division in AMG Realty Co., the Fair Housing Act vests in COAH the responsibility for determining whether identifiable financially-needy households are to be considered in the calculation of indigenous or regional need for affordable housing.

2. Prospective Need.

The Substantive Rules of the Council on Affordable Housing, which became effective in August 1986, see 18 N.J.R. 1124(b), 18 N.J.R. 1527(a), and expire in February 1996, provide that the prospective regional need for low- and moderate-income housing be calculated primarily on the basis of the estimated regional population growth from 1987, the base year for determining present need, to 1993. The population figures are divided into eight age groups or cohorts: Less than twenty-five years; twenty-five to twenty-nine years; thirty to thirty-four years; thirty-five to forty-four years; forty-five to fifty-four years; fifty-five to sixty-four years; sixty-five to seventy-four years; and seventy-five years and over. Using 1980 county-based statistics, the population growth projections are converted into estimates of growth in households, based on “headship” rates, which predict the expected percentages of household formation in each age group. See AMG Realty Co., supra, 207 N.J.Super. at 403, 504 A.2d 692. After the projected number of households in each age cohort has been calculated for both 1987 and 1993, statewide 1980-based income statistics are used to project the percentage of households in each age cohort that will qualify for low- and moderate-income housing. The difference between the projected number of low- and moderate-income households in each age cohort for 1987, as compared with 1993, constitutes the projected county-wide need by age group for low- and moderate-income housing units. The county-wide figures are combined in order to arrive at the prospective need for each of the State’s six regions. N.J.A. C. 5:92-5.6; 5:92 — Appendix A at 92-44 to 92-46.

County-wide data constitute the statistical basis from which each region’s prospective need for low- and moderate-income housing has been calculated. Because prospective need is defined as the difference between the projected number of low- and moderate-income households in 1987 and 1993, and COAH’s regulations became effective in 1986, the calculation of the estimated number of 1987 and 1993 low- and moderate-income households necessarily relied' in part on statistical projections. The source data include the following studies: State of New Jersey, Dep’t of Labor, Div. of Planning and Research, Office of Demographic and Economic Analysis, Population Estimates for New Jersey, July 1, 1984 (Trenton, NJ: Div. of Planning and Research, Sept. 1985) (1984 Population Estimates); State of New Jersey, Department of Health, New Jersey State and County Population Estimates by Age, Sex, and Race (Trenton, NJ: Center for Health Statistics, Oct. 1985) (1985 Population Estimates); State of New Jersey, Dep’t of Labor, Div. of Planning and Research, Office of Demographic and Economic Analysis, Population Projections — New Jersey and Counties: 1990 to 2020 (Trenton, NJ: Div. of Planning and Research, November 1985) (1990-2020 Population Projections). See N.J.A. C. 5:92 — Appendix A at 92-44, -45, and -58.

The 1984 Population Estimates rely on actual 1980 census data for municipalities and counties, and include provisional estimates as of July 1, 1984, for both municipalities and counties, but without breakdowns by age cohorts. The 1985 population estimates are county-wide only, but include age-group breakdowns. The 1990-2020 Population Projections include population projections at five-year intervals on a county-wide basis only, and include breakdowns by age groups. (COAH’s projections for 1993 are based on three-fifths of the difference between the 1990 and 1995 county-wide population projections.) The calculation of projected household formations is also based on county-specific headship rates. N.J.A.C. 5:92 — Appendix A at 92-45. Examination of the specific underlying data .used by COAH leads inescapably to the conclusion that the calculation of each region’s prospective need, based on anticipated population growth and household formation by age cohort from 1987 to 1993, represents the cumulative county-wide population and household projections within each region. The underlying data do not contain municipal-population projections beyond 1984.

3. Allocation of Projected Need and Reallocated Present Need.

Each municipality’s share of the region’s prospective need for low- and moderate-income housing is based on four factors, the first two of which relate to municipal need for and responsibility to provide affordable housing, and the latter two relate to municipal capacity to provide such housing. Although all four factors, weighted equally, determine the allocation of prospective need, only the first three factors are applied in distributing reallocated present need. N.J.A.C. 5:92-5.3; 5.4.

The allocation factors are the following:

(a) Regressed annual covered employment change within a municipality from 1977-1984, as a percentage of regional regressed annual covered employment change for the same period. (Covered employment refers to employees covered by the New Jersey Unemployment Compensation Law, L.1936, c. 270, as amended. For an explanation of the difference between an arithmetic and a regression measurement of employment growth, see AMG Realty Co., supra, 207 N.J.Super, at 441-42, 504 A.2d 692.);

(b) Covered employment in a municipality as a percentage of regional covered employment (1984);

(c) Municipal land located in growth areas as a percentage of growth area in the region;

(d) Municipal 1983/1984 aggregate per-capita income as a percentage of 1983/1984 regional aggregate per-capita income. N.J.A.C. 5:92 — Appendix A at 92-46.

We note that fifty percent of the weight accorded to the prospective-need-allocation factors derive from the municipality’s proportionate share of the region’s employment and the municipality’s proportionate share of the region’s employment growth. That emphasis on employment recognizes that an important objective of the Mount Laurel doctrine is to provide workers with housing in the vicinity of their place of employment. Mount Laurel II, supra, 92 N.J. at 256, 456 A.2d 390; AMG Realty Co., supra, 207 N.J.Super, at 412-14, 504 A.2d 692. Counter-balancing the weight accorded employment in the allocation formula, however, are the municipality’s physical capacity to accommodate, new housing units — as reflected by the growth-area factor — and its financial capacity to absorb infrastructure costs incidental to high-density development— measured by the comparison between the municipality’s and the region’s per-capita income. By according a municipality’s physical and financial capacity to accommodate affordable housing a significance equal to that assigned to the demand for housing generated by local employment, CO AH has developed an allocation formula that avoids concentrating affordable housing in less affluent industrial communities in favor of a formula that encourages the construction of affordable housing in communities with the financial capacity and adequate growth-area land to absorb high-density housing. See AMG Realty Co., supra, 207 N.J.Super. at 433-34, 504 A.2d 692.

When we compare the method of calculation with the method of allocation, we discern that COAH deems housing demand generated by municipal employment to be more relevant for purposes of allocating prospective regional need than it is for purposes of calculating the extent of that need, presumably on the basis that workers are likely to reside in the region and therefore have already been taken into account by the population projections used to forecast prospective need. That relevanee is diluted, however, by COAH’s decision to accord equal weight to a municipality’s physical and fiscal capacity to accommodate affordable housing.

C. Federal and State Anti-Discrimination Laws.

The Public Advocate also contends that the occupancy preference, because of the disproportionately-low minority-resident and-worker population in the six municipalities, has the effect of favoring eligible white households as occupants of the newly-constructed affordable-housing units and virtually excluding minorities from the units eligible for the preference. The Public Advocate asserts that the occupancy preference’s discriminatory impact on minorities violates federal and state anti-discrimination statutes.

Using the Borough of Bloomingdale to illustrate its contentions, the Public Advocate notes that although African-Americans and Hispanics constitute only 1.5% of Bloomingdale’s population, they comprise 20.9% of the residential population of the Northeast region. In the aggregate, African-American and Hispanic households constitute 50.5% of the region’s low- and moderate-income housing population, and are approximately 3.7 times as likely as white households to be categorized as eligible for low- and moderate-income housing. Bloomingdale’s fair share of the region’s prospective and reallocated present need is 117 units, which is to be satisfied by rezoning land to permit the development of 118 units of low- and moderate-income housing. Fifty-nine units will be eligible for the occupancy preference. Because COAH did not conduct an evidentiary hearing on the issue, the record does not reveal the number of minority workers in Bloomingdale eligible for low- and moderate-income housing, but the Public Advocate asserts that virtually all workers in Bloomingdale are residents. Accordingly, the Public Advocate assumes that because of the extremely small number of African-American and Hispanic residents and workers in Bloomingdale, all or nearly all of the fifty-nine units eligible for the occupancy preference will be allocated to white households. The Public Advocate observes that if those units were not subject to an occupancy preference and were allocated to minority households, in the same proportion as such households are represented in the region’s eligible low- and moderate-income population (50.5%), approximately thirty of the fifty-nine units would be occupied by minority households. Based on the current average size of Bloomingdale households, minority occupancy of those thirty units would represent an increase of approximately 81% in Bloomingdale’s minority population.

The Appellate Division addressed the Public Advocate’s contentions in its Warren Township opinion, supra, 247 N.J.Super. at 174-77, 588 A.2d 1227, focusing primarily on whether the occupancy preference violated the federal Fair Housing Act, which prohibits discrimination in the sale or rental of housing “because of race, color, religion, sex, familial status or national origin.” 42 U.S.C.A. § 3604(a). The Appellate Division reasoned that

the federal Act does not deal with a preference for government sponsored housing which is extended on the basis of present residence or place of employment. Therefore, unless an occupancy preference for government sponsored housing is adopted as a subterfuge for discrimination on the basis of race or another ground proscribed by 42 U.S.C.A. § 3604(a), which is not alleged in this case, it does not violate the federal Fair Housing Act.

[247 NJ.Super. at 176, 588 A.2d 1227.]

Moreover, because the Appellate Division had determined that the municipalities had legitimate reasons for granting the occupancy preference, that court concluded that even if disparate racial impact had been established, the occupancy preference would violate neither the federal Fair Housing Act nor the Law Against Discrimination, N.J.S.A. 10:5-1 to -42. The Appellate Division observed that “the occupancy preference furthers a ‘legitimate, bona fide governmental interest’ and the Public Advocate has failed to suggest any alternative provision which ‘would serve that interest with less discriminatory effect.’ ” Id. at 177, 588 A.2d 1227 (quoting Huntington Branch, NAACP v. Town of Huntington, 844 F.2d 926, 936 (2d Cir.), affd, 488 US. 15, 109 S.Ct. 276, 102 L.Ed.2d 180 (1988)).

The Public Advocate contends that the Appellate Division’s application of the federal Fair Housing Act to the occupancy preference erroneously requires proof of a racially-discriminatory purpose, noting that proof of a racially-disparate impact is sufficient to establish a violation of the Act. Moreover, the Public Advocate asserts that the Appellate Division improperly relied on governmental justifications for the occupancy preference, observing that the alleged justifications had never been the subject of an evidentiary hearing and the Public Advocate had been given no forum in which to rebut their validity. The governmental justifications on which the Appellate Division relied were: (1) the interest of municipalities in providing affordable housing for existing residents who encounter financial downturns; (2) the desire to preserve a municipality’s social fabric by providing affordable housing for residents with roots in the community; (3) the desirability of encouraging adoption of fair-share plans likely to meet with approval by community residents, thereby promoting voluntary compliance with the Act. Warren Township, supra, 247 N.J.Super, at 172-73, 588 A.2d 1227. The Public Advocate contends that the governmental justifications cited by the Appellate Division are factually unsupported and legally insufficient.

We note the possible parallel or overlap concerning the governmental justifications that may be relevant to whether the occupancy preference results in impermissible discrimination and the factors that determine whether such preferences can be reconciled with the Fair Housing Act and Mount Laurel doctrine. However, our disposition of these appeals will focus essentially on the grounds implicating the State’s affordable-housing policy and will not rest primarily on the alleged violation of Title VIII. Further, the ultimate resolution of federal and State discrimination issues requires, in our view, a broader and more detailed record than that before the Court. Consequently, we choose not to delve extensively into the federal case law that has applied the provisions of Title VIII to a diverse variety of allegations of discrimination in the provision of housing. A summary of the prevailing principles, however, is instructive.

The cases interpreting Title VIII uniformly hold that a facially-neutral law or policy that results in a discriminatory effect on the sale or rental of housing will establish a prima facie violation of Title VIII, even if unaccompanied by evidence of discriminatory intent. See, e.g., Familystyle of St. Paul v. City of St. Paul, Minn., 923 F.2d 91, 94 (8th Cir.1991); Edwards v. Johnston County Health Dep't 885 F.2d 1215, 1223 (4th Cir.1989); Huntington Branch NAACP, supra, 844 F.2d at 933-36; Betsey v. Turtle Creek Assocs., 736 F.2d 983, 986-87 (4th Cir.1984); Smith v. Town of Clarkton, N.C., 682 F.2d 1055, 1065 (4th Cir.1982); Robinson v. 12 Lofts Realty, Inc., 610 F.2d 1032, 1036-38 (2d Cir.1979); Resident Advisory Bd. v. Rizzo, 564 F.2d 126,146-48 (3d Cir.1977), cert, denied sub nom. Whitman Area Improvement Council v. Resident Advisory Bd., 435 U.S. 908, 98 S.Ct. 1457, 55 L.Ed.2d 499 (1978); Metropolitan Hous. Dev. Corp. v. Arlington Heights, 558 F.2d 1283, 1288-90 (7th Cir.1977), cert, denied, 434 U.S. 1025, 98 S.Ct. 752, 54 L.Ed.2d 772 (1978); United States v. City of Black Jack, Mo., 508 F.2d 1179, 1184-85 (8th Cir.1974), cert, denied, 422 U.S. 1042, 95 S.Ct. 2656, 45 L.Ed.2d 694 (1975).

The Second Circuit’s opinion in Huntington Branch, supra, appears to reflect the dominant view of the prevailing standards of proof in Title VIII cases. ■ The Huntington Branch of the National Association for the Advancement of Colored People (NAACP), Housing Help, Inc. (HHI), and two African-American residents of Huntington had instituted suit alleging that Huntington had violated Title VIII by restricting private construction of multi-family housing to a narrow urban-renewal area in which fifty-two percent of the residents were minority, and by refusing to rezone a parcel in a virtually all-white neighborhood where plaintiffs wished to build multi-family housing. 844 F.2d at 928. HHI had intended to sponsor a racially-integrated project, and had determined that because ninety-five percent of Huntington’s residents were white, the project could achieve racial integration only if it were located in a white neighborhood. Id. at 929-30. The Town Board of Huntington rejected the rezoning request based on an alleged lack of public transportation, the potential creation of traffic hazards, and the threatened disruption of residential patterns in the neighborhood, and recommended that the federal Department of Housing and Urban Development reject HHI’s request for funding. Id. at 932.

Reversing the district court, which had applied an intent-based standard for the disparate impact claim, 668 F.Supp. 762, 786 (E.D.N.Y.1987), the Second Circuit relied on the legislative history of Title VIII, the parallel between Title VIII and Title VII of the Civil Rights Act of 1964, and what it termed “practical considerations” to conclude that proof of discriminatory impact alone can sustain a violation of Title VIII. 844 FI 2d at 934-35. Observing that “an intent requirement would strip the statute of all impact on de facto segregation,” the court noted that the Third Circuit in Rizzo, supra, had found significant the rejection by the Senate of an amendment requiring “proof of discriminatory intent to succeed in establishing a Title VIII claim.” 844 F.2d at 934 (quoting Rizzo, supra, 564 F. 2d at 147). The court also noted that both Title VIII and Title VII were "part of a coordinated scheme of federal civil rights laws enacted to end discrimination * * 844 F.2d at 935 (citing Griggs v. Duke Power Co., 401 U.S. 424, 429-36, 91 S. Ct. 849, 852-56, 28 L.Ed.2d 158, 163-67 (1971), as authority for the principle that proof of discriminatory effect is sufficient to establish a prima facie violation of Title VII). 844 F.2d at 935. (Although the United States Supreme Court’s decision in Wards Cove Packing Co. v. Antonio, 490 U.S. 642,109 S.Ct. 2115, 104 L.Ed.2d 733 (1989), significantly eviscerated the holding in Griggs v. Duke Power Co., supra, the Civil Rights Act of 1991, Pub.L. No. 102-166, 105 Stat. 1071, overruled much of the diluting effect of Wards Cove, and section 105(a) of that Act amends Title VII specifically to provide that disparate impact discrimination is an “unlawful employment practice.” Cuello Suarez v. Puerto Rico Elec. Power Auth., 798 F.Supp. 876, 883 n. 5 (D.P.R.1992); see Charles Sullivan et ah, Special Release on the Civil Rights Act of 1991 22 (1992) (supplementing Employment Discrimination 2d ed.)). Finally, the Second Circuit expressed the practical concern that facially-neutral rules “bear no relation to discrimination upon passage, but develop into powerful discriminatory mechanisms when applied.” 844 F.2d at 935. The Eighth Circuit also acknowledged that concern in City of Black Jack, supra, in recognizing that

"the arbitrary quality of thoughtlessness can be as disastrous and unfair to private rights and the public interest as the perversity of a willful scheme.” [508 F.2d at 1185 (quoting Hobson v. Hansen, 269 F.Supp. 401, 497 (D.D.C. 1967), affd sub nom. Smuck v. Hobson, 408 F.2d 175 (D.C.Cir.1969) (en banc)).]

In its assessment of the strength of the plaintiffs’ proofs of discriminatory impact, the Huntington court distinguished between “adverse impact on a particular minority group and harm to the community generally by the perpetuation of segregation,” 844 F.2d at 937, and concluded that the Town’s refusal to amend its zoning ordinance to permit privately-built multifamily housing at the proposed site “significantly perpetuated segregation” in Huntington. Id. at 938. Having determined that the proof of discriminatory impact was sufficient to establish a prima facie case, the court adopted the Third Circuit’s formulation in Rizzo, supra, requiring defendant to prove “that its actions furthered * * * a legitimate, bona fide governmental interest and that no alternative would serve that interest with less discriminatory effect.” Id. at 936 (citing Rizzo, supra, 564 F.2d at 148-49). The court also expressed its agreement with the principle that in weighing the governmental justification against the discriminatory impact, “the balance should be more readily struck in favor of the plaintiff when it is seeking only to enjoin a municipal defendant from interfering with its own plans rather than attempting to compel the defendant itself to build housing.” Id. at 940 (citing Arlington Heights, supra, 558 F.2d at 1293). The court concluded that the strong demonstration of discriminatory impact substantially outweighed the Town’s justifications for refusing to rezone, id. at 940-41, and entered judgment directing the Town to rezone the plaintiff’s proposed site for multi-family housing.

The Public Advocate asserts that the occupancy preference also violated the LAD, which prohibits racial discrimination in the provision of housing. N.J.S.A. 10:5-4, -12. We often have observed that the LAD “ ‘is aimed at fulfilling provisions of the state constitution guaranteeing civil rights.’ ” Andersen v. Exxon Co., 89 N.J. 483, 492, 446 A.2d 486 (1982) (quoting Goodman v. London Metals Exch., 86 N.J. 19, 30-31, 429 A.2d 341 (1981)); accord Jones v. Haridor Realty Corp., 37 N.J. 384, 392-93, 181 A.2d 481 (1962); Levitt & Sons v. Division Against Discrimination, 31 N.J. 514, 524, 158 A.2d 177, appeal dismissed, 363 U.S. 418, 80 S.Ct. 1257, 4 L.Ed.2d 1515 (1960). Although our disposition of these appeals does not require that we decide the issue, our precedents persuasively suggest that proof of discriminatory impact alone, without proof of discriminatory intent, would be sufficient to establish a prima facie violation of the LAD. See, e.g., Countiss v. Trenton State College, 77 N.J. 590, 595, 392 A.2d 1205 (1978); Lige v. Montclair, 72 N.J. 5, 11, 367 A.2d 833 (1976); Giammario v. Trenton Bd. of Educ., 203 N.J.Super. 356, 361, 497 A.2d 199 (App.Div.), certif. denied, 102 N.J. 336, 508 A.2d 212 (1985), cert, denied, 475 U.S. 1141, 106 S.Ct. 1791, 90 L.Ed.2d 337 (1986); cf. United Bldg. & Constr. Trades Council v. Camden, 88 N.J. 317, 329, 443 A.2d 148 (1982) (upholding Camden requirement that 40% of all public works contract employees be city residents as consistent with objectives of LAD, but observing that underlying legislative purpose would have been frustrated if minority population of Camden were less than that of surrounding area), rev’d on other grounds, 465 U.S. 208, 104 S.Ct. 1020, 79 L.Ed.2d 249 (1984).

We also take note of the Public Advocate’s contention that the occupancy preference violates article 1, pars. 1 and 5 of the New Jersey Constitution, and the view of the dissenting member of the Appellate Division that the occupancy preference not only lacked statutory authority, 247 N.J.Super, at 183, 588 A.2d 1227, and “disserv[ed] the purposes of our Mount Laurel holdings,” id. at 186, 588 A.2d 1227, but also violated the fundamental freedom to travel protected under the federal constitution, id. at 183-86, 588 A.2d 1227 (citing Shapiro v. Thompson, 394 US. 618, 89 S.Ct. 1322, 22 L.Ed.2d 600 (1969)). In view, however, of our disposition of these appeals on other grounds, we decline to resolve or address those issues.

Ill

The long-standing and well-established principles governing judicial review of agency action require that we accord an administrative regulation a presumption of reasonableness and validity. In re Adoption of N.J.A.C. 7:26B, 128 N.J. 442, 449, 608 A.2d 288 (1992); Medical Soc’y of N.J. v. New Jersey Dep’t of Law and Pub. Safety, 120 N.J. 18, 25-26, 575 A.2d 1348 (1990); In re Amendment of N.J.A.C. 8:31B-3.31, 119 N.J. 531, 543, 575 A.2d 481 (1990); Smith v. Director, Div. of Taxation, 108 N.J. 19, 25, 527 A.2d 843 (1987); In re Barnert Memorial Hosp. Rates, 92 N.J. 31, 39, 455 A.2d 469 (1983); New Jersey Guild of Hearing Aid Dispensers v. Long, 75 N.J. 544, 561; 384 A.2d 795 (1978). “Our strong inclination, based on the principle that the coordinate branches of government should not encroach on each other’s responsibilities, is to defer to agency action that is consistent with the legislative grant of power.” Lower Main St. Assocs v. N.J. Hous. & Mortgage, 114 N.J. 226, 236, 553 A.2d 798 (1989); accord A.A. Mastrangelo, Inc. v. Department of Envtl. Protection, 90 N.J. 666, 687, 449 A.2d 516 (1982); New Jersey Guild of Hearing Aid Dispensers v. Long, supra, 75 N.J. at 562, 384 A.2d 795. As we observed in Williams v. Department of Human Services, “Courts * * * act only in those rare circumstances when it is clear that the agency action is inconsistent with the legislative mandate.” 116 N.J. 102, 108, 561 A.2d 244 (1989).

This principle of judicial deference to agency action is particularly well-suited to our review of administrative regulations adopted by COAH to implement the Fair Housing Act, “a new and innovative legislative response to deal with the statewide need for affordable housing.” Van Dalen v. Washington Township, 120 N.J. 234, 246, 576 A.2d 819 (1990). In sustaining the constitutionality of the Fair Housing Act, we described its unique attributes:

The Act that we review and sustain today represents a substantial effort by the other branches of government to vindicate the Mount Laurel constitutional obligation. This is not ordinary legislation. It deals with one of the most difficult constitutional, legal and social issues of our day — that of providing suitable and affordable housing for citizens of low and moderate income. In Mount Laurel II, we did not minimize the difficulty of this effort — we stressed only its paramount importance — and we do not minimize its difficulty today. But we believe that if the Act before us works in accordance with its expressed intent, it will assure a realistic opportunity for lower income housing in all those parts of the state where sensible planning calls for such housing.

[Hills Dev. Co., supra, 103 N.J. at 21, 510 A.2d 621.]

We also took note of the broad powers entrusted to COAH in implementing the statutory goals:

The Act creates an administrative agency (the Council on Affordable Housing) with power to define housing regions within the state and the regional need for low and moderate income housing, along with the power to promulgate criteria and guidelines to enable municipalities within each region to determine their fair share of that regional need. The Council is further empowered, on application, to decide whether proposed ordinances and related measures of a particular municipality will, if enacted, satisfy its Mount Laurel obligation, i.e., will they create a realistic opportunity for the construction of that municipality’s fair share of the regional need for low and moderate income housing. Southern Burlington County N.A.A.C.P. v. Mount Laurel, 92 N.J. 158, 208-09 [456 A.2d 390] (1983). The agency’s determination that the municipality’s Mount Laurel obligation has been satisfied will ordinarily amount to a final resolution of that issue; it can be set aside in court only by “clear and convincing evidence” to the contrary.

[Id. at 19-20, 510 A.2d 621.]

In recognition of the unprecedented goals of the Fair Housing Act, and COAH’s broad responsibility for implementing those goals, we recently observed that

[b]ecause the legislative scheme is novel, the implementation of its goals is necessarily an evolving process. Accordingly, COAH is entitled to a reasonable, degree of latitude, consistent with the legislative purpose, in its effort to ascertain which planning and statistical studies best serve the long-term statutory objectives.

[Van Dalen, supra, 120 N.J. at 246, 576 A.2d 819.]

The breadth of COAH’s discretion in selecting methodologies to implement the Fair Housing Act, however, does not dilute COAH’s duty to adopt regulatory methods that are consistent with the statutory goals.

In reviewing administration actions, the judicial role is ordinarily confined to three inquiries: (1) whether the agency’s action violates enabling acts, express or implied legislative policy; (2) whether there is substantial evidence and records to support the findings upon which the agency based application of the legislative policies; and (3) whether, in applying the legislative policies to the facts, the agency clearly erred by reaching a conclusion that could not reasonably have been made upon a showing of the relevant factors. Williams, supra, 116 N.J. at 108, 561 A.2d 244. These inquiries guide and inform our review of the occupancy-preference regulation.

Notwithstanding the deference to which COAH is entitled in adopting regulations to implement the expansive goals of the Fair Housing Act, we are unable to find on this record that the occupancy preference furthers those legislative policies. That incompatibility is most graphically evidenced by the fundamental inconsistencies between the effect of the occupancy preference and the detailed regulations that determine the regional obligation for affordable housing. COAH’s adoption of the occupancy preference, in the context of its existing methodology for calculating and allocating indigenous, present, and regional need for affordable housing, is invalid because it is inconsistent with and undermines the methodology adopted by COAH for calculating and allocating regional fair share of low- and moderate-income housing. In that framework the occupancy preference does not comport with the Fair Housing Act’s central purpose of providing affordable housing on a regional basis consistent with both sound planning concepts and the Mount Laurel doctrine. N.J.S.A. 52:27D-303. Further, as applied, and contingent on the development of an adequate record detailing the racial composition of non-resident workers in these six municipalities, the occupancy preference poses a significant risk of violating the anti-discrimination provisions of Title VIII of the Civil Rights Act of 1968 and the provisions of the New Jersey Law Against Discrimination. We find it particularly incongruous that a regulation promulgated to implement a statute designed to address the practice of exclusionary zoning would itself be challenged as violating federal and state anti-discrimination laws.

A. The Occupancy Preference and State Affordable-Housing Policy

The occupancy-preference regulation itself cannot be sensibly reconciled with the overall regulatory scheme that has been adopted by CO AH to implement the Fair Housing Act and therefore cannot coexist within the present regulatory framework. There is a fundamental inconsistency between the occupancy preference and COAH’s methodology for calculating regional fair share and allocating that fair share among the region’s municipalities. As noted, supra at 16-17, 622 A.2d at 1264-1265, regional prospective need for 1987-1993 is calculated primarily on the basis of county-wide population forecasts, county-wide age cohort statistics, county-based statistics estimating household formations, and statewide income statistics projecting the percentage of households in each age cohort that will be eligible for low- and moderate-income housing. The affordable-housing needs of municipal residents are reflected only in the calculation of substandard housing units, in order to determine each municipality’s indigenous need, and no party has challenged the principle that all housing units built or rehabilitated to satisfy indigenous present need are to be allocated to eligible residents. But to the extent that financially-needy local residents, currently living in standard housing units or otherwise not identified by COAH’s equating of indigenous need with substandard-housing stock, are eligible to occupy low- and moderate-income housing units, that pool of residents has not been included directly in the calculation of regional prospective need. Supra at 14-15, 622 A 2d at 1263-1264. To the contrary, COAH’s methodology for calculating such need relies almost exclusively on county-wide statistics. By implication, to the extent the occupancy preference favors local residents, the likelihood is that the housing needs of those who benefit from the preference were not considered in calculating the prospective regional need for affordable housing.

The inconsistency between the occupancy preference and COAH’s methodology may be illustrated simply by reference to Bloomingdale. That municipality’s regional fair-share obligation of 117 units is to be satisfied by construction of 118 new units, fifty-nine of which are reserved for local residents and workers. To the extent that the reserved units are occupied by financially-needy Bloomingdale residents not considered in the calculation of Bloomingdale’s fair share, the preference will defeat COAH’s formula by reducing below its fair-share obligation the number of housing units actually available in Bloomingdale to meet the regional need.

COAH’s current methodology also ignores eligible residents in its allocation formula but assigns fifty percent of the weight in the allocation formula to each municipality’s proportionate share of the region’s employment and its proportionate share of the region’s employment growth. Offsetting the weight accorded to employment-related data are the municipality’s growth-area land in proportion to that of the region, and the municipality’s per-eapita income as a percentage of the region’s per-capita income. The result is that COAH’s allocation methodology balances job-related housing demand with a municipality’s physical and financial capacity to absorb affordable housing. Supra at 17-19, 622 A 2d at 1265-1266.

In the aggregate, we conclude that the weight COAH has assigned to housing demand by eligible local residents and eligible local workers in calculating and allocating prospective regional need and reallocated present need does not justify a fifty-percent occupancy preference for eligible residents and workers. Although job-related housing demand is significant in the allocation formula, it is unidentified in the prospective-fair-share calculation. Nor has CO AH attempted to relate the grant of an occupancy preference for workers to its fair-share calculation or allocation methodology. That local workers are taken into account in the allocation to a municipality of its regional fair share neither compels nor justifies a regulation that accords a preference to both residents and workers for housing constructed to fulfill a regional need, although we acknowledge that a limited preference only for workers would be less vulnerable to challenge than the resident and worker preference in its present form.

Absent evidence of the agency’s rationale or attempted correlation between the occupancy preference and the fair-share methodology, we conclude that both the authorization and the extent- of the occupancy preference are inconsistent with COAH’s methodology for calculating regional fair share. Consequently, we conclude that the occupancy-preference regulation is invalid and cannot survive in its present context as a valid exercise of agency rule-making.

That conclusion indicates that the occupancy preference as currently adopted does not comport with the Fair Housing Act. The underlying theme of the Fair Housing Act is found in the Legislature’s acknowledgment that the zoning ordinance of every growth-area municipality must provide “a realistic opportunity for a fair share of its region’s present and prospective needs for housing for low and moderate income families.” N.J.S.A. 52:27D-302a. The Act is declared by the Legislature to “comprehend[ ] a low and moderate income housing planning and financing mechanism in accordance with regional considerations * * * which satisfies the constitutional obligation enunciated by the Supreme Court.” N.J.S.A. 52:27D-303. The evident focus of the Act is regional, implying that consideration and accommodation of local needs for affordable housing must be reconciled or integrated with the meeting of regional needs.

We also noted at the outset of this opinion, supra at 9-12, 622 A.2d at 1261-1262, that the Mount Laurel doctrine derived from this Court’s concern that a municipality like Mount Laurel, through its zoning ordinances, had made it “physically and economically impossible to provide low and moderate income housing in the municipality for the various categories of persons who need and want it * * * thereby * * * excluding] such people from living within its confines because of the limited extent of their income and resources.” Mount Laurel I, supra, 67 N.J. at 173, 336 A.2d 713. The Mount Laurel doctrine was intended to redress the widespread use of exclusionary zoning in New Jersey. To that end, this Court rejected the view, characterized by the Mount Laurel ordinance, that a municipality adequately discharged its zoning burden by providing housing only for its own low-income population. We held that “a developing municipality’s obligation to afford the opportunity for decent and adequate low and moderate income housing extends at least to * * * the municipality’s fair share of the present and prospective regional need therefor.” Id. at 188, 336 A.2d 713.

Now, seventeen years after this Court’s decision in Mount Laurel I, some 23,000 low- and moderate-income housing units have been scheduled for production and approximately 14,000 units have been completed. See Martha Lamar et al., Mount Laurel at Work: Affordable Housing in New Jersey, 1983-1988, 41 Rutgers L.Rev. 1197, 1209 (1989) (Mount Laurel at Work); Future Uncertain for Mount Laurel Affordable Housing, The Star-Ledger, Nov. 15, 1992, at 1. Although COAH has not maintained statistical records concerning occupants of existing Mount Laurel housing, the available information, although limited, suggests that such housing may not be serving regional needs. See Mount Laurel at Work, supra, 41 Rutgers L.Rev. at 1264. COAH’s occupancy-preference regulation may be a factor in the existing pattern of absorption of Mount Laurel housing.

The Mount Laurel doctrine is generally and correctly understood as prohibiting economically exclusionary zoning practices. See John M. Payne, Title VIII and Mount Laurel: Is Affordable Housing Fair Housing?, 6 Yale L. & Pol’y Rev. 361 (1988). Although the Court in Mount Laurel I acknowledged that the plaintiffs represented minorities that had been excluded from housing opportunities, we observed that the class affected by exclusionary zoning ordinances was much broader:

Plaintiffs represent the minority group poor (black and Hispanic) seeking such quarters. But they are not the only category of persons barred from so many municipalities by reason of restrictive land use regulations. * * * We will, therefore, consider the case from the wider viewpoint that the effect of Mount Laurel’s land use regulation has been to prevent various categories of persons from living in the township because of the limited extent of their income and resources.

[67 N.J. at 159, 336 A.2d 713 (footnotes omitted).]

However, our Mount Laurel jurisprudence permitted no room for doubt that the urban poor were of special significance among the class of persons disadvantaged by exclusionary zoning practices:

Much industry and retail business, and even the professions, have left the cities. Camden is a typical example. The testimonial and documentary evidence in this case as to what has happened to that city is depressing indeed. For various reasons, it lost thousands of jobs between 1950 and 1970, including more than half of its manufacturing jobs (a reduction from 43,267 to 20,671, while all jobs in the entire area lábor market increased from 94,507 to 197,037). A large segment of retail business faded away with the erection of large suburban shopping centers. The economically better situated city residents helped fill up the miles of sprawling new housing developments, not fully served by public transit. In a society which came to depend more and more on expensive individual motor vehicle transportation for all purposes, low income employees very frequently could not afford to reach outlying places of suitable employment and they certainly could not afford the permissible housing near such locations. These people have great difficulty in obtaining work and have been forced to remain in housing which is overcrowded, and has become more and more substandard and less and less tax productive. There has been a consequent critical erosion of the city tax base and inability to provide the amount and quality of those governmental services — education, health, police, fire, housing and the like — so necessary to the very existence of safe and decent city life. This category of city dwellers desperately needs much better housing and living conditions than is available to them now, both in a rehabilitated city and in outlying municipalities. They make up, along with the other classes of persons earlier mentioned who also cannot afford the only generally permitted housing in the developing municipalities, the acknowledged great demand for low and moderate income housing.

[Id. at 172-73, 336 A.2d 713.]

In assessing realistically the impact of the occupancy preference on the entire class of economically-disadvantaged persons affected by exclusionary zoning practices — including minorities and the urban poor — we take into account the viewpoint often expressed that the pool of suburban poverty substantially exceeds the projected numbers of low- and moderate-income households used in COAH’s fair-share formulas. See Payne, supra, 6 Yale L. & Pol’y Rev. at 368. Another commentator, expressing concern that eligible African-Americans will not benefit from the construction of Mount Laurel housing, speculates that because of the large number of eligible urban white households, “selection criteria left unmonitored could result in apparent significant success for the Mount Laurel mandate without accommodating a single black family.” Robert C. Holmes, A Black Perspective on Mount Laurel II: Toward a Black “Fair Share, ” 14 Seton Hall L.Rev. 944, 950 (1984).

In sum, the validity of the occupancy preference cannot be determined out of context, and the relevant context is both pragmatic and historical. The practical argument for sustaining the preference is that because the affordable housing constructed in a municipality serves the regional need, those eligible households whose members reside or work in the municipality should be given first choice, rather than be forced to move elsewhere. A related argument — but not one that can be based on the current occupancy preference — is that there may be persons among a municipality’s resident poor who have special needs or concerns pertaining to age, health, or responsibilities to dependents that may justify a local housing preference tailored specifically to such persons. The pragmatic argument against the preference is that as it addresses the housing needs of local residents and workers, it simultaneously excludes from even the chance to compete for the reserved units all eligible households in the region whose members, by virtue of their own poverty, are equally deserving of affordable housing, but neither reside nor are employed in the municipality. The region’s urban poor — white and minority — are among those excluded from applying for the affordable housing units subject to the preference in these six municipalities.

Those considerations indicate that the occupancy preference in its current form does not comport with the doctrine of Mount Laurel. Under prevailing zoning practices prior to Mount Laurel I and II, suburban municipalities frequently excluded the poor through ordinances that did not permit construction of affordable housing. In Mount Laurel I, we held that the State Constitution requires that local zoning ordinances permit construction within each municipality of affordable-housing units sufficient to provide not only for any local need, but also for a municipality’s fair share of the region’s need. That constitutional objective was ratified by the Fair Housing Act, and implemented by COAH’s fair-share methodology. Thus, the affordable-housing units constructed to address regional needs within these six municipalities reflect each municipality’s compliance with the constitutional and statutory prohibition against exclusionary zoning, because their evolution defines them as housing units built for the region’s poor, constituting the municipality’s fair share of the region’s need. In that historical context, the occupancy preference in its present form cannot apply to housing units built to comply with the Fair Housing Act and the Mount Laurel doctrine, because the preference without any standards excludes from eligibility for a portion of the municipality’s low- and moderate-income housing members of the class for whose benefit the obligation to construct that housing was established.

We understand and appreciate the concern of these and other municipalities that long-standing residents, because of economic circumstances, may be unable to continue to reside in their communities. Without any occupancy preference, such residents are eligible to compete on an equal footing for a municipality’s affordable-housing units constructed to satisfy its regional obligation. State affordable housing policy, however, does not or need not constitute an absolute bar against dealing with the affordable housing needs of local residents. A municipality. whose local need for affordable-housing units is not fulfilled by the construction of the units mandated by COAH’s methodology may, if feasible, zone for additional affordable housing and, subject to state and federal anti-discrimination laws, reserve such units to address exclusively local needs, provided that its regional obligation pursuant to the Fair Housing Act is not diluted. In addition, COAH is free to consider alternative means by which to recognize the affordable housing needs of eligible local residents. For example, if COAH’s methodology for calculating regional need took into account financially-needy households in the respective municipalities, COAH could consider authorization of an occupancy preference tailored to eligible local households with special needs, on the condition that such a preference did not frustrate compliance with a municipality’s regional obligation to provide affordable housing. Infra, at 40-41, 622 A.2d at 1277-1278. Those approaches, aside from their feasibility, would be consistent with the main principle of our affordable-housing policy with its focus on regional need because they would not diminish a municipality’s regional obligation. Considerations of alternative approaches to address local needs for affordable housing, however, impel us to emphasize that we do not rule today on the validity of other possible local preferences for lower-income residents, or on the extent to which the Fair Housing Act and the Mount Laurel doctrine would permit an occupancy preference tailored to meet genuine special and cognizable needs of residents otherwise eligible for affordable housing, irrespective of a municipality’s regional obligation. The complexities and varieties of that issue suggest caution in resolving such possibilities in the abstract. Accordingly, we decide only the invalidity of the preference before us.

B. The Occupancy Preference: Title VIII and the Law Against Discrimination.

Earlier in this opinion, supra at 19-20, 622 A.2d at 1266, we took note of the Public Advocate’s reliance on the data indicating that Bloomingdale’s minority population, as a percentage of its total population, was 1.5%, whereas its region’s minority population was 20.9%. Similarly compelling statistical data are before us comparing the minority population percentage in each of the other five municipalities with that of its region: Denville 1.1%/Region 52%; Hillsborough 2.8%/Region 8.9%; Holmdel 1.7%/Region 8.6%; Roseland 1%/Region 52%; Warren 1.3%/Region 8.9%. The Public Advocate contends that because the minority resident and-worker population in those municipalities is disproportionately low, compared to their respective regions, the occupancy preference has a discriminatory impact on minorities, in that it favors occupancy of affordable-housing units by eligible white households and perpetuates the pattern of minority exclusion that exists in those municipalities.

The Appellate Division expressed the view that Title VIII “does not deal with a preference for government sponsored housing * * * extended on the basis of present residence or place of employment.” 247 N.J.Super. at 176, 588 A.2d 1227. Accordingly, that court concluded that unless the occupancy preference was “adopted as a subterfuge for discrimination on the basis of race,” ibid., the Act is not violated. In addition, the Appellate Division adverted to the governmental justifications advanced to support the occupancy preference:

(1) the interest of municipalities in providing affordable housing for existing residents who encounter financial downturns;

(2) the desire to preserve a municipality's social fabric by providing affordable housing for residents with roots in the community;

(3) the desirability of encouraging adoption of fair-share plans likely to meet with approval by community residents, thereby promoting voluntary compliance with the Act.

[Warren Township, supra, 247 N.J.Super. at 172-173, 588 A.2d 1227.]

Based on those justifications, the Appellate Division held that the occupancy preference furthered a “ ‘legitimate, bona fide governmental interest,’ ” noting that the Public Advocate had failed to suggest any alternative provision that “ ‘would serve that interest with less discriminatory effect.' ” Id. at 177, 588 (quoting Huntington Branch, NAACP v. Town of Huntington, supra, 844 F.2d at 936).

We conclude that the Appellate Division erred in its summary disposition both of the Public Advocate’s Title VIII claims and the related claims under the LAD. Its holding that Title VIII does not deal with residential preferences for government-sponsored housing fails to take into account cases such as United States v. Housing Authority of Chickasaw, 504 F.Supp. 716 (S.D.Ala.1980). There, the district court invalidated defendant’s imposition of a residency requirement for its governmentally-sponsored housing units, noting that the City of Chickasaw had virtually no minority residents whereas the adjacent cities of Pritchard and Mobile had African-American populations of 50.5% and 32.3% respectively. Id. at 718. The court concluded that the residency requirement “perpetuates segregation,” and that because of its discriminatory impact it violated Title VIII even in the absence of proof of discriminatory intent. Id. at 732.

Moreover, the Appellate Division’s conclusion that the occupancy preference does not violate Title VIII unless “adopted as a subterfuge for discrimination,” 247 N.J.Super, at 176, 588 A.2d 1227, suggests that a discriminatory motive is an essential element of a Title VIII violation. As the federal cases have demonstrated, see supra at 22-25, 622 A.2d at 1267-1269, proof of discriminatory impact is sufficient to establish a prima facie violation of Title VIII. On the sparse record presented, the disparity between the percentage of minority population in the six municipalities as compared with that of their respective regions would be sufficient to present a significant possibility that the occupancy preference could cause a discriminatory impact on minorities. Because COAH did not hold an evidentiary hearing, the issue is not before us on an adequate record, no evidence whatsoever having been presented with respect to the percentage of minority workers in each of the municipalities. The inference is compelling that unless the percentage of minority workers proved to be sufficient to compensate at least in part for the significant disparity in the local and regional percentages of minority residents, the occupancy preference would exert a discriminatory effect on minority residents in the region eligible for affordable housing.

Nor can we readily accept the Appellate Division’s summary assessment of the validity of the asserted governmental justifications for the occupancy preference, or its conclusion that no alternative provision would adequately serve those interests. In our view, the conclusion that a prima facie violation of Title VIII has been rebutted adequately by governmental justifications cannot be reached without the benefit of a full evidentiary record that includes evidence establishing and contesting the legitimacy of the governmental interests and the existence of reasonable and non-discriminatory alternatives. Assessed at face value, however, the governmental justifications proffered to support the occupancy preference do not appear to have sufficient weight to counterbalance proof that the occupancy preference may have a discriminatory impact on minorities eligible to occupy low- and moderate-income housing.

IV

In determining that COAH’s occupancy-preference regulation is flawed, we note that COAH has not demonstrated how the preference, in the context of COAH’s fair-share methodology, can be reconciled with the regulatory scheme that has been adopted to implement the State’s affordable-housing policy. The occupancy preference as currently formulated cannot be reconciled with COAH’s own methodology, which calculates the regional need for affordable housing without reference to the financially-needy local residents who are eligible to benefit from the preference. To that extent the occupancy preference does not further the statutory goals of the Fair Housing Act in implementing the Mount Laurel doctrine. To repeat, that doctrine recognizes that each community bears an obligation to provide its fair share of housing for the region. To establish a regional goal and then, without sound basis and comprehensive standards, to exclude the region’s poor collides with the basic goals of the Fair Housing Act.

We have recognized the enormous responsibility that the Legislature has reposed in COAH to effectuate the goals and standards of the Fair Housing Act. Acknowledging the complexity that surrounds implementation of the Fair Housing Act, this Court has accorded COAH wide leeway in fashioning regulatory approaches to meet fair housing goals, and, at the same time, insisted that COAH exercise that regulatory responsibility and do so in a reasoned and accountable way. See Holmdel Builders Ass’n v. Township of Holmdel, 121 N.J. 550, 579-80, 583 A.2d 277 (1990). We do not in this case foreclose the exercise of that responsibility.

Earlier in this opinion we acknowledged that municipalities understandably may desire to provide affordable housing for long-standing residents otherwise unable to remain in their community, and that specific categories of eligible local residents having special needs might justifiably be considered for entitlement to a preference in the allocation of affordable housing. Supra at 34, 35-36, 622 A.2d at 1274-1275. Consistent with a municipality’s regional obligation to provide affordable housing, COAH could consider revision of its fair-share methodology to address more comprehensively the affordable-housing needs of the residents of a municipality. For example, such a revision could modify or expand the concept of present need or regional need to take into account to some extent the category of financially-needy households that require affordable-housing units. A fair-share methodology that included identifiable financially-needy households in municipalities within a region would at least afford COAH a statistically-valid base against which to consider authorization of an occupancy preference for specific categories of local residents having special needs. The authorization of a preference to accommodate such households would have to be carefully and soundly integrated into the statutory and regulatory scheme in order not to frustrate a municipality’s regional obligation. Its validity would depend less on mathematieally-precise compliance with regional fair-share calculations than on whether such a preference was consistent with COAH’s overall responsibility to fulfill the legislative objectives of the Fair Housing Act that serve to implement the Mount Laurel doctrine. Whether such a limited occupancy preference would be consistent with the mandate of the federal Fair Housing Act or the Law Against Discrimination would depend on municipally-specific statistical data indicating the effect of the preference on minorities eligible for affordable housing within the region. As noted, supra at 36, 622 A.2d at 1275, we do not now attempt to consider or resolve the validity of alternative approaches to establish an occupancy preference designed to address special and cognizable needs of local residents.

In the context of administrative agency adjudicative determinations, we have uniformly insisted that “an agency must set forth basic findings of facts supported by the evidence and supporting the ultimate conclusions and final determinations so that parties in any reviewing tribunal will know the basis on which the final decision was reached.” Riverside Gen. Hosp. v. New Jersey Hosp. Rate-Setting Comm’n, 98 N.J. 458, 468, 487 A.2d 714 (1985). In the rule-making setting, we impose the analogous requirement that the agency demonstrate at a minimum that its action can be understood to be consistent with the underlying legislative mandate. Williams, supra, 116 N.J. at 108, 561 A.2d 244. Texter v. Department of Human Servs., 88 N.J. 376, 387, 443 A.2d 178 (1982). Under COAH’s present methodology for calculating regional-fair share, we cannot discern how the local occupancy preference as currently formulated is consistent with and advances the purpose of the regional obligation. The central goal of the Fair Housing Act is to implement the Mount Laurel doctrine. As presently structured, the occupancy preference adopted by COAH conflicts with the fundamental precept of the Mount Laurel doctrine, as explicated through the Fair Housing Act and implemented by COAH regulations, because it excludes from eligibility for up to fifty percent of the low- and moderate-income housing constructed within a municipality members of the class for whose benefit the obligation to construct that housing was established.

V

The judgment of the Appellate Division is reversed and the matter is remanded to the Council on Affordable Housing for further proceedings consistent with this opinion.

APPENDIX A

1 A brief summary of each of the municipal plans is set forth in Appendix A.