Summary
holding unconstitutional a mobile-home zoning set-aside that imposed an unfair burden in form of discriminatory subsidies on middle-income residents
Summary of this case from Holmdel Builders Ass'n v. Township of HolmdelOpinion
Decided December 6, 1984.
Carl S. Bisgaier for plaintiffs ( Bisgaier Pancotto, attorneys). Alfred J. Villoresi for defendant.
OPINION
This is a suit by a developer (Van Dalen) attacking the validity of the zoning ordinance of Washington Township (Washington) in Morris County. The primary ground of attack is that Washington's zoning fails to afford a realistic opportunity for the construction of low and moderate income housing, as required by the Mount Laurel doctrine. See Southern Burlington Cty. N.A.A.C.P. v. Mount Laurel Tp., 92 N.J. 158 (1983) ( Mount Laurel II).
It has been determined by oral opinion dated February 9, 1984 that a corridor of Washington adjacent to Hackettstown lies in an area designated "growth" in the State Development Guide Plan (SDGP) and that this designation is valid. Under Mount Laurel II, such a designation means that Washington has a responsibility to provide a realistic opportunity through its zoning for the construction of housing for its own inadequately housed lower income residents and also for its fair share of the regional need for lower income housing.
Before discussing the magnitude of Washington's Mount Laurel obligation, it is appropriate to note that another of the judges specially appointed to hear Mount Laurel cases, Judge Eugene D. Serpentelli, has issued a comprehensive opinion in AMG Realty Co. v. Warren Tp., ___ N.J. Super. ___ (Law. Div. 1984), dealing with the calculation of a municipality's Mount Laurel obligation. That opinion adopts a methodology which evolved out of discussions among a group of planners relating to settlement of Urban League of Greater New Brunswick v. Borough of Carteret, Docket No. C-4122-73, and is generally referred to as the Urban League methodology. In deciding the present case, this court has followed the Urban League methodology in some respects and departed from it in others. To avoid undue length, this opinion refers to parts of the discussion of the Urban League methodology contained in AMG Realty.
As noted before, there are two components of Washington's Mount Laurel housing obligation; first, the needs of its own residents who are inadequately housed, and second, its fair share of the regional need for lower income housing. The calculation of Washington's indigenous need for lower income housing is set forth in section I of this opinion and its regional fair share obligation in section II. The steps purportedly taken by Washington to satisfy its Mount Laurel obligation are considered in section III.
I
Van Dalen's expert, Geoffrey Wiener, utilized the Urban League methodology to determine the number of lower income residents of Washington who currently occupy deficient housing. This methodology relies upon housing data generated by the 1980 census. Specifically, it obtains from published census data an unduplicated count of housing units which lack complete plumbing ( i.e., where residents of the unit do not have a toilet, sink and bathing facilities for their exclusive use), are overcrowded ( i.e., have 1.01 or more occupants per room) or lack central heating ( i.e., have no heating or are heated by fireplace, stove or room heaters with no flue). AMG Realty at ___.
In an initial report Wiener relied upon a slightly different methodology for calculating indigenous present need. This generated an indigenous need for Washington of 137, while the Urban League methodology results in an indigenous need of 124.
The Census Bureau also publishes data regarding other negative housing characteristics. See Center for Urban Policy Research, Rutgers — The State University of New Jersey, Mount Laurel II: Challenge Delivery of Low-Cost Housing at 111 (1983) ( Rutgers Report). This court has concluded in other litigation that the most appropriate method for determining the extent of present housing need takes seven negative housing characteristics reported by the census into account. Countryside Properties, Inc. v. Ringwood, 203 N.J. Super. 291 (Law Div. 1984). However, neither party to this case presented census data reflecting negative housing characteristics other than lack of complete plumbing, overcrowding and lack of central heating. This court is satisfied that use of these three housing characteristics provides a sufficiently reliable means of identifying deficient units to justify its use in the present litigation.
Wiener observed that the census does not directly report housing dilapidation. Therefore, except for overcrowded units, which are an independent category of inadequate housing, the negative housing characteristics reflected by the census and utilized by the Urban League methodology are referred as "indicators" or "surrogates" of physically dilapidated housing. In other words, a housing unit may lack central heating or complete plumbing for the exclusive use of its residents and yet be structurally sound and possess the other qualities of satisfactory housing. Conversely, a housing unit may not exhibit any of the negative characteristics revealed by census data and yet have broken windows and doors, a failed roof and collapsing exterior structure, and hence be "dilapidated." Nonetheless, Wiener expressed the opinion that there is a sufficient degree of correlation between lack of complete plumbing or central heating and actual physical dilapidation to justify the use of these surrogates to measure the extent of present need for lower income housing.
The final step in calculating indigenous need under the Urban League methodology is to determine the percentage of deficient units occupied by lower income persons. AMG Realty at ___. The source used for this information is a 1978 report published by the Tri-State Regional Planning Commission, entitled People, Dwellings and Neighborhoods ( Tri-State Report), which states that 82% of all inadequate housing units are occupied by lower income households.
Applying this methodology, Wiener concluded that Washington contains 151 deficient housing units. This total is comprised of 17 units that lack complete plumbing, 35 that are overcrowded, and 99 that lack central heating. Using the 82% figure derived from the Tri-State Report, Washington's indigenous need for lower income housing is 124 units (82% X 151).
Washington's expert, P. David Zimmerman, took issue with the Urban League methodology for calculating indigenous need in three respects. First, he contended that overcrowding should not be a component of indigenous need. In his view, overcrowding is reflective not of housing condition but rather of the occupants' economic and social condition. Hence, it is a problem which should be addressed through means other than zoning for construction of new units. Second, while agreeing generally with the use of lack of complete plumbing or central heating to determine the extent of dilapidated housing, Zimmerman contended that the use of these surrogates in Washington produces inflated results. To support this conclusion, he relied upon a study by the municipal tax assessor which purported to disclose that the extent of dilapidated housing in Washington is far less than the 116 units which the census reported to lack complete plumbing or central heating. Zimmerman also expressed the opinion that it is common in rural areas for housing units which are in satisfactory condition to be heated by wood stoves, and therefore that the lack of central heating surrogate should exclude such units. This exclusion would reduce the total number of dilapidated units to 72. Third, Zimmerman contended that the assumption that 82% of deficient housing units are occupied by lower income persons is erroneous as applied to Washington and that a lower percentage should be used.
This court is satisfied that, except for its use of 82% to determine the percentage of deficient housing units occupied by lower income persons, the Urban League methodology represents a reasonably reliable means for determining indigenous need.
With respect to Zimmerman's first objection to the Urban League methodology, this court concludes that Mount Laurel II mandates the inclusion of overcrowded units as a component of present need and that Zimmerman's contrary opinion must therefore be rejected. Although the Mount Laurel II opinion states at one point that the present need obligation relates to "resident poor who now occupy dilapidated housing," 92 N.J. at 214, it describes this need at another point as generated by "present dilapidated or overcrowded lower income units." Id. at 243; emphasis supplied. It is unlikely that the Court's reference, in the disjunctive, to "dilapidated or overcrowded" units was inadvertent. Therefore, the Court appears to have concluded that present indigenous need includes not only housing units which are dilapidated but also units which are physically adequate but overcrowded.
Second, the surrogates of lack of complete plumbing and central heating provide a reasonably accurate indication of the extent of physically dilapidated housing, and the survey conducted by the municipal assessor failed to impugn the accuracy of the numbers generated by these surrogates. Although a properly conducted physical survey of all housing in a municipality undoubtedly would provide a more reliable count than census data, Washington's survey was flawed in so many respects that no meaningful conclusion may be drawn from it. The survey ostensibly involved the review of property record cards to determine which specific housing units lack plumbing for exclusive use or central heating, and then physical observation to determine the quality of those units. However, the evidence failed to establish that the property record cards accurately reflect the existence of lack of complete plumbing or central heating. The assessor testified that where entry into a house could not be gained, the revaluation official would simply "estimate" the house or use information gathered a decade earlier. Furthermore, the results of the assessor's inspections of certain houses strongly suggested that the notations placed on the record cards were simply erroneous. The magnitude of this erroneous reporting is indicated by the fact that whereas the census data indicated the existence of at least 116 units which lack complete plumbing or central heating and are occupied year-round, the property record cards revealed only 70 such units. In addition, there were interior observations made of only 20 of the 119 units included in the survey and questionnaires received from the owners of only 41 properties. Assessments of the quality of the remaining units were based solely on exterior observations, which can fail to reveal serious interior deficiencies. Most important, even if the data were reliable, the survey is based on the false premise that the only dilapidated housing in Washington is that which lacks complete plumbing or central heating. However, as explained previously, these negative housing characteristics are simply surrogates for dilapidation. There are some units exhibiting one or the other of these characteristics which are not dilapidated, but by the same token there are some dilapidated units which do not exhibit either of these characteristics. Therefore, a properly conceived physical survey of housing to determine the extent of dilapidation would not be restricted to units which lack complete plumbing or central heating.
The census actually indicates that there are a total of 187 housing units in Washington occupied year-round which lack complete plumbing or central heating. However, under the Urban League methodology, units heated by a room heater with a flue are excluded from the count of units lacking central heating. It appears probable that such units would be counted as lacking central heating on Washington's property record cards. If this is so, it would indicate an even greater magnitude of undercounting on the property record cards.
The only respect in which this court agrees with Zimmerman that the Urban League methodology fails to provide a sufficiently reliable indication of the extent of Washington's indigenous need is in its reliance upon the 82% figure derived from the Tri-State report for determining the percentage of deficient housing occupied by lower income persons. The "82%" is contained in a single sentence in the Tri-State Report, which reads as follows: "3. Low and Moderate Incomes. This includes one-third of all households, and it also includes almost all 82 percent of the households experiencing inadequate housing conditions." Tri-State Report at 15. The report contains no indication of how the "82%" figure was established. Furthermore, since the Tri-State Commission has been dissolved, it has not been possible to locate any person or background document which might reveal the source of this statistic.
While the specific source of the "82%" figure has not been determined, an overall review of the Tri-State Report indicates that there are a number of problems in using the data contained therein to determine the percentage of New Jersey households occupying deficient housing who are lower income. First, the study encompassed the entire New York metropolitan area, including the Connecticut and New York suburbs and New York City. Consequently, the size of New York City's population causes its housing characteristics to dominate the statistics. Second, the primary source of data appears to have been the 1970 census, which is now 14 years old. Third, the term "inadequate" housing may have included not only physically deficient and overcrowded units but also units occupied by persons who pay a disproportionate percentage of income for housing or who commute an excessive distance to work. See Tri-State Report at 8. Since the Urban League methodology does not include these latter two characteristics as part of present need, this is yet another reason to conclude that the "82%" figure is unreliable.
Furthermore, evidence was presented of an alternative method of determining the percentage of deficient housing units occupied by lower income persons, which has none of the weaknesses of the Tri-State Report statistic. The Rutgers Report, mentioned earlier in this opinion (at 312, n. 3), calculates the extent of deficient housing in New Jersey, based upon negative housing characteristics reported in the census of 1980 rather than 1970. That report also calculates the percentage of deficient units actually occupied by lower income persons on a regional basis within New Jersey. Consequently, the Rutgers Report has the decided advantage, compared with the Tri-State Report, of providing data which is based on the 1980 rather than the 1970 census and which relates solely to northwestern New Jersey rather than the entire New York metropolitan area.
As noted in footnote 2, the Rutgers Report uses seven housing characteristics to determine the number of deficient units rather than the three used under the Urban League methodology. However, the results under the two methodologies are sufficiently similar to justify use of the percentage of deficient units occupied by lower income households contained in the Rutgers Report, at least in the absence of a computer calculation of a percentage using the Urban League surrogates.
The Rutgers Report indicates that 67.5% of the deficient housing units in northwestern New Jersey are occupied by lower income persons. Applying this percentage to the 151 deficient units calculated by the Urban League methodology, Washington's total indigenous need is 102 units.
II
The regional Mount Laurel obligation of a municipality located in a "growth" area is to provide a realistic opportunity for the construction of its "fair share" of the regional need for lower income housing. 92 N.J. at 238-239. There are two components of regional need. The first is the region's present need, which is the current need for adequate housing in excess of the average need in the region. See 92 N.J. at 214-215. The second component is prospective need, which is the projected future need for lower income housing caused by growth in regional population and employment. See 92 N.J. at 256.
A three-step process is required to establish a municipality's fair share of regional present and prospective need. First, the relevant region must be identified. Second, regional need must be determined. Third, that need must be allocated among the municipalities in the region. See 92 N.J. at 248.
Both parties employed the Urban League methodology as their basic framework for determining Washington's regional Mount Laurel obligation. Therefore, it is appropriate to describe briefly the Urban League methodology and then to identify the respects in which the experts for each party departed from that methodology.
This description is summary in nature. A more complete description and the justifications for the Urban League methodology may be found in the AMG Realty opinion.
The Urban League methodology establishes two different regions, one for present and the other for prospective need. See AMG Realty, ___ N.J. Super. at ___-___. The present need regions are "fixed line" regions; the one in which Washington is located consists of 11 counties in northern New Jersey. The prospective need regions are "commutershed" regions which are determined separately for each municipality. Specifically, a determination is made of how far a person can travel in 30 minutes from the center of a municipality, and this establishes its commutershed. The whole of every county penetrated by this 30-minute commutershed is the prospective need region of that municipality. Washington's prospective need region, using this methodology, consists of Hunterdon, Morris, Somerset, Sussex and Warren counties.
To determine regional present need, the percentage of housing units in the present need region which are deficient and occupied by lower income persons is calculated first. The percentage of such units in each municipality is then determined and to the extent any individual municipality's percentage exceeds the regional percentage, the excess is added to the excesses from other municipalities with high present need to arrive at the regional present need. AMG Realty at ___, ___-___. The total regional present need for the 11-county region in which Washington is located is 35,014 units. However, under the Urban League methodology, this overall need is to be met in three six-year stages, so that only one third of the total, or 11,671 units, must be satisfied between 1984 and 1990.
To determine regional prospective need, it is necessary to project the number of new lower income households that will be established between 1980, the year in which present need is determined, and 1990. AMG Realty at ___-___, ___-___. The Urban League methodology does this by averaging two different sets of county population projections prepared by the Office of Demographic and Economic Analysis in the New Jersey Department of Labor. One set is based upon a method known as the Economic Model and the other on a method known as the Demographic Model, both of which project growth by age group. To translate population data into household data, the methodology relies upon the Rutgers Report, which uses census data to predict the extent of new household formation for each age group. To project the percentage of new households which will be lower income, it relies upon the statewide figure of 39.4% referred to in Mount Laurel II, 92 N.J. at 221 n. 8. Thus, applying the Urban League methodology, the prospective regional need for the five-county region in which Washington is located is 37,049 housing units.
The allocation of fair share of regional need among municipalities in the present and prospective need regions is made in accordance with two mathematical formulas, one for present and the other for prospective need. AMG Realty at ___-___.
With respect to present need, the formula involves two factors, the percentage of growth area in the region contained in each municipality and the percentage of regional employment in the municipality. These two percentages are averaged, and the average is modified by taking into consideration the median income in the municipality compared to that in the region. The total present need of the region then is multiplied by the resulting percentage to arrive at the municipality's fair share of the regional present need. AMG Realty at ___-___. There are then two further adjustments. First, each municipality's fair share is increased by 20% to reflect the fact that some municipalities in any region have insufficient vacant land or other circumstances preventing them from accommodating their fair shares. Second, it is increased by another 3% to provide for a vacancy rate which will facilitate mobility in housing choice. AMG Realty at ___-___. Applying this formula to Washington, it was determined that the township has a regional present need obligation of 13 housing units.
The formula by which median income is taken into account, which is rather complicated, is explained in the AMG Realty at ___-___ and ___.
The formula for allocating prospective need is substantially the same as the present need formula, with one additional factor. The municipality's percentage of regional employment growth from 1972 to 1982 is factored in with its percentage of regional employment and growth area. The three percentages are averaged, and the remainder of the formula is the same as that for present need. The median income modifier is applied, the resulting percentage is multiplied times the total prospective need for the region, and that number is subject to the 20% and 3% upward adjustments discussed in the preceding paragraph. AMG Realty at ___-___, ___-___.
Applying this formula to Washington, it was determined that the township has a regional prospective need obligation of 172 units. Added to the 13-unit present need obligation, this would result in a total regional fair share obligation of 185 housing units.
As noted previously, the experts for both parties generally accepted the Urban League methodology as a reasonable means of determining a municipality's fair share obligation. However, each party contends that there are certain circumstances peculiar to Washington which require adjustments to be made in the methodology.
Van Dalen's experts accepted the Urban League delineations of the present and prospective need regions and of present and prospective regional need. They also accepted the allocation formulas for present and prospective regional need, but contended that strict application of those formulas to Washington produces numbers which do not properly reflect that municipality's fair share obligation. Consequently, they contended that additional factors should be included in the allocation formulas. These proposed adjustments to the formula led one of Van Dalen's experts, Wiener, to conclude that Washington's total regional fair share obligation should be 729 units, and the other, Alan Mallach, to conclude that it should be 326 units.
Washington's expert, Zimmerman, accepted the Urban League allocation formulas in their entirety. However, it was his opinion that Washington's location and the commuting patterns of its residents require its present and prospective need regions to be drawn differently than prescribed by the Urban League methodology. Specifically, he urged that the long commuting patterns of Washington's residents justify a larger prospective need region than that drawn by the Urban League methodology, or in the alternative that its location in a rural center justifies a smaller prospective need region. He also urged that the present need region should be smaller than the 11-county region used under the Urban League methodology. Use of one or the other of the regions proposed by Zimmerman would result in a regional fair share obligation for Washington of between 122 and 142 units.
This court accepts for purposes of this litigation those parts of the Urban League methodology which are not in controversy between the parties. Therefore, the remainder of the discussion in this section is addressed to those aspects of the fair share determination which were controverted at trial. For convenience the discussion is divided into subsections reflecting the three components of a fair share determination — region, regional need and fair share allocation.
(A) Region
In determining the appropriate region for Washington it is necessary to consider its location, its SDGP designation and the significance assigned to the SDGP in Mount Laurel litigation. Washington is a large municipality located in the southwestern corner of Morris County. Although the SDGP designates a major portion of eastern Morris County and the corridor along Route 80 as growth areas, the southwestern part of the county is a non-growth area. Thus, all of Chester Township, Chester Borough, Mendham Township and Mendham Borough, which lie immediately to the east of Washington, the part of Mount Olive which lies immediately to the north, and the parts of Tewksbury and Lebanon which lie immediately to the south in Hunterdon County, are in limited growth areas. Ninety-eight percent of Washington also is designated limited growth or agricultural. The 2% of the municipality designated growth is located in a narrow corridor adjacent to Hackettstown.
Hackettstown and the area immediately surrounding it have been designated in the SDGP as a "rural center." The function of rural centers and the reason for designating them growth areas is described as follows in the SDGP:
Other designated Growth Areas are proposed not because of their accessibility to metropolitan regions, but because of their function as service centers for designated Agricultural Growth Areas. Such centers were originally established as places where farmers could purchase supplies and sell their produce. . . .
These towns continue to serve as centers, although some of their residents may commute to jobs well beyond the immediate area. All are served by at least one major highway. Freight rail service is also available in some locations although such service may be reduced by proposed consolidations. As established settlements, these towns are served by public water supply and sewer systems and often include regional hospitals, educational facilities and other special services for surrounding areas.
The Guide Plan recognizes the important function these centers play as regional service centers by designating them as Growth Areas, within larger areas designated as Agricultural or Limited Growth. By encouraging future growth within these areas, pressures to develop in Agricultural or Limited Growth Areas may be relieved with future growth better faciliting (sic) the efficient use of public services and financial and energy resources. [at 63-64.]
Therefore, Washington's small growth area received that designation not because the drafters of the SDGP viewed it, like the municipalities along the Rockaway and Clinton corridors, as linked to the employment centers in northeastern New Jersey and New York City, but rather because it is part of a rural center to which growth should be channeled to relieve development pressures on nearby agricultural and limited growth areas. Stated another way, but for its adjacency to Hackettstown, all of Washington would lie outside any growth area.
The Supreme Court in Mount Laurel II assigned critical significance to the planning designations of the SDGP. It characterized the SDGP as providing "a statewide blueprint for future development" and determined that reliance upon the SDGP designations ". . . will ensure that the imposition of fair share obligations will coincide with the state's regional planning goals and objectives." 92 N.J. at 225.
The Court observed that a central theme of the SDGP was to channel development in order to avoid suburban sprawl:
The Plan notes that "a major challenge in the coming years will be to provide a variety of housing opportunities in appropriate locations for New Jersey's expanding population." SDGP at 7. Referring to the "suburbanization process" that followed World War II as "expensive and wasteful," the Plan notes "a need now in New Jersey to alter this unplanned pattern of spread development. A compact development pattern for the future can serve to promote the utilization of the existing infrastructure and service system in an economical way. . . . It is now suggested that a major portion of the State's development efforts should be directed to areas within and contiguous to existing development." SDGP at 25. [ 92 N.J. at 231.]
It would be inconsistent with this policy of discouraging suburban sprawl to place Washington in a housing region which would link it with the northeastern New Jersey metropolitan area and require it to zone in order to accommodate a "fair share" of lower income housing need generated by that area. Rather, to maintain consistency with the planning objectives of the SDGP, the housing region of a municipality located in a rural center should be defined in light of its rural center designation.
Theoretically, such a region would consist of the municipalities in the rural center itself and the agricultural and limited growth area for which it functions as a "service center." See SDGP at 63. In this case, a region drawn in this manner probably would include 10 to 15 municipalities in southwestern Morris and northeastern Warren counties. However, certain data required for a fair share determination, such as population projections, is either unavailable or unreliable at any unit of measure smaller than a whole county. Therefore, it appears that a region for purposes of Mount Laurel litigation must be drawn using county boundary lines.
The region composed of whole counties, which encompasses the entire Hackettstown rural center and the area for which it functions as a service center, consists of Morris and Warren counties. It must be acknowledged that this region encompasses the part of the Rockaway growth corridor in Morris County as well as the rural centers of Phillipsburg and Washington Township in Warren County, and hence data pertaining to this two-county region necessarily will reflect development pressures and other circumstances which are operative beyond the Hackettstown rural center area. However, the Court recognized in Mount Laurel II that fair share allocations cannot be made with "scientific accuracy." 92 N.J. at 257. It also must be acknowledged that a region encompassing Morris and Warren counties may be somewhat smaller than the region which would be generated by strict application of the Court's definition of region as "that general area which constitutes, more or less, the housing market area of which the subject municipality is a part, and from which the prospective population of the municipality would substantially be drawn, in the absence of exclusionary zoning." 92 N.J. at 256 [quoting Oakwood at Madison, Inc. v. Tp. of Madison, 72 N.J. 481, 543 (1977).] However, adherence to the planning objectives of SDGP with respect to rural centers constitutes a "special circumstance," which justifies a deviation from this definition of "region." See id.
Accordingly, this court concludes that a two-county region consisting of Warren and Morris counties will most closely reflect the present and prospective regional housing obligation of Washington deriving from its presence in the Hackettstown rural center.
(B) Regional Need
While the experts testified to significantly different total regional needs, their differences resulted primarily from their different positions concerning the region in which Washington is located. Apart from those disagreements there do not appear to be substantial differences between the parties concerning the determination of regional need.
With respect to present need, Wiener and Zimmerman agreed that regional present need consists of the total excess need, excess being defined as the extent to which the present need in any municipality exceeds the regional average of present need. Both experts agreed to use the three surrogates employed under the Urban League methodology ( i.e., overcrowding, lack of complete plumbing and lack of central heating). They also agreed that only one third of the total regional need should be required to be satisfied between 1980 and 1990. In short, both followed the Urban League methodology for calculating present need.
Although Zimmerman excluded overcrowded units and certain units lacking central heating from his calculation of Washington's indigenous need, he did not follow those modifications in his regional present need calculations.
Van Dalen argues in his post-trial brief that all regional present need should be satisfied by 1990. That argument is rejected because not supported by appropriate expert opinion.
With respect to prospective need, the experts also both essentially followed the Urban League methodology. They agreed that prospective need should be calculated to 1990 and not beyond. Zimmerman expressed some reservation concerning the population projections on which the methodology relies, but his ultimate conclusion was that averaging of the Economic and Demographic model population projections of the Office of Demographic and Economic Analysis is a reasonable method for estimating population growth. In any event, even if parts of Zimmerman's testimony suggest that those population projections are too high, this court is satisfied from all the evidence, including in particular Wiener's rebuttal testimony, that the use of those averaged projections is an appropriate means of predicting the 1990 populations of Morris and Warren counties.
Zimmerman departed in one respect from the Urban League methodology for calculating regional prospective need. While the Urban League methodology postulates that the percentage of new lower income households in any region will be the statewide average of 39.4%, Zimmerman contended that it would be more appropriate to use the actual percentage of lower income households in the Morris-Warren region adjusted to reflect an upward trend towards the statewide average. He suggested that this be done by averaging the percentage of lower income households in the two-county region (31.5%) with the statewide percentage (39.4%) to arrive at 35.5%.
This adjustment in the Urban League methodology appears reasonable. The prospective need calculation should be essentially descriptive rather than normative in nature; in other words, it should represent the best estimate of the actual number of new lower income households which will come into existence in the region between 1980 and 1990. While the fact that Washington's current percentage of lower income households is lower than the statewide average may be partly the result of exclusionary zoning, it also would appear to be partly the result of existing job opportunities and other social influences. Therefore, it is not realistic to expect that the need for new lower income housing units in this region from 1980 to 1990 will precisely mirror the current statewide average of lower income households.
With this one modification to the Urban League methodology, the total prospective need to 1990 in the Morris-Warren region is 16,245 units.
(C) Fair Share Allocation
As noted previously, both parties accept the general reasonableness of the Urban League allocation formula. However, Van Dalen's two experts contended that the Urban League formula assigns an unfairly low fair share of regional need to Washington. Specifically, they pointed to the fact that Washington has a large amount of vacant land suitable for residential development and that it has experienced a substantial amount of new residential development over the last decade. They concluded that a significant amount of new development may be expected over the next six years as well, and contended that a larger portion of that development should be for lower income persons than would result from strict application of the Urban League allocation factors. Therefore, they urged that allocation criteria in addition to median income and regional percentages of growth area, employment, and employment growth should be employed in determining Washington's fair share. Wiener urged that the percentage of single-family building permits issued by a municipality compared to the total for the region should be added to the Urban League allocation factors. Mallach urged that the area of Washington which has experienced substantial growth over the last decade should be treated as if it were in a growth area. This would increase the municipality's percentage of growth area and thus its fair share allocation.
To evaluate these proposed modifications of the Urban League allocation criteria, it is necessary to refer again to the SDGP designation of Washington and Van Dalen's challenge to that designation. As noted previously, the SDGP places only a narrow corridor of Washington on the fringe of the Hackettstown growth area. Van Dalen contended that a more extensive area should have been designated growth because the rate of development in Washington over the last decade had been more rapid than in any other municipality in Morris County. This challenge to the SDGP was rejected. This court's opinion sustaining the SDGP designation of Washington observed that one of the significant holdings of Mount Laurel II was the abandonment of the purely descriptive "developing" municipality test for assigning regional fair share obligations and the shift instead to reliance upon the SDGP's normative judgments concerning where development ought to occur. 92 N.J. at 224. The opinion concluded that following sound planning principles it was reasonable for the drafters of the SDGP to conclude that the significant development which has occurred in Washington represents undesirable suburban sprawl and that similar further development should not be encouraged.
To a substantial extent the proofs upon which Van Dalen's experts relied in urging that there should be an upward adjustment in Washington's fair share as determined by the Urban League methodology represented a recycling of the same proofs presented in support of Van Dalen's previous unsuccessful attack upon Washington's SDGP designation. Washington's rapid growth during the last decade, the availability of infrastructure such as water and sewer facilities and the commuting patterns of its residents again were relied upon to support the conclusions that Washington is basically a suburb on the perimeter of the New York metropolitan area and that because it has substantial vacant developable land it should accommodate a greater fair share of lower income housing than the Urban League formula would assign to it.
Wiener also argued more generally that the Urban League allocation criteria fail to take into account the amount of land suitable for development, and that the number of residential building permits issued is a form of "surrogate" which indicates the availability of developable land. However, almost all the building permits issued in Washington recently have been for development outside the growth area. Therefore, the number of building permits issued in the municipality is not indicative of the amount of land suitable for development in the growth area; and no significant weight should be given to the existence of developable land outside the growth area, since Mount Laurel II holds that regional lower income housing needs should be satisfied within growth areas whenever possible. Under these circumstances this court concludes that issuance of building permits should not be taken into account in determining Washington's fair share. This does not foreclose consideration of this factor under other circumstances.
The flaw in the argument is that the SDGP does not characterize Washington in those terms. The SDGP places the perimeter of the New York metropolitan area to the east of Washington and it does not place this municipality in either the Rockaway or Clinton growth corridors. This court has upheld the SDGP designation of Washington. It also has concluded that the significance of that designation in Mount Laurel litigation is not limited to whether Washington has any regional fair share obligation but extends as well to the nature and magnitude of that obligation. Therefore, it would not be appropriate to make an upward adjustment in Washington's fair share allocation predicated on a characterization of the municipality which is inconsistent with its designation in the SDGP. Accordingly, this court rejects the modifications in the Urban League allocation factors urged by Van Dalen's experts.
Zimmerman testified that applying the Urban League allocation factors to the regional present and prospective need in the two-county region would make Washington's fair share of regional present need 3 units and of regional prospective need 122 units. The accuracy of these mathematical calculations was not disputed by Van Dalen and therefore is accepted by this court.
Zimmerman's testimony did not precisely dovetail with this court's conclusions concerning regional present need. His calculations followed the Tri-State Report in estimating the percentage of deficient units occupied by lower income persons at 82%. Recalculation using the Rutgers Report estimate of 67.5% would slightly decrease Washington's fair share of regional present need. He also used the percentage of Mount Laurel housing units in the five-county present need region which he advocated (3.9%) in calculating the extent of present regional need which must be reallocated. Use of the percentage of Mount Laurel housing in the Warren-Morris region (3.5%) would slightly increase Washington's fair share. Since one deviation in Zimmerman's calculations would tend to decrease regional present need slightly and the other would tend to increase it slightly, they basically offset each other and would not affect the ultimate conclusion that Washington's regional present need obligation is three units.
This means Washington's total regional fair share obligation is 125 units which, when added to its indigenous need of 102, establishes a total Mount Laurel obligation of 227 units.
III
Washington relies upon the zoning in three districts to satisfy its Mount Laurel obligation. In one, the Planned Unit Development (PUD) Zone, single-family houses, apartments, town-houses and patio homes are permitted uses. This zone has a 15% set-aside requirement; that is, property owners are required to make 15% of all new units affordable to lower income households. Washington's expert testified that 763 housing units could be built in this zone, of which 114 would be affordable to lower income persons. In the second zone, the R2A zone, mobile homes are permitted at a density of six units per acre and 30% of the units must be affordable to lower income persons. Washington's expert projected that 300 mobile homes could be placed in this zone, of which 90 would be affordable to lower income persons. In the third zone, the RCH zone, mobile and modular homes are permitted at a density of eight units per acre, with a 50% set-aside requirement. It is projected that 192 units could be placed in this zone, of which 96 would be affordable to lower income persons.
If all the housing postulated by Washington's expert were in fact constructed, there would be a total of 300 units affordable to lower income persons. This would more than satisfy Washington's Mount Laurel obligation of 227 units. However, Van Dalen's expert Mallach testified that in all three of the zones there are serious questions as to the feasibility of erecting lower income housing. Therefore, it is necessary to review in detail the suitability of each of the zones for lower income housing.
Before undertaking this review, a threshold legal issue must be addressed. During the pendency of this litigation, Washington amended the ordinance provisions governing the three zones. Washington takes the position that judgment should be entered in its favor if the amended ordinance satisfies the requirements of Mount Laurel. On the other hand, Van Dalen urges that he should prevail if the ordinance in effect when his complaint was filed violated Mount Laurel, and that the newly amended ordinance should be considered solely at the remedial stage of the case.
It is well established that when a zoning ordinance is amended during the pendency of a lawsuit challenging its validity a court ordinarily will decide the validity of the ordinance in its amended form. Kruvant v. Cedar Grove, 82 N.J. 435 , 440 (1980); Hohl v. Readington Tp., 37 N.J. 271 , 279 (1962). This is sometimes referred to as the "time of decision" rule. Kruvant v. Cedar Grove, supra at 440. One purpose of this rule is "to avoid rendering an advisory opinion on a moot question." Id. A related purpose is to provide an early judicial declaration as to the validity of the new zoning provision.
Although the Court indicated in Mount Laurel II that there may be circumstances in which "vindication of the constitutional obligation requires that compliance with Mount Laurel be determined on the basis of . . . prior ordinances," 92 N.J. at 200 n. 1, this court is satisfied that the "time of decision" rule should be followed in the present case. See Oakwood at Madison, Inc. v. Madison Tp., 72 N.J. 481 , 491-492 (1977). Washington offers plausible reasons for not adopting the amendments in question sooner. It did not receive direct notification that Van Dalen contended part of the municipality was in a growth area until September 1983, and it was not until the issuance of this court's decision on February 9, 1984 that the extent of growth area was determined. Therefore, Washington could not assess the precise extent of its Mount Laurel obligation until that date.
Furthermore, while the zoning amendments were adopted on the eve of trial, both parties had ample opportunity to present evidence on their validity. There are persuasive considerations of judicial efficiency and of avoiding the imposition of unnecessary expense upon the parties for ruling upon those amendments. On the one hand, if the ordinance were valid, the municipality would have an interest in learning so immediately, in order to avoid the expense of further amending it, including very likely the cost of a court-appointed master. Lower income persons in need of housing also would have an interest in such a declaration, since it would pave the way for early construction of Mount Laurel housing. On the other hand, if the present zoning fails to satisfy Washington's Mount Laurel obligation, the court's ruling to that effect would obviate the need to rule upon the validity of the old ordinance and would provide guidance concerning the steps needed to achieve compliance with Mount Laurel. Therefore, it is appropriate to review Washington's zoning in light of the recent amendments.
Even where the court determines that it should pass on the validity of a newly amended zoning ordinance, a plaintiff still may argue that it is entitled to a "builder's remedy," if the zoning in effect at the time of the filing of its complaint failed to satisfy its Mount Laurel obligation. In view of the Court's strong statement that "builder's remedies must be made more readily available to achieve compliance with Mount Laurel," 92 N.J. at 279, a builder's remedy arguably may be awarded where the ordinance in effect when a complaint was filed violated Mount Laurel but amendments adopted during the pendency of the action bring the ordinance into compliance with Mount Laurel. However, it is unnecessary to pursue the point here since this court concludes that even the amended zoning ordinance in this case fails to comply with Mount Laurel.
In the PUD district, ownership of at least 100 acres is a precondition for obtaining authorization for development. Only one property in the district, which consists of approximately 200 acres, satisfies this condition. One third of that property is occupied by a golf course and another third has steep slopes which impose constraints upon development. There is a longstanding controversy between the owner and the township over plans for development of the property. The owner proposes a development which would allow him to continue operation of the golf course. However, Washington's plan for compliance with Mount Laurel assumes that housing will be developed in that area. As of the time of trial, the impasse between the owner and the municipality had not been resolved. This situation creates doubt whether this property ever will be developed with high-density housing and, if it is, how many units will be constructed.
Van Dalen also presented evidence that the provision governing the PUD district contains unnecessary cost-generating provisions such as low maximum density standards and excessive buffer requirements. However, many other cost-generating provisions formerly in effect in this district were repealed on the eve of trial. Moreover, the 15% set-aside requirement is less than the 20% most frequently used in Mount Laurel compliance plans and hence would require less internal subsidization of low cost housing than a larger set-aside. In addition, there appears to be a strong demand in the Washington area for the market units proposed to be constructed in the PUD. Therefore, it cannot be concluded on the existing record that the cost-generating provisions attacked by Van Dalen would impede the construction of low cost housing in the PUD zone. See 92 N.J. at 259. However, since this court has concluded that Washington has failed to achieve compliance with Mount Laurel through its existing zoning and will have to be rezoned, these cost-generating features should be subject to further examination in the rezoning process.
Therefore, while the proposed set-aside units in the PUD zone may appropriately be taken into account, it cannot be concluded, as urged by Washington, that this zone will provide 114 units of Washington's total obligation of 227. Rather, the uncertainties over construction of lower income housing in the PUD zone lead to the conclusion that in this case "a realistic opportunity to provide the municipality's fair share (requires) overzoning, i.e., zoning to allow for more than the fair share. . . ." 92 N.J. at 270 (emphasis in original).
The R2A zone consists of 50 acres, all in common ownership. Mobile homes have been a permitted use in this district for a substantial period of time. By a 1983 amendment to the zoning ordinance, a requirement was added that 30% of any new units be affordable to lower income households. The same amendment reduced the permitted gross density from eight to six units per acre. Washington projects the placement of 265 mobile homes in this zone which, added to the 35 already there, would result in 300 units, of which 90 would be affordable to lower income households.
Van Dalen's expert, Mallach, expressed the opinion that it was unlikely any substantial number of additional mobile homes would be placed in the R2A zone. He pointed out that mobile homes had been a permitted use in this district for many years at an allowable density of eight units per acre with no set-aside requirement. Nonetheless, only 35 mobile homes have been placed there, and Mallach described the present condition of the park as "a mess." The primary obstacle to placement of any significant number of additional mobile homes in the district is the difficulty of providing adequate sewage disposal facilities. The site has no access to any sewage treatment facility. Although the existing homes are serviced by septic systems, only 25 more homes can be treated in this fashion since the ordinance requires a "central sanitary sewer system" before gross density may exceed 1.2 units per acre. Any central sewer system would require the approval of the New Jersey Department of Environmental Protection (DEP) and, in Mallach's opinion, the only type of system likely to receive such approval, a tertiary package treatment plant, would not be economically feasible for a 300-unit mobile home park. Other less expensive sewage treatment systems, such as a common septic system, probably would not be approved by DEP, and in the unlikely event one were approved it would occupy approximately half of the 50-acre tract.
Washington's experts, Zimmerman and Dengler, testified that it would be feasible to provide sewage for the R2A tract by a common septic system, that it would not require a substantial amount of acreage and that it probably would be approved by DEP. However, the basis for these conclusions was vague and the contrary testimony of Mallach was far more persuasive. Furthermore, while Washington's experts testified that the owners of the tract had explored further development with both the planning board and DEP and received encouragement, no representative of the owner was produced as a witness and no details concerning their possible development plans were presented.
Consequently, this court accepts Mallach's conclusion that it is not reasonably likely that more than 25 additional units will be developed in this zone. Only eight of these units would be required to be affordable by lower income persons. Therefore, the R2A zone provides only limited assistance to Washington in achieving compliance with Mount Laurel.
The third zone through which Washington claims to have achieved compliance with Mount Laurel, the RCH zone, is a 24-acre "floating" mobile home park district within a larger commercial zone, all also in common ownership. Development at up to eight units per acre is permitted, of which 50% must be affordable to lower income persons. If developed at maximum density, there would be a total of 192 units in the RCH zone including 96 affordable to lower income persons.
Mallach expressed the opinion that this tract could not practically be developed at a density of eight units per acre. More important, he concluded that it is not economically feasible to develop the tract at all for mobile homes with a 50% set-aside requirement. He pointed out that the tract is not currently serviced by sewers and that there would be significant off-site improvement costs in constructing sewer lines to the property. He set forth all the costs which a lower income person would incur for purchase and installation of a mobile home and rental of a site from the mobile home park owner and demonstrated that mobile homes in the RCH zone would not be affordable to lower income persons without substantial internal subsidization. Specifically, he concluded that it would be necessary to raise the monthly pad rental fee paid by other occupants of the park from $255 to $388 in order for the set-aside mobile homes to be affordable to lower income persons. However, there is no market for mobile home sites renting at above $300 a month. Consequently, Mallach concluded that the RCH zone could not be developed profitably with a 50% set-aside requirement.
Washington's expert, Zimmerman, did not dispute that the rental fee for the non- Mount Laurel mobile home sites would have to be increased for the set-aside units to be affordable. However, he calculated a much smaller shortfall between the cost of the sites and what lower income households could afford than that determined by Mallach. He determined that the non- Mount Laurel residents would have to pay $2,333 extra for their mobile home sites to provide the required internal subsidization for the Mount Laurel units. Zimmerman concluded that there was a sufficiently strong demand for middle income housing in the Washington area to support the site rental fees required to cover this shortfall.
Zimmerman's calculations were expressed in terms of the total required subsidization per market unit, whereas Mallach's conclusions were expressed in terms of extra monthly rentals. Therefore, their calculations are not easily compared. However, the required internal subsidization calculated by Zimmerman appears to be roughly one-fourth that calculated by Mallach.
Washington also presented testimony that the owner of the RCH tract has expressed interest in development with a 50% set-aside. However, no application for site plan approval has been filed. Furthermore, the owner of the tract was not called as a witness and there was no indication that a concept plan has been prepared to demonstrate the feasibility of development. Therefore, this testimony is not entitled to any significant weight.
This court is convinced that the internal subsidization required for development in the RCH zone is closer to the order of magnitude determined by Mallach than that estimated by Zimmerman, and that the market will not support rentals for mobile home sites in excess of $300 per month as would be required for such development to be economically feasible. Consequently, the present zoning of the RCH zone does not create a realistic opportunity for the construction of lower income housing.
In any event, even assuming development of the R2A and RCH zones were economically feasible, this court would still conclude that the zoning of these districts fails to satisfy the requirements of Mount Laurel. As noted, the experts for both parties agreed that mobile homes in these districts would not be affordable by lower income persons without internal subsidization, and that this internal subsidization would take the form of the mobile home park owners charging inflated site rental fees to non-lower income residents.
There is nothing in Mount Laurel II to indicate that the Supreme Court contemplated that families who are only one short step higher on the economic ladder than lower income families may be required to pay substantially more for housing in order for a municipality to achieve compliance with Mount Laurel. The Court observed at one point that "the construction of lower income housing is practically impossible without some kind of governmental subsidy." 92 N.J. at 263. On the other hand, it was more optimistic in its discussion of mandatory set-asides:
There may very well be no "subsidy" in the sense of either the landlord or other tenants bearing some burden for the benefit of the lower income units: those units may be priced low not because someone else is subsidizing the price, but because of realistic considerations of cost, amenities, and therefore underlying values. [ 92 N.J. at 267-268 n. 30.]
The theory of mandatory set-asides without subsidization is that the requirement will be accompanied by higher-density zoning than normally would be permitted. Such higher-density zoning permits a developer to construct more units and thereby to cover any shortfall between what lower income households can afford and the real costs of construction. In effect, an increase in permitted densities generates any subsidies necessary for lower income housing, without cost either to the developer or non-lower income residents of the development. See Rutgers Report, supra, at 352-353; Fox and Davis, "Density Bonus Zoning to Provide Low and Moderate Cost Housing," 3 Hastings Const. L.Q. 1015, 1027-28 (1976); Kleven, "Inclusionary Ordinances — Policy and Legal Issues in Requiring Private Developers to Build Low Cost Housing," 21 U.C.L.A.L. Rev. 1432, 1478-81 (1974).
The record in this case demonstrates that this theory of burdenless set-asides does not accurately reflect the economic realities of Washington's two mobile home park zones. The zoning of these two districts simply does not generate the excess economic value required for internal subsidization from a source other than non-lower income residents of the parks. In the R2A zone the allowable density actually has been decreased from eight to six units per acre. This decrease in density, coupled with the sewage disposal problems, makes it doubtful that any significant number of additional mobile homes could be placed in the park even without a set-aside requirement. In the RCH zone, high off-site improvement costs in developing the site for mobile homes gives it only marginal development potential even without a set-aside requirement. Therefore, compliance with any set-aside requirement for lower income families in either mobile home zone could be achieved (assuming it were economically feasible at all) only by increasing the site rental charges imposed upon other occupants of the parks.
The Supreme Court of New Jersey has reached varying results in dealing with regulatory provisions under which added costs may be imposed upon one group of consumers in order to benefit another group. In Property Owners Ass'n of North Bergen v. Twp. of North Bergen, 74 N.J. 327 (1977), the Court invalidated a rent control ordinance that froze the rents of all senior citizens with incomes below a specified level. Pointing out that the ordinance's practical effect would be to shift the burden of paying the additional rents to other tenants, the Court said:
Imposition of such cost on the other tenant or tenants residing in the apartment building may not pass constitutional muster. For example, suppose that nine of ten tenants are Senior Tenants so that the financial burden would be imposed on the one remaining tenant. The relationship between this tenant and his co-tenants does not justify imposing on him the duty of assisting in the payment of their rent. This is quite distinct from the public recognizing and assuming an obligation to assist the elderly or where the cost is spread over a large number of people so that the effect is minimal. [ Id. at 337; footnote omitted.]
On the other hand, in New Jersey Ass'n of Health Care Facilities v. Finley, 83 N.J. 67 (1980), the Court upheld regulations of the New Jersey Department of Health that required nursing homes to make a reasonable number of beds available to indigent patients, even though it recognized that the probable effect of the regulations would be to increase the rates charged other patients. The Court stated:
The rates which a nursing home charges its private patients are unregulated. Undoubtedly, many nursing homes will increase their rates to private patients to offset the net cost of maintaining indigent patients. We agree with the Appellate Division that there is nothing invidious in the notion as such. [ New Jersey Assoc. of Health Care Facilities v. Finley,] 168 N.J. Super. [152] at 167. In its amicus brief, American Health Care Association concedes that there is no federal impediment to subsidization by private patients as long as it is not excessive. See Property Owners Assn. of N. Bergen v. Tp. of N. Bergen, 74 N.J. 327 (1977); Borland v. Bayonne Hospital [ 72 N.J. 152 (1977)], supra. Any judgment on this score will have to await a case involving specific facts and circumstances. [ Id. at 83-84.]
In Mount Laurel II the Court stated that "(m)andatory set-asides do not give rise to the legal issues treated in Property Owners Ass'n of N. Bergen v. Twp. of N. Bergen, 74 N.J. 327 (1977)." 92 N.J. at 267 n. 30. In reaching this conclusion it expressed the view, noted previously, that subsidies very well might not be necessary at all. In addition, the Court seemed to assume that any subsidies required to make set-aside units affordable to lower income families would be paid by the builder and not by other occupants of the development. It stated: "The builder who undertakes a project that includes a mandatory set-aside voluntarily assumes the financial burden, if there is any, of that condition." Id. Therefore, the Court did not directly confront the legal issues presented by a mandatory set-aside that has the practical effect of increasing housing costs for non-lower income residents of a development. Since the record in this case establishes that imposition of set-aside requirements in Washington's mobile home park zones would have such an effect, the issue must now be faced.
The best indication of how the Supreme Court would deal with this issue is provided by New Jersey Ass'n of Health Care Facilities v. Finley, supra. This case was decided after Property Owners Ass'n of North Bergen v. North Bergen, supra. Moreover, Mount Laurel II may be read as limiting to some extent the precedential impact of North Bergen. Assuming New Jersey Ass'n of Health Care Facilities v. Finley is the controlling authority, its holding appears to be that a governmental regulation which has the practical effect of shifting part of the cost of goods or services from indigents to other persons is not necessarily invalid but that such subsidization may not be "excessive." See id., 83 N.J. at 84. Therefore, it appears unlikely the Supreme Court would invalidate a mandatory set-aside based simply upon a showing that it would result in some increase in the housing costs of non-lower income households.
However, this does not mean the Court would sustain a plan for compliance with Mount Laurel under which a significant economic burden would be placed upon non-lower income mobile home park residents. Although the Court indicated that a mandatory set-aside may be appropriate in a mobile home zone to assure that a particular percentage of units actually is occupied by lower income persons, see 92 N.J. at 308-309, it did not envision that non-lower income residents of mobile home parks would have to pay more for their housing as a result. On the contrary, it observed that "mobile homes apparently can be built that are affordable without subsidy by lower income families." Id. at 275 n. 35. However, as demonstrated above, this does not hold true in the Washington mobile home zones. Substantial internal subsidies would be required to accommodate lower income households.
The appropriateness of such subsidies must be determined in light of the constitutional basis for Mount Laurel, which the Court characterized as "substantive due process," "equal protection," "the constitutional obligation to zone only in furtherance of the general welfare" and "concepts of fundamental fairness in the exercise of governmental power." 92 N.J. at 208-209. Although the Mount Laurel obligation relates solely to lower income housing, Id. at 211, the substantive due process and equal protection provisions of the state constitution, as well as the related concepts of "general welfare" and "fundamental fairness," extend protection to every class of citizen. Taxpayers Ass'n of Weymouth Tp. v. Weymouth Tp., 80 N.J. 6 , 42-44 (1976), cert. den. sub nom. Feldman v. Weymouth Tp., 430 U.S. 977, 97 S.Ct. 1672, 52 L.Ed.2d 373 (1977); Southern Burlington Cty. N.A.A.C.P. v. Mount Laurel Tp., 67 N.J. 151 , 173-181 (1975), cert. den. 423 U.S. 808, 96 S.Ct. 18, 46 L.Ed.2d 28 (1975). Therefore, a municipality may not seek to achieve compliance with Mount Laurel by means which infringe upon the constitutional rights of non-lower income citizens.
The record in this case supports the conclusion that non-lower income residents of mobile home parks are middle class, generally with incomes only slightly higher than those of the households which qualify as low or moderate income. Such middle income families often experience substantial economic obstacles to obtaining decent housing. Indeed, they often "choose" to reside in mobile homes only because other housing in the area is not affordable. In Washington, for example, most available housing units are single-family residences sold at prices in excess of $100,000. Nonetheless, under Washington's plan for complying with Mount Laurel, the more affluent persons who purchase single-family homes and other more expensive housing, with the possible exception of those who acquire units in the PUD development, will be free of any extra financial burden at the same time that a significant added burden is imposed upon persons of more modest means who rent mobile home sites.
This court is convinced that such zoning is discriminatory and unfair and therefore denies prospective middle income residents of Washington's mobile home parks equal protection and due process of law. Compliance with Mount Laurel is a legitimate — indeed constitutionally mandated — governmental interest. However, that interest may not be pursued by means which impose an excessive and unfair burden upon middle income households when there are other suitable means of achieving this objective. See Mount Laurel II, 92 N.J. at 258-278. These means include government-subsidized housing, Id. at 262-265, density bonuses, Id. at 266-267, and mandatory set-asides that impact more equitably on other parties, Id. at 267-274. Zoning for mobile home parks, with a mandatory set-aside requirement, also may be a suitable part of a Mount Laurel compliance plan, so long as it does not impose an excessive or unfair economic burden upon non-lower income residents of the parks. However, since Washington's mobile home zones do not equitably apportion the economic burden of achieving compliance with Mount Laurel and other fairer means of compliance are available, the township's present zoning violates the constitutional principles on which the Mount Laurel doctrine rests.
Therefore, the RCH and R2A zones as they now stand may not be utilized to satisfy Washington's obligation to provide for low income housing. However, they may be utilized to satisfy at least a portion of Washington's moderate income housing obligation, since the record supports the conclusion that at most only small internal subsidies will be needed to make mobile homes in those zones affordable to moderate income persons. Such small subsidies either can be absorbed by the park owner or will impose a minimal burden upon other residents of the park. Reliance upon the RCH and R2A zones must be accompanied by sufficient over-zoning unless the uncertainties concerning development in these zones are resolved before the rezoning is completed.
Thus, Washington's present zoning fails to provide a realistic opportunity for construction of the 227 housing units it is obliged to provide. An order will be entered declaring Washington's zoning invalid under the Mount Laurel doctrine and directing Washington to rezone, with the assistance of a master, within 90 days of today.
The conclusion that Washington's zoning ordinance is invalid under Mount Laurel makes it unnecessary to consider Van Dalen's alternative claim that the Municipal Land Use Law, N.J.S.A. 40:55D-1 to 92, requires a "reasonable mix of housing types" in each municipality. In any event, the Supreme Court rejected a substantially similar argument in Mount Laurel II, characterizing the Municipal Land Use Law as "a procedural decree" and "not a legislative effort aimed at creating balanced housing opportunities in New Jersey." 92 N.J. at 320.
The rezoning must retain the provision that at least half of the 227 units be affordable by and exclusively available to low income households. It also must correct other deficiencies in the existing ordinance identified by Mallach, by requiring a reasonable mix of bedroom sizes in both low and moderate income units and establishing income standards which make Mount Laurel units affordable to a reasonable range of lower income households rather than only those at the ceiling of the low and moderate income categories.