Respondents appeal a judgment granting the petition to annul two resolutions issued by respondent New York City Board of Standards and Appeals (BSA) for zoning variances to permit construction of two buildings containing both residential and commercial premises in the SoHo neighborhood in lower Manhattan.
The two properties in issue, presently unimproved and used as open-air parking lots, are located on the north side of the SoHo district and also within the SoHo Cast Iron Historic District. The latter was created in 1973 in recognition of its location as “the world’s largest collection of buildings with full and partial cast iron facades.”
SoHo is a community of artists-in-residence. SoHo, reflecting its manufacturing past, is still zoned for light manufacturing, excluding residential development. However, reflecting New York City’s historic legislative commitment to preserve and encourage the artistic community that found fertile ground in SoHo’s lofts and spacious former manufacturing premises, SoHo has an additional zoning use designation of M1-5A, allowing mixed-use joint living and working quarters for artists (NY City Zoning Resolution [ZR] § 41-11) as a narrow exception to the district’s allowable manufacturing uses. The Zoning Resolution, therefore, prohibits new residential buildings in the SoHo district, except for “joint living-work quarters” for artists (artist lofts).
In 1997, respondent developer and site owners submitted an application to the New York City Department of Buildings for permits to construct two new residential apartment buildings at the site. One building would contain up to nine stories with 60 one-bedroom and studio apartments, and the other would have up to eight stories with 43 one-bedroom and studio apartments. Each building would also contain ground floor retail stores which are also prohibited by zoning restrictions.
On August 11, 1997, the Department of Buildings denied the permits. On August 12, 1997, the owners and developer filed applications for zoning variances with respondent BSA. Petitioners, including community organizations and residents of the SoHo neighborhood, opposed the application, arguing *72that the proposal was unprecedented and would have a negative impact on the character of SoHo as a community of artists.
On April 21, 1998, BSA, after some rocky travel along the administrative path, granted respondents’ application for variances. Petitioners commenced this CPLR article 78 proceeding to annul the BSA resolutions which granted the variances, contending that the determination that the requirements for a variance had been satisfied was arbitrary and capricious and that BSA had erred in ruling that an Environmental Impact Statement was not required.
The motion court granted the petition and annulled the resolutions. The court held that the administrative determination was unsupported by substantial evidence, and rejected the nature and quality of the record evidence putatively supporting uses and bulk that were prohibited by zoning for this community. Supreme Court also noted that, at the least, further environmental review should have been undertaken prior to issuance of the variances. Respondents appeal.
Zoning analysis, quintessentially, starts with conditions at the site and the nature of the neighborhood. The two properties in issue extend from Mercer Street to Wooster Street along the south side of West Houston Street, within the SoHo boundaries. Respondent HS Corporation owns 19/35 West Houston Street, which runs 200 feet along West Houston Street, and is 50 feet deep on Mercer Street and 25 feet deep on Greene Street. This latter site was acquired in 1994. Respondent Housing Garage Corporation owns 55 West Houston Street, which also runs 200 feet along West Houston Street, and is 45 feet deep on Wooster Street and 20 feet deep on Greene Street. As previously stated, both sites are now used as street-level parking lots..
The SoHo neighborhood is characterized by two major zoning features that bear on construction at these sites. Local zoning seeks to preserve SoHo’s artistic character, limiting the use to which neighborhood properties may be put, and the location of the proposed development is within an historic district seeking to preserve 19th century architectural features derived from the area’s historic use. The SoHo Cast Iron Historic District overlaps, but is not entirely contiguous with, the zoning district, but for present purposes, any distinction is not material. Any development in this district requires the issuance of a certificate of appropriateness from the Landmarks Preservation Commission ([LPC] Administrative Code of City of NY § 25-307), which was granted in this case. The proposed *73buildings’ facades, modified to LPC specifications, will conform to the special district’s architectural character. But their size will still be out of conformity with that of many other structures throughout the neighborhood, a factor noted by the City Planning Commission (CPC), infra. Although the majority notes the LPC approval, the only significance is that LPC engaged in an architectural review, which has no ramifications regarding the proposed use and bulk of the project relevant to the issuance of the zoning variances.
After evaluating the district’s character, zoning analysis turns on the legislative mandate. Zoning Resolution § 41-11, which covers the SoHo district, provides in relevant part: “Ml Light Manufacturing Districts
“These districts are designed for a wide range of manufacturing and related uses which can conform to a high level of performance standards. Manufacturing establishments of this type, within completely enclosed buildings, provide a buffer between Residence (or Commercial) Districts and other industrial uses which involve more objectionable influences. New residential development is excluded from these districts, except for joint living-work quarters for artists in M1-5A and M1-5B Districts” (italics added), preventing yet further categorization of manufacturing districts.
Regulations governing Ml Districts apply to M1-5A and M1-5B Districts (ZR § 42-131) unless certain special use regulations (ZR § 42-14 [D]), relating to joint living-work quarters for artists, allow otherwise.
Hence, the designation for M1-5A districts, which covers the site in issue, notably does not create a residential community or otherwise authorize residences as such; it only allows artists to live where they work as a means of enhancing the very trait that underlays the zoning in the first place — the existence of an artists’ district as a corollary to manufacturing uses.
The key term “joint living-work quarters for artists” is defined as one or more rooms, inclusive of cooking and bathroom facilities, in a nonresidential building, the occupant of which is certified to be an artist by the Department of Cultural Affairs (ZR § 12-10). Again, the legislative proscription of residential living, as such, is clear, as is the requisite correlation between living quarters and artistic activities and production. Hence, a use variance is necessary to allow non-artist residential development.
In M1-5A Districts, for all buildings occupying more than 3,600 square feet of lot area, commercial and retail uses are *74generally prohibited below the second floor level (ZR § 42-14 [D] [2]), unless approval is authorized by the City Planning Commission regarding joint living-work quarters for artists (ZR § 42-141) and severe restrictions are placed on the use of any building for eating and drinking establishments (ZR § 42-14 [D] [3]). As noted, the applicants in the present case are not incorporating joint living-work quarters for artists in the subject proposed buildings.
Generally, “[t]he law * * * views nonconforming uses as detrimental to a zoning scheme, and the overriding public policy of zoning in New York State and elsewhere is aimed at their reasonable restriction and eventual elimination” (Matter of Toys “R” Us v Silva, 89 NY2d 411, 417). Hence, in New York City, existing nonconforming uses are intended to expire (ZR § 51-00). Additional intrusion of nonconforming uses, then, is disfavored, especially in view of the legislative goal of preserving the character of districts “in the light of their peculiar suitability to particular uses” (ZR § 51-00). Hence, nonconforming residential uses in Ml Districts tend to retain their viability only by being grandfathered into existing buildings (ZR § 52-45).
Ml zoning resolutions also regulate area requirements of building within the zoning district. The Zoning Resolution (ZR § 43-17) provides that buildings containing joint living-work quarters for artists in M1-5A and M1-5B districts may not be subdivided into quarters of less than 1,200 square feet, except for stories containing no more than one joint living-work quarters. The floor area ratio (FAR) for the district is 5.0.
The applicants’ proposed project consists of two lots, requiring two applications for variances, but the lots are intended to be joined in a single project. As noted above, the applicants submitted their building permit application to the Department of Buildings in 1997, setting forth a proposal for a- nine-story, 60-unit building, with an FAR of 7.95, and an eight-story, 43-unit building, with an FAR of 6.84. The residential units would consist of one-bedrooms and studios, averaging 770 square feet each. Each building was designed for 9,700 square feet of ground floor retail and other commercial space. Since the proposed development violated use as well as area limitations, the Department of Buildings refused to issue building permits.
The applicants then applied for variances with BSA (see, NY City Charter § 666 [6]). BSA is required to “preserve coherent land use determinations and adherence to the zoning plan itself’ (Matter of Bella Vista Apt. Co. v Bennett, 89 NY2d 465, *75471). In furtherance of that goal, the Zoning Resolution requires BSA to undertake a five-part analysis, and mandates that every one of the five criteria be satisfied, before a variance may be granted (ZR § 72-21). In order to grant the variance, BSA must make, and substantiate, all of the following findings:
“(a) that there are unique physical, conditions, including irregularity, narrowness or shallowness of lot size or shape, or exceptional topographical or other physical conditions peculiar to and inherent in the particular zoning lot; and that, as a result of such unique physical conditions, practical difficulties or unnecessary hardship arise in complying strictly with the use or bulk provisions of the Resolution * * * [which are] not due to circumstances created generally by the strict application of such provisions in [that lot’s] neighborhood or district * * *
“(b) that because of such physical conditions there is no reasonable possibility that the development of the zoning lot in strict conformity with the provisions of this Resolution will bring a reasonable return, and that the grant of a variance is therefore necessary to enable the owner to realize a reasonable return from such zoning lot * * *
“(c) that the variance, if granted, will not alter the essential character of the neighborhood or district * * *
“(d) that the practical difficulties or unnecessary hardship claimed as a ground for a variance have not been created by the owner * * *
“(e) that within the intent and purposes of this Resolution the variance, if granted, is the minimum variance necessary to afford relief.”
In support of the variances, the applicants advanced several arguments. Some of those arguments, on their face, invite some preliminary commentary. For instance, the applicants insisted that the context in which the variances should be considered should take into account the sites’ location adjacent to the West Houston Street traffic “corridor.” Inferentially, they viewed the lots’ location as being only marginally in SoHo, so that SoHo zoning requirements could be diluted without impacting the neighborhood, especially since West Houston Street already detracted from the SoHo qualities of the sites. This argument ignores the legislative purpose underlying the district’s M1-5A designation. The applicants also argued that the variance applications should take into account the lots’ location in the SoHo Cast Iron Historic District, which, they *76argued, imposed landmarking restrictions, and hence undue hardship, on development of these lots. The “neighborhood” was described by applicants as containing institutional, commercial and residential uses, with New York University’s (NYU) residence halls and faculty housing, actually located in Greenwich Village, inferentially included in the “neighborhood.” The applicants placed:the proposed buildings in a context of nearby buildings with significant bulk, such as a 30-story NYU complex north of Houston Street, and a nearby 12-story Chase bank building. They argued that urban planners favor the view that major thoroughfares can accommodate more development than narrower local streets — a point which, of course, does not address the specific zoning requirements extant for the south side of this thoroughfare. They also contended that some 40% of SoHo buildings exceed the allowable FAR. However, leaving aside the question whether the statistic is valid, one cannot help observing that, in view of the Zoning Resolution’s explicit grandfathering of many of these historic buildings, and considering the Zoning Resolution’s clear disfavoring of excessive bulk in other buildings, the statistic does not greatly support the case for these variances. The applicants, returning to the impact a nonconforming use will have, contended that the residential population of SoHo is increasing, and that many non-artists now reside in SbHo’s lofts, contrary to City and State policies. But these very observations, if accurate, would seem to argue more for enforcement of existing policies rather than for allowing even more incremental erosion of neighborhood character.
An extensive economic analysis, disputed in salient parts by various of the petitioners, was submitted by the applicants with the application to demonstrate why construction within zoning requirements presented an economic hardship correlating with below-market returns. The applicants argued that keeping the project within the 5.0 allowable FAR is not economically feasible due to the lot’s “unique conditions,” putatively consisting of the lot’s narrowness relative to its width, requiring “single-loaded corridors [in the narrow portion] resulting in a substandard ratio of net rentable to gross buildable area.” The applicants also point to landmarking requirements and a subway easement burdening a portion of the site as also purportedly constituting “unique conditions.” The applicants’ consultant posited that, in order to realize a reasonable return, the applicants would have to make $300 per square foot. Petitioners submitted evidence, which the ap*77plicants disputed, that in SoHo the average selling price was $403 per square foot. The applicants’ study, although purporting to analyze “several as-of-right scenarios,” in fact examined only two: hotel and office use. Notably missing from the analysis in either of two studies was the economic feasibility of development of joint living-work quarters for artists; in fact, the only mention was in responsive documentation by the consultant submitted to BSA stating in conclusory terms that it was not feasible. Also missing from the applicants’ submissions generally was reference to the economic return from the current use. In any event, the applicants’ analysis was intended to illustrate impediments to profits on such construction if maintained within zoning requirements. They then leaped from that illustration to a conclusion that only large-scale residential use would be profitable.
The applicants’ economic study had certain features that the court, though not BSA, noticed and that should be highlighted. The study relied on many buildings for comparables as to construction costs and economic returns that were not within the SoHo district. The majority urges that we reject the proposition that comparables from outside the district are impermissible. While I do not dispute the point, I would further suggest that such a proposition is a straw man in this case. Nor do I read the motion court’s decision as advancing such a proposition. The problem here, as I see it, lies more in degree than in kind in that the applicants relied so heavily on out-of-district comparables. The applicants’ device in this case necessarily expanded the boundaries of the study area to well beyond the SoHo community. While this, by itself, might not fatally compromise the study, the obvious skewing of prices and returns (and also projections on the impact arising from density) warrants scrutiny. For instance, a first feasibility study, using 1996-1997 sales prices, listed 70 sales, of which 27 were not in SoHo. The average selling price was stated to be $262 per square foot — below the $300 per square foot that the applicants claimed was necessary to realize a reasonable return. However, the size of those units was not specified. As a consequence, those comparables were not useful as regards the more typical SoHo type units comprising the 1,200 minimum square footage. A second study added 16 additional units, of which only one was located in SoHo, with the others identified as being in “adjacent areas.” All of these units approximated 1,200 square feet, except for the SoHo unit, which was 1,850 square feet. Curiously, the average selling price of these comparables— *78which would be conforming in terms of unit size — was $317.64 per square foot, and the SoHo unit sold for $383.78 per square foot. The conclusion to be rationally drawn is that these larger units exceeded the minimum price that the applicants claimed they needed but that 1,200-square-foot units, purportedly, could not realize. The implication, though not extant in the studies, is that building with significant square footage units in SoHo could well be profitable.
The applicants expanded the study area for neighborhood impact outside the SoHo district into Greenwich Village, including NYU housing, and Little Italy. Although they contend that it is permissible to view extra-neighborhood impact, for which proposition they refer to the CEQR Technical Manual, that is not quite what was done: the intent, rather, seems to have been to use more densely populated extra-neighborhood areas to dilute the intra-neighborhood impact arising from increased density of use.
The applicants further noted that Houston Street is the northern boundary of the neighborhood. They suggested that the properties’ location thus marginalizes neighborhood impact and allows for inclusion of adjacent neighborhood areas within the impact study area (see, infra). However, Houston Street at this point is a six-lane thoroughfare, effectively an urban highway providing for traffic flow between the FDR Drive and Williamsburg Bridge on the East Side of Manhattan and the West Side Drive/Holland Tunnel complex on the West Side. As such, Houston Street effectively provides a well-defined and distinct boundary, setting SoHo off from neighborhoods on the northern side of the divide which thus seem less than adjacent to SoHo. In this context, northern SoHo may be distinguished from the City’s numerous other neighborhoods that indistinctly merge with adjacent neighborhoods where City planning boundary lines often exist as little more than lines on a map.
Additionally, though, the general points urged by the majority, that district borders should more easily be relieved of zoning restrictions, are problematic. The natural conclusion to be drawn is that if nonconforming development may be accommodated at this district’s border, a zoning variance can also be granted to sites located in SoHo’s other boundaries. Where will this process end? The majority seeks to avoid this conclusion by noting the relative scarcity of other developable sites in SoHo. However, the premise — that there are only six other sites, so the concern for nonconforming development at the borders is unfounded — strikes me as being speculative itself. *79Moreover, one may further surmise that the majority’s decision in this case, and all that it portends, may itself become the springboard for further nonconforming boundary development or conversion in SoHo and elsewhere, effectively undermining the manifest legislative intent of the Zoning Resolution.
Geography aside, SoHo also is distinct in zoning characteristics that are directly derived from the neighborhood’s historical uses and appearance, often manifested in the famous SoHo lofts, that provide an additional sharp distinction with nearby neighborhoods. BSA’s determination, focusing on the West Houston Street location of the properties, gave inadequate attention to these factors.
The applicants also argued that bringing in construction materials will be more inefficient, and thus more costly, as a consequence of the narrowness of the side-street access to the building site. As noted above, they also contended that a subway easement burdens the lot, purportedly increasing construction costs for footings and requiring a more extensive permitting process.
Petitioners, opposing, argued that the development should be geared toward joint living-work quarters for artists, as per the City’s articulated legislative preference, that lot size would not impede the designated zoning use for joint living-work quarters for artists for which the same corridor floor area would not be required, and that, viewing each lot in separate components, 25 by 100 lots are common in SoHo. They disputed the contention that construction-related access would be restricted to narrow side streets rather than utilizing the wider, but less used, south side of Houston Street; noted that maps indicated that only a “small sliver” of the site is affected by the Transit Authority easement, and even challenged the purported construction costs relied on by the developer in its schema; challenged the economic analysis’s reference to commercial zoning districts, not in the vicinity of the site, on which to base residential conclusions and on properties outside the district for comparable rents; noted the wide disparity between neighborhood bulk, FAR and room size zoning standards, and the neighborhood average FAR, and those proposed for the development; argued that removal of the present parking lots but replacing them with residential apartment buildings would create a “parking nightmare” for SoHo; speculated that the developer intended to convert the units into student or faculty housing for NYU, which would further detract from SoHo character of an artistic community resided in by working art*80ists; and noted that, except for hotel and office use, no other uses were explored, leaving unsubstantiated the claim that the variance was the minimum necessary to ensure a reasonable return. Petitioner Freed, a member of the City Council, purportedly speaking on behalf of the SoHo community, noted that the proposal allowed for the construction of a mixed-use high-density building “totally out of character” with, and “unprecedented” in, this manufacturing district. She also decried what she characterized as “glaring inaccuracies and hyperbole” in the developer’s supporting materials. Parenthetically, she also noted that the owner was realizing almost a 7% annual return by merely maintaining the site as a parking lot, and that the projected return for developing the 55 West Houston Street site at a 7.96 FAR was only 6.9%. She thus questioned the very basis of the developer’s plea that a conforming use could not yield a reasonable return.
While the BSA proceeding was pending, the City Planning Commission reviewed the variance request. CPC, as a party to the proceeding (NY City Charter § 668 [e]), reviewed the proposals and made its own recommendations. Since CPC is the agency responsible for regulation of uses and bulk with regard to districts (Administrative Code §§ 25-110, 25-111) and for over-all City development (NY City Charter § 192 [f]), its views have a necessary resonance for land use matters. Although it recommended approval, CPC’s misgivings were apparent in the manner in which the recommendation was phrased. CPC described the requests as “very large and significant. The use variance for residential portions of the two buildings is highly exceptional, constituting the first use of the variance procedure for such type of development in the M1-5A District in SoHo. The proposed floor area of one of the buildings * * * is almost 60 percent more than allowed for conforming commercial developments * * * while the second building would have a floor area 35 percent higher than as-of-right developments * * *. Additionally, the proposed dwelling units would average more than one-third smaller in floor area than the required minimum size of 1,200 square feet for joint living-work quarters for artists in converted buildings in the neighborhood.” CPC further noted that “in virtually all cases, such extensive departures from zoning requirements are unwarranted and will result in building types and forms so out of context with their surroundings that the essential character of the neighborhood will be impaired * * *. Such land use changes which are as extensive as those posed in these ap*81plications should typically occur on the basis of well-considered reviews and revisions to the zoning framework.” That is, by legislation rather than by administrative fiat. These “well-considered reviews,” of course, closely reflect the comprehensive and targeted legislative underpinnings of the Zoning Resolution.
However, CPC, in contrast to the BSA determination under review, was able to closely analyze the potential for nonconforming development, albeit development carefully circumscribed by these clear zoning goals. Hence, CPC, with significant hedging in its analysis, suggested that the particular location had unique features where flexible treatment might not automatically lead to repetition in other parts of SoHo, especially since the lots had “idiosyncratic * * * configurations that are generally not duplicated in other parts of the M1-5A District,” and had remained unimproved since the 1963 widening of West Houston Street. These “singular conditions” — note that CPC did not state “unique physical conditions” or in that context address hardship — “could constitute a rare and limited situation where the significant departures from zoning controls would not have the potential to adversely alter the character of the surrounding neighborhood if the proposals were modified to markedly reduce the maximum floor area ratio to allow no more than 6.5 FAR, substantially increase the average size of the dwelling units” (italics added), and to provide additional amenities. Although the buildings, as modified, could actually enhance the community character, by providing a “street wall presence,” the point was repeated that a harmonious relationship between the buildings and the adjacent cast-iron loft buildings, which range from five to seven stories with FARs from 4.4 to 6.4, depended on reducing the variance request to a maximum of 6.5. CPC also underscored that the density of the proposed units was out of character with the SoHo neighborhood, emphasized that flexible live-work arrangements allowed by larger units is an important neighborhood characteristic, and stated that the minimum size of individual units in the proposal should be substantially increased to at least a range between 900 to 1,200 square feet minimum average. CPC, noting the scarcity of open space in the neighborhood and the zoning requirement that rooftop space be dedicated to recreational amenities, urged adherence to that requirement. Finally, CPC noted a concern that the proposal would result in the loss of air and light to adjacent converted buildings, urged that the proposal be carefully reviewed to ensure that the variance, as *82modified, and the configuration proposed not escalate this impact, but conceded that the site dimensions were problematic in this regard.
Significantly, CPC’s recommended modifications were not incorporated into the developer’s proposal. As such, the final CPC recommendation of approval provides doubtful support for the majority’s position.
BSA conducted public hearings, conducted an environmental assessment and subsequently issued a negative declaration indicating a finding of no significant environmental impact. By resolutions dated March 31, 1998 and April 21, 1998, upon a divided vote, BSA issued variances basically consistent with those sought by the applicants.- Since the BSA determination is challenged as being arbitrary and capricious and unsupported by substantial evidence, the points addressed, or left unaddressed, by the determination are critical to our review.
The BSA resolution basically sidestepped many of the points articulated in opposition. It also basically ignored the CPC recommendation. For instance, the BSA resolution stated in con-clusory terms that “these” unique conditions — alluding to size and configuration without explaining how they were “unique”— created “practical difficulties and unnecessary hardship” — the key words from the Zoning Resolution’s variance provisions— for developing the site in a conforming manner. BSA adopted the applicants’ documentation to conclude that conforming development would not yield a reasonable return. No reference was made to the economic challenges to the applicants’ documentation. BSA also adopted the applicants’ conclusion that the lot configuration imposed distinct construction requirements — such as a single-loaded corridor requiring additional floor, area for elevators and stairs — without addressing the argument that conforming use might avoid this problem. BSA also simply adopted the unsubstantiated claim that 40% of SoHo buildings exceed FAR standards, and seemingly adopted the logic that this was a basis to grant a variance in the present case. BSA, again parroting the phrasing of the variance provisions without an explication of its findings, found that the proposal would not alter the essential character of the neighborhood, and that the proposal was the minimum necessary to afford the owner relief.
The IAS Court noted the CPC concerns regarding FAR and the size of the apartments. The court found that the applicants’ BSA submissions repeatedly ignored the boundaries of the zoning district, even relying on examples from outside the district *83that were egregiously misrepresentative of the district and that “make a mockery of the very presupposition of zoning regulation.” The court also rejected salient parts of the applicants’ economic support on the basis that a significant number of buildings offered as comparables for construction and sale prices were not even in the district, and thus the study too heavily relied on adjacent areas in the analysis of neighborhood impact. Even if the court had not annulled the BSA determinations, it would have found the agency action procedurally flawed as a consequence of the negative declaration, in view of the failure of the Environmental Assessment Form to consider the particular art-orientation of SoHo families, the failure to consider the impact of the projected 500 daily patrons of the proposed commercial space, and, yet again, a conflation of SoHo with other neighborhoods when addressing neighborhood impact.
In these several regards, and more, I agree that the BSA determinations failed to consider critical evidence and could not have reasonably reached the present result in view of the evidence in toto. On the basis of this record, particularly the applicants’ skewed neighborhood impact analysis, its inadequate economic proof in support of the hardship claim, BSA’s failure to adduce evidence directly supporting the categorical requirements allowing for departure from the legislated zoning requirements, and especially in view of the substantial departure from these standards for which there is inadequate justification, I would affirm the ruling under review.
Local authority to enact zoning ordinances such as the Zoning Resolution is derived from New York’s Town Law. The standards pertinent to granting variances are presently codified in section 267-b. As a practical matter, much case law on zoning matters addresses local ordinances as measured against those general standards. Our immediate task, as was BSA’s, is to enforce the words and intent of the New York City Zoning Resolution. The Zoning Resolution has the force of statute in New York City, and neither its terms nor its over-all intent may lightly be cast aside by an administrative agency. Judicial review of a zoning board’s determination, which is administrative rather than quasi-judicial, is governed by a standard of rationality, and evaluated on the basis whether it is arbitrary and capricious or, where there was an administrative hearing, whether it is supported by substantial evidence in the record (Matter of Sasso v Osgood, 86 NY2d 374). An administrative authority interpreting its own promulgated regulations, or ap*84plying the statute that it administers, as the BSA does for the Zoning Resolution (NY City Charter § 666 [6]), is thus entitled to substantial judicial deference under standard canons of article 78 review (Matter of Bella Vista Apt. Co. v Bennett, supra; Matter of Fuhst v Foley, 45 NY2d 441), but that deference extends only so far as BSA hews to the law’s stated requirements, a factor relevant in this case. Moreover, where “the question is one of pure legal interpretation of statutory terms, deference to the BSA is not required” (Matter of Toys “R” Us v Silva, supra, at 419). Since the variances sought have both use and area ramifications, represent significant departures from zoning norms for the district at that, and depend on the specific application of statutory terms, they warrant particular scrutiny to assure compliance with the well-articulated legislative mandate. The Court of Appeals has noted that “since a prohibited use, if permitted, will result in a use of the land in a manner inconsistent with the basic character of the zone, a heavier burden is placed on the applicant” (Matter of Consolidated Edison Co. v Hoffman, 43 NY2d 598, 606).
An analysis of whether a variance was properly granted or denied is necessarily two-fold. First, it must be established what kind of a variance is sought, either use or area (termed “bulk” in the Zoning Resolution), since either type has its own goals and requirements. In the present case, the developers seek both. Second, the Zoning Resolution provides a specific catechism of findings that must be made, and supported by the record, before BSA may lawfully grant the variance. For reasons set forth below, I conclude that use variances were not permissible in this case on the basis of the applicants’ studies, and the BSA findings were, in several respects, unsupported by this record.
At the outset, the developer and the BSA resolution both combine what should be distinct lines of analysis for use and area variances. As noted, a use variance requires satisfaction of a heavier burden of proof than does the mere relaxation of area limitations (Matter of Village Bd. v Jarrold, 53 NY2d 254; Matter of Consolidated Edison Co. v Hoffman, supra; cf., Matter of Townwide Props. v Zoning Bd. of Appeals, 143 AD2d 757). Municipal ordinances often gloss over the distinction (see, 2 Anderson, New York Zoning Law and Practice § 23.33 [3d ed]), which was not even specified in State statutory law until a 1991 amendment to the Town Law (compare, Town Law former § 267 [5] with Town Law § 267-b [3]). The distinction is important insofar as a use variance traditionally required a *85more enhanced showing of hardship of use, whereas an area variance typically would be granted if a permissible use posed financial or other practical difficulties, a more generous standard for the applicant. Hence the alternative standards of “unnecessary hardship” and “practical difficulties,” respectively. It is clear based on case law that each showing traditionally was exclusive to a particular variance (Matter of Sasso v Osgood, supra, distinguishing the standards; see, e.g., Matter of Zagoreos v Conklin, 109 AD2d 281 [proof of unnecessary hardship required for use variance; proof of mere practical difficulties insufficient]; Matter of Gianchetta v Wilens, 122 AD2d 317 [area variance requires showing of practical difficulties]; accord, Matter of Delmarco v Zoning Bd. of Appeals, 204 AD2d 447; Matter of Shaughessy v Roth, 204 AD2d 333 [variance based on practical difficulties does not necessarily require showing of economic injury]; see generally, 2 Anderson, op cit., § 23.33, and citations within).
Before analyzing the Zoning Resolution’s criteria (unique physical condition; reasonable return; minimum relief, etc.), and before applying traditional use and area analysis to those criteria, there is a preliminary point that in this case I believe to be important. One facet of this case that concerns me is that the variances as to area were sought, and were given, in substantial part because of the proposed nonconforming use. Basically, the request for more bulk was bootstrapped onto and justified by the request to allow a specifically nonpermissible use, which seems to me a very dubious application of basic zoning principles. Although coupling use and area variances is not prohibited, when the need for a nonconforming use is what drives the need for substantial extra bulk, the application warrants the greatest circumspection, especially to ensure that the variances, in the aggregate, do not trammel the purposes of the zoning law. This is a case for which a conforming use, joint living-work quarters for artists, was not seriously analyzed. Traditionally, variances for increased height or bulk turned on the applicant’s demonstration that area restrictions imposed practical difficulties for permissible uses (Matter of Putnam Materials Corp. v Zoning Bd. of Appeals, 188 AD2d 651 [height variance denied; although applicant’s preferred manufacturing use might be thus impeded, other permissible manufacturing uses were still practicable, which did not require area variance]; cf., Matter of Boyadjian v Board of Appeals, 136 AD2d 548 [area variance does not involve a prohibited use]).
Turning, then, to the threshold criterion in Zoning Resolution § 72-21 (a), the applicant for a variance must demonstrate *86the existence of “unique physical conditions.” This criterion is described in nonexclusive terms as including irregularity, narrowness or shallowness of lot size. Although the applicant may also assert “exceptional topographical or other physical conditions peculiar to and inherent in” the lot, the applicants here rely on the putative narrowness of the lots and their “L”-shaped configuration. Hence, our focus is limited to how “unique” the specific site is in these regards. Typically, such unique physical conditions (i.e., not imposed by zoning restrictions) consisted of geologic or topological factors (Matter of Douglaston Civic Assn. v Klein, 51 NY2d 963), although the definition is not so restrictive. Applying a use analysis, BSA would have had to find that factors inherent in the site, a “singular disadvantage” (Hickox v Griffin, 298 NY 365, 370), rather than only a choice of operations (Matter of Putnam Materials Corp. v Zoning Bd. of Appeals, supra; see generally, 2 Anderson, op cit., § 23.25), precluded an as-of-right use. Uniqueness may be demonstrated by multiple conditions (Matter of Supkis v Town of Sand Lake Zoning Bd. of Appeals, 227 AD2d 779), but even viewing the applicants’ claims in the aggregate, the requirement of unique physical conditions is not satisfied. Here, both sites are “L” shaped. However, the “L” configurations are modestly sized, a point to which I will return, and neither site is inordinately narrow, as that characteristic is understood for Manhattan real estate. The record does not suggest any noticeable slope, let alone steep slopes or some similar impediment to construction, such as surface bedrock, groundwater interference or another of the typical features presenting uniqueness and unmitigable difficulties often factoring into variance applications. Nor is access to the sites impeded by their locations or configurations, as is obvious by the frontage dimensions. Both sites occupy corner lots, so that all parts of each lot are accessible to street frontage.
Upon proof of such “uniqueness,” under Zoning Resolution § 72-21 (a), BSA was further obliged to determine whether those factors create “practical difficulties or unnecessary hardship,” discussed below, to justify departures from strict zoning limitations. As an additional caveat, though, relating back to uniqueness, the applicants must further show that such alleged practical difficulties or the unnecessary hardship “are not due to circumstances created generally by the strict application of such provisions [i.e., use and area restrictions] in the neighborhood or district” where the site is located. This analysis is derived from the landmark ruling in Otto v Stein-*87hilber (282 NY 71, 76) that “the plight of the owner [must be] due to unique circumstances and not to the general conditions in the neighborhood which may reflect the unreasonableness of the zoning ordinance itself.” In other words, if the present lot sizes or configurations are not particularly distinct when compared with other neighborhood lots, the applicants are hard-pressed to demonstrate why they deserve special treatment.
This latter consideration, that the condition impinging on permissible uses must also be unique in terms of general neighborhood conditions, has some resonance in this case. An applicant’s plea of uniqueness necessarily requires a comparison between similarly situated lots in the neighborhood (i.e., with minimum widths of 25 feet or 20 feet) and the applicant’s lot (Matter of Douglaston Civic Assn. v Klein, supra, at 965). Again, though, the impediment must be physical and must be peculiar to the subject lot; for instance, the mere location in an historic district, also common to numerous other SoHo properties, does not suffice. It has long been recognized that if the hardship reflects the unreasonableness of the ordinance itself or that it has become unreasonable in view of changed conditions, the remedy is to change the zoning law rather than to ask the zoning authority to circumvent it by issuing ad hoc variances (Clark v Board of Zoning Appeals, 301 NY 86, 91, cert denied 340 US 933). Legislative action is preferable to piecemeal exemptions that could ultimately defeat the purpose of the ordinance (Otto v Steinhilber, supra, at 77). In fact, New York City’s recent proposal to amend the Zoning Resolution underscores the viability of a legislative remedy.
Although BSA, and even the parties on appeal, focused on the economics of the alternative uses of the site, I find it particularly interesting that BSA failed to address itself to some basic geography. Each of these lots is 200 feet long — and there is no requirement that the entire length must be built— with a minimum depth of 25 feet (extending to 50 feet) for one lot and 20 feet (extending to 45 feet) for the other lot. The majority finds this unusual length in relation to width to constitute sufficient uniqueness. However, the standard is not whether the shape is unusual, but whether it singularly precludes conforming use. In each case, the putative narrowness affects only a portion of the lot, leaving significant building room otherwise. The applicants have not demonstrated inordinate narrowness for a permissible use. Rather, as indicated in the evidence submitted to BSA, to the contrary, *88this depth is fairly common, undermining, at the outset, much of the applicants’ plea that these lots are being particularly burdened. In any event, neither narrowness (Matter of Colonna v Board of Stds. & Appeals, 166 AD2d 528 [narrow, irregular, Staten Island lot could still be used for permissible residential use]) nor irregularity of configuration (supra, Matter of Marchese v Koch, 120 AD2d 590 [large, trapezoidal lot did not preclude permissible use]), simply by virtue of those characteristics, equates with uniqueness, especially where other neighborhood lots have similar characteristics (see, e.g., Matter of Faham v Bookman, 151 AD2d 665 [31-foot-wide Brooklyn residential lot not unusual for area, and did not preclude permissible use]; Matter of Kallas v Board of Estimate, 90 AD2d 774, affd 58 NY2d 1030 [Brooklyn Heights lot shallower than adjoining lots, but not unusually shallow for the neighborhood]; Matter of Kingsley v Bennett, 185 AD2d 814 [shallowness and irregular configuration of Staten Island lot not unique for neighborhood]; Matter of Marchese v Koch, supra) and, as noted, the greater portions of these lots are not particularly narrow, and the configurations, while irregular, and possibly, inconvenient, do not preclude permissible use. Although CPC referred to these configurations as being unusual for the neighborhood, the context of that statement addressed whether variances here would be repeated elsewhere; finding not many other “L” configurations, CPC concluded in the negative. However, that very different focus is not tantamount to a finding by CPC that these configurations precluded as-of-right development.
Moreover, narrowness or irregularity, absent compelling economic consequences, cannot justify a use variance. The applicant is required to demonstrate that the unique site conditions upon which the applicant relies to establish the physical impracticability of a permissible use also impose a compelling economic hardship, defined as the probability that a permissible use will not generate a reasonable return (ZR § 72-21 [b]). The term “reasonable,” of course, is not further clarified. However, as one commentator has noted, “[o]nly in rare circumstances is property zoned in a manner such that it cannot realize a reasonable return for any of the uses permitted in the zoning district. * * * [A]n applicant possesses an extraordinary threshold to satisfy and must fully demonstrate a lack of reasonable return by competent evidence” (Rice, Practice Commentaries, McKinneys Cons Laws of NY, Book 61, Town Law § 267-b, 1999 Pocket Part, at 142). There are two aspects to *89this standard: whether the applicant’s showing has considered all permissible uses before seeking a use variance, and whether the return is “reasonable.” The initial test, then, is not the relative profitability of alternative uses, but whether any conforming use will yield a reasonable return (Matter of Douglaston Civic Assn. v Klein, supra; Matter of Town Bd. v Zoning Bd. of Appeals, 161 AD2d 647, lv denied 76 NY2d 709; Matter of Colonna v Board of Stds. & Appeals, supra; Matter of Marchese v Koch, supra). Specifically, a use variance allowing otherwise prohibited residential use in a manufacturing district should not be granted on the basis that residential use would earn a higher return (Matter of 35 Broadway Co. v Bennett, 161 AD2d 767). Although the present developer analyzed two conforming uses, hotels and office space, finding them to have practical difficulties arising from the lot sizes and configurations, it failed to analyze why joint living-work quarters for artists, the use for which the district was especially zoned, were thereby precluded. However, an applicant for a use variance is entitled to relief only if, by “dollars-and-cents proof,” he cannot realize a reasonable return by any conforming use (Matter of Wheeler v City of Elmira, 63 NY2d 721; Matter of Village Bd. v Jarrold, supra). Hence, “[t]o establish unnecessary hardship the applicant must demonstrate * * * for each and every use permitted under the zoning regulations, that the property in question cannot yield a reasonable return” (Matter of Park Hill Residents’ Assn. v Cianciulli, 234 AD2d 464; accord, Matter of Aiello v Saladino, 132 AD2d 1002; Cortese v Avis Rent A Car Sys., 167 AD2d 940; Matter of Drake v Zoning Bd. of Appeals, 183 AD2d 1031). If the property is presently being used for a nonconforming use, the applicant must even demonstrate the infeasibility of continuing that use as well as conforming uses before seeking an additional nonconforming use (Matter of Rostlee Assocs. v Amelkin, 121 AD2d 725, lv denied 69 NY2d 603), which underscores the parsimony with which use variances are considered. In the present case, continuing the sites as parking lots, while, perhaps, aesthetically undesirable, demonstrably remains an economically viable use, while development of the sites for joint living-work quarters for artists is not, on this record, economically precluded. If the applicant fails to demonstrate an inability to realize a reasonable return under any permissible use, there is no rational basis to grant the variance (Matter of Park Hill Residents’ Assn. v Cianciulli, supra). That principle applies here.
The majority, noting a responsive comment by applicants’ consultant regarding loft development, finds joint living-work *90quarters for artists to have been explicitly considered. However, this option appears to have been avoided by the applicants until pressed by an opposing architect during the comment period. The response consisted of a brief comment that the architect’s alternative proposal, consisting of only two lofts per floor, averaging 3,500 square feet each, was too costly. This, of course, is not tantamount to a study fully analyzing, and then rejecting, the general proposed use as not being economically viable. Elsewhere, the consultant submitted a responsive memorandum contending that the necessary rent for units exceeding 1,200 square feet would exceed rent that could be anticipated for that market and location. However, while that factor, briefly and only responsively noted, might be relevant to area consideration and if accurate might warrant flexibility on area requirements, the responsive memorandum does not explain why artistic use is necessarily compromised.
Turning to what constitutes a reasonable return, case law provides some parameters, although, of course, caution argues against making too close a comparison of possibly disparate locations and time periods. A use variance was properly denied when the applicants’ evidence indicated that, although the nonconforming use could earn 10% to 11% the conforming use would still earn 5.7% (Matter of Ryan v Miller, 164 AD2d 968). In another case, a 3.6% return for a conforming residential use was found not to be unreasonable, although contrasted with a projected greater return for nonconforming commercial use (Matter of Kingsley v Bennett, supra). It bears repeating that, using the applicants’ own financials, current use for parking purposes generates returns of 6% or 7% for the sites. The majority reads my observation as a determination that this return is reasonable as a matter of law. I do not. It is merely an observation that the matter was never directly analyzed, with the return from current use simply ignored by BSA. The mere fact that a reference to the present return can be found within a voluminous record is not, in my view, tantamount to BSA analyzing the issue. I remain open-minded to an explanation from the agency specializing in these matters why this is, or is not, a reasonable rate of return.
On the basis of the present record, BSA could not have rationally concluded that any permissible use Was economically untenable. As such, allowing a drastic departure from the Zoning Resolution’s use designations for this neighborhood cannot be sustained.
The evidence also does not satisfy the remaining criteria set forth in Zoning Resolution § 72-21 as preconditions to the issu-*91anee of the variances. Under section 72-21 (c), community impact must be evaluated. The majority contends that this project will not “destroy” neighborhood character. This is an overstatement of the standard. As a general proposition, the greater the deviation, the more likely it is that the impact on community character will be severe (Matter of Consolidated Edison Co. v Hoffman, 43 NY2d 598, supra; Matter of National Merritt v Weist, 41 NY2d 438). The recent trend is to look at the over-all effect of multiple significant variances to evaluate the likelihood of a substantial effect (Matter of Tetra Bldrs. v Scheyer, 251 AD2d 589 [area variances; request for departures of 12% of total area and 70% of back yard area; cumulative effect warranted denial]; accord, Matter of Becvar v Scheyer, 250 AD2d 842). CPC itself characterized the proposal as presenting “extensive departures from zoning requirements,” a characterization supported by the record. The significantly changed use and expanded bulk requirements in this case beg the question how community character, even if measured only by the scale of this project, will not be affected. The question was unanswered by BSA. Rather, in contrast to the CPC suggestions as to how to minimize the adverse impact on community character, the applicants, looking across the roadway to a different community, simultaneously contended that the impact on SoHo would be minimal, in that the proposed development is in line with the study’s comparables and that, in any event, the SoHo community character has already changed. The effort is disingenuous.
Similar reasoning compels the conclusion that the applicants have failed to demonstrate that the variances sought are the minimal necessary to afford relief (ZR § 72-21 [e]) — relief being defined as the need to generate a reasonable financial return.
Further, one may posit that the problem is self-created (ZR § 72-21 [d]). Property purchasers, especially those seasoned in the real estate field, are charged with knowledge of, and are bound by, applicable zoning restrictions at the time of the purchase (Matter of Weisman v Zoning Bd. of Appeals, 260 AD2d 487). As the Court of Appeals stated, “[w]e could end this opinion at this point by saying that one who thus knowingly acquires land for a prohibited use, cannot thereafter have a [use] variance” (Clark v Board of Zoning Appeals, supra, at 89). These particular applicants, then, in purchasing, and seeking to develop, these lots, cannot claim injury arising from the fact of those zoning restrictions (Matter of Kingsley v Bennett, supra [realtor purchased property with knowledge of zoning *92restrictions; injury self-inflicted]; accord, Matter of Drake v Zoning Bd. of Appeals, supra [residential property also used by owner for his construction business; knowledge of zoning restrictions at time of purchase makes injury self-inflicted]). If the strategy was to purchase an underdeveloped lot at an attractive price and then to plead an injury imposed by existing zoning, that strategy should not be rewarded. Nor should a use variance be a means of mitigating a possibly poor business decision with regard to purchase or operation of the property which is the actual genesis of the applicant’s own hardship (Matter of Carriage Works Enters. v Siegel, 118 AD2d 568, 570).
Departing from this point, in view of the sophistication of these real estate operatives, the lack of ambiguity in the zoning requirements, and CPC’s own, apparently more neutral, evaluation of the scale of the project, it is difficult to see how these applicants, on this record, are entitled to the relief they seek. Again, that relief seems driven more by the desire to maximize economical use of the property rather than their being precluded from permissible uses.
When use and area variance applications are coupled together, if the use variance must be denied, it moots the area variance application (Matter of Delmarco v Zoning Bd. of Appeals, supra). However, my review of the BSA resolution with a view toward the area variances, considered apart from use, can be briefly summarized. First, as noted above, the site configurations are hardly unique in terms of the proposed use. The variance is substantial in relation to the zoning requirement (see, e.g., Matter of Tetra Bldrs. v Scheyer, supra). The substantially increased bulk will likely produce a substantial change in neighborhood character. There is a feasible nonvariance method to alleviate the practical difficulties: adhere to conforming use and/or reduce the bulk as suggested by CPC. Agencies also look to whether, in view of the manner in which the problem arose, the interest of justice will be served by allowing the variance (Matter of Gersten v Cullen, 203 AD2d 744). Courts, in evaluating the urgency of the applicant’s putative practical difficulties, have also looked to whether the impact on the public interest outweighs the financial impact on the applicant (Matter of Licari v Scheyer, 193 AD2d 604). Again, the sophistication of these owners and the developer in the real estate field necessarily must be factored into any interest of justice analysis as well as any balancing of impact. These factors are not dispositive, but they support the conclusion *93that requiring the owners and developer to adhere to neighborhood zoning does not work an injustice. Of course, if flexibility on area requirements had been sought with regard to a permissible use, in line with the CPC recommendation, the applicants might have come closer to achieving more beneficial use of the site, but that is not the application before us and any result in such a case remains speculative. Finally, even where an applicant demonstrates some entitlement to an area variance, we may validly consider whether “the spirit of the ordinance will be observed notwithstanding the variance” (Matter of Gianchetta v Wilens, supra, at 319), which, where the legislative mandate is basically being debunked, warrants a negative conclusion.
The BSA resolution fails to account for these deficiencies. BSA could not rationally have concluded that the developer’s proposal and supporting materials satisfied each and every one of the five findings mandated by Zoning Resolution § 72-21.
Accordingly, I would affirm the judgment.
Sullivan, P. J., and Mazzarelli, J., concur with Friedman, J.; Tom and Rubin, JJ., dissent in a separate opinion by Tom, J.
Judgment, Supreme Court, New York County, entered January 19, 1999, reversed, on the law, without costs, the petition denied, and the resolutions of the Board of Standards and Appeals granting the variances reinstated.