Twenty years ago, widespread racial segregation threatened to rip civil society asunder. In response, Congress adopted broad remedial provisions to promote integration. One such statute, Title VIII of the Civil Rights Act of 1968, 42 U.S.C. §§ 3601-3631 (1982 & Supp. Ill 1985) (“Fair Housing Act”), was enacted “to provide, within constitutional limitations, for fair housing throughout the United States.” 42 U.S.C. § 3601. Today, we are called upon to decide whether an overwhelmingly white suburb’s zoning regulation, which restricts private multi-family housing projects to a largely minority “urban renewal area,” and the Town Board’s refusal to amend that ordinance to allow construction of subsidized housing in a white neighborhood violates the Fair Housing Act.
The Huntington Branch of the National Association for the Advancement of Colored People (NAACP), Housing Help, Inc. (HHI), and two black, low-income residents of Huntington appeal from an adverse judgment of the United States District Court for the Eastern District of New York (Glasser, J.), following a bench trial, in their suit against the Town of Huntington (the Town) and members of its Town Board. Appellants allege that the Town violated Title VIII by restricting private construction of multi-family housing to a narrow urban renewal area and by refusing to rezone the parcel outside this area where appellants wished to build mul-ti-family housing.1 Specifically, appellants sought to construct an integrated, multifamily subsidized apartment complex in Greenlawn/East Northport, a virtually all-white neighborhood. The Town’s zoning ordinance, however, prohibited private construction of multi-family housing outside a small urban renewal zone in the Huntington Station neighborhood, which is 52% minority. Thus, appellants petitioned the Town to revise its code to accommodate the project. When the Town refused, appellants brought this class-action2 to compel the change under Title VIII.
This dispute has been before this court before, in response to a district court determination that appellants lacked standing to bring their complaint. Huntington Branch NAACP v. Town of Huntington, 530 F.Supp. 838 (E.D.N.Y.1981). We reversed, holding that standing could not be denied because of the lack of United States Department of Housing and Urban Development (HUD) Section 8 funds in a particular year. Huntington Branch NAACP v. Town of Huntington (Huntington I), 689 F.2d 391 (2d Cir.1982), cert. denied, 460 U.S. 1069, 103 S.Ct. 1523, 75 L.Ed.2d 947 (1983).3 In so doing, we expressly held that, contrary to the Town’s assertion, appellants were not required to exhaust local remedies by filing a formal application for rezoning. Huntington I, 689 F.2d at 393 n. 3.
In the case currently appealed, Huntington Branch NAACP v. Huntington, 668 F.Supp. 762 (E.D.N.Y.1987), the district court refused to invalidate the zoning restriction. The district judge, however, incorrectly employed an intent-based standard for the disparate impact claim asserted here both in analyzing the showing of effect and in scrutinizing the validity of the Town’s reasons for rejection. Accordingly, we reverse and, finding a Title VIII viola*929tion, grant appellants’ request for site-specific relief. Although prior opinions exhaustively document the facts, a brief statement of the pertinent points will aid in the understanding of this case.
Huntington is a town of approximately 200,000 people located in the northwest corner of Suffolk County, New York. In 1980, 95% of its residents were white. Blacks comprised only 3.35% of the Town’s population and were concentrated in areas known as Huntington Station and South Greenlawn. Specifically, 43% of the total black population lived in four census tracts in Huntington Station and 27% in two census tracts in the South Greenlawn area. Outside these two neighborhoods, the Town’s population was overwhelmingly white. Of the 48 census tracts in the Town in 1980, 30 contained black populations of less than 1%.
The district court found that the Town has a shortage of affordable rental housing for low and moderate-income households. The Town’s Housing Assistance Plan (HAP), which is adopted by the Town Board and filed with HUD as part of Huntington’s application for federal community development funds, reveals that the impact of this shortage is three times greater on blacks than on the overall population. Under the 1982-1985 HAP, for example, 7% of all Huntington families required subsidized housing, while 24% of black families needed such housing.
In addition, a disproportionately large percentage of families in existing subsidized projects are minority. In Gateway Gardens, a public housing project built in 1967, 38 of 40 units were occupied by blacks and Hispanics in 1984. Seventy-four percent of those on the project’s waiting list were minority. In Whitman Village, a 260-unit HUD subsidized development built in 1971, 56% of the families were minority in 1984. Lincoln Manor, which was built in 1980, is a 30-unit HUD Section 8 project. Thirty percent of the households and 45% of those on the waiting list were minority in 1984. Under a HUD Section 8 program, lower income families can obtain certificates to supplement their rent. Each family, however, must locate its own apartment. In January 1984, 68% of families holding certificates and 61% of those on the waiting list were minority.
Although a disproportionate number of minorities need low-cost housing, the Town has attempted to limit minority occupancy in subsidized housing projects. Michael Miness, the Director of Huntington’s Community Development agency and responsible for developing the Town’s low-cost housing, and Angela Sutton, Executive Director of the Huntington Housing Authority, repeatedly told whites opposing the Lincoln Manor project that they would impose a racial quota on occupancy. When HUD reviewed the project’s management plan which established 5% minority occupancy, however, it advised the Huntington Housing Authority that it would not permit a racial quota at Lincoln Manor. The Town similarly attempted to impose racial quotas on occupancy at a proposed 150-unit subsidized housing project in Huntington Station on the Melville Industrial Associates (MIA) site. When Alan H. Wiener, HUD’s Area Director, wrote Kenneth C. Butterfield, Town Supervisor, that “limitations on minority occupancy of housing on the Huntington Station site are not justifiable and will not be permitted,” (Letter of June 19, 1981, E-18), the Town Board unanimously passed a resolution withdrawing its support for the project because they could not “ensure a particular ethnic mix.” (Huntington Town Board Resolution re: Huntington Station Urban Renewal Project, June 23, 1981, E-17.)
Under the Town’s zoning ordinance, mul-ti-family housing is permitted only in an “R-3M Apartment District.” The relevant portion of section 198-20(A) provides:
Use regulations. In the R-3M Apartment District, a building or premises shall be used only for the following purposes:
(1) Any use permitted in the R-80, R-15 and R-5 Residence Districts.
(2) Multiple-family dwellings which constitute an approved public housing project to be owned, maintained and op*930erated by the Housing Authority of the Town of Huntington.
(3) Multiple-family dwellings where such dwellings constitute an element in a formally approved land use or a use plan for all or part of an urban renewal area which has been designated as such under the provisions of Article 15 of the General Municipal Law.
On its face, then, this provision limits private construction of multi-family housing to the Town’s urban renewal area, where 52% of the residents are minority.4 It does permit the Huntington Housing Authority (HHA) to build multi-family housing town-wide. But HHA’s only project, Gateway Gardens, is in the urban renewal zone. The private housing projects are also in or nearby the urban renewal area. Whitman Village is adjacent to Gateway Gardens in census blocks that are over 40% minority. Lincoln Manor, only a few blocks from the projects in the urban renewal area, is also in a racially impacted census block.
The Town’s zoning ordinance also includes a special category for multi-family housing for senior citizens called “R-RM Retirement Community District.” Only one such development — Paumanack Village — has been built in Huntington. It is the only multi-family housing for low income people which is situated in an overwhelmingly white neighborhood. The development itself is largely white, having a black occupancy of 3%.
Only one vacant parcel of land in Huntington currently is zoned R-3M and thus would be eligible for the appellants’ proposed development: the MIA site, which is at the northeast corner of Broadway and New York Avenue, is partially zoned C-6 and partially zoned R-3M. The Town in 1980 requested pre-approval for 150 units of Section 8 housing on this site.5
In response to the great need for subsidized housing in the Town, HHI decided to sponsor an integrated housing project for low-income families. HHI determined that the project could foster racial integration only if it were located in a white neighborhood outside the Huntington Station and South Greenlawn areas. This decision eliminated consideration of the MIA site, the only vacant R-3M property located in the urban renewal area.
In its effort to create racially integrated, low-cost housing, HHI actively sought the assistance of Town officials. Specifically, HHI’s Executive Director, Marianne Gar-vin, and HHI Board members met repeatedly with Michael Miness. In response to Miness’s suggestion that HHI pursue rehabilitating existing structures before fo-cussing on new construction, HHI commissioned a study in 1979 to assess whether any of the vacant schools were suitable for the housing project. After narrowing the possibilities to the Green Meadow School, HHI determined that this location was inappropriate for a low-cost housing development. Throughout 1979, Miness assured HHI representatives that existing zoning should not impede their efforts because the Town Board would amend the zoning ordinance if it supported the organization’s project.
After a lengthy search, HHI determined that a 14.8 acre parcel located at the corner of Elwood and Pulaski Roads in the Town was well suited for a 162-unit housing project. This flat, largely cleared and well-drained property was near public transportation, shopping and other services, and immediately adjacent to schools.
*931Ninety-eight percent of the population within a one-mile radius of the site is white. HHI set a goal of 25% minority occupants. The district court found that “a significant percentage of the tenants [at Matinecock Court] would have belonged to minority groups.” Huntington, 668 F.Supp. at 785. HHI officials determined that the property was economically feasible and offered a lengthy option period.
Prior to purchasing the option for the property, Garvin asked Miness to visit the property and evaluate it. Garvin testified that, although Miness told Garvin he would not give an opinion before HHI secured an option, he assured her that the property’s R-40 designation (single family homes on one-acre lots) should not be an obstacle because the Town Board, if it supported the project, would simply amend the zoning ordinance.
HHI obtained its option to purchase the Elwood-Pulaski parcel on January 23, 1980. Garvin again called Miness and invited him to visit the site. She testified that he responded that he was familiar with the property and believed it was a good location for development.
Throughout 1980, HHI sought to advance its project by gaining the approval of the Town Board to rezone the property to R-3M from its R-40 designation. The opinion below recites the multitudinous disputes which color the parties’ versions of the events of those months, especially concerning the apparently “illegal” closed meeting between the Board and HHI on February 22,1980. Certain facts, however, are not in contention. Robert Ralph, a director of HHI, addressed the Town Board on February 26, 1980, at a public hearing. The district court found that he filed a document requesting “a commitment by the Town to amend the zoning ordinance to allow multi-family rental construction by a private developer.” Huntington, 668 F.Supp. at 784 (quoting Presentation to the Huntington Town Board at the Community Development Block Grant Hearing 2 (February 26,1980)). In August 1980, HHI and National Housing Partnership, an owner-manager of federally subsidized housing, filed a joint application with HUD for Section 8 funding for the project.
At the time HHI applied for the Section 8 funding, Huntington had a Housing Assistance Plan, which had been approved by HUD. Pursuant to the provisions of the Housing and Community Development Act of 1974, 42 U.S.C. §§ 5301-20 (1982 & Supp. Ill 1985), when a town has such a plan, HUD must refer a Section 8 application to the Town for comment. In an October 14, 1980, letter to Alan H. Weiner, HUD Area Manager, Town Supervisor Kenneth C. Butterfield set forth seven reasons why Huntington opposed the project. It reads, in pertinent part, as follows:
The Town’s professional staff in the Planning, Legal and Community Development Departments have reviewed the proposal and have submitted the following comments:
1. The HUD-appróved Housing Assistance Plan (both the three-year goal submitted with the Community Development Block Grant 1979-80 application and the annual goal submitted with the 1980-1981 Community Development Block Grant) contains no “new construction” units as a program goal.
2. The plan for development cannot be carried out within the existing single family R-40 (1 acre) zoning.
3. The development is located at the intersection of two heavily trafficked streets.
4. The site plan presents a poor parking plan in terms of location with respect to the units, substandard in size and the lack of streets results in very poor fire protection access.
5. The development is located adjacent to both the Long Island Railroad as well as a ’LILCO substation. This is in addition to the heavy traffic conditions.
6. The site plan shows recreation and/or play areas very inadequate for the number and type of dwelling units being proposed.
7. The three and four-bedroom units are quite undersized; have poor layout; bedrooms are much too small; liv*932ing space is unrealistic; no storage; one full and two half-baths for a family of 6 to 8 is not realistic.
In conclusion, I do not recommend HUD approval of this proposal based on the material reviewed and the comments presented above.
When the proposal became public, substantial community opposition developed. A group called the Concerned Citizens Association was formed, and a petition containing 4,100 signatures against the proposal was submitted to the Town Board. A protest meeting in November drew about 2,000 persons. Supervisor Butterfield was the principal speaker and assured the audience of his opposition to the project. Mati-necock Court came before the Town Board at a meeting on January 6, 1981. The Board rejected the proposed zoning change and adopted the following resolution:
WHEREAS, it has been proposed by HOUSING HELP, INC., a private nonprofit group, that Huntington’s zoning code be changed in order to build 162 federally-subsidized apartments for low to moderate income people at Elwood and Pulaski Roads in the Elwood section of the Town of Huntington; and
WHEREAS, the Town Board has studied the various aspects of the proposal for a zoning change for 162 apartments at the said location of Elwood and Pulaski Roads;
NOW, THEREFORE,
THE TOWN BOARD finds that although favoring housing for the senior citizens and others, in appropriate areas, that the location referred to herein is not an appropriate location due to lack of transportation, traffic hazard and disruption of the existing residential patterns in the Elwood area and requests that the Department of Housing and Urban Development (HUD) reject the application by HOUSING HELP, INC.
Huntington Town Board Resolution re: Housing at Elwood and Pulaski Roads (January 6, 1981).
The district court based its refusal to order rezoning on three alternative grounds: (1) appellants never formally applied for rezoning; (2) even if they had applied, they failed to make the requisite prima facie showing of discriminatory effect; and (3) even if they had demonstrated discriminatory effect, the city had rebutted it by articulating legitimate, non-pretextual justifications. We now consider each ground separately.
The district court, in its first alternative holding, found that appellants failed to exhaust their available remedies because they did not formally apply to rezone their parcel. Specifically, Judge Glasser noted that HHI never filed an application to rezone its option property to R-3M status. Relying on Robinson v. 12 Lofts Realty, Inc., 610 F.2d 1032 (2d Cir.1979), he held that appellants must demonstrate that they actually applied. Quoting from testimony of Robert Ralph, an HHI director, and Marianne Garvin, HHI’s Executive Director, he concluded that no formal application was ever in fact made. Huntington, 668 F.Supp. at 784-85. This court has already rejected this reasoning. Huntington I, 689 F.2d at 393 n. 3.
Despite the myriad of factual disputes surrounding HHI’s dealings with the Town Board, two facts are crystal clear: HHI requested the Town Board to amend the code to allow R-3M zoning townwide at a formal meeting on February 26, 1980, and the Board, in its resolution of January 6, 1981, deemed this application a “propos[al] ... that Huntington’s zoning code be changed in order to build 162 federally-subsidized apartments for low to moderate income people at Elwood and Pulaski Roads_” During the numerous contacts with HHI directors and members, Town officials never mentioned that HHI should be pursuing a different application process. Both parties thus clearly understood that an application for a zoning change had been made. As this court determined in Huntington I, no further petitions, formal or informal, were necessary. Huntington I, 689 F.2d at 393 n. 3. Moreover, the Town’s refusal to amend the zoning code rendered meaningless a request to change the zoning on the Elwood-Pulaski property, as R-3M classifications were re*933served for property within the urban renewal area, and there were no other multifamily housing designations.
In its second holding, the court adopted the four-prong disparate impact test set out in Metropolitan Housing Dev. Corp. v. Village of Arlington Heights, 558 F.2d 1283, 1287-90 (7th Cir.1977), cert. denied, 434 U.S. 1025, 98 S.Ct. 752, 54 L.Ed.2d 772 (1978) (Arlington Heights II), and concluded that, even if appellants applied for a rezoning change, they had failed to make out a prima facie case. The court considered:
(1) how strong is the plaintiffs showing of discriminatory effect; (2) is there some evidence of discriminatory intent, though not enough to satisfy the constitutional standard of Washington v. Davis [, 426 U.S. 229, 96 S.Ct. 2040, 48 L.Ed.2d 597 (1976) ]; (3) what is the defendant’s interest in taking the action complained of; and (4) does the plaintiff seek to compel the defendant to affirmatively provide housing for members of minority groups or merely to restrain the defendant from interfering with individual property owners who wish to provide such housing.
Huntington, 668 F.Supp. at 781 (quoting Arlington Heights II, 558 F.2d at 1290). On the first prong, the court found that the showing of discriminatory effect was “not particularly strong.” Huntington, 668 F.Supp. at 786. Although the judge held that a shortage of rental housing existed, that a disproportionately large percentage of the households using subsidized rental units are minority, and, accordingly, that a “significant percentage” of Matinecock Court tenants would be minority, 668 F.Supp. at 785, he compared the larger absolute number of white poor (22,160) with minority poor (3,671) and concluded that the beneficiaries “might not come disproportionately from minority groups.” Huntington, 668 F.Supp. at 786. On the second factor, Judge Glasser found no proof of segregative intent, deeming this a plus in the Town’s favor. In so holding, he determined that appellants had failed to prove that the Town was motivated by segregative intent when it confined subsidized housing to the urban renewal area. The third prong of Arlington Heights II, he concluded, was satisfied by “legitimate, nondiscriminatory reasons for [the Town’s] conduct.” Huntington, 668 F.Supp. at 786. He deemed the fourth factor to cut in favor of appellants because they were not asking the Town to provide housing. Nevertheless, because the first three factors weighed in favor of appellees, he held that the appellants had failed to demonstrate a prima facie case.
In its third rationale, the court applied the test set forth in McDonnell Douglas Corp. v. Green, 411 U.S. 792, 93 S.Ct. 1817, 36 L.Ed.2d 668 (1973), as a final determination on the merits for Title VII disparate treatment cases. According to this formula, if plaintiffs establish a prima facie case of disparate treatment, the “burden shifts to the defendant to articulate some legitimate, nondiscriminatory reason for the employee’s rejection.” McDonnell Douglas, 411 U.S. at 802, 93 S.Ct. at 1824. If defendants meet this burden, plaintiffs must show that the legitimate justifications offered were pretextual and not the employer’s true reasons. McDonnell Douglas, 411 U.S. at 804, 93 S.Ct. at 1825. Applying this test, the court below found that, even if appellants had demonstrated a pri-ma facie showing of discriminatory effect, the Town’s justifications for rejecting the project were legitimate and non-discriminatory reasons which “have not been exposed as pretextual.” Huntington, 668 F.Supp. at 787.
We find it convenient to discuss Judge Glasser’s second and third holdings together. In considering them, we start by pointing out that this case requires what has been called “disparate impact” or “disparate effects” analysis, not “disparate treatment” analysis. A disparate impact analysis examines a facially-neutral policy or practice, such as a hiring test or zoning law, for its differential impact or effect on a particular group. Sobel v. Yeshiva Univ., 839 F.2d 18, 28 (2d Cir.1988). Disparate treatment analysis, on the other hand, involves differential treatment of similarly situated persons or groups. The *934line is not always a bright one, e.g., Sobel, 839 F.2d at 28-29, but does adequately delineate two very different kinds of discrimination claims.
Here, appellees would collapse the distinction between disparate impact and disparate treatment by characterizing this as a “mixed” impact and treatment case. Thus, they argue, “treatment” analysis should be applied to the Town’s refusal to rezone the Matinecoek Court site, while “impact” analysis should be applied to the zoning ordinance’s restriction of multi-family housing to the urban renewal area. Under appellees’ methodology, however, every disparate impact case would include a disparate treatment component. This cannot be the case. There is always some discrete event (refusal to rezone property, refusal to hire someone because he did not graduate from high school) which touches off litigation .challenging a neutral rule or policy.
The prima facie standard for Title VIII disparate impact cases involving public defendants is a question of first impression in this circuit. In Boyd v. Lefrak Org., 509 F.2d 1110 (2d Cir.), rehearing denied, 517 F.2d 918 (2d Cir.), cert. denied, 423 U.S. 896, 96 S.Ct. 197, 46 L.Ed.2d 129 (1975), a divided panel, considering a Title VIII case against a private defendant, doubted the relevance of the methodology used in Title VII litigation. 509 F.2d at 1114. Since then, however, we have pointedly accepted the relevance of Title VII cases to Title VIII cases. See United States v. Starrett City Assoc., 840 F.2d 1096, 1101 (2d Cir.1988); Robinson, 610 F.2d at 1036-37. Thus, even if the views expressed in Lefrak still apply in a Title VIII case against a private defendant, a matter of considerable uncertainty, the disparate impact approach of Title VII cases is fully applicable to this Title VIII case brought against a public defendant.
Under disparate impact analysis, as other circuits have recognized, a prima fa-cie case is established by showing that the challenged practice of the defendant “actually or predictably results in racial discrimination; in other words that it has a discriminatory effect.” United States v. City of Black Jack, 508 F.2d 1179, 1184-85 (8th Cir.1974), cert. denied, 422 U.S. 1042, 95 S.Ct. 2656, 45 L.Ed.2d 694 (1975). The plaintiff need not show that the decision complained of was made with discriminatory intent. United States v. Yonkers Board of Education,6 837 F.2d 1181, 1217 (2d Cir.1987); Robinson, 610 F.2d at 1036-38; Resident Advisory Board v. Rizzo, 564 F.2d 126, 146-48 (3d Cir.1977), cert. denied, 435 U.S. 908, 98 S.Ct. 1457, 55 L.Ed.2d 499 (1978); Smith v. Anchor Bldg. Corp., 536 F.2d 231, 233 (8th Cir.1976); Black Jack, 508 F.2d at 1184-85. Refusal to require intent in disparate impact cases is entirely consistent with our prior decisions. See Starrett City, 840 F.2d at 1100; Yonkers, 837 F.2d at 1217; Robinson, 610 F.2d at 1036-38. In determining whether discriminatory effect is sufficient, we look to congressional purpose, as gleaned from the legislative history of Title VIII, related Title VII jurisprudence, and practical concerns. Although none of these considerations is alone determinative, taken together they strongly suggest that discriminatory impact alone violates Title VIII.
The Act’s stated purpose to end discrimination requires a discriminatory effect standard; an intent requirement would strip the statute of all impact on de facto segregation. Comment, Justifying a Discriminatory Effect Under the Fair Housing Act: A Search for the Proper Standard, 27 UCLA L.Rev. 398, 406 (1979) (hereafter UCLA Comment). Congress appears not to have resolved this precise question. Nonetheless, the legislative history provides some indication that an intent standard was not contemplated. The Rizzo court attached significance to the Senate’s rejection of an amendment that would have required “proof of discriminatory intent to succeed in establishing a Title VIII claim.” *935Rizzo, 564 F.2d at 147. The amendment, however, was far less sweeping than Rizzo suggests because it applied only to a single-family owner-occupied house. UCLA Comment, 27 UCLA L.Rev. at 406-7 n. 45. Nevertheless, its rejection does underscore congressional willingness to broaden Title VIII to encompass segregation resulting from the application of facially neutral rules, even in the absence of discriminatory intent.
More persuasive is the parallel between Title VII and Title VIII noted by both courts and commentators. E.g., Starrett City, 840 F.2d at 1101; Smith v. Town of Clarkton, 682 F.2d 1055, 1065 (4th Cir.1982); Rizzo, 564 F.2d at 146-48; Arlington Heights II, 558 F.2d at 1288-89; United States v. Hunter, 459 F.2d 205, 216-17 & n. 12 (4th Cir.), cert. denied, 409 U.S. 934, 93 S.Ct. 235, 34 L.Ed.2d 189 (1972); Comment, Applying the Title VII Prima Facie Case to Title VIII Litigation, 11 Harv.C.R.-C.L.L.Rev. 128, 158-60 (1976). The two statutes are part of a coordinated scheme of federal civil rights laws enacted to end discrimination; the Supreme Court has held that both statutes must be construed expansively to implement that goal. See, e.g., Trafficante v. Metropolitan Life Ins. Co., 409 U.S. 205, 211-12, 93 S.Ct. 364, 367-68, 34 L.Ed.2d 415 (1972) (Fair Housing Act must be generously construed to foster integration); Griggs v. Duke Power Co., 401 U.S. 424, 429-36, 91 S.Ct. 849, 852-56, 28 L.Ed.2d 158 (1971) (Title VII should be interpreted broadly to achieve equal employment opportunity). Courts and commentators have observed that the two statutes require similar proof to establish a violation. See, e.g., Rizzo, 564 F.2d at 146-48; United States v. West Peachtree Tenth Corp., 437 F.2d 221, 226-27 (5th Cir.1971); Hunter, 459 F.2d at 217-18; Schwartz, The Fair Housing Act and “Discriminatory Effect”: A New Perspective, 11 Nova L.J. 71, 73 (1987); Harv.C.R.-C.L.L.Rev. Comment at 158-60. Thus, just as the Supreme Court held that Title VII is violated by a showing of discriminatory effect, Griggs, 401 U.S. at 429-36, 91 S.Ct. at 852-56, we hold that a Title VIII violation can be established without proof of discriminatory intent.
Practical concerns also militate against inclusion of intent in any disparate impact analysis. First, as this court noted in Robinson, “clever men may easily conceal their motivations.” 610 F.2d at 1043 (quoting Black Jack, 508 F.2d at 1185). This is especially persuasive in disparate impact cases where a facially neutral rule is being challenged. Often, such rules bear no relation to discrimination upon passage, but develop into powerful discriminatory mechanisms when applied. Second, inclusion of intent undermines the trial judge’s inquiry into the impact of an action. The lower court’s insistence on probing the “pretextual” nature of appellees’ justifications vividly demonstrates the extent to which an intent-based standard can infect an analysis and draw it away from its proper focus. Accordingly, we will not require proof of discriminatory intent to establish a prima facie disparate impact case under Title VIII.
Confusion concerning the content of a prima facie disparate impact case under Title VIII has been engendered by the tendency of some courts to consider factors normally advanced as part of a defendant’s justification for its challenged action in assessing whether the plaintiff has established a prima facie case. That appears to have occurred in this case when Judge Glasser analyzed the factors set forth in Arlington Heights II in the course of concluding that a prima facie case was not established. See also Keith v. Volpe, 618 F.Supp. 1132, 1148 (C.D.Cal.1985) (Pregerson, C.J., sitting by designation). Though, as will shortly appear, we are not persuaded to adopt precisely the formulation of the Arlington Heights II factors, we agree with the Third Circuit that factors such as those mentioned in Arlington Heights II are to be considered in a final determination on the merits rather than as a requirement for a prima facie case. See Rizzo, 564 F.2d at 148 n. 32. Nothing in Arlington Heights II indicates the court saw its test as anything but a final determination on the merits. Furthermore, treating the four factors as steps necessary to *936make out a prima facie case places too onerous a burden on appellants. The legislative history of the Fair Housing Act, although sparse, argues persuasively against so daunting a prima facie standard.
As Senator Mondale, the bill’s author, said, the proposed law was designed to replace the ghettos “by truly integrated and balanced living patterns.” 114 Cong. Rec. 3422 (1968) quoted in Trafficante, 409 U.S. at 211, 93 S.Ct. at 368. In Traffi-cante, the Supreme Court held that Title VIII should be broadly interpreted to fulfill this congressional mandate. 409 U.S. at 212, 93 S.Ct. at 368. Moreover, both the majority and the thoughtful dissent in a recent Title VIII case in this circuit, Star-rett City, agree: Congress intended that broad application of the anti-discrimination provisions would ultimately result in residential integration. Employing the test in Arlington Heights II as a prima facie hurdle would cripple Title VIII.
Once a prima facie case of adverse impact is presented, as occurred here, the inquiry turns to the standard to be applied in determining whether the defendant can nonetheless avoid liability under Title VIII. The Third Circuit in Rizzo and the Seventh Circuit in Arlington Heights II have both made useful contributions to this inquiry. Both circuits essentially recognize that in the end there must be a weighing of the adverse impact against the defendant’s justification. As phrased by the Third Circuit, the defendant must prove that its actions furthered, in theory and in practice, a legitimate, bona fide governmental interest and that no alternative would serve that interest with less discriminatory effect. Rizzo, 564 F.2d at 148-49. We agree with that formulation. Furthermore, according to the Third Circuit, “Title VIII criteria [would] emerge, then, on a case-by-case basis.” 564 F.2d at 149. The Seventh Circuit adds two other factors that can affect the ultimate determination on the merits. One factor is whether there is any evidence of discriminatory intent on the part of the defendant. Though we have ruled that such intent is not a requirement of the plaintiff's prima facie case, there can be little doubt that if evidence of such intent is presented, that evidence would weigh heavily on the plaintiff’s side of the ultimate balance. The other factor is whether the plaintiff is suing to compel a governmental defendant to build housing or only to require a governmental defendant to eliminate some obstacle to housing that the plaintiff itself will build. In the latter circumstance, a defendant would normally have to establish a somewhat more substantial justification for its adverse action than would be required if the defendant were defending its decision not to build.
In this case, we are obliged to refine the standard for assessing a Title VIII defendant’s justification somewhat beyond what was said in either Rizzo or Arlington Heights II. In Rizzo, two of the defendants offered no justification for the adverse decision, 564 F.2d at 149, and the municipal defendant offered only the entirely unacceptable apprehension of violence, 564 F.2d at 150. The Third Circuit therefore did not have anything of substance to weigh on the defendants’ side. In Arlington Heights II, the consideration of the defendant’s justification scarcely moved past inquiring whether the municipal defendant was acting within the scope of zoning authority granted by state law. 558 F.2d at 1293.
In considering the defendant’s justification, we start with the framework of Title VII analysis. When an employer’s facially neutral rule is shown to have a racially disproportionate effect on job applicants, that rule must be shown to be substantially related to job performance. See Griggs v. Duke Power Co., 401 U.S. 424, 91 S.Ct. 849, 28 L.Ed.2d 158 (1971). In a zoning case, the facially neutral rule is the provision of the zoning ordinance that bars the applicant and, in doing so, exerts a racially disproportionate effect on minorities. The difficulty, however, is that in Title VIII cases there is no single objective like job performance to which the legitimacy of the facially neutral rule may be related. A town's preference to maintain a particular zoning category for particular sections of the community is normally based on a variety of circumstances. The complexity of *937the considerations, however, does not relieve a court of the obligation to assess whatever justifications the town advances and weigh them carefully against the degree of adverse effect the plaintiff has shown. Though a town’s interests in zoning requirements are substantial, see Village of Belle Terre v. Boraas, 416 U.S. 1, 94 S.Ct. 1536, 39 L.Ed.2d 797 (1974); Village of Euclid v. Ambler Realty Co., 272 U.S. 365, 47 S.Ct. 114, 71 L.Ed. 303 (1926), they cannot, consistently with Title VIII, automatically outweigh significant disparate effects.
A district court’s findings of fact may not be set aside “unless clearly erroneous.” Fed.R.Civ.P. 52(a); Yonkers, 837 F.2d at 1217-18. But Rule 52(a), the Supreme Court has held, “does not inhibit an appellate court’s power to correct errors of law, including those that may infect a so-called mixed finding of law, or a finding of fact that is predicated on a misunderstanding of the governing rule of law.” Bose Corp. v. Consumers Union of United States, Inc., 466 U.S. 485, 501, 104 S.Ct. 1949, 1960, 80 L.Ed.2d 502 (1984); Pullman-Standard v. Swint, 456 U.S. 273, 287, 102 S.Ct. 1781, 1789, 72 L.Ed.2d 66 (1982); Inwood Laboratories, Inc. v. Ives Laboratories, Inc., 456 U.S. 844, 855 n. 15, 102 S.Ct. 2182, 2189 n. 15, 72 L.Ed.2d 606 (1982). With these principles in mind, we review Judge Glasser’s findings in two areas:7 the strength of the discriminatory effect and the import of the Town’s justifications.
The discriminatory effect of a rule arises in two contexts: adverse impact on a particular minority group and harm to the community generally by the perpetuation of segregation. Arlington Heights II, 558 F.2d at 1290. In analyzing Huntington’s restrictive zoning, however, the lower court concentrated on the harm to blacks as a group, and failed to consider the seg-regative effect of maintaining a zoning ordinance that restricts private multi-family housing to an area with a high minority concentration. Yet, recognizing this second form of effect advances the principal purpose of Title VIII to promote, “open, integrated residential housing patterns.” Otero v. New York Housing Authority, 484 F.2d 1122, 1134 (2d Cir.1973).
Seventy percent of Huntington’s black population reside in Huntington Station and South Greenlawn. Matinecock Court, with its goal of 25% minorities, would begin desegregating a neighborhood which is currently 98% white. Indeed, the district court found that a “significant percentage of the tenants” at Matinecock Court would belong to minority groups. Huntington, 668 F.Supp. at 785. The court, however, failed to take the logical next step and find that the refusal to permit projects outside the urban renewal area with its high concentration of minorities reinforced racial segregation in housing.8 This was erroneous. Similarly, the district court found that the Town has a shortage of rental housing affordable for low and moderate-income households, that a “disproportionately” large percentage of the households using subsidized rental units are minority citizens, and that a disproportionately large number of minorities are on the waiting lists for subsidized housing and existing Section 8 certificates. Huntington, 668 F.Supp. at 785. But it failed to recognize that Huntington’s zoning ordinance, which restricts private construction of multi-family housing to the largely minority urban renewal area, impedes integration by restricting low-income housing needed by mi*938norities to an area already 52% minority.9 We thus find that Huntington’s refusal to amend the restrictive zoning ordinance to permit privately-built multi-family housing outside the urban renewal area significantly perpetuated segregation in the Town.
On the question of harm to blacks as a group, the district court emphasized that 22,160 whites and 3,671 minorities had incomes below 200% of the poverty line, a cutoff close to the Huntington Housing Authority’s qualification standards. Huntington, 668 F.Supp. at 786. Thus, the district court focussed on the greater absolute number of poor whites compared with indigent minorities in Huntington. The district court, however, did not analyze the disproportionate burden on minorities as required by Griggs v. Duke Power Co., 401 U.S. 424, 91 S.Ct. 849, 28 L.Ed.2d 158 (1971).10 By relying on absolute numbers rather than on proportional statistics, the district court significantly underestimated the disproportionate impact of the Town’s policy. Thus, the district court perceived facts through a misapprehension of the applicable law and we must make our own findings at least as to the significance of the undisputed underlying facts. Bose, 466 U.S. at 501, 104 S.Ct. at 1959.
The parties have stipulated that 28% of minorities in Huntington and 11% of whites have incomes below 200% of the poverty line. What they dispute is the meaning of these statistics. Judge Glasser found that, as the Town contends, there is no showing of discriminatory effect because a majority of the victims are white. We disagree for reasons analogous to those the Supreme Court enumerated in Griggs. The disparity is of a magnitude similar to that in Griggs, where the Court found discriminatory an employer’s policy of hiring only high school graduates because 12% of black males in North Carolina had high school diplomas while 84% of white males were high school graduates. But the plaintiffs presented even stronger evidence reflecting the disparate impact of preventing the project from proceeding. Under the Huntington HAP for 1982-1985, 7% of all Huntington families needed subsidized housing, while 24% of the black families needed such housing. In addition, minorities constitute a far greater percentage of those currently occupying subsidized rental projects compared to their percentage in the Town’s population. Similarly, a disproportionately high percentage (60%) of families holding Section 8 certificates from the Housing Authority to supplement their rents are minorities, and an equally disproportionate percentage (61%) of those on the waiting list for such certificates are minorities. Therefore, we conclude that the failure to rezone the Matinecock Court site had a substantial adverse impact on minorities.11
In sum, we find that the disproportionate harm to blacks and the segregative impact on the entire community resulting from the refusal to rezone create a strong prima facie showing of discriminatory effect — far more than the Rizzo test would require. Thus, we must consider the Town’s asserted justifications.
*939Once a plaintiff has made a prima facie showing of discriminatory effect, a defendant must present bona fide and legitimate justifications for its action with no less discriminatory alternatives available. Riz-zo, 564 F.2d at 149. Following McDonnell Douglas, a disparate treatment case, Judge Glasser held that if appellees articulated a legitimate, nondiscriminatory reason for their conduct, appellants must show that the reason is a “pretext.” Huntington, 668 F.Supp. at 782. He went on to list the seven reasons in Butterfield’s October 14, 1980, letter to HUD and found them “legitimate, nondiscriminatory” reasons which “have not been exposed as pretextual.” Huntington, 668 F.Supp. at 787. The McDonnell Douglas test, however, is an intent-based standard for disparate treatment cases inapposite to the disparate impact claim asserted here. No circuit, in an impact case, has required plaintiffs to prove that defendants’ justifications were pretextual. In Black Jack, 508 F.2d at 1185, for example, the court required defendants to show that their conduct was necessary to promote a “compelling governmental interest.” The Third Circuit in Rizzo rejected this standard for its own test of legitimate and bona fide concerns.
The Rizzo approach has two components: (1) whether the reasons are bona fide and legitimate; and (2) whether any less discriminatory alternative can serve those ends. For analytical ease, the second prong should be considered first. Concerns can usually be divided between “plan-specific” justifications and those which are “site-specific.” “Plan-specific” problems can be resolved by the less discriminatory alternative of requiring reasonable design modifications. “Site-specific” justifications, however, would usually survive this prong of the test. Those remaining reasons are then scrutinized to determine if they are legitimate and bona fide. By that, we do not intend to devise a search for pretext. Rather, the inquiry is whether the proffered justification is of substantial concern such that it would justify a reasonable official in making this determination. Of course, a concern may be non-frivolous, but may not be sufficient because it is not reflected in the record.
Appellants challenge both the ordinance which restricts privately-built multifamily housing to the urban renewal area and th'e Town Board’s decision to refuse to rezone the Elwood-Pulaski site. All the parties and the district court judge, however, focussed on the latter issue. Indeed, appellees below simply relied on the existence of the Housing Assistance Plan and the zoning ordinance and failed to present any substantial evidence indicating a significant interest in limiting private developers to the urban renewal area. On appeal, appellees now contend that the ordinance is designed to encourage private developers to build in the deteriorated area of Huntington Station. Although we believe that the Town’s failure to raise this argument below precludes its consideration here, we briefly address this contention. The Town asserts that limiting multi-family development to the urban renewal area will encourage restoration of the neighborhood because, otherwise, developers will choose to build in the outlying areas and will bypass the zone. The Town’s goal, however, can be achieved by less discriminatory means, by encouraging development in the urban renewal area with tax incentives or abatements. The Town may assert that this is less effective, but it may actually be more so.
Developers are not wed to building in Huntington; they are filling a perceived economic void. Developments inside the urban renewal area and outside it are not fungible. Rather, developers prevented from building outside the urban renewal area will more likely build in another town, not the urban renewal area. Huntington incorrectly assumes that developers limit their area of interest by political subdivision. In fact, the decision where to build is much more complex. Hence, if the Town wishes to encourage growth in the urban renewal area, it should do so directly through incentives which would have a less discriminatory impact on the Town.
We turn next to the Town’s reasons rejecting the Elwood-Pulaski site. The *9401980 letter written by Town Supervisor Butterfield detailed seven justifications for the Town’s refusal to rezone: (1) inconsistency with the Town’s Housing Assistance Plan; (2) inconsistency with zoning; (3) traffic considerations; (4) parking and fire protection problems; (5) proximity to the railroad and Long Island Lighting Company substation; (6) inadequate recreation and play areas; and (7) undersized and unrealistic units. As the judge below noted, the first two beg the question because appellants are challenging the Town’s zoning ordinance. Huntington, 668 F.Supp. at 786. More significantly, as we have already indicated, the Town simply relied on the existence of the Housing Assistance Plan and the zoning ordinance and failed to present any substantial evidence indicating why precluding plaintiff from building a multi-family housing project outside the urban renewal area would impair significant interests sought to be advanced by the HAP and the ordinance. The fourth, sixth and seventh problems are “plan-specific” issues which could presumably have been solved with reasonable design modifications at the time appellants applied for rezoning of the parcel. The fifth concern also is largely plan-specific because proper landscaping could shield the project from the railroad and substation.
Thus, only the traffic issue and health hazard from the substation are site-specific. At trial, however, none of Huntington’s officials supported these objections. But-terfield, for example, was primarily concerned that the Matinecock Court project would “torpedo” the Town’s plan to develop the site at Broadway and New York Avenue in the urban renewal area in Huntington Station. (Testimony of Kenneth C. Butterfield, JA-1752.) Moreover, Huntington’s only expert, planner David Portman, set forth entirely different problems than were contained in Butterfield’s letters. Specifically, he noted sewage concerns, lack of conformity with the low density of the surrounding neighborhood, and inaccessibility of the site to public transportation (Testimony of David J. Portman, JA-1508-11.) Once during his testimony, he did mention “the relationship [of the site] to the power station.” (JA-1511) Never, however, did he raise any concern about a health hazard from the proximity to the substation. Indeed, appellees do not broach this issue in their brief to this court. Accordingly, we find the reasons asserted are entirely insubstantial.
The sewage problem was first raised at trial by appellees’ expert Portman. Appel-lees now advance it as an additional concern. The district court, however, chose not to consider it. We agree. Post hoc rationalizations by administrative agencies should be afforded “little deference” by the courts, Securities Indus. Ass’n. v. Board of Governors, 468 U.S. 187, 143-44, 104 S.Ct. 2979, 2982-83, 82 L.Ed.2d 107 (1984), and therefore cannot be a bona fide reason for the Town’s action. Moreover, the sewage concern could hardly have been significant if municipal officials only thought of it after the litigation began. If it did not impress itself on the Town Board at the time of rejection, it was obviously not a legitimate problem. In sum, the only factor in the Town’s favor was that it was acting within the scope of its zoning authority, and thus we conclude that the Town’s justifications were weak and inadequate.
In balancing the showing of discriminatory effect against the import of the Town’s justifications, we note our agreement with the Seventh Circuit that the balance should be more readily struck in favor of the plaintiff when it is seeking only to enjoin a municipal defendant from interfering with its own plans rather than attempting to compel the defendant itself to build housing. As the Arlington Heights II court explained, “courts are far more willing to prohibit even nonintentional action by the state which interferes with an individual’s plan to use his own land to provide integrated housing.” 558 F.2d at 1293. Bearing in mind that the plaintiffs in this case seek only the freedom to build their own project, we conclude that the strong showing of discriminatory effect resulting from the Town’s adherence to its R-3M zoning category and its refusal to rezone the Mati-necock Court site far outweigh the Town’s *941weak justifications. Accordingly, to recapitulate, we find that the Town violated Title VIII by refusing to amend the zoning ordinance to permit private developers to build multi-family dwellings outside the urban renewal area. We also find that the Town violated Title VIII by refusing to rezone the Matinecock Court site. We thus reverse the district court and direct entry of judgment in appellants’ favor.12
Appellees argue that we should deny site-specific relief because there are 64 “community development” sites available for low-cost multi-family housing in Huntington. They claim that a 1978 letter from Town Attorney Ronald Glickman to Michael Miness, Director of the Community Development Agency, established that any land within the community development areas can be rezoned as R-3M property. The record, however, makes clear that Glickman’s interpretation of the zoning code was neither codified nor represents current Town policy. As late as 1983, Town Supervisor Butterfield stated that the ordinance restricted private developers to the urban renewal zone. (Deposition of Kenneth C. Butterfield, JA-2045-46.) The Town has maintained in its papers submitted to the district court and to the United States Supreme Court in its 1982 petition for certiorari in Huntington I that the R-3M district “.... provided for multiple dwellings in urban areas, or; if owned by the Housing Authority; within or without such an area ...” Moreover, not only are 63 of the “community development” parcels not presently zoned for multi-family housing, but most of the sites are not in fact vacant but are “under-developed.” Therefore, there is only one site, not 64 sites, zoned and available for private low-cost multifamily housing. However, even as to the one site — the MIA site in Huntington Station — by the time of trial, HUD had determined it was in an area with a high concentration of minorities and therefore an inappropriate location for a federally subsidized housing de velopment. (Testimony of HUD official, William Miecuna, JA-421-22.)
Ordinarily, HHI would not be automatically entitled to construct its project at its preferred site. The Town might well have legitimate reasons for preferring some alternative site to the one preferred by HHI. On the other hand, the Town would not be permitted to select a site that suits the Town’s preference if that site imposed undue hardships on the applicant, such as distance from public transportation or other services. Thus, we would ordinarily remand this case to the district court to afford the appellees an opportunity to identify an alternative site, outside the urban renewal area, that would be appropriate for HHI’s project and eligible for the same financial arrangements and assistance available at the Matinecock Court site. If the Town identified such a site, it would then have the burden of persuading the district court that there were substantial reasons for using its preferred site and that those reasons did not impose undue hardships on the appellants. If the district court was not persuaded on balance of the benefits of an alternative site, it would then enter an appropriate judgment to enable HHI to proceed with its project at the Matinecock Court site.
This case, however, is not ordinary. First, we recognize the protracted nature of this litigation, which has spanned over seven years. Further delay might well prove fatal to this private developer’s plans. Second, other than its decision in December 1987 to build 50 units of low-income housing in the Melville section, the Town has demonstrated little good faith in assisting the development of low-income housing. After the Town began receiving federal community development funds, *942HUD found it necessary to pressure the Town continually to include commitments for construction of subsidized family housing in the Town’s HAPs. Because of the Town’s lack of progress in constructing such housing, HUD imposed special conditions on the Town’s community development grants for the 1978 fiscal allocation. Thereafter, HUD continued to express its dissatisfaction with the Town’s performance. This history, while it does not rise to a showing of discriminatory intent, clearly demonstrates a pattern of stalling efforts to build low-income housing.
Third, the other 63 parcels outside the urban renewal area are not presently zoned for multi-family housing and, indeed, the zoning ordinance presently forbids rezoning of these properties. Thus, this situation differs from Arlington Heights II, where 60 tracts currently zoned for multifamily housing were available and, accordingly, the Seventh Circuit remanded the case to the district court to determine if one of those sites were suitable. Arlington Heights II, 558 F.2d at 1295. Appel-lees cannot argue, as they do now, that the zoning ordinance does not now limit private builders to the urban renewal area when the Town Board in its January 6, 1981, resolution refused to amend the ordinance to delete the restriction of such housing to the urban renewal area, and appellees throughout this litigation have defended that decision. We therefore refuse to remand this case to the district court to determine the suitability of the 63 sites outside the urban renewal area. Rather, we find that site-specific relief is appropriate in this case.
Accordingly, we direct the district court to include in its judgment provision ordering the Town to rezone the 14.8 acre Mati-necock Court site located at the corner of Elwood and Pulaski Roads in Huntington Township to R-3M status. The judgment should also order the Town to strike from its R-3M zoning ordinance that portion which limits private multi-family housing projects to the urban renewal area.
. Appellants have abandoned their equal protection, 42 U.S.C. §§ 1982 and 1983, and state law claims pressed below.
. The district court certified a class of "[a]ll black, Hispanic and lower income persons in need of lower cost housing opportunities in Huntington and surrounding areas and who would qualify for residency in the proposed Matinecock Court and other Section 8 projects in Huntington, and who seek to reside in and insure opportunity for racially and economically integrated housing in Huntington.” "Section 8” refers to a federal program that provides subsidies for newly-constructed and substantially-rehabilitated housing. 42 U.S.C. § 1437f (1982 and Supp. Ill 1985).
.Although the original complaint also named HUD asweir'as the Huntington defendants, ap-ngllerifshave not challenged the dismissal of the rederal defendant.
. The Town claims that, in rezoning the Lincoln Manor site in 1978, the Town Board relied on a broadened interpretation of the R-3M provision. That legal opinion was contained in a letter from Town Attorney Ronald Glickman to Michael Miness, the Director of the Huntington Community Development Agency dated August 22, 1978. Glickman wrote that the community development program superseded the urban renewal designations, so that "so long as the subject premises is within a community development area” it can be rezoned as R-3M property. The Town claims it informed HHI of this interpretation. The letter, however, was not made public until 1984 in the midst of this litigation.
. Although pre-approval was granted, the project was delayed by community opposition and by an attempt by Butterfield to reserve 30 units for the elderly and to set a limit on black participation of 10%. In June 1981, Alan Weiner, HUD’s Area Manager, suspended pre-approval.
. In Yonkers, this court affirmed the district court’s finding that the Town purposefully maintained housing and educational segregation and thus violated the equal protection clause of the Fourteenth Amendment. Yonkers, 837 F.2d at 1218; See also Taylor v. Board of Education, 191 F.Supp. 181 (S.D.N.Y.), aff'd, 294 F.2d 36 (2d Cir.1961).
. Because we hold that we will no longer require a showing of discriminatory intent in Title VIII disparate impact claims, we do not review Judge Glasser's findings on intent to discriminate.
. There would likely also be a desegregative effect on Huntington Township as a whole in comparison to the region, given the tight housing market throughout the area. See, e.g., Black Jack, 508 F.2d at 1186; United Farmworkers of Florida Housing Project, Inc. v. City of Delray Beach, 493 F.2d 799, 810 (5th Cir.1974). Appel-lees assert that there is no statistical disparity between the Town and its surrounding region, the Nassau-Suffolk Standard Metropolitan Statistical Area. We need not reach this question, however, because we find sufficient desegrega-tive impact within Huntington itself from the project.
. The lower court opinion's sole reference to the zoning ordinance’s perpetuation of segregation is contained in its discussion of the Town’s discriminatory intent. Huntington, 668 F.Supp. at 786. Judge Glasser failed to consider the impact of the zoning ordinance on the perpetuation of segregation.
. Although the Arlington Heights II court compared the absolute numbers of blacks and whites affected by the Village’s refusal to rezone, it also examined whether the decision had a disproportionate impact on nonwhite people. 558 F.2d at 1291; see also Rizzo, 564 F.2d at 149; Black Jack, 508 F.2d at 1186.
. Appellees correctly note that Title VII case law subsequent to Griggs requires some showing that statistics based on the general population bear a proven relationship to the actual applicant flow. New York City Transp. Auth. v. Beazer, 440 U.S. 568, 99 S.Ct. 1355, 59 L.Ed.2d 587 (1979); Townsend v. Nassau County Medical Center, 558 F.2d 117 (2d Cir.1977), cert. denied, 434 U.S. 1015, 98 S.Ct. 732, 54 L.Ed.2d 759 (1978). The district court in this case, however, made a factual finding that a "significant” percentage of Matinecock Court tenants would be minority. Huntington, 668 F.Supp. at 785. Indeed, the court went further in noting that minority persons made up 74% of the Gateway Gardens waiting list, 45% of the Lincoln Manor waiting list and 61% of the waiting list for existing Section 8 certificates. Huntington, 668 F.Supp. at 785.
. In its brief, the Town asks us to take judicial notice of a new proposal, approved by the Town Board in December, 1987, to build 50 units of subsidized housing in the Melville section. Although this proposed action is indeed laudable, we do not consider it. It is entirely speculative and smacks of a mid-litigation effort to demonstrate that the Town is acting in good faith. Moreover, the Town’s action does not address the appellants’ chief concern: that the zoning code perpetuates racial segregation in housing by restricting privately-built multi-family dwellings to the urban renewal area.