MEMORANDUM OF DECISION AND ORDER
This is an action brought by neighborhood residents and community organizations seeking to preserve a racially integrated neighborhood on Staten Island. In this case, a racially mixed and balanced community is attempting to prevent not only the undue concentration and segregation of minorities in the area, but also the ultimate destruction of the community. See Otero v. New York City Housing Auth., 484 F.2d 1122 (2d Cir. 1973). It is the first time, to this court’s knowledge, that the National Association for the Advancement of Colored People (“NAACP”) has sought legal action to oppose the development of low-income housing designed to aid minority groups.1 Compare Jones v. Tully, 378 F.Supp. 286 (E.D.N.Y.1974), aff’d, 510 F.2d 961 (2d Cir. 1975).
The plaintiffs are individuals living near a proposed housing project and community and homeowner associations which represent the interests of area residents. They seek to enjoin the use of federal funds for the construction of a low-income highrise2 apartment building. They allege that the development will further the creation of an area of minority concentration and upset the delicate racial balance-which presently exists.3
Defendants are Patricia Harris, Secretary of the Department of Housing and Urban Development (HUD), and the developers of the proposed complex. The jurisdiction of this court is based on the Thirteenth and Fourteenth Amendments to the United States Constitution, the Fair Housing Act of 1968, 42 U.S.C. § 3601, et seq., the National Environmental Policy Act of 1969, 42 U.S.C. § 4321, et seq., and 28 U.S.C. 1343.4 Plaintiffs (“homeowners” or “area residents”) sue as a class comprised of all New York State residents who live near the proposed development.5 Initially, the plaintiffs sought a preliminary injunction. During the hearing on the preliminary injunction, the court found a sufficient basis to proceed to a trial on the merits. Fed.R. *831Civ.P. 65.6 The court must now determine whether to issue a permanent injunction.
FINDINGS OF FACT
A. The Proposal
The proposal in issue involves federally funded low-income housing. On June 3, 1977 HUD published a Notification of Fund Availability (“NOFA”) for 1,000 family dwelling units to be funded under the Section 8 Housing Assistance Payments Program (“§ 8”).7 On June 22,1977 the developer, in response to the NOFA, submitted a preliminary proposal for the construction of new § 8 low-income housing.8 The proposed development was a 96 unit six story apartment house to be located at the corner of Hill Street and Tompkins Avenue on Staten Island (“Tenhill”).
On August 10, 1977 HUD informed the developer that the preliminary proposal was approved and that $719,568 in federal funds was reserved for the project. HUD directed the developer to furnish information for the final proposal. HUD also notified the local New York City authorities, requesting objections to the proposal. Apparently, no objections were filed. On July 6, 1978 HUD approved the final proposal for low-income housing.
HUD had previously reviewed a proposal submitted by the developer for housing for the elderly at the same site. That proposal was disapproved by both the Equal Opportunity (“EO”) and Housing Management (“HM”) Divisions of HUD.9 The EO specialist indicated a likelihood of increased minority population in census tract 29 discussed infra, not indicated by the 1970 census data. He found that the complex would create an undue concentration of subsidized housing in the area. The HM analysts had similar objections. They recommended a housing survey to ascertain the need for additional housing in the area because several projects in the vicinity were having difficulty renting their units. John O’Rourke, an HM analyst, indicated in his supplemental report that the proposed project would adversely affect the other subsidized housing. He indicated a generally questionable and unsavory neighborhood with a high occurrence of muggings and drug use. He indicated that he feared for the safety of the elderly and believed the site very inappropriate for any housing.
The various HUD branches10 also considered the preliminary and final proposal for low income housing. All but the Economic and Market Analysis Division (“EMAD”) found the preliminary proposal to be acceptable. The EMAD, however, disapproved the proposal because the site was impacted with subsidized housing.11 The analysis also indicated that the 1970 *832census data did not reflect the current condition of the neighborhood. However, HUD decided to approve the final proposal on the basis of the recommendations of its other branches, and filed its report. HUD’s technical checklist on that report indicated that the preliminary proposal was approved by only four of its six branches. The HM and EMAD recommendations were not listed. The report indicated, however, that the proposal was approved on the basis of a11 reviews.
The final HUD report on the proposal was submitted after the institution of this action. This court had remanded the matter of HUD to reconsider the site proposal and to review the EMAD report. An EO specialist, Eli Forman, prepared the final report. His report and findings were limited to census tract 29. His conclusions concerning the racial composition of the census tract were based mainly upon the 1970 census data. Nevertheless, he found a high percentage of minorities in the Stapleton Houses project, which is a public housing project in census tract 29 adjacent to the proposed site. He indicated that the only updated material available showed a dramatic increase in the number of minorities in the housing project. He found a need to deescalate the racially sensitive housing market in the census tract, and recommended a public relations program to overcome the high crime stigma of the neighborhood. The proposal was reapproved by HUD over any previously filed divisional objection and the matter was brought before the court for a hearing on the preliminary injunction.
B. The Area
Tenhill is located in census tract 29 (“Tract”). Tract 29 is an area encompassing parts of the Stapleton-Concord section of Staten Island. Its designated boundaries are Tompkins Avenue to the East, Vanderbilt Avenue to the South, Broad Street to the North, and Van Duzer Street to the West. Tenhill is approximately 300 feet from Vanderbilt Avenue and is adjacent to Tompkins Avenue. Tompkins Avenue is the boundary between Tracts 29 and 27.12 Vanderbilt Avenue is the dividing line between Tracts 29 and 40.13
HUD considered Tract 29 to be the relevant area for estimating total minority concentration. HUD relied exclusively upon the census data compiled in 1970 for Tract 29 in making that determination. In 1970 that data indicated a total population of 4,623 in Tract 29. There were 4,000 white persons and 623 minority persons. Thus in 1970 the minority population for Tract 29 was approximately 12.8% of the total population. HUD considered the census data to be accurate, even though it recognized that the 1970 census data was outdated. The evidence supports the conclusion that because the 1970 census data was clearly outdated it was an insufficient indicator of minority and low-income concentration in the area.
The technical boundaries of Tract 29 do not represent the relevant neighborhood for any consideration of minority or low-income population. The evidence at trial indicates the difficulty with confining the neighborhood to Tract 29. Based on the testimony of area residents, the proximity of the housing complexes in Tracts 29 and 40, and the natural and man-made boundaries in the area, the court finds that the relevant neighborhood extends beyond the imaginary boundaries of Tract 29. The relevant community includes Tract 27 and parts of Tracts 29 and 40. It extends along Canal and Water Streets in Tract 27, Broad Street in Tract 29 up to Targee Street. Targee Street is the western boundary from Broad Street to Mary Street in Tract 40. The railroad track which represents the southeastern border of Tract 40 is the outer boundary of the community. That boundary extends to Bay Street. The eastern boundary is Bay Street from Greenfield Av*833enue to Water Street.14 The center of this community is concentrated with high rise apartment complexes in the immediate vicinity of Tenhill. This core of the neighborhood shows an excessive concentration of a low-income, minority population.15
The community contains mostly low density one and two family dwelling units. In the context of Staten Island housing, and this neighborhood, a six story apartment complex represents a dominant feature in the community. Undoubtedly Tenhill, if constructed, will have a significant impact on the community.
Tract 29 is dominated by the Stapleton Houses Project. Stapleton Houses is a 693 unit public housing development. Its physical plant extends over a large segment of the land area of Tract 29. It has a population of more than 2,500 people, more than half of the 1970 census tract population. Stapleton Houses is situated across the street from Tenhill and is within 100 feet of the proposed site. In 1970 80.7% of the units were occupied by whites. By 1977 that figure had decreased to 34.1% and by 1978 whites occupied 25.1% of the units. In 1970 only 36 families were receiving public assistance and only 63 families were headed by a single parent. In 1977 these figures were 243 and 280 respectively, and by 1978 there were 285 and 317 such families respectively. In 1972 the project had a crime rate of 5.6% per 1,000 population. In 1977 that figure had changed to 25.2%, the highest of all public housing projects on Staten Island.
Stapleton Houses and five other complexes in Tract 40 are the center of the neighborhood. That these complexes represent part of one neighborhood is evidenced by the distance between the projects; the community attitudes about the projects; the similar nature of the housing and its contrast to the low density housing in the area; the accessibility to each complex from any other; the proximity of shopping centers, recreation centers, houses of religious worship and schools; and the racial and economic composition of the complexes. In combination they represent a dominant aspect of the neighborhood.
This series of high-rise dwelling complexes is in the immediate vicinity of Stapleton *834Houses and Tenhill. They lie along the Vanderbilt Avenue border of Tracts 29 and 40. They are between 1500 feet16 and 3200 feet17 from Tenhill. HUD did not consider the impact of these complexes in reaching its decision to approve Tenhill and did not investigate the minority concentration in them because HUD considered only those complexes within the boundaries of Tract 29 to be relevant. The court finds that these complexes are a significant aspect of the community and should be viewed as part of the relevant neighborhood.
The relevance of these complexes is apparent from their proximity to Tenhill and their statistical prominence. The Park Hill apartments is an 806 unit complex approximately 2,000 feet from the proposed site. As of June 1977, of 749 units occupied, only 5 were occupied by whites. Minorities occupied the remaining 744 units and there were 50 vacancies. There has been a consistently high pattern of turnovers each year. The statistics demonstrate a dramatic increase in the number of minority families since 1970. In 1970, 401 of 800 units were occupied by whites. By 1977, of 749 units occupied, only 5 were occupied by whites. Moreover, 701 of 749 units were under the § 8 Housing Assistance Plan, including the 5 white units. Thus, by 1977 the minority population had doubled to 99% of the total unit population.
The Fox Hill complex is approximately 1,500 feet from Tenhill. The minority population is 99% of the 360 residential units occupied. The vacancy rate was constantly between 6% and 10% of the available units. Moreover, of 291 individuals receiving income, 125 were on welfare and 25 were unemployed. The St. George Plaza complex consists of 305 units with a 9% white population for 1978. The Seaview Arms has 84 units with a white population of 5%.18 Thus, with due consideration for the nature of Staten Island housing and demographics, these complexes are a dominant feature in the area.
The remainder of Tract 29 is a mixture of vacant lots, operating or abandoned business facilities, and one or two family houses. The apartment complexes overshadow the low density housing. No statistics were presented by HUD concerning the racial composition of the residents in these houses. No survey was conducted to ascertain the racial mixture in the houses before approving Tenhill. However, the evidence indicates that the housing around the complexes in Tract 29 is racially mixed.19 The New York City tax records indicate a relatively stable low-income community with an average assessed housing valuation between $3,000 and $8,000, with a large concentration in the $4,000 to $5,000 range.
The main business district in Tract 29 has deteriorated significantly during the period from 1970 to the present. Numerous businesses, including two supermarkets and a bank branch, have closed not only because of the rising criminal activity in the area, but also because of the inability of the large low-income population to generate sufficient capital to support business enterprises. A supermarket adjacent to Tenhill has remained closed since December 1975. It and other abandoned buildings evidence the continuing urban blight in the area.
A significant cause of the deterioration is the increased criminal activity in the neigh*835borhood. Tenhill is located in the 120th precinct. Most of the criminal activity in the 120th precinct centers around the Stapleton Houses and Fox Hill complexes. In 1970 the 120th precinct ranked 24th of 78 precincts in New York City in terms of felony complaints. The precinct is divided into 22 sectors. The Stapleton and Fox Hill apartment complexes are located in sectors C and E respectively. These two sectors extend into Tracts 29 and 40. From 1970 to 1977 the precinct experienced a total crime increase of 60.6%. The Stapleton-sector C area consistently ranked first or second of the 22 sectors in terms of criminal activity. The Park Hill-sector E area ranked 9th of 22 in 1970 and has grown progressively worse. It ranked 2nd of 22 in 1974 to 1977, and 3rd of 22 in 1978. The Stapleton sector experienced a 72.8% increase in crime from 1970 to 1977. The Park Hill sector experienced a 161.6% crime increase during that period. Thus the proposed site is centered between the areas of highest criminal activity on Staten Island. Moreover, while crime has risen, the city’s financial crises have forced cutbacks in the police workforce for the area,20 raising the likelihood of sustained criminal activity in the area.
The statistics produced by the New York City Fire Department indicated a strained delivery system. While the fire companies servicing the area rank in the bottom half of all city fire companies in terms of activity, they are very active in terms of the Staten Island community. While manpower has decreased, there have been a great number of fires and false alarms in the area. Significantly, the companies serve a relatively small area of Staten Island. However, because of the number of runs made, the companies ranked 1st and 5th in terms of activity of the 16 fire units on Staten Island in 1977.
The public schools servicing the area near Tenhill are presently minority and low-income concentrated. Public School (P.S.) 14 is approximately 25 feet from Tenhill. In 1974 it had a minority population of 44%. This steadily increased to 73% in 1978. The school is 98% utilized at present. Moreover, 84% of the children at P.S. 14 are eligible for the free lunch program, which indicates their low-income background. The school ranked 38 of 39 on Staten Island in terms of reading level achievement.
Similarly, the minority attendance at Junior High School (J.H.S.) 49 increased from 28% in 1974 to 51% in 1978. This influx of minorities into the schools reflects the general influx of minorities into the area. The other neighborhoods or housing projects which contribute students to J.H.S. 49 are predominantly white and would not account for the increased minority attendance. J.H.S. 49 is 85% utilized. It is within 500 feet of the proposed site. Two other schools, P.S. 13 and P.S. 5721 also service the neighborhood residents. It was unlikely, however, that children from Tenhill would attend either school.22 Thus, younger children from Tenhill would attend either J.H.S. 49 or P.S. 14.
The evidence established that over the years the area has experienced rapid deterioration. Area residents believe the community is presently racially balanced, but attribute the dramatic deterioration to the apartment complexes in the area. It is clear that an increase in low-income housing units will create an intolerable situation for both black and white neighborhood residents. The creation of Tenhill will insure the ghettoization of the area. While the court does not use that phraseology lightly, it finds that its use here is appropriate. The neighborhood surrounding the proposed *836building site has deteriorated into a Staten Island ghetto with an undue concentration of low-income, minority residents.
Tenhill will not be attractive to middle-income white or black families. The area is in a state of deterioration with no signs of revitalization. The area’s black and white community stands ready to abandon the area if Tenhill is built. Moreover, the area is already concentrated with high-rise dwelling units with undue concentrations of low-income, minority families.
Tenhill will not provide a balanced tenancy. It is a § 8 project. Therefore, at least 30% of its population must be in the very low income range. Even if the builder is the best in his business, it is doubtful that he could make the remaining 70% of the apartment complex appealing to middle class whites or blacks. The facts establish a correlation between low-income families and minority families. The other § 8 projects have attracted mostly minority residents. For example, the Park Hill statistics indicated that 701 of 749 § 8 units were occupied by minority families. The Sea-view Arms complex indicated that 182 of 221 persons were under § 8 assistance in an apartment complex with a 95% minority population. Thus, Tenhill will be populated by low-income families with a high correlation of minorities.
Moreover, an additional low-income housing project is not needed in the community. It will not alleviate the low-income housing shortage in New York City. Tenhill will provide 96 more units in an area concentrated with apartment complexes which have a high turnover rate and a high vacancy rate. For example, in June 1977, when this project was initially proposed, Staple-ton Houses had a 121% turnover rate. Park Hill had 50 vacant units, and 389 of 749 units were turned over during the year. Fox Hill had 35 vacant units. Seaview Arms and St. George Plaza had 11 vacancies. Thus, there were 96 vacant units in June, 1977 in only four of the six relevant apartment complexes.23 Therefore, no urgency for low-income housing existed in the area at the time of HUD’s approval. That condition has persisted to the present.
CONCLUSIONS OF LAW
I. REVIEW OF AGENCY ACTION
Plaintiffs seek to set aside HUD’s approval of Tenhill on numerous grounds. The parties agree that the proper standard for judicial review in this case is found in Section 706(2)(a) of the Administrative Procedure Act, 5 U.S.C. § 706(2)(A). The standard was set forth in Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 91 S.Ct. 814, 28 L.Ed.2d 136 (1971):
The court is first required to decide whether the Secretary acted within the scope of his authority. Schilling v. Rogers, 363 U.S. 666, 676-677, 80 S.Ct. 1288, 4 L.Ed.2d 1478 (1960). This determination naturally begins with a delineation of the scope of the Secretary’s authority and discretion . . . Scrutiny of the facts does not end, however, with the determination that the Secretary has acted within the scope of his statutory authority. Section 706(2)(A) requires a finding that the actual choice made was not “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” 5 U.S.C. § 706(2)(A) (1964 ed., Supp. V). To make this finding the court must consider whether the decision was based on a consideration of the relevant factors and whether there has been a clear error of judgment, (citations omitted) Although this inquiry into the facts is to be searching and careful, the ultimate standard of review is a narrow one. The court is not empowered to substitute its judgment for that of the agency. Id. at 415-16, 91 S.Ct. at 823.
Accord, Schicke v. Romney, 474 F.2d 309 (2d Cir. 1973). Thus § 706 precludes judicial de novo review or a review based on the “substantial evidence” standard.
*837Plaintiffs contend that under this standard, HUD’s approval of Tenhill was arbitrary and capricious and violated HUD’s duty to integrate the neighborhood. The court finds that HUD’s action was arbitrary and capricious and did not comply with its legal obligations in several ways.
II. HUD’S DUTY TO INTEGRATE
[2] Any review of Tenhill must focus on HUD’s authority. It is now well settled that HUD has an affirmative duty to promote racial integration through its housing policy. Karlen v. Harris, No. 78-7147, 590 F.2d 39 (2d Cir. Dec. 14, 1978); Shannon v. HUD, 436 F.2d 809 (3d Cir. 1970); Trinity Episcopal School Corp. v. Romney, 387 F.Supp. 1044 (S.D.N.Y.1974), aff’d, 523 F.2d 88 (2d Cir. 1975). See also Otero v. New York City Housing Auth., 484 F.2d 1122 (2d Cir. 1973). This duty is mandated by the stated policies of the Fair Housing Act of 1968, 42 U.S.C. § 3601, et seq.,24 the Housing and Development Act of 1974, 42 U.S.C. § 5301, et seq.,25 the Section 8 Housing Assistance Payments Program, 42 U.S.C. § 1437f, and by the National Environmental Policy Act of 1969, 42 U.S.C. § 4321, et seq.26 Either separately or in combination these acts place a heavy burden on HUD to further national housing policy by' avoiding the concentration of minority or low-income families in the same community. See Karlen v. Harris, supra. See also Otero v. New York City Housing Auth., supra. As noted by the court in Otero:
We do not view that duty [to integrate] as a “one-way street” limited to introduction of non-white persons into a predominantly white community. The Authority is obligated to take affirmative steps to promote racial integration even though this may in some instances not operate to the immediate advantage of some nonwhite persons. Id. at 1125. An authority may not, for instance, select sites for projects which will be occupied by nonwhites only in areas already heavily concentrated with a high proportion of nonwhites. . . . Not only may such practices be enjoined, but affirmative action to erase the effects of past discrimination and desegregate housing patterns may be ordered. Id. at 1133 (emphasis added).
Thus, HUD must affirmatively seek to provide housing in areas that will further these goals and avoid the continued and unnecessary area concentration of low-income and minority families. To this end, site selection has surfaced as the major vehicle for insuring that the purposes of the Acts are accomplished. See Karlen v. Harris, supra; Otero v. New York City Housing Auth., supra.
HUD’s site and neighborhood standards substantially comply with the policies of the Acts. HUD’s site selection standards are set forth in 24 C.F.R. 880.112. That section provides in relevant part:
Proposed sites for new construction projects must be approved by HUD as meeting the following standards:
(b) The site and neighborhood shall be suitable from the standpoint of facilitating and furthering full compliance with *838the applicable provisions of Title VI of the Civil Rights Act of 1964, Title VIII of the Civil Rights Act of 1968, Executive Order 11063, and HUD regulations issued pursuant thereto.
(c) The site shall not be located in:
(1) An area of minority concentration unless (i) sufficient, comparable opportunities exist for housing for minority families, in the income range to be served by the proposed project, outside areas of minority concentration, or (ii) the project is necessary to meet overriding housing needs which cannot otherwise feasibly be met in that housing market area. (An “overriding need” may not serve as the basis for determining that a site is acceptable if the only reason the need cannot otherwise feasibly be met is that discrimination on the basis of race, color, religion, creed, sex, or national origin renders sites outside areas of minority concentration unavailable.)
(2) A racially mixed area if the project will cause a significant increase in the proportion of minority to non-minority residents in the area.
(d) The site shall promote greater choice of housing opportunities and avoid undue concentration of assisted persons in areas containing a high proportion of low-income persons, (emphasis added)27
Thus, HUD has recognized its continuing duty to prevent the creation of concentrated areas of low-income and minority persons. HUD’s rules require it to scrutinize both an area of racial concentration containing a minority population of 30% to 40% of the total population and an area of low-*839income concentration without regard to minority population. In delineating the relevant area of racial or low-income concentration, HUD analysts are not required to look to any particular area with definite pre-determined boundaries. Nor, in light of the national housing policies contained in the various housing Acts, could it properly limit the scope of its inquiry to any one area where that would distort HUD’s determination. See Karlen v. Harris, supra; Trinity Episcopal School Corp. v. Romney, supra; Otero v. New York City Housing Auth., supra; Shannon v. HUD, supra. With this duty in mind, the court finds that HUD’s approval of Tenhill was arbitrary and capricious.
A. Reliance on Census Tract 29
HUD considered the relevant “area” of inquiry to be only Tract 29. HUD’s analysts indicated that the use of the census tract was the most reliable and convenient method for ascertaining the racial or economic composition of an area in the context of its rules and the Acts. The court finds that reliance on census tract boundaries in vacuo in order to delineate a relevant area of minority or low-income concentration was arbitrary and capricious and in violation of HUD’s statutory duty to integrate. Delineating neighborhood boundaries requires a detailed examination of numerous objective and subjective factors. Initially, the problem was brought into focus by the racial imbalance existing in the nation’s schools. The school desegregation cases demonstrate the perplexing conceptual barriers to an accurate determination and the difficulty with providing an adequate remedy in each case. Those cases indicate the necessity for an ad hoc determination in order to achieve a satisfactory resolution. See, e. g., Swann v. Charlotte-Mecklenburg Bd. of Educ., 402 U.S. 1, 91 S.Ct. 1267, 28 L.Ed.2d. 554 (1971), rehearing denied. 403 U.S. 918, 91 S.Ct. 2200, 29 L.Ed.2d 689 (1971); Green v. County School Bd., 391 U.S. 430, 88 S.Ct. 1689, 20 L.Ed.2d 716 (1968). See also Fiss, Racial Imbalance in the Public Schools: The Constitutional Concepts, 78 Harv.L.Rev. 564 (1965).
In recent years thé focus has shifted to housing patterns as the underlying cause of segregation. See Karlen v. Harris, supra; Otero v. New York City Housing Auth., supra; Shannon v. HUD, supra. Those cases indicate the problems which confront any housing proposal in a community. To accommodate the peculiar patterns of urban neighborhoods, any approach to a definition of neighborhood requires imagination, innovation, and flexibility. Such an approach can best assure integration. Since each neighborhood is uniquely affected by community perceptions, any neighborhood analysis must be made on an ad hoe basis. In general, a neighborhood represents any section of a region or city, having indefinite boundaries, and which is drawn together by the shared perceptions of its residents as to what constitutes their neighborhood, by the facilities generally available for their use, by their social and economic status, and by natural or man made physical boundaries.
The court finds that imaginary census tract boundaries alone reflect neither a community’s perception of itself nor the social cohesiveness of a given area. While census tracts may provide HUD with a general indication of residential patterns, they are inadequate as the sole indicators of the racial or economic composition of housing in a neighborhood. Other courts have not hesitated to ignore imaginary or even physical boundaries where those boundaries hinder the affirmative obligation of government authorities to integrate. See Gautreaux v. City of Chicago, 480 F.2d 210 (7th Cir. 1973), cert. denied, 414 U.S. 1144, 94 S.Ct. 895, 39 L.Ed.2d 98 (1974); Henry v. Clarksdale Municipal Separate School Dist., 409 F.2d 682 (5th Cir. 1969), cert. denied, 396 U.S. 940, 90 S.Ct. 375, 24 L.Ed.2d 242 (1969); Resident Advisory Bd. v. Rizzo, 425 F.Supp. 987 (E.D.Pa.1976), modified, 564 F.2d 126 (3d Cir. 1977); Blackshear Residents Org. v. Housing Auth. of City of Austin, 347 F.Supp. 1138 (W.D.Tex.1972); Banks v. Perk, 341 F.Supp. 1175 (N.D.Ohio 1972) , aff’d in part, 473 F.2d 910 (6th Cir. 1973) . In Henry, the court refused to allow a railroad track to operate as an effective *840boundary when it had a segregative and discriminatory impact. In Banks v. Perk, the census tracts were so drawn as to perpetuate black and white segregation in the city. The court found that such artificial boundaries prevented integrated housing and chose instead to define the relevant areas on an ad hoc basis after an evidentiary hearing.
The court finds this precedent to be especially persuasive in the context of this case. HUD erroneously considered Vanderbilt Avenue to be the standard neighborhood boundary. The testimony at trial indicated that all of the area residents considered Vanderbilt Avenue to be part of their neighborhood. While their perceptions of the neighborhood varied slightly, each resident included the four apartment complexes discussed in this court’s factual findings as part of the neighborhood. Thus, the racial composition and impact of those complexes was relevant to HUD’s inquiry. By excluding them from its determination on the basis of an arbitrary and imaginary boundary, HUD distorted the true racial composition of the neighborhood. HUD, in effect, concluded that individuals on the Tract 40 side of Vanderbilt Avenue would not go into Tract 29. The court finds that the residents of the complex will not be constrained by the imaginary census tract boundaries. It is clear that the complexes in Tract 40 along Vanderbilt Avenue and Tenhill are part of the same neighborhood. They were relevant to HUD’s inquiry. Thus, the court finds that HUD acted arbitrarily and capriciously in ignoring the substantial concentration of minority families in those complexes. Accordingly, HUD may not in the future limit its site selection process solely to census tract boundaries.
Defendants contend that if any part of Tract 40 is considered, all of Tract 40 must be considered. The court rejects this argument for the same reasons given above. Tract 40, like Tract 29, is composed of imaginary boundaries. The evidence indicates that the Tract 40 boundaries do not accurately depict the relevant neighborhood. Since some parts of Tract 40 are outside of the relevant neighborhood, the court rejects any contention that it must adhere rigidly to imaginary lines.
Moreover, defendants themselves point to the complexes in Tract 40 in order to demonstrate the insignificance of a 96 unit project in Tract 29. They contend that such a development will have little or no impact on an area with a large number of developments in the immediate area. Defendants may not attempt to minimize the impact of Tenhill by using or ignoring census tract boundaries when convenient. That inconsistency in position is at odds with HUD’s affirmative duty to integrate. Therefore, if these complexes are relevant to minimize Tenhill’s impact, they are relevant to amplify its impact as well.
In conclusion, census tract boundaries are not an adequate indicator of neighborhood boundaries. Where,'as here, such boundaries distort the racial composition of a neighborhood and lead to an undue concentration of minorities in a small area, they may not be relied upon by HUD in approving a low-income housing project. Accordingly, HUD’s use in this case of those boundaries alone was arbitrary and capricious.
B. Reliance on Census Data
HUD relied almost exclusively upon 1970 census data to determine the racial concentration of Tract 29. That data indicated a 12.8% minority population in 1970. At trial, HUD admitted that the 1970 data was clearly outdated. Yet, in a startling revelation, HUD refused to depart from its practice of using census data because HUD considered it to be reliable as an area indicator and found it to be readily available.
Plaintiffs produced overwhelming evidence to demonstrate the weakness of HUD’s position. That evidence was readily accessible and easily compiled in preparation for litigation by a small law firm. It indicated the drastic changes that occurred in and around Tract 29 from 1970 to the present. The area experienced a state of rapid decline both physically and economically. The tremendous increase in the num*841bers of minority families in the subsidized housing surrounding the proposed site was even more startling. HUD’s own analysts on several occasions questioned the suitability of the proposed site for more low-income housing. The Court finds, therefore, that HUD acted arbitrarily and capriciously when it based its determination on outdated census statistics when readily accessible information clearly disputed the 1970 census figures.
HUD argues, however, that the rapid increase of minorities in the apartment complexes near Tenhill has had and will have little or no effect on the area. HUD bases this contention on an assumption that apartment complexes, such as Stapleton Houses, are self-contained units. The court finds such a conclusion to be incredible.
HUD has, in effect, ignored the totality of Staten Island housing and living patterns by its assumption. HUD indicated that Staten Island was a low density area. The area overtaken by the apartment complexes is comprised of one and two family homes. The evidence indicates that while a six-story structure will have little impact in the context of Manhattan housing, it will have a definite impact if located at the Tenhill site on Staten Island. The impact of the larger developments, such as Stapleton Houses, Park Hill, and Fox Hill, is even more pronounced. For example, Stapleton Houses accounted for more than half of the census tract population and covered a large portion of the total land mass in Tract 29. In terms of the area occupied and population represented it is the dominant feature of the area. Moreover, residents in all of the complexes must enter the other parts of the community for work, recreation, travel, and shopping. Thus, it cannot be assumed that the population of these complexes will have no impact on the areas surrounding Tenhill.
HUD also contends that it is statistically improbable that a given area will shift from an area of low minority concentration to one of high minority concentration in a seven year period. HUD again assumes too much. HUD is required to take affirmative steps to determine the racial makeup of an area. Here, plaintiffs established the availability of data which seriously undermined HUD’s conclusions concerning Tenhill. Moreover, community attitudes reflected their awareness of a growing racial imbalance. HUD was obligated both to test those attitudes and to compile further, updated information about the area. Its failure to do so led HUD to ignore the relevant housing conditions in the part of the neighborhood immediately outside the Tract 29 boundaries.
Thus, reliance on outdated census data as the primary indicator of racial and low-income concentration is arbitrary and capricious. Mere administrative convenience cannot justify the use of such data where the effect is to prolong and intensify an unhealthy situation. HUD is vested with a statutory duty to provide integrated housing and avoid low-income and minority concentrations, and cannot ignore blatant indicators of concentration. The court appreciates the difficulty HUD may encounter in its attempts to compile relevant data. However, that burden is part of HUD’s statutory duty. HUD may not await the 1980 census data which will inevitably reflect the changes in this community. That “wait-and-see” attitude would place the area residents in an unenviable position. As a result, HUD must look beyond census statistics and arbitrary area boundaries in attempting to ascertain the existence of a neighborhood and the changes that have occurred there.28
III. TIPPING AND RACIAL CONCENTRATION
Plaintiffs contend that HUD violated its statutory duty to integrate by approving a low-income housing project at the proposed *842site. They claim that an increase in the number of low income housing units in a neighborhood already concentrated with such housing will have a “tipping” effect on the community. They base this claim on the anticipated departure of both black and white residents in the neighborhood if another low-income housing project is built in the area. Thus, they claim that the area will tip from its present racial balance into an area of minority and low-income concentration.
A. Tipping
In Trinity Episcopal School Corp. v. Romney, 387 F.Supp. 1044 (S.D.N.Y.1974), aff’d on tipping, 523 F.2d 88 (2d Cir. 1975), the court set forth the standard for “tipping.” The court stated:
We conclude for the purposes of this litigation that the tipping point of a community is that point at which a set of conditions has been created that will lead to the rapid flight of an existing majority class under circumstances of instability which result in the deterioration of the neighborhood environment. The criteria to determine whether an area has reached or is approaching this point are: (1) the gross numbers of minority group families in a measurable economic or social group which are likely to affect adversely Area conditions; (2) the quality of community services and facilities; and (3) the attitudes of majority group residents who might be persuaded by their subjective reactions to the first and second criteria to leave' the Area. Id. at 1065-66.
Under that standard, plaintiffs are required to produce convincing evidence that construction at the proposed site will cause the area to “tip.” See Otero v. New York City Housing Auth., 484 F.2d 1122 (2d Cir. 1973). The court finds that plaintiffs have sustained their burden of proof on the issue of tipping.
First, the area surrounding Tenhill is concentrated with minority families. The evidence indicates a rapid influx of low-income, minority families within the last eight years. The evidence convincingly demonstrates that the 1970 estimate of minority population in the area is outdated and does not reflect the current condition of the neighborhood. Moreover, Tenhill is at the center of a tightly patterned group of apartment complexes with excessive concentrations of low-income minority families. Tenhill represents the seam of a closely drawn minority “pocket ghetto” which will inevitably tear the fabric of a racially integrated community. Thus, consistent with this court’s determination that census tract boundaries alone are not accurate yardsticks, the court finds that Tenhill is in an area of excessive minority concentration.
Second, community services and facilities have deteriorated and will not adequately support further low-income housing. Both residents and analysts indicated a blighted area, without prospect for any economic recovery. A crime and fire hazard exists and is complicated by the decrease in municipal services in the area. Most of the criminal activity centers around Tenhill and the other projects.29 The schools are low-income and minority concentrated, and are marked by poor scholastic achievement. The businesses in the area have been hard hit by rapid and persistent deterioration. Many businesses have closed, leaving the community residents with inadequate or distant shopping facilities. Moreover, travel costs for Staten Island residents are high in relation to other parts of the city, and will add a further burden to low-income area residents.
Third, the community has voiced vehement objections to the construction of another low-income housing complex. The residents who testified indicated that many have left or are in the process of leaving the neighborhood. They stated that they would also leave if Tenhill was built. They *843and their neighbors see this as the “last straw” that will insure the ghettoization of area. This attitude is shared by both black and white residents who have decided to abandon the neighborhood because they refuse to live in a ghetto. Those individuals who testified or presented stipulated testimony represent civic and homeowner groups in the area as well as themselves. No residents who approved of Tenhill were produced to rebut their assertions.
The “tipping” context presented here is, therefore, clearly distinguishable from that found in Trinity. In Trinity, the court noted:
The assertion that a community might tip because of the unsupported fears of its residents demonstrates the inherent. weakness of that tipping concept. In large measure, the tipping point is a subjective and prejudicial reaction of whites to minority group encroachment and to that extent is subject to a wide range of subjective factors. Id. at 1072 (emphasis added).
Here, in contrast no racial majority clearly exists. While the area contains a disproportionate concentration of minority families, the white population has not stood by itself in protest of the continued ghettoization of the area. Instead, this integrated commuobjected to the further increase in low-income housing in the area.
Finally, HUD has failed to indicate the absence of more appropriate sites on Staten Island. The evidence indicates that few parts of Staten Island outside of this area are concentrated with low-income or minority concentration. There is no rational basis for the continued concentration of low-income housing complexes in this area. HUD has failed to establish that the housing needs of New York’s low-income minority population could not be better served by funding alternate, more desirable proposals.30
Accordingly, the court finds that Tenhill will inevitably result in the “tipping” of the neighborhood, and its construction with federal funding must be enjoined.31 The conclusion to be drawn from the evidence is that this is not the neighborhood which HUD assumed it to be. The court must therefore alleviate the condition before it grows worse. As the court noted in Karlen v. Harris, supra at 45:
No construction has yet been commenced. It would be folly not to heed the most modern thinking on the subject where a major undertaking is involved. After all, what doctor would fail to use the most recently discovered curative medicine merely because the patient had become ill prior to its discovery?
B. Undue Concentration of Low-Income Housing
Tenhill, as a § 8 project, will further the concentration of low-income housing in the community. In Karlen v. Harris, No. 78-7147, 590 F.2d 39 (2d Cir. 1978), the Court of Appeals for the Second Circuit enjoined the construction of a low-income housing project in an area containing an already high concentration of low-income housing. The Court found that such concentration, regardless of its impact on a “tipping” analysis, was contrary to HUD’s affirmative duty to develop integrated communities through its housing policies. The Court rejected New York City’s low-income housing shortage as a justification for continued concentration of low-income housing in one area. It noted:
*844The importance of this decision lies in the fact that once a high-rise low-income apartment is built on Site 30, the character of the neighborhood is permanently established — at least for the life of the structure — if as represented, it adds to an already concentrated group of low-income buildings. This construction will not be a temporary stopgap to enable the City to solve its problem for adequate housing for its low-income population. The Congress did not authorize the use of federal resources for this single purpose. Id. at 44 (emphasis added).
The court finds this statement even more relevant to this case, given the low density housing in the community surrounding Ten-hill, and the general environment of Staten Island.
Furthermore, Karlen requires HUD to consider communities and the impact of low-income housing on an ad hoc basis. The court, relying on its decision in Otero v. New York City Housing Auth., supra, held:
Once the area is committed to a high concentration of low-income housing, the Congressional purpose of racial and economic integration is thwarted. Id. at 45 (emphasis added).
Therefore, there is no requirement that HUD limit its scope of inquiry to artificial census tract boundaries when determining the concentration of low-income housing in an area. Where, as here, the immediate area surrounding Tenhill has demonstrated a definite pattern of low-income housing and that housing problem is further compounded by high minority concentration, the project must be enjoined.32
Defendants argue that evidence of low-income concentration in a community is inappropriate for a “tipping” analysis according to the holding in Trinity Episcopal School Corp. v. Romney, supra. In Trinity, plaintiffs had failed to demonstrate any correlation between low-income and minority populations. Thus, the Court found any evidence of low-income population unconvincing in a tipping situation. In this case, the evidence demonstrates the correlation that exists between the low-income and minority populations in the New York City area. While such a correlation may not be drawn in many other areas of the country, the testimony and exhibits at trial indicated that the correlation may be drawn and used as probative evidence of the issue presented. Thus, the correlation, if established, may add some weight to a “tipping” analysis.
Even if the correlation was not established, or if established not usable as any evidence of “tipping”, the racial elements of “tipping” are convincingly shown in this case. Moreover, under the analysis used in Karlen v. Harris, supra, there is convincing evidence that the area is concentrated with low-income housing. Therefore, without making any correlation between low-income and minority populations, the evidence supports plaintiffs’ claims. Accordingly, the court finds that the use of federal funds for the development of Tenhill must be enjoined.
IV. RELIEF
Plaintiffs seek to enjoin the use of federal funds for the construction of Tenhill.33 They have adequately shown that they will suffer immediate, irreparable harm and that adequate remedies at law are not available. O’Shea v. Littleton, 414 U.S. 488, 94 S.Ct. 669, 38 L.Ed.2d 674 (1974). If Tenhill is built, plaintiffs will be deprived of the benefits associated with an integrated community. Trafficante v. Metropolitan Life Ins. Co., 409 U.S. 205, 93 S.Ct. 364, 34 L.Ed.2d 415 (1972).
*845Accordingly, the permanent injunction must issue. HUD’s approval of the proposed site for low-income housing is stricken. HUD is permanently énjoined from providing or reserving federal funds for the construction of a low-income housing project at the proposed site.
This decision shall constitute the court’s findings of fact and conclusions of law pursuant to Fed.R.Civ.P. 52. The decision shall constitute the final judgment of the court. Accordingly, the Clerk of the Court is directed to enter a final judgment pursuant to Fed.R.Civ.P. 54, 58.
So ordered.