United States v. Yonkers Board of Education

KEARSE, Circuit Judge:

Defendants City of Yonkers (the “City”), Yonkers Community Development Agency (“CDA”), and Yonkers Board of Education *1184 (the “Board”) appeal from a judgment entered in the United States District Court for the Southern District of New York following a trifurcated bench trial before Leonard B. Sand, Judge, holding the City liable for racial segregation of housing in Yonkers, holding both the City and the Board liable for racial segregation of the Yonkers public schools, and ordering each defendant to take steps to remedy the segregation for which it was found liable. The district court held that the City, by its pattern and practice of confining subsidized housing to Southwest Yonkers, had intentionally enhanced racial segregation in housing in Yonkers, in violation of Title VIII of the Civil Rights Act of 1968 (“Title VIII” or the “Fair Housing Act”), 42 U.S.C. § 3601 et seq. (1982), and the Equal Protection Clause of the Fourteenth Amendment to the United States Constitution. The court held that the actions of the Board, including its decisions relating to individual schools, faculty assignments, and special education, and its selective adherence to a neighborhood-school policy in light of the City’s segregative housing practices, combined with its failure to implement measures to alleviate school segregation, constituted intentional racial segregation of the Yonkers public schools, in violation of Titles IV and VI of the Civil Rights Act of 1964, 42 U.S.C. § 2000c et seq. (“Title IV”) and § 2000d et seq. (“Title VI”) (1982), and the Equal Protection Clause. The court held that the City had contributed to the segregation of the Yonkers public schools by means of, inter alia, its segregative housing practices, and that its segregative intent was revealed by the foreseeable effects of its housing practices, its direct involvement with certain schools, and the mayor’s appointments to the Board of persons firmly committed to maintaining the segregated state of the schools that both reflected and enhanced the segregated residential patterns. The court thus found the City liable for intentional racial segregation of the schools in violation of Title IV and the Equal Protection Clause.

To remedy the segregation in housing, the district court ordered principally that the City provide sites for 200 units of public housing in nonminority areas; the order stated that if the City did not identify sites the court would do so. The court ordered that the City reallocate at least a substantial portion of its federal housing grant funds for the next several years to a fund to be used to foster the private development of low- and moderate-income housing in a way designed to advance racial integration.

To remedy the school segregation, the court ordered the Board to take steps toward the desegregation of each school within specified numerical parameters by the 1987-88 school year. To this end, the Board was ordered to create magnet schools and implement a program in which it would assign each student to a school from among those nominated by his or her parents. The court ordered the City to fund the school desegregation plan.

On appeal, the City and the Board raise a variety of objections to the district court’s rulings on liability and remedies. The City contends principally that the court (1) improperly imposed an affirmative duty on the City to build public housing outside of the City’s predominantly minority neighborhoods; (2) erroneously found (a) that Yonkers’s segregated housing patterns were the result of the City’s intentional discrimination, and (b) that the City’s housing decisions were a cause of school segregation; and (3) improperly considered the mayor’s Board appointments in holding the City liable for school segregation. The Board contends principally that (1) the court erred in considering the City’s deliberately segregative housing practices as a factor relevant to the Board’s liability for school segregation, and (2) the court’s finding of segregative intent on the part of the Board was clearly erroneous.

We conclude that the district court properly applied the appropriate legal principles, that its findings of fact are not clearly erroneous, and that its remedial orders are within the proper bounds of discretion. We therefore affirm the judgment in all respects.

*1185

A. BACKGROUND

The present litigation, unique in its conjoined attack on the actions of state and municipal officials with respect to segregation in both schools and housing, brings into question acts, omissions, policies, and practices of the City and the Board of Education over five decades. The case was commenced by the United States in December 1980, with the filing of a complaint alleging, inter alia, that the City and CDA had intentionally engaged in a pattern of selecting sites for subsidized housing that perpetuated and aggravated residential racial segregation, and that the City and the Board had, by their intentionally discriminatory acts and omissions, caused and perpetuated racial segregation in the schools. In June 1981, the Yonkers Branch of the National Association for the Advancement of Colored People (“NAACP”) and an individual minority student, by her next friend, were allowed to intervene as plaintiffs on behalf of themselves and all others similarly situated, see 518 F.Supp. 191, 201-03 (S.D.N.Y.1981), and the action was subsequently certified as a class action.

Trial on the liability issues was held over a period of some 14 months in 1983 and 1984. During the 90 trial days, evidence was heard from 84 witnesses; depositions of 38 additional witnesses were introduced; and thousands of documents were received in evidence. In November 1985, in an exhaustive and well documented opinion reported at 624 F.Supp. 1276 -1553, Judge Sand found the City and CDA liable for housing segregation and found the City and the Board liable for school segregation. Following hearings as to the appropriate remedies for these violations, the court ordered system-wide, comprehensive remedies. See Parts A.I.C. and AII.G. below.

In view of the challenges made in these appeals to the sufficiency of the evidence to support the district court’s findings of intentional discrimination and the contentions that the remedies ordered are overly broad, we summarize at some length the evidence supporting both the findings and the imposition of system-wide remedies.

I. Housing Segregation

The City of Yonkers, New York, is a section of Westchester County roughly 4 to 6 miles long by 3 to 3-V2 miles wide, just north of New York City’s Bronx County. For purposes of this suit Yonkers is regarded as consisting of three basic geographic areas, referred to as East Yonkers, Northwest Yonkers, and Southwest Yonkers. Southwest Yonkers, which comprises less than one-quarter of the City’s land mass, is the City’s most densely populated and urban area. Characterized as containing the “downtown” or “inner city” area, it is the only section having any significant amount of industrialization.

At trial, there was little dispute that, at least as of 1980, when this suit was commenced, the residents of Yonkers were largely segregated by race, with the minorities concentrated in Southwest Yonkers. United States Census figures for 1980 showed that minorities, defined as blacks or hispanics, made up 18.8% of Yonkers’s total population; minorities made up 40.4% of the population of Southwest Yonkers but only 5.8% of East and Northwest Yonkers. Southwest Yonkers, while housing only 37.5% of Yonkers’s total population, housed 80.7% of Yonkers’s minority population.

The minority population of Yonkers grew to 18.8% in 1980 from 2.9% in 1940. During this period, the concentration of minorities in Southwest Yonkers increased as follows:

Total Minority Percentage Minority Percentage of Southwest Minority Percentage Outside of Southwest

1940 2.9 3.5 2.0

1950 3.2 4.5 1.6

1960 4.5 6.7 2.8

1970 10.2 19.8 3.9

1980 18.8 40.4 5.8

Concentration has also been evident within the Southwest itself. In 1940, when minorities constituted only 2.9% of Yonkers’s total population, two of the 10 census tracts in Southwest Yonkers had minority populations between 10% and 50%. In 1980, when minorities constituted 18.8% of Yonkers’s total population, four of the 10 *1186 Southwest tracts had minority populations between 20 and 50%, and five had minority populations of more than 50%. A census-tract map showing the 1980 concentrations is attached to this opinion as Appendix A.

Northwest Yonkers and East Yonkers contained 14 census tracts in 1980, divided into 32 sub-tracts. Of the 32, only two had minority populations of 7% or more. One, located in Northwest Yonkers, had a minority population of 28.6%, most of whom lived in a neighborhood abutting a Southwest Yonkers tract that had a minority population of more than 50%. The other, a neighborhood in East Yonkers known as Runyon Heights, had a minority population of 79.8%. Runyon Heights was a middle-income community founded early in the century on a large tract of land owned by a state senator who regularly brought busloads of black residents from Harlem for picnics at which he auctioned off parcels of land to them. Runyon Heights is bounded to the north by a white neighborhood called Homefield. The original deeds for many Homefield properties contained restrictive covenants prohibiting the sale of such properties to minorities, and as Runyon Heights developed, the Homefield Neighborhood Association purchased and maintained a four-foot strip of land as a barrier between the streets of the two neighborhoods. “To this day, Runyon Heights streets terminate in a dead-end just below this strip.” 624 F.Supp. at 1410 .

The current location of low-income subsidized housing in Yonkers corresponds largely to its concentrations of minority residents. As of 1982, the City had 6,800 units of subsidized housing; of these, 6,566 units, or 96.6%, were located in or adjacent to Southwest Yonkers. A map showing the City’s subsidized housing sites is attached to this opinion as Appendix B. Only two subsidized housing projects were not in or adjacent to Southwest Yonkers. One was a family project located in Runyon Heights; the other, also in East Yonkers, was a project for senior citizens, the majority of whose residents had been expected to be, and were, white. Block-by-block maps for 1950-1980, showing more detail than the census tracts and sub-tracts, revealed that all sites approved by the City for low-income or low-and-middle-income family housing were in or very near neighborhoods that already had high percentages of minority residents.

Given the facts as to Yonkers’s segregated housing patterns, most of the trial evidence on housing issues concerned whether the City’s subsidized housing decisions bespoke a racially segregative intent.

A. Evidence as to the City’s Subsidized Housing Decisions

During the pertinent periods, Yonkers’s governing body was its City Council (“Council”), comprising the mayor, elected in a City-wide election, and 12 councilmen, each elected by one of the City’s 12 wards. The Yonkers Planning Board (“Planning Board”) consisted of seven nonpaid citizens appointed by the mayor. The Yonkers Municipal Housing Authority (“MHA”), a public corporation organized in the 1930’s pursuant to New York State’s Public Housing Law, was the entity authorized to propose, construct, and operate public housing in Yonkers.

Under state law, federal funding could not be requested for a site proposed by MHA until the site was either (1) approved by a majority vote of both the Planning Board and the Council, or (2) approved by at least three-quarters of the Council if less than a majority of the Planning Board approved. According to the testimony of one member of the Council, the opposition of any councilman to a project proposed for his own ward was routinely honored by the other Council members.

1. Housing Decisions in 1948-1958

Prior to 1949, the City had erected two housing projects, both in Southwest Yonkers. The second came about apparently as community leaders’ response to concerns expressed in the late 1930’s about difficulties blacks were encountering in obtaining decent and affordable housing in the private market. Thus, “the City resolved to build a public housing project ‘for Negroes’ and set about finding a suitable site on *1187 which to do so.... Various sites were rejected on the ground that the level of minority concentration there was not sufficiently high, and the site eventually selected in 1940 was in one of the most heavily minority areas of Southwest Yonkers.” 624 F.Supp. at 1312 .

In 1949, pursuant to the National Housing Act of 1949 (“1949 Housing Act”), ch. 338, 63 Stat. 413 (codified, as amended, at 42 U.S.C. § 1441 et seq. (1982)), which provided federal funds for urban renewal, the City applied for the reservation of funds to build 750 units of low-income housing. Its application was approved, but it was not to receive the funds until it had officially designated specific sites and these were approved by the federal Public Housing Administration (a predecessor of the United States Department of Housing and Urban Development (collectively “HUD”)). The City’s initial deadline for submitting approved sites was August 31, 1950. In February 1950, MHA began proposing sites for the construction of these units.

MHA’s first proposed site was a vacant, largely City-owned, parcel of land located in an overwhelmingly white area of Northwest Yonkers. The City’s ownership and the nonuse of the land would have made it a relatively inexpensive building site and avoided any residential displacement and relocation problems. Neighborhood groups, however, swiftly opposed designation of this site, stating that the new housing would be occupied by persons coming from slum areas and that the old slums would continue to exist. The groups recommended clearance of the existing slum areas and the construction of new housing on those sites. The Planning Board rejected MHA’s proposed site, citing the parcel’s nonconformity with planning standards such as sufficiency of school and shopping facilities.

The next two sites proposed by MHA in 1950 were located in white neighborhoods of Southwest Yonkers. Initially, the councilmen of the two wards in which these sites were located recommended approval. As to one site, however, residents of the area appeared at a Planning Board meeting to express their opposition on the ground that the terrain was irregular and that the presence of such housing would tend to harm property values in the area; their councilman withdrew his support for the project, and the site was not approved. The other proposed site was initially approved by both the Planning Board and the Council. However, when an attempt was made to enlarge the approved area, community groups opposed both the enlargement and the original site designation, principally citing the likely deterioration of property values. Eventually, the councilman from this ward withdrew his support, the Planning Board voted unanimously to disapprove the requested expansion, and MHA abandoned its proposal for even the originally approved project.

By December 1950, the City had approved just one project, to which there had been no community opposition, for 274 units. Its site, previously zoned for industrial use, was in a section of Southwest Yonkers having one of the highest concentrations of minorities.

After all of the other MHA-proposed sites had been rejected, a federal official warned that the City would lose its reservation of funding for the remaining 476 units unless it acted to put additional units into development immediately. The City’s response was to expand the previously approved Southwest Yonkers project to 415 units, notwithstanding a prior Planning Board recommendation that no more than 250 units be placed on any site.

In the period 1951 to 1953, MHA proposed 9 more sites for subsidized low-income housing in predominantly white neighborhoods, four in Southwest Yonkers and five in Northwest and East Yonkers. Eight of these proposals prompted vigorous opposition by community civic and social groups, who sent petitions and resolutions to the Planning Board and the councilmen, contending that such projects in their areas would “lead to the eventual deterioration of the surrounding community by the element which they attract.” None of MHA’s proposed sites was approved by the City.

*1188 In the meantime, between 1,200 and 3,000 applications had been received for the 415 units that had been approved. Notwithstanding recognition by the Planning Board and the public of the “desperate need” for additional subsidized housing, no other sites were approved. The City thereby lost allocation of federal funds for the remaining 335 units of its original 750-unit allocation under the 1949 Housing Act.

In 1956, the City was able to renew its reservation of funds for 335 units, and MHA promptly proposed four new sites. One of these was quickly rejected because it was in the path of a proposed highway. The remaining three prompted strong community opposition. Two of these, including one described by HUD as “extremely desirable” for subsidized housing, were in all-white neighborhoods of East Yonkers. The residents of both areas vigorously voiced their opposition at rallies, in petitions, by telegram, and by attending Council meetings in numbers ranging from 400 to 1,000. The City rejected these two sites.

The fourth proposed site was in Runyon Heights, the predominantly black community in East Yonkers. Representatives of the neighborhood opposed the building of low-income housing there on the ground that predominantly white communities had successfully opposed having such projects in their neighborhoods and Runyon Heights should not be the only community in which such a project would be built. They contended that it would be preferable to integrate Runyon Heights into the communities surrounding it and that the placement of low-income housing in Runyon Heights would have the contrary effect of enhancing its racial isolation. The City rejected this site as well.

At least four other sites for low-income housing were formally considered in 1957; none was approved by the City.

In 1958, MHA proposed five sites, four new ones plus one that had previously been rejected because of conflicting highway plans. An MHA official described the sites to the Planning Board as “ ‘the least objectionable’ of those surveyed” but nonetheless predicted that there would be “ ‘a lot of objections on the grounds of race or age in certain sites.’ ” 624 F.Supp. at 1299 .

Two of MHA’s proposed sites in Southwest Yonkers — one in a predominantly white area, and the other in a predominantly minority area — were disapproved by the Planning Board because they lay in the paths of proposed highways. The Council, however, by a three-fourths vote, overrode the Planning Board’s opposition to these two sites; it approved family housing units for the site in the predominantly minority neighborhood and senior citizen units for the site in the predominantly white neighborhood.

The other three sites proposed by MHA in 1958 were approved by the Planning Board. Two of these sites were in overwhelmingly white neighborhoods, one in East Yonkers and described by the City’s Planning Director as “ideal” in terms of transportation, shopping, recreation, and schools, and the other in Southwest Yonkers; the third site was in Runyon Heights. All met with opposition from the residents of their respective neighborhoods. From the two white areas, taxpayer and civic groups wrote their councilmen shortly before the Council was to vote, describing their general opposition as follows:

We personally prefer a public referendum with time to acquaint each and every citizen with the full facts on public housing. Where will these tenants come from? How will we provide schools? How much will it cost us over the years? What safeguards do we have against our having to absorb the overflow from Puerto Rico or Harlem?

The Council voted to reject the sites proposed for the white neighborhoods. It approved the project proposed for Runyon Heights.

Thus, in 1958, the City finally approved sufficient family housing sites to use the remainder of the 750 units that had been allocated to it for 1949. All 750 units were constructed in neighborhoods of high minority concentration; the City had rejected all sites proposed for family housing in any neighborhood not already having a high minority concentration.

*1189 2. Housing Decisions in 1958-1967

For the next several years, MHA and the City concentrated on finding sites for senior citizen housing. The councilmen and the public equated senior citizen housing with housing for whites, and in fact, few of the residents of Yonkers’s senior citizen housing projects have been minorities.

Such housing, so long as not denominated “low-income,” was not perceived as being for minorities and met with little or no community opposition. In 1961, for example, the City approved a senior citizen housing site for 300 units in a minority neighborhood of Southwest Yonkers; though the site abutted a predominantly white neighborhood, the only opposition came when expansion of the project was proposed and residents complained of area overcrowding. In 1963, however, when MHA proposed eight senior citizen sites, four in East Yonkers and four in white neighborhoods of Southwest Yonkers, a local news article, headlined “8 Possible Sites Picked for Low-Rent Housing,” reported that these locations might also be considered to house families displaced by urban renewal. Public protests followed, including a letter from a community association representing more than 2,000 families expressing concern that “[t]o penetrate the community with subsidized housing would tend to deteriorate realty values and adversely affect the character of th[e] community.” Six of the proposed sites were withdrawn.

In 1964, the City sought federal funds to begin a new stage of urban renewal. When its application was rejected due to its poor record with respect to building subsidized housing for displaced residents, the City began once again to look for suitable sites for family housing. In 1965, MHA proposed eleven sites, including five in East Yonkers or white areas of Southwest Yonkers and four in minority areas of Southwest Yonkers. Protests and petitions were lodged against the five white-area sites on grounds of potential overcrowding and the effect on property values. A news report quoted one resident of East Yonkers as complaining that the City wanted to put in her neighborhood “ ‘everything [her family had] tried to get away from’ ” by moving from urban areas to East Yonkers, and another resident as saying “ ‘it wasn’t that she didn’t believe in racial or social or economic integration ... but [that] those people from Yonkers would feel so out of place here ... it would not be fair to them.’ ” 624 F.Supp. at 1303 . The Planning Director supported the East Yonkers sites; the Planning Board approved only the four sites that were in minority areas of Southwest Yonkers.

These four minority-area sites were then approved by a committee of the Council and one was approved by the Council itself. Before any of the sites could be formally submitted to HUD, however, HUD wrote the City suggesting “scattered sites” instead of site concentration in Southwest Yonkers because “[Relocation feasibility, even though quantitatively adequate, falls short of acceptability if racial containment will result from the proposed provision of relocation housing.” In response, a subcommittee of CDA, the coordinating agency for all of Yonkers’s urban renewal projects, compiled a list of 19 sites scattered throughout Yonkers; however, when this list was made public it caused “alarm in the community.” According to one news report, at a meeting of Yonkers housing agencies, “fear was expressed by several speakers that the public is not yet ready to accept the federal government’s plan for racial and economic integration on a citywide basis.” None of the 19 sites was approved.

In 1967, the Council finally approved three sites from among those proposed by MHA in 1965. Despite the Council’s awareness of the federal preference for scattered sites, the three sites approved were located in densely occupied, heavily minority sections of Southwest Yonkers. HUD refused to approve the sites.

3. Housing Decisions in 1968-1974

During the period 1968 to 1974, the City turned to other federal programs for subsidized housing. CDA sought out private sponsors for a combination of low-and-mod *1190 erate-income family projects; it focused its efforts solely on sites in Southwest Yonkers.

Proposed sites that were in the Southwest’s predominantly white areas drew heated community opposition. Notwithstanding the view expressed by former councilman Edward O’Neill that race played no role in site selections — because “ ‘nothing was ever expressed for the record to indicate that it did play a role,’ ” 624 F.Supp. at 1311 — several City officials testified that race was a factor. Some stated that their constituents tended to equate low-income housing with minorities. Others “publicly identified the issue before them as being whether the residents of Yonkers were ‘ready’ for the economic and racial integration being urged upon the City” by HUD and groups such as the NAACP and the Council of Churches. Id. at 1310 .

CDA’s director, Walter Webdale, testified to his view that the high level of emotionalism exhibited at public meetings indicated that residents were concerned about far more than mechanical matters such as the size of the street or the availability of public utilities, and that “racial considerations d[id] come into play.” He gave as an example the reaction to a site proposal for the northern end of Southwest Yonkers which, though just a few blocks from a predominantly minority area, was immediately surrounded by a white neighborhood. A Catholic Church group, led by their pastor, opposed use of this site for family housing and urged that it be used for a senior citizen project instead. The group told Webdale they opposed family housing because they “feared an influx of blacks into the neighborhood.”

Another proposed site called Rockledge, located in a predominantly white area of the Southwest, was initially supported by the ward councilman, Dominick Iannacone. Iannacone testified, however, that he received “flack” from his constituents. Some complained about the loss of the proposed site as a parking facility; others, “who knew him better,” stated that “they didn’t want the housing because they didn’t want any blacks there.” 624 F.Supp. at 1321 . Thereafter, concerned that he would not be reelected if he supported Rockledge, Ianna-cone withdrew his support, citing his constituents’ concern about loss of parking. Using the informal veto power enjoyed by any councilman in whose ward a project was proposed, he “buried” the matter in a Council committee of which he was chairman. At trial, he “acknowledged that his publicly stated reasons for opposing the project were pretextual, and that his opposition in fact was in response to his constituents’ racially influenced opposition.” Id. at 1322 .

In the end, CDA’s efforts resulted in the construction of eight low-and-moderate-income family projects; all were in Southwest Yonkers and all were in or close to that area’s predominantly minority neighborhoods.

Other City activities included consideration in 1969 of subsidized housing for the relocation of 1,000 families from Southwest to other parts of Yonkers; the City’s goal was to ensure plant expansion space in Southwest Yonkers for one of the City’s largest employers, which threatened to move out of Yonkers. A private consulting firm surveyed 98 possible sites, 76 of which were located in East or Northwest Yonkers. A City Council agenda noted that consideration of sites in nonminority neighborhoods had generated a “great deal of controversy”; neighborhood opposition was expressed by citizens’ committees and the presentation of petitions by more than 3,000 residents. Proposals from local businesses for different sites, some “located deep in Yonkers’[s] ghetto areas,” prompted “a passionate debate over racism.”

Alfred Del Bello, mayor of Yonkers from 1970 to 1974, testified that he abandoned the 98-site survey and focused instead on four sites within a five-block radius of the predominantly minority downtown section of Southwest Yonkers. The State Urban Development Corporation agreed to sponsor these sites despite the known concern of the Planning Board that the locations chosen were inconsistent with the goal of commercial and industrial revitalization of *1191 Yonkers; construction was begun without consultation with the Planning Board. Del Bello testified that he had settled on the four sites in minority areas because he “was dedicated to producing housing, and [he] had to find a political course that would allow us to get it constructed.” He stated that “race was definitely a consideration in many of the demonstrations and visible opposition that we had.”

In 1971, HUD warned the City that Yonkers would lose millions of dollars in federal funding unless it provided a more balanced distribution of its subsidized family housing. City efforts to find sites acceptable to HUD included some dozen meetings in nonminority neighborhoods. One official described these meetings as chaotic and carrying a pervasive feeling of “strong fear” on the part of the residents; his perception was that “racial” motivations were “very thick in the air.”

Eventually, in 1972, the City approved construction of 334 units of subsidized housing on a site that was bordered on the north by a heavily minority area and on all other sides by neighborhoods that were predominantly white. This site was approved over the opposition of residents of the predominantly white neighborhoods, the only minority housing site approved over such opposition. Shortly thereafter, the common view being that the councilman in whose ward that site was located had little chance for reelection, the councilman resigned to take an appointed City position. In 1973, a new mayor, Angelo Martinelli, was elected, having promised during his campaign to impose a moratorium on all subsidized housing in Yonkers. The 334 units approved in. 1972 were the last subsidized housing for families constructed in Yonkers.

4. Housing Decisions in 1974-1982

In 1974, the Housing and Community Development Act (“1974 Housing Act”), Pub. L. No. 93-383, 88 Stat. 633 (codified, as amended, in scattered sections of 42 U.S.C.), replaced previous federal urban renewal programs. Designed in part to expand housing opportunities for minorities, this statute allowed a community, inter alia, to receive certificates (called “Section 8 Certificates”) to be distributed to eligible families or individuals who could then choose an apartment in any participating building and have part of the rent subsidized by the federal government. See 42 U.S.C. § 14371 In 1975, the Yonkers Department of Development, an agency formed in 1971 during HUD’s pressure for scattered sites for public housing, applied for 100 Section 8 Certificates, 50 for senior citizens and 50 for families. HUD reserved these certificates for the City, pending approval by the Council.

The Council, however, refused to approve use of Section 8 Certificates by families. Two City officials who attended a Council meeting at which the certificates were discussed testified that many councilmen had been “concerned about the possibility that members of the minority community would, in fact, seek and probably find units on the east side of the city.”

Accordingly, during the next several years, the City either applied for no Section 8 Certificates for families, or applied for and received family certificates but used few of them, or was denied further certificates because of its nonuse of prior certificates. In 1981, after MHA, at the urging of HUD, applied to HUD for Section 8 Certificates for both families and senior citizens, the Council passed a resolution forbidding MHA to apply for certificates for families. To the extent that the City allowed minority families to use any of the family certificates it had received, it referred those families only to buildings that were located in Southwest Yonkers; only white families used certificates in East or Northwest Yonkers.

The 1974 Housing Act also allowed a community to receive funds for housing construction. During the period 1974 to 1979, the City built four senior citizen housing projects using such funds. All four were in Southwest Yonkers.

In 1975, an additional senior citizen project was proposed by a private developer for East Yonkers. It was supported by the Planning Board as “well suited for. *1192 Housing for the Elderly vis-a-vis public transportation, shopping, recreation, etc. as well as its location in the eastern half of the city.” The developer, however, had filed a fair housing statement with HUD, expressing his hope to attract elderly blacks and hispanics from Southwest Yonkers and achieve a 20% minority representation in the project. Local residents opposed the project on the ground that it contained the “seeds of a ghetto,” and the project was killed by the refusal of the City’s Zoning Board to grant minor zoning variances for parking, and by the Council, which criticized the project on the ground— squarely contradicted by the planning experts — that it was unsuitable for senior citizen housing because, inter alia, there was an “unsightly car lot” nearby. The project was not built.

In June 1980, HUD advised the City that continued receipt of federal funding would be conditioned on the City’s taking “all actions within its control” to construct 100 units of subsidized housing for families “outside of areas of minority concentration.” Although the City signed a contract with HUD containing such an undertaking, and several sites were thereafter proposed, no such housing was built. One such site was disapproved by the Council after receiving the “[cjustomary community opposition.” Three others, out of a list of 14 submitted to HUD by CDA, were tentatively found acceptable by HUD, but their use for low-income housing was thwarted by Council zoning actions. For one site, the Council approved a zoning change so that it eventually became a shopping center instead. For another, the Council refused to approve a zoning change to a category consistent with development as subsidized housing. The third site tentatively approved by HUD was the site of School 4, which had been closed in 1976 and remained vacant, costing the City $40,000 to $50,000 per year in maintenance; this site was already in a zoning category that would permit a housing project. It was also in an area that was 98% white. In 1979, as soon as the School 4 property was mentioned as a possible site for low-income housing, the Council voted to remove it from the multifamily zoning category in order “to ‘give the community some peace of mind.’ ” 624 F.Supp. at 1359 .

In 1982, a developer expressed interest in the School 4 site for luxury condominiums priced at more than $100,000. The Council bypassed the Planning Board and took the unprecedented step of creating a citizens’ committee, composed of five white residents of the area, to assess proposals for the use of the property. Four of the five committee members had no experience in planning or zoning, and the committee was not advised to consult the Planning Board. The committee recommended the sale because condominiums priced at $100,000 would attract the kind of people “that we would like to live in the neighborhood.”

Prior to Council action on the proposed sale, a councilman whose ward was near School 4 wrote his constituents urging them to attend the Council meeting, explaining that the NAACP opposed the sale on the ground that low-income housing should be built instead. At the meeting, a videotape of which is in the record, the predominantly white audience overflowed the room. The discussion was emotionally charged, with frequent references to the effect that subsidized housing would have on the “character” of the neighborhood. The final speaker from the audience, a white proponent of the sale, stated that the Bronx had been ruined when blacks moved there and that he supported the condominium proposal because he did not want the same thing to happen in Yonkers. The audience responded with an ovation. During the discussion that followed, when one councilmember pointed out that the current zoning of the site was inconsistent with the condominium proposal (the Council having, as noted above, removed the site from the multifamily zoning category as soon as it was suggested for low-income housing), another councilmember responded, “ ‘we will change that zone when the concept fits the people, not before.’ ” 624 F.Supp. at 1363 .

The Council voted 11-2 to sell the site for luxury housing. A majority of those who voted for the sale stated that “the will of the community” should be honored. Con *1193 summation of the sale has been delayed pending resolution of this suit.

B. The District Court’s Findings as to Housing

After an extensive review of the evidence, Judge Sand ruled that, in view of the “consistent and extreme” segregative effect of the City’s actions, which catered consistently to community positions that were in significant part racially motivated, plaintiffs had sustained their burden of proving that the segregated housing pattern in Yonkers had been caused or exacerbated by the City’s pattern and practice of discrimination on the basis of race in its decisions on the location of subsidized housing. Id. at 1369-73 . He found that this pattern had begun with the City’s first selection of subsidized housing sites under the 1949 Housing Act and had continued through its 1982 attempt to sell the School 4 property for luxury housing. Id. at 1373 .

The court rejected each of the City’s arguments that persons and factors other than the City had been the cause of Yonkers's segregated housing pattern. It found that the cause was not HUD encouragement of subsidized housing construction in Southwest Yonkers, id. at 1328-30 ; rather, HUD had urged scattered construction sites, and the City had repeatedly risked the loss of federal funding by its refusal to select more widely distributed sites, e.g., id. at 1323, 1347,1356 . Nor was the cause a lack of private developer interest in areas outside Southwest Yonkers, id. at 1330-31 ; CDA had sought out developers only for Southwest Yonkers, id., and the City had thwarted the efforts of a developer who sought to build an East Yonkers project intended to attract 20% of its residents from minority groups, id. at 1350-51 . Nor could the housing patterns be attributed to the desire of minority communities for concentration of subsidized housing in Southwest Yonkers; minority groups had begun at least as early as 1956 to express concern about the segregative effects of locating subsidized housing in heavily minority areas and had expressed a desire to "hav[e] the opportunity to live elsewhere in Yonkers.” Id. at 1332-33 . Nor was there, as the City contended, a lack of suitable sites in East Yonkers, id. at 1333-37 ; some of the sites rejected by the Council had been considered by the planners to be “ideal,” e.g., id. at 1300 .

The court also rejected the City’s argument that its site-selection decisions were made pursuant to a race-neutral “legitimate planning strategy" for urban renewal, id. at 1337-42 , for the City’s site selections, far from revitalizing Southwest Yonkers, had brought revitalization efforts to a halt, id. at 1310, 1337 . Rather, the court found that whenever a site was proposed for a predominantly white area, strong community opposition emerged. Id. at 1369 . Though this opposition was not “based wholly upon race,” race was “a significant factor,” id. at 1371 (emphasis in original); the opposition was “based, at least in significant part, upon fear of an influx of minorities into what were (and remain today) overwhelmingly white neighborhoods,” id. at 1313 . The court found that “City officials consistently responded to that opposition.” Id. at 1371 . The inference that racial animus was a significant element in the community opposition to which City officials were responding was drawn from, inter alia, direct testimony to that effect, evidence of overtly racist comments, the racially divided quality of private housing in Yonkers, and a general pattern in which only sites proposed in the predominantly white Northwest or East Yonkers or the white areas of Southwest Yonkers engendered opposition. Id. at 1311-12 . The court found that City officials “came to view racially influenced opposition to subsidized housing in East Yonkers as a ‘fact of life,’ ” id. at 1316 , and made “conscious decisions” to concentrate on “ ‘politically feasible’ ” sites, id. at 1313 . In addition, the court found that “numerous City officials not only responded to, but, in the words of the campaign literature of some, ‘led the fight against subsidized housing in East Yonkers.’ ” Id. at 1373 .

The court found further evidence of the City’s intent to preserve segregation in housing in its conduct with regard to Sec *1194 tion 8 Certificates. Its cut-off of applications for family certificates and its failure to use any already obtained family certificates for minority families outside of Southwest Yonkers were found “inexplicable except by reference to the anticipated race of the certificate holders,” id. at 1347 , i.e., “inexplicable except on the basis of fear that minorities might use the certificates to relocate to East Yonkers,” id. at 1373 . Similarly, with respect to the City’s 1982 attempt to sell School 4 for luxury housing, the court found that the procedural innovations and the nature of the debate made it “difficult to imagine a clearer case of an action taken for a discriminatory purpose.” Id. at 1363 ; see also id. at 1518-21.

In sum, Judge Sand concluded that “the extreme concentration of subsidized housing that exists in Southwest Yonkers today is the result of a pattern and practice of racial discrimination by City officials, pursued in response to constituent pressures to select or support only sites that would preserve existing patterns of racial segregation, and to reject or oppose sites that would threaten existing patterns of segregation.” Id. at 1373. The court emphasized that its finding of the City’s segre-gative intent rested not on a failure to act, but on “a thirty-year practice of consistently rejecting the integrative alternative in favor of the segregative — a practice that had the unsurprising effect of perfectly preserving, and significantly exacerbating, existing patterns of racial segregation in Yonkers.” Id. at 1368.

The court concluded that the conduct of the City and CDA violated the Equal Protection Clause and that their conduct since 1968 violated the Fair Housing Act as well.

C. The Housing Remedy

Having found the City and CDA liable for statutory and constitutional violations, the court held a six-day hearing as to appropriate remedies. In an order published at 635 F.Supp. 1577 (1986) (“Housing Order”) and an unpublished Modification to Housing Remedy Order (“Modification Order”), dated July 8, 1986, the court permanently enjoined the City from, inter alia, intentionally promoting racial residential segregation in Yonkers and ordered that certain affirmative steps be taken toward a wider distribution of public housing.

The court noted that the City had already committed itself to providing sites for 200 units of public housing in order to receive its 1983 Community Development Block Grant (“Development Grant”) funds but had never fulfilled that commitment; the City also had entered into a Consent Decree with HUD that provided that HUD would reduce Development Grant funding if the City did not submit for preapproval sites for at least 140 of the 200 public housing units. The court ordered the City to submit an acceptable Housing Assistance Plan to HUD and execute a grant agreement with HUD, in order to receive, the Development Grant funds for 200 units of subsidized housing, 635 F.Supp. at 1580 ; Modification Order at 2-4, and to “submit to HUD for preapproval at least two sites for 140 [of the agreed 200] units of family public housing,” 635 F.Supp. at 1580 .

The Housing Order provided that if the City did not submit two such sites within 30 days of the court’s order, the City would be deemed to have submitted the sites of three closed schools in East Yonkers, i.e., School 4, School 15, and the Walt Whitman School, or such other sites as might be proposed by plaintiffs and approved by the court. Schools 4 and 15, closed in 1976, had been returned to the City in 1982; Walt Whitman had been closed in 1983, and the court ordered the Board of Education to return that school to the City as well. The court also ordered the City to submit sites selected from a specific list for the remaining 60 public housing units. Id. at 1581 .

In addition, the court ordered the City to create an Affordable Housing Trust Fund for the encouragement of private development of low-and moderate-income housing, to be funded initially with at least 25% of the Development Grant funds allocated to the City by HUD. Id. at 1581-82 ; Modification Order at 1-2. It also ordered the City to establish a Fair Housing Office *1195 with prescribed responsibilities, to seek HUD approval for transfer of the administration of the Section 8 Certificate program to MHA, and to develop a plan for more subsidized family housing units in areas outside of Southwest Yonkers. 635 F.Supp. at 1577-82 .

II. School Segregation

Management and control of the Yonkers school district were entrusted to defendant Yonkers Board of Education. The Board, an independent municipal corporation subject to the control of New York State’s Board of Regents and Commissioner of Education, consisted of nine members appointed by the mayor for staggered five-year terms. Its budget was subject to review by the Yonkers City Council.

At the liability trial, plaintiffs sought to show that students in Yonkers schools were segregated and that that segregation had been caused or enhanced principally by (1) the Board’s general adherence to a neighborhood-school policy, with awareness of the City’s practice of maintaining segregated neighborhoods; (2) other segregative actions of the Board with respect to (a) school openings, closings, and boundary changes, (b) faculty assignments, (c) special education classes, and (d) vocational programs; and (3) the Board’s failure to take any of a number of recommended or otherwise appropriate steps to alleviate the growing school segregation.

Plaintiffs contended also that the segre-gative housing practices of the City were designed in part to achieve and preserve segregation in the schools. They sought to show that the City helped to maintain such school segregation also by, inter alia, the mayor’s appointing to the Board persons known to advocate preservation of the segregated neighborhoods and neighborhood schools.

A. Racial Composition of Each School’s Student Population

As of the 1980-81 school year, Yonkers had 23 elementary schools for grades K-5 or K-6; four middle schools for grades 6-8 or 7-8; two combined elementary and middle schools; four general academic high schools; and one vocational high school. In a number of these schools, special education classes were conducted for students with learning disabilities or emotional disturbances.

1. The General Student Population

In 1980, the student enrollment in Yonkers public schools was approximately 37% minority. The percentage of minority enrollment had approximately doubled from 1970 to 1980, due in part to an increase in minority enrollment and in greater part to a decline in white enrollment:

Yonkers Public School Student Population White White Minority % Minority

1967 . 25,875 85 4,421 15

1970 . 25,049 82 5,583 18

1975 . 21,514 72 8,195 28

1980 . 13,840 63 8,023 37

In 1980, only two of Yonkers’s schools, one an elementary school located in Southwest and the other a middle school in Northwest, had student populations whose racial compositions approximated that of the system as a whole. The next most balanced schools had student populations that were, respectively, 21%, 45%, and 47% minority. The great majority of the schools were either disproportionately white or disproportionately minority.

At the elementary level, although 61% of the students were white, in 19 of Yonkers’s 25 elementary schools the student populations were either more than 80% white or more than 80% minority. Some 85% of Yonkers’s minority elementary school students attended nine schools in Southwest Yonkers. In addition, one elementary school in Northwest Yonkers had an 88% minority population. These 10 schools enrolled 92% of all of Yonkers’s minority elementary school students. More than 55% of Yonkers’s minority elementary school students attended just five Southwest schools, whose minority populations were 75%, 81%, 90%, 98%, and 98%.

Sixteen elementary schools were located outside of Southwest Yonkers. Of these, 14 had student populations that were at least 90% white; more than 70% of Yonkers’s white elementary school students at *1196 tended these 90%-white schools. Of the 11 elementary schools in East Yonkers, only one had a minority student population of more than 7%.

In Yonkers’s middle schools, 62% of the students were white. Two of the six middle schools were located in East Yonkers and together enrolled only 62 minority students, or 5% of Yonkers’s total middle school minority population; these two schools were, respectively, 94% and 96% white. Three middle schools were located in Southwest Yonkers and had minority student populations of 62%, 69%, and 94%. Nearly 80% of Yonkers’s middle school minority students attended the three Southwest schools. Another 15% attended a middle school in Northwest.

About 70% of the students attending Yonkers public high schools, including the vocational high school {see Part AII.A.3. below), were white. Of the four academic high schools, two were located in East Yonkers, one in Southwest, and one in Northwest. The two located in East Yonkers had student populations that were 91% and 98% white. The high school in Southwest had a student body that was 62% minority; it enrolled nearly two-thirds of all Yonkers minority students attending academic high schools.

2. Special Education Classes

The Yonkers special education program provided special classes for students with mental or physical handicaps, including those with learning disabilities or emotional disturbances. Beginning in the 1960’s, there was a growing and disproportionate number of minority students in special education classes. These classes, especially those for the emotionally disturbed, were viewed by many teachers, school officials, and community members as a “dumping ground for black children.” In general, white children would be placed in a special class only after having been referred first to a school psychologist for an evaluation, then to the principal for review of that evaluation, then to the school district’s special education screening committee on the handicapped for a final decision as to what type, if any, special program was appropriate. A black child whose teacher considered him or her “disruptive,” however, would often (“for the sake of discipline”) be consigned immediately by the teacher and the principal to a class for the emotionally disturbed, without prior reference to a psychologist and with no effort to determine whether other options might meet the child’s needs.

As a result, in 1961, when regular classes in Yonkers elementary schools had a system-wide minority population of 10%, minorities made up 22% of the special education classes. By the 1971-72 school year, when the system-wide minority population was 20%, the minority children made up 40% of all special education classes and more than 70% of the classes for those with emotional disturbances.

Location of the special education classes did not follow the Board’s usual neighborhood-school policy; rather, these classes were placed in schools that had space available to accommodate them. Since most of the schools with high minority populations tended to be more crowded, most of the available space was found in schools having virtually all-white student populations. The principals of many of the latter schools resisted the placement of special education classes in their schools for reasons that, in the opinion of a former director of the program, were race-related. Nonetheless, most of the special education classes were placed in schools having few other minority students. In 1972, for example, classes for some 78% of the children classified as emotionally disturbed were conducted in schools whose regular student populations were at least 97% white. Three-quarters of the students in these special classes were minorities.

In most of the schools, there was no mainstreaming of the special education classes into the general school population. Because special education assignments were made without regard to residence, the students were often bused long distances, often well over an hour’s trip, and sometimes up to two hours, in each direction. Thus they arrived at school later than the *1197 regular students and departed earlier. In some instances they entered the school through separate entrances and were kept in classrooms located in secluded areas of the school. In one school, for example, they had to file down two flights below ground and pass through a boiler room to reach their classroom in the subbasement. Special education students also generally took their lunch, gym classes, and recesses separately from the regular students. To the extent that school officials allowed contact between the two groups, the interaction was often purposely negative. One witness who had been a regular student at a 98%-white elementary school in the late 1960’s recalled her perception that all special education students were black and that they were held up to the regular students as examples of “poor, bad behavior.” Thus the special education students were perceived as “different” and “bad.” Another witness, a parent and PTA president, testified that her children had thought the words “retard” and “nigger” were interchangeable because the children’s only knowledge of blacks was of special education students bused into their school.

Nor was the negative reaction to special education students limited to the school's other students. One of the special education teachers and coordinators testified that parents and community members had thrown rocks at her car and shouted “Take your niggers and get out.”

In 1972, the Board hired Dr. Gary Car-man, a special education expert, to direct the program. At trial, he testified that Yonkers, by busing its special education students long distances and physically segregating them from the regular student population, “had the most inhumane program for handicapped children [he] had ever seen anywhere.” Dr. Carman “knew of no causes, medical causes, social causes, biological causes that could possibly account” for the disproportionate number of minorities placed in the classes for the emotionally disturbed. The disproportionate referral of minority students to special education classes eventually prompted an investigation by state and federal education officials. The conclusion of the United States Department of Education was that the Yonkers special education program subjected minority students to discrimination and violated their civil rights.

From 1972 to 1975, Dr. Carman attempted to improve the special education program by reducing the amount of busing, returning some special education students to regular classes, to an extent mainstreaming the special education students into the general school population, and reducing the incidence of virtually all-minority special classes in virtually all-white schools. After Dr. Carman left in 1975, however, these efforts lapsed and the system reverted to one of long-distance busing and placement of blocs of minority special education students in virtually all-white schools. Dr. Carman testified that where the total experience of white children with blacks was their exposure to those in special education classes, the white children would view the special education children as “less worthy” and could well “generalize that to all blacks.”

3. Vocational High Schools

Prior to 1974, Yonkers had two specialized vocational high schools, Saunders Trade and Technical High School (“Saunders”), and the High School of Commerce (“Commerce”). Saunders offered technical courses such as auto mechanics, carpentry, and electricity; Commerce, which was closed in 1974, offered courses such as stenography, bookkeeping, cosmetology, food trades, and dressmaking. Both schools were located in Southwest Yonkers. Neither was subject to the Board’s neighborhood policy and each accepted students from anywhere in the City.

Although precise statistics with regard to vocational school enrollment by race are not available for years prior to 1967, the trial testimony indicated that, prior to 1958, Saunders had a large minority enrollment. From the 1930’s until approximately 1958, it had a reputation as “a school for problem kids” or for “academically retarded pupils,” or as a “dumping ground for minority students.” Many black students from Runyon Heights attended Saunders or *1198 Commerce instead of Roosevelt, the school nearest their homes, often encouraged by their guidance counselor to do so even if they wanted an academic program. Similar steering usually did not occur with respect to academically undistinguished white students.

In 1958, the Board decided to establish entrance requirements for Saunders and Commerce based on grades, achievement and aptitude test scores, recommendations, and discipline records. The criteria for admission were not precise, however, and final decisions lay within the discretion of the respective principals. Apparently these entrance requirements had the effect of changing the community’s perception of the schools as inferior, and by the early 1970’s, Saunders, whose capacity was roughly one-half that of the smallest academic high school, was receiving nearly twice as many applications as it could accept.

At the same time, Saunders’s minority enrollment was decreasing substantially, due in part to the heightened entrance requirements, the acknowledged inferiority of the educational programs available in Southwest Yonkers schools, the subjectivity of the school officials’ evaluation of the applicants’ credentials, and the absence of any effort on the part of the Board to see that minority students, most of whom attended schools in Southwest Yonkers, had an equal opportunity to get into Saunders. Robert Alioto, the school system’s superintendent from 1971 to 1975, and other school district officials believed that Saunders’s selection process “ ‘appeared to systematically exclude minority youngsters.’ ” 624 F.Supp. at 1450 . The Board, “though aware of the systematic exclusion of minorities which resulted from the Saunders admissions process, did relatively little until the late 1970’s to eliminate the discriminatory impact of the methods by which students were chosen.” Id. at 1452.

B. Facility and Faculty Disadvantages of the Predominantly Minority Schools

In support, of their contention that Yonkers’s segregated school system provided minorities with lower quality education than was given to whites, plaintiffs offered evidence of inferior and generally overcrowded facilities at schools with high minority populations, and of high faculty turnover and a lower overall level of teacher experience in such schools.

1. Plant Facilities

School officials testified that adequate facilities at a school are important not only to a student’s physical development but also to his ability to benefit from the instructional aspects of the educational process. Inadequate physical facilities, including space for recreation, can cause disciplinary problems and cause the community to perceive the school as inferior. According to Alioto, the Southwest Yonkers schools “had probably the worst facilities that one could imagine.”

The predominantly minority schools had smaller buildings and sites, particularly in the amount of playground and recreation areas for each school, than the predominantly white schools. For example, the site size of the five most heavily minority elementary schools averaged 1.83 acres; the average site size of the nine most heavily white elementary schools was 4.84 acres. At the minority schools averaging 1.83 acres, the average school population was 413 students. At the white schools averaging 4.84 acres, the average school population was 308 students.

The three predominantly minority middle schools, all in Southwest Yonkers, were located on property totaling 7.2 acres. The two predominantly white middle schools located in East Yonkers were on a total of 19 acres. The total number of students attending each group of schools was nearly identical: 1299 in the Southwest schools, and 1312 in the East Yonkers schools. During the 1970’s, crowded conditions forced one Southwest middle school to use storage closets as classrooms.

The 62% minority high school in Southwest Yonkers was located on 8.0 acres. The high school in Northwest Yonkers, 47% minority, was located on 6.38 acres. The *1199 two high schools in East Yonkers, averaging 95% white student populations, were located on 12.64 and 23.41 acres respectively. A total of some 350 fewer students attended these two East Yonkers schools than attended the Northwest and Southwest schools.

2. School Staffing

Educators testified that it is generally desirable for a school to have a balance of experienced and newer teachers on its faculty and for its staff to be relatively stable from year to year. Relatively high rates of turnover and low levels of faculty experience are factors that contribute to a school’s lower level of educational effectiveness. The evidence regarding the Yonkers public school system revealed that the predominantly minority schools in Southwest Yonkers had low levels of faculty stability, lower levels of teacher experience than the system-wide average, and produced the students with the lowest academic achievement test scores in the system. These schools also had much higher than average concentrations of minority staff as a result of a Board practice of race-based assignments.

The first minority teachers employed by the Yonkers school system, hired between 1946 and 1950, were assigned to School 1, then the only predominantly minority school in the system (91% minority student population in 1950). Until the late 1960’s, the system had few minority teachers and no minority principals. The Board then began to recruit minorities, and the number of minority staff members (i.e., teachers, principals, and assistant principals) rose from 95 in 1967 (out of a total of 1416), to 174 by 1975. Consistently over the years, most of the minority staff members were assigned to the schools having the highest percentages of minority students. For example, in the 1967-68 school year, Yonkers had 28 elementary schools; seven of the eight with the highest percentages of minority students were assigned 40% of the minority staff members. In the 1972-73 school year, Yonkers had 30 elementary schools, including six whose student populations were predominantly minority. The Board assigned 61% of its minority staff members to these six schools. In the 1975-76 school year, Yonkers had 31 elementary schools, including nine whose minority student populations ranged from 60% to 98%. These schools enrolled 29% of all elementary students; they were assigned 75% of all elementary level minority teachers.

Similar patterns were evident in the middle and high schools. For example, in the 1972-73 school year, Yonkers had seven middle schools; the three that had the highest percentages of minority students had 34% of the City’s total middle school enrollment but were assigned 69% of the Board’s middle school minority staff members. In 1975-76, the City had eight middle schools; the four having the highest percentages of minority students, though enrolling only 43% of all middle school students, had assigned to them 81% of all middle school minority teachers.

The Board followed a similar practice in its assignments of minority principals. For example, at the elementary level in the 1973-74 school year, the City had six minority principals; four were assigned to schools whose minority student populations ranged from 68% to 96%. In the 1974-75 and 1975-76 school years, the City had five minority elementary school principals; in 1975-76 it also had one minority assistant principal; all of these persons were assigned to schools having minority student populations of 66% or higher.

While at no time was the faculty of any Yonkers school predominantly staffed by minority teachers, the disproportionate assignment of minority staff to schools having predominantly minority student populations increased the identification of those schools in terms of race. And to the extent that minority teachers were assigned to the virtually all-white schools of East Yonkers it was often to teach the special education classes, which themselves had become known as dumping grounds for minority students. The minority special education teachers “were deliberately assigned to such schools because of the disproportionate number of minority students in Special Education classes.” 624 F.Supp. at 1465 .

*1200 Not surprisingly, in view of the assignment of a disproportionate number of the more recently hired minority teachers to the predominantly minority schools, the average level of teaching experience at those schools was usually lower than the system-wide average. In the year 1967-68, the system-wide average level of teacher experience was 8.45 years. In the elementary schools having minority student enrollments of 40% or higher, the teacher experience level averages ranged from 5.61 to 7.88 years. The only schools whose teachers averaged more than 10 years in experience were schools having 11% or less minority enrollment, four of which were less than 4% minority.

The disparity in teacher experience levels was aggravated in 1969 when the Board entered into a new collective bargaining agreement with the teachers’ union. Notwithstanding the already clear trend of concentration of minority teachers in schools having predominantly minority student bodies, the Board agreed that before assigning any teacher hired from outside the school district to any vacant position within the system, teachers already employed within the system would be given the option, in order of their seniority, of transferring to the vacant position. Thus, as positions became available in East Yonkers schools, the most experienced teachers in Southwest Yonkers schools could, and often did, opt to change schools.

The effects in terms of minority staff concentration, staff turnover, and teacher experience levels were predictable. For example, School 10 was opened in 1972 as a predominantly minority, physically inferior elementary school in Southwest Yonkers (see Part AII.E.2. below). Of the original 17 teachers, 15 were white; within two years, 14 had left the school. In the period 1971 to 1975, the total number of minority staff members employed by the City increased from 133 to 174; but in none of the 17 elementary schools having white student populations in excess of 90% did the number of minority teachers increase. Indeed, in 10 of these schools, the number of minority teachers actually declined; and the four schools that had had no minority teachers prior to 1971 still had none.

In 1971-72, when the system-wide average teaching experience was 7.15 years, the average levels of experience at six of the seven elementary schools having minority student enrollments of 40% or higher ranged from 3.33 to 6.19 years. In contrast, only two of the 13 elementary schools having white student enrollments of more than 95% had below-average teacher experience levels; four of the 13 had staffs averaging more than 10 years’ experience. The disparity in teaching experience levels was, to an extent, decreased in 1976 when, because of the City’s fiscal crisis, the Board laid off 250 teachers, a great number of whom were relatively inexperienced. But even by the school year 1979-80, when the system-wide average was 14.2 years, the average levels at the predominantly minority elementary schools ranged from 9.9 to 13.4 years.

In the 1969 collective bargaining agreement that gave teachers an option to transfer, on the basis of seniority, to vacant positions elsewhere in the system, the Board had reserved the right to compel a teacher to change schools, in certain enumerated circumstances, “when judged to be in the best interest of the school system.” The Board never sought to use this provision in order to decrease the concentration of minority teachers in schools with predominantly minority student populations. Indeed, in 1977, the Board agreed to additional limitations on its right to implement involuntary transfers of teachers.

As a result of the Board’s race-based assignment practices, the eastward flow of the more senior teachers, and the Board’s failure to take any steps to halt that flow or to correct the imbalance of its assignments, by 1980 most of the City’s minority staff members were concentrated in one-quarter of the system’s 36 schools. Of the City’s 25 elementary schools, five that had minority student populations of 75-98% were assigned at least half of the system’s elementary level minority teachers; no minority teachers whatever were assigned to five other schools, all of whose white-stu *1201 dent enrollments exceeded 92%. Of the City’s six middle schools, the three in Southwest Yonkers, which had minority student populations ranging from 62-94% and accounted for 42% of the total number of middle school students in the system, had 62% of the system’s middle school minority teachers. Of the five high schools, the two that had the highest minority enrollments (47% and 62%) accounted for 46% of all the high school students in the system but had 77% of the system’s high school minority teachers.

C. The Board’s Decisions as to School Closings, Openings, and Attendance Zone Changes

During the decades on which this litigation focused, the Board made many decisions with regard to opening and closing schools and realigning their attendance zones. Plaintiffs sought to show that many of these decisions evinced an intent to create or maintain segregation in the Yonkers public schools.

1. Attendance Zone Changes

Among the attendance zone changes were several affecting Schools 16 and 25, elementary schools located in Northwest Yonkers, less than one mile apart. Between 1953 and 1968, the Board redrew the boundary between these two schools four times. In 1953, School 25 had a minority student population of 4%; that of School 16 was 0%. The 1953 boundary change resulted in the reassignment of 35 white students, and no minority students, from School 25 to School 16. Ten years later, the minority population of School 25 had risen to 14%; School 16 still had no minority students. A 1963 boundary change resulted in the reassignment of nine white students, and no minority students, from School 25 to School 16. In the following year, a boundary change resulted in the reassignment from School 25 of 23 white students and nine minority students, thereby bringing the minority population of School 16 to 2%. By 1968, Yonkers’s system-wide percentage of minority students was about 15%, and School 25 had a minority population of 42%. A boundary change in that year resulted in the reassignment of six of its white students, and no minority students, to School 16. School 16’s minority population was 1%.

The Board argued that these changes had been designed to avoid having the reassigned children traverse a steep hill between their homes and school. Board reassignments in other sections of the City, however, had been made though they forced the reassigned students to cope with similar topographical conditions, and in fact two of the boundary changes between School 25 and School 16 made the trip to school harder, not easier, for the students who were reassigned. No other explanation was offered by the Board.

A 1963 attendance zone change between Southwest Yonkers Schools 9 (15% minority) and 12 (42% minority) was similarly unexplained by race-neutral criteria. The attendance zone for School 9 was directly north of that for School 12. In 1963, the Board moved the boundary line farther north. While this change slightly lowered the minority percentage enrolled in School 12, it substantially lowered School 9’s 15% minority percentage as it reassigned some 40% of School 9’s minority students to School 12; even prior to the reassignment, School 12 had had the second-highest minority concentration in Yonkers. This boundary change was contraindicated by the relative student-capacities of the two schools. According to the Board’s figures, prior to the change, only 77% of the capacity of School 9 was utilized; School 12 was 96% full. The boundary change caused School 12 to be overcrowded.

2. The Failure To Close or Rezone Longfellow

The Longfellow Middle School, located in Southwest Yonkers, has long been the Yonkers middle school with the highest percentage of minority students. In 1950, though only 12% of its students were minorities, these students constituted 41% of the City’s entire minority middle school population. Housed in a relatively small facility with no outdoor recreational space, *1202 by 1969 the school had become underutilized as Burroughs Middle School was opened one mile away and the attendance zone for Longfellow shrank. The drawing of the attendance zone line between Longfellow and Burroughs decreased the number of white students attending Longfellow, and the increasing minority population of Southwest Yonkers led to increasing numbers of minority students. In 1967, Longfellow’s student population was 38% minority; after the opening of Burroughs in 1969, Longfellow became 50% minority. By 1973, Longfellow had become 79% minority.

The combination of its disproportionately high minority student population, its inferi- or physical facilities, and its underutilization caused many education officials and community leaders to urge repeatedly, beginning at least as early as 1967, that Longfellow be closed. The Board rejected all proposals either to close Longfellow and transfer its students to other schools that were less heavily minority, or to expand Longfellow’s attendance zone so as to achieve a desegregative influx of nonmi-nority students. For example, in 1977, when the Board planned to close the nearby Burroughs as a middle school, the Longfellow PTA urged the Board to return to Longfellow the predominantly white area that had been rezoned from Longfellow to Burroughs in 1969; such a realignment would have made use of Longfellow’s excess capacity and had a desegregative effect. The Board rejected this suggestion, deciding instead to reassign the Burroughs students — even those who lived within one mile of Longfellow — to Emerson Middle School, two miles away near the northwest corner of the City, or to Whitman Middle School, four miles away near the northeast corner of the City. Though the Board initially reached this decision while an overall school reorganization plan recommending the closing of Longfellow was under consideration, it adhered to the decision after the reorganization plan had been rejected, stating that Longfellow might still be closed.

Other proposals recommended closing Longfellow and reassigning its students to Mark Twain Middle School, located in the southeast corner of East Yonkers, some three miles from the site of Longfellow. The proposal had both fiscal and desegre-gative merit, for Twain was operating at less than its stated capacity, and it had only a 2% minority population. The Board refused, however, stating that the distance the students would have to travel to reach Twain would be too great and that Longfellow students’ parents would not have the ability to carpool their children or pay for the necessary transportation. In fact, however, many students already within the Twain attendance zone were required to travel some 2lk miles to school, and a one-way distance of some four miles had not deterred the Board from reassigning some Burroughs students to Whitman. Further, though the Board had arranged transportation several times in other circumstances, it made no effort to explore this possibility with respect to the proposed reassignment of Longfellow students to Twain. Finally, the net cost of providing transportation for reassigned Longfellow students would have been relatively low, both because the Board could have saved some $500,000 per year in operating and faculty costs by closing Longfellow, and because under New York law the state would have provided 90% reimbursement for transportation expenses incurred for purposes of school desegregation.

In sum, from the late 1960’s, the Board rejected proposal after proposal for the reassignment of more white students to Longfellow or of Longfellow minority students to schools with lower percentages of minorities. It declined to desegregate Longfellow on the ground that the school might be closed; but Longfellow was not closed, even in 1976 when the City’s well publicized fiscal crisis required the Board to close several schools. At the time this suit was commenced, Longfellow remained in inferior physical facilities, operated at 31-40% of its capacity, and had a minority population of 94%.

3. The Opening of Commerce Middle School

In 1973, in conjunction with the closing of the High School of Commerce, located in *1203 Southwest Yonkers a few blocks from the downtown area, the Board opened a new Commerce Middle School (“Commerce Middle”). Its student body consisted of junior high school students who theretofore had attended Gorton, a combined junior and senior high school located in the southern part of Northwest Yonkers. The initial enrollment in Commerce Middle was 53% minority.

Prior to deciding on Commerce Middle’s attendance zone, the Board had been presented with a number of proposals that would have avoided this creation of yet another predominantly minority school in Southwest Yonkers. These proposals principally involved Emerson, a combined elementary and middle school in Northwest Yonkers located about V-k miles north of Gorton. Emerson then had a middle school minority population of 8%. One proposal was to assign to Commerce Middle the middle school students from Emerson who lived in the southernmost part of the Emerson attendance zone. There was strong opposition from white residents, however, to any relocation of white students to form an integrated Commerce Middle, opposition that the Board perceived as grounded principally in racial concerns. The Board was also well aware that transferring Gorton students to the proposed new Commerce Middle without reassigning students from any other school would have a distinctly segregative effect: memoranda assessing this alternative noted, “Commerce may become an all-black school”; “Commerce could be all black”; “Commerce becoming basically a black school”; “Racial Distribution — all black.” It decided to assign to Commerce Middle no students other than those from Gorton.

It also rejected proposals to reassign the Gorton junior high school students — 41% minority — to Emerson instead of to Commerce Middle, a course that apparently was both feasible in terms of Emerson’s capacity and consistent with repeated proposals from school officials and community members to convert Emerson from a combined elementary and middle school to an exclusively middle school. The Board declined to reassign Gorton students to Emerson, on the ground that “tensions” would be created, apparently a reference to racial concerns, for in 1973, one-third of the Emerson’s middle school minority students were transferred to Burroughs

in response to race-related concerns of the Emerson community regarding the presence of minority students at the school. According to [school administration officials], this transfer was effectuated for the purpose of insuring the safety of minority students who had been enrolled at the school in light of altercations which had occurred between students at the school and the Emerson community’s opposition to the attendance of minority students at Emerson.

624 F.Supp. at 1481 .

After opening Commerce Middle as a 53% minority school in 1973, instead of expanding Commerce Middle’s attendance zone northward to draw in any predominantly white neighborhoods, the Board redrew the zone boundary farther south, thereby reassigning to Commerce Middle students from Longfellow and another predominantly minority school. Commerce Middle’s minority population thus increased to 70% in 1974 and to 77% in 1975. In 1976, the school was closed as part of the Board’s response to the City’s fiscal crisis.

4. Other Board Actions

Other Board decisions challenged by plaintiffs included the early rezoning and 1954 closing of School 1 in Runyon Heights, the 1969 opening of the Martin Luther King, Jr., School in Southwest Yonkers, and the fiscal-crisis-related closings of several schools in 1976.

School 1 was located in Runyon Heights, the predominantly black community in East Yonkers. For a time in the 1930’s it was attended by students from the Homefield section immediately to the north as well as by students from other largely white neighboring areas; white students then made up one-half to two-thirds of its student population. In 1938, however, the Board redrew the School 1 zone to correspond more precisely with the boundaries of Runyon *1204 Heights. Students from Homefield were reassigned to School 22, increasing the distance of their trip but sending them to a virtually all-white school; students south of Runyon Heights were sent to the then-virtually all-white School 5. By 1950, School 1 was 91% minority; at the time of its closing in 1954, it was 99% minority.

As a result of the 1938 rezoning, described by the court as “deliberate, racially motivated gerrymandering, done in a manner which carefully incorporated privately created residential segregation,” 624 F.Supp. at 1411 , the School 1 zone was the smallest in the City, and the school operated at less than 42% of its capacity. Meanwhile, two nearby virtually all-white schools, Schools 8 and 22, became overcrowded. Runyon Heights community members sought to have the Board expand the School 1 boundaries in order to draw in students from the surrounding areas, thereby decreasing its underutilization, relieving the surrounding schools’ overcrowding, and having a desegregative effect on School 1. Instead, in 1954 the Board decided to close School 1 and send its students to Schools 5 and 24, which had a desegregative effect on those schools. None of the Runyon Heights students were sent to School 22, which remained virtually 100% white, thereby “preserving] an all-white school experience for Homefield students, consistent with the Board’s deliberately segregative attendance zone boundary changes of prior years.” 624 F.Supp. at 1413 .

With respect to the Martin Luther King, Jr., School (“King”), the court found that the initial hope of the Board was, unlike its segregative intent in rezoning School 1, that the opening of King would serve as a significant step toward correcting racial imbalance in the schools of Southwest Yonkers. King was opened in 1969 for grades 4-6 with students reassigned from Schools 6 and 12, both of which were overcrowded and predominantly minority. The population of King at this point was 57% minority. The following year, in accordance with the Board’s original plan, students from the predominantly white School 9 were added, thereby decreasing the minority population of King to 49%.

The assignment of children who had attended School 9 prompted a December 1970 petition signed by 434 of their parents to have the Board restore the prior attendance zones. The Board held fast for a year and then relented. In the interim, white students from the School 9 area began to withdraw from King, apparently either relocating or entering private schools, reducing the number of white students at King from 392 in 1970-71 to 224 in 1971-72.

In 1972, School 9 was eliminated as a King feeder school, and third-graders who would otherwise have gone on to King for fourth grade remained at School 9. Some 60% of this group were white. In 1973, King was converted from a grade 4-6 school to a K-5 school; its students came from the predominantly minority areas previously served by Schools 6 and 12, but not the predominantly white areas of School 9. King’s minority enrollment rose from 49% in 1970, to 70% in 1971, to 78% in 1972, to 87% in 1973. By the time of this lawsuit, it had a minority student population of 98%.

Although the district court viewed the consequences of some of the Board’s decisions with regard to King as “foreseeably segregative,” 624 F.Supp. at 1402 , it was unpersuaded, in light of the surrounding circumstances and the Board’s initial de-segregative intent, that the later decisions of themselves bespoke a segregative intent.

The court explored Board decisions with respect to opening and closing other schools, including those closed in 1976 in response to the City’s fiscal crisis. Most of these decisions had some segregative and some desegregative effects and the court was unpersuaded that the decisions themselves demonstrated a Board intent to preserve segregation. Rather, the court concluded that a major indicator of segre-gative intent was the Board’s failure to adopt any proposal or plan to alleviate the segregated patterns its prior actions had achieved.

D. The Board’s Rejection of All Proposals Involving Desegregation

The first significant official recognition of the need to address the racial imbalance *1205 of the Yonkers public schools occurred during the 1968-1970 superintendency of Paul Mitchell, who expressed his concern that the racial segregation of the schools prevented equality in educational opportunity. During his tenure came the opening of King and School 10 (see Part AII.E.2. below) in Southwest Yonkers, both of which, though they quickly became minority schools, had been planned by the Board as racially integrated schools. The Board also conducted a series of human relations workshops and sought the assistance of state education officials in addressing the problem of racial imbalance. Mitchell’s successor, Alioto, hired a special consultant to serve as a liaison between school officials and community members, with particular emphasis on communicating the concerns of the minority community to school officials and alleviating the tensions at the racially troubled Gorton School.

Nonetheless, while Alioto recognized the increasing racial imbalance in the schools and the inequality of educational opportunity within the system, particularly with respect to the inadequate facilities and inexperienced teachers that characterized many of the Southwest Yonkers disproportionately minority schools, he and other officials noted that there was strong community opposition to desegregation. For example, the education specialist sent to Yonkers by New York State described a “very hostile audience” at one PTA meeting in East Yonkers and testified that white parents had stated explicitly, “ ‘We don’t want desegregation, I don’t want my children going to school with black children.’ ” Accordingly, Alioto, having instructed his special consultant to gather information on the extent of racial imbalance in the schools, instructed him to cease work in this area because Alioto believed it would be politically infeasible to proceed with desegrega-tive efforts in the schools at that time. The state specialist testified that Alioto informed him that “there was great community resistance and that it was unfeasible to try to develop a desegregation plan and then implement it.” A former Board member testified, “There is no question [Alioto] said it and he said it to many people. He said it could never be sold in the Yonkers community. Any kind of totally city-wide racially balanced program would be politically infeasible.”

1. The NYU Report and the 1973 Reorganization

In October 1971, the Board commissioned a study of the Yonkers public school system by the New York University (“NYU”) School of Education’s Center for Educational Research and Field Services. The study team was not asked to address the issue of racial imbalance.

The NYU Report, delivered in 1972, made several recommendations, some of which, though not addressing racial issues directly, had desegregative implications. In this category were recommendations to (1) reorganize all schools into a uniform K-5, 6-8, and 9-12 grade configuration, one facet of which would involve a potentially desegregating school attendance zone change for students from Homefield; and (2) decentralize the vocational education program by (a) closing the High School of Commerce and having a new set of courses offered at Saunders, (b) having two complete sets of the existing vocational courses taught in the academic high schools, one set divided between the two high schools located in the northern part of the City and the other set divided between the two high schools located in the southern part of the City, and (c) allowing a student to take any of the vocational courses taught either in his own school or in the paired school located to the east or west (the “variable access plan”).

The NYU Report prompted strong community opposition to any revision of the vocational program that would either cause the predominantly white students from East Yonkers to have to attend classes in the disproportionately minority high schools in the western half of the City or allow the minority students from the west to attend classes at the 94-97% white high schools in East Yonkers. School officials characterized these objections as reflecting a “[f]ear of racial encro[a]chments.”

*1206 Two weeks after the last public hearing on the NYU Report, Alioto presented his 1973 Reorganization Plan to the Board. In general, substantially as a result of community opposition to the desegregative facets of the NYU recommendations, the plan included the most segregative proposals that had been made either in the NYU Report itself or in the ensuing alternative suggestions. Thus, the plan adopted the suggestion to decentralize the Saunders vocational programs, but only in part: It rejected the east-west pairing-and-sharing proposal of the NYU Report, and instead incorporated the significantly more expensive approach of duplicating certain of Saunders’s vocational courses in each of the four academic high schools. The opening of Commerce Middle as a predominantly minority school, discussed in Part AII.C. 3. above, was also part of this proposed 1973 Reorganization Plan. The Board promptly adopted the plan as recommended by the superintendent.

The only potentially desegregating feature of the NYU recommendations that was adopted was that part of the suggestion to standardize the grade configurations which entailed reassigning students from the predominantly white Homefield neighborhood, then attending the overcrowded Roosevelt High School (then 6% minority) in East Yonkers, to the soon-to-be-underutilized Gorton (high school population 24% minority). This recommendation was adopted over opposition of Homefield parents that school officials inferred was partly race-related. However, the major desegregative effect of even this change was delayed, as in the first year thereafter the Board permitted nearly half of the 132 reassigned Homefield students to remain at Roosevelt; later some Homefield students began using false addresses to avoid having to attend Gorton. In all,

the evolving segregation of the district’s schools remained substantially unaltered. No student movement between the district’s regular high schools was effectuated despite the recognition that racial integration would be an advantageous result of the variable access plan. The Saunders facility remained intact despite the realization that the school’s physical inadequacies and screening process [were] presently resulting in the inaccessibility of vocational and occupational education opportunities to many minority students. The racially balanced High School of Commerce was closed and was replaced by a predominantly minority middle school. No desegregative reorganizations were effectuated at the elementary school level, as would have occurred under some of the NYU Report proposals.

624 F.Supp. at 1475-76 .

2. Phase II

A serious official proposal for the desegregation of the Yonkers public schools was made in 1977 by then-superintendent Joseph Robitaille. In late 1975, in response to concerns expressed by the Yonkers NAACP over the increasing racial imbalance in the schools, the Board had established a Task Force for Quality Education (“Task Force”) to explore the system’s problems, including declining enrollment, underutilization of school facilities, and fiscal constraints. Announcement of the initial formation of the Task Force omitted any mention that the group would explore racial problems, an omission designed to avoid arousing community hostility. Nonetheless, public resistance quickly materialized, with East Yonkers residents expressing concern that transfer of western Yonkers students into their schools would lead to a decline in educational standards and student achievement and create disciplinary problems; they took the position that the Task Force should be more concerned with improving the schools’ overall educational quality than with correcting racial imbalance. Nonetheless, the Task Force’s final report, concluding that the Yonkers schools were “racially and ethnically segregated ... due to segregated housing patterns, socio-economic deprivation, and systematic racism,” made a number of remedial recommendations.

In August 1977, Robitaille issued his Phase II School Reorganization Plan, which recognized the interrelationship among the *1207 system’s fiscal, enrollment, utilization, and racial problems, and incorporated some of the Task Force’s recommendations. The principal changes proposed in the Phase II plan were (1) the reorganization of the below-high-school grade configuration to K-6 and 7-8; (2) the closing of three middle schools, Longfellow, Fermi, and Burroughs; (3) the relocation of Saunders to the to-be-vacated Burroughs facility; (4) the closing of Southwest Yonkers’s School 6, then 98% minority, and reassignment of its students to underutilized elementary schools to the north, with a view to improving racial balance; and (5) the “Yonkers Plan” for school desegregation. The Yonkers Plan was essentially to limit the size of each elementary and each middle school, drawing its attendance zone accordingly, and to bus students residing outside the redrawn zone lines to other schools in a pattern that would improve the overall racial balance of the system. It was anticipated that no more than 20% of the students would have to be bused and that the greater efficiencies would result in savings to the City, over a 10-year period, of nearly $29 million.

Phase II in general, and the Yonkers Plan in particular, met with overwhelming community opposition. Many statements from residents of East Yonkers focused on the loss of neighborhood schools, the lack of any planned improvement in the quality of education, and the failure to present possible alternatives to busing, such as the use of magnet schools. Residents of Southwest Yonkers objected to the plan because of the loss of neighborhood schools and because the burdens of traveling to school by bus would be borne disproportionately by the minority students from that area.

Some statements from East Yonkers residents presented explicitly race-related opposition, including flyers protesting the busing of East Yonkers students and busing of “the black children (3,000 in number) to our neighborhood schools”; a letter from a community group that was “unalterably opposed” to “compulsory (non-voluntary) busing for racial purposes as an end in itself”; a letter expressing concern that busing “ ‘blacks & hispanics’ into our east side schools” would be detrimental to the neighborhood, and suggesting that the Task Force be renamed “ ‘Racist Force us’ to take our children and go!”; and a letter from a neighborhood association stating the residents’ desire to “preserve the nature of our neighborhoods” and their opposition to “mov[ing] children about for the sole purpose of ethnic and racial mixing” (emphasis in original).

Similarly, at community meetings in East Yonkers, school officials were presented with comments expressing concern that the plan would result in Yonkers’s becoming “another Bronx,” referring to the perceived community deterioration and slum-like conditions that speakers associated with the increase of minority population in that New York City borough. The audience punctuated these and similar statements by local residents with cheers and applause. In contrast, proponents of Phase II were booed and hissed upon introduction, upon mentioning such matters as the inferior books used in Southwest Yonkers schools, and throughout their presentations. One elderly black woman, upon mentioning the prospect of busing students from west to east and stating that children should learn from one another, was booed and shouted at to such an extent that a recess had to be called.

While no explicit racial epithets were used by persons making public statements at the hearings, several trial witnesses testified that community members made specific racial slurs both inside and outside the hearing room, such as, “they are going to send blacks, and they are going to send niggers and they are going to send spicks out here,” and “we don’t want those children.”

Without ever taking a formal vote, the Board unanimously disapproved of all of the desegregative aspects of Phase II. The Yonkers Plan was rejected; School 6 was not closed; Longfellow was not closed; no students were bused.

The stated basis for the rejection of Phase II’s desegregative components was *1208 the Board’s preference for the use of magnet schools and open enrollment plans for achieving voluntary desegregation. Although it appears that all of the Board members acknowledged that at least some of the community opposition to Phase II was racially motivated, and some believed that racism was the principal basis of that opposition, there was no express discussion by Board members of the race-related community opposition to Phase II except by Quentin Hicks and Anne Bocik, members whose recent appointments to the Board had been extremely controversial, see Part AII.E.3. below. Hicks, a black whose appointment had been protested by members of the black community because he did not represent their interests, stated that black parents were concerned about having their children transported out of their neighborhoods into a “white jungle.” Bocik, a former principal who had been forced to retire in part because of “her use of racial slurs and other racially insensitive behavior toward minority students,” 624 F.Supp. at 1507 , stated that minority students and administrators from minority schools “would like to be with their own.”

Notwithstanding its stated preference for voluntary methods of desegregation, the Board took no steps to develop or implement any of the desegregative alternatives suggested by its own members or by members of the community. Thus, despite its professed enthusiasm for magnet schools or open enrollment, no magnet school, open enrollment, or other voluntary plan for desegregation was implemented at any time.

As a result, in 1980, the schools of East Yonkers, many of which were operating at less than 60% of their planned capacities, remained predominantly (overall 95%) white in student population, with superior and spacious physical plants, and experienced faculties. The schools of Southwest Yonkers remained predominantly (overall 67%) minority in student population, some overcrowded and some seriously underutilized, housed in inferior physical facilities, staffed with less-experienced staff members and more than half of the minority teachers employed by the school system, and providing their students with concededly inferior educational opportunities.

E. The City’s Activities With Respect to School Segregation

In contending that the City as well as the Board should be held liable for segregation in the Yonkers public schools, plaintiffs pointed to, inter alia, the interrelationship between housing segregation and school segregation, the City’s control over school budgeting and plans, and the mayor’s appointments to the Board of persons opposed to desegregative action.

1. The Interrelationship Between Schools and Housing

In an effort to refute the contention that its actions in concentrating subsidized low-income housing in Southwest Yonkers had had the effect of enhancing school segregation, the City offered a study that concluded that if none of the subsidized housing projects in Southwest Yonkers had been built and each of the project sites had remained vacant, the racial balance in Southwest Yonkers’s schools would not have differed significantly from the actual 1980-81 figures. In contrast, plaintiffs’ expert testified that building low-income housing to be occupied principally by minority families tends to create a school that, while not necessarily showing an immediate dramatic increase in minority students, soon becomes identified as a “minority school.” Such an identification encourages resident white families to move out of the neighborhood and discourages other white families from moving in.

The relationship between schools and housing was hardly lost on the City while it was making its various decisions as to whether and where to construct subsidized housing. One Council member testified that nearly all of the East Yonkers councilmen had indicated that their constituents objected to subsidized low-income housing partly because “[i]n order to keep the schools nice, you know, you’d have to keep out the minorities.” Further, as described in the previous section, a common theme of *1209 East Yonkers residents’ opposition to the Yonkers Plan for school desegregation was the desire to “preserve the nature of our neighborhoods.” As described in Part AII.E.3. below, Mayor Martinelli explicitly opposed desegregation of the schools by busing in part because it would diminish the stability of the residential patterns.

There was also evidence that City officials had requested that the Board make several school attendance zone changes that would have enhanced segregation at the schools to be affected. For example, in 1974, Martinelli urged that a small nonmi-nority area of a neighborhood be moved from the attendance zone of an elementary school that was 60% minority to one that was 88% white. A few months later a Council member suggested that several predominantly white blocks be redistricted from a school that was 28% minority to one that was 97% white. In 1976, another City official made a similar request at the behest of a landlord who had complained that his ability to attract tenants was detrimentally affected by the location of his property within the zone of a school that had a substantial minority enrollment. The Board declined to implement any of these requested changes.

2. City Influence on the Board

Under state law, the Yonkers school district is fiscally dependent upon the City, and the Board’s annual budget is subject to approval, line by line, by the Council. N.Y. Educ.Law § 2576 (McKinney 1981). Because of the Council’s fiscal control over the Board, “in the public mind there [were] two boards of education actually operating,” with citizens often looking directly to the Council in school matters. There was no evidence, however, that the Council in any particular instance disapproved a school budget that included a desegregation plan; there could be no such evidence because the Board never sought to implement a plan that had any significant de-segregative elements.

The Board’s willingness to put specific proposals before the Council was not constrained solely by fiscal considerations. For example, in 1973, the superintendent recommended to the Board, and the Board submitted to the Council, recommendations for vocational program modifications that were more expensive than the pairing-and-sharing proposal of the NYU Report. Both the East Yonkers community and a number of Council members had publicly opposed the NYU Report’s recommendation. The Board’s spurning of the less expensive NYU proposals “was influenced by the perceived infeasibility of obtaining City Council approval.” 624 F.Supp. at 1506 . Similarly, in the Phase II proposals, the school closings and the state’s substantial subsidization of transportation costs would have resulted in a net reduction of the school system’s annual expenditures, and fiscal concerns thus could not explain the Board’s rejection of those proposals. As a practical matter, however, East Yonkers community opposition to Phase II was strong, Council members and the mayor had publicly expressed their opposition, and the Board always had an eye on what was “politically,” not just fiscally, feasible. As one Board official put it, “we, in essence, had to convince another series of people, most of whom were elected by the community, and to the extent that the community resisted the idea, any idea, it seems to me that that would have some impact upon the people who owed election to those same individuals.”

The City’s influence on the Board was also visible in certain decisions as to school sites and configurations. For example, in the late 1960’s the Board commenced plans for School 10, which it intended to open as an integrated elementary school in Southwest Yonkers, drawing students from School 3 (then 34% minority), School 19 (then 68% minority), and School 27 (then 5% minority). Planned as an experiment in the “open school” concept, in which the interior space would be flexible, unstructured, and without walls, the building was to be located on a five-acre site having a general openness of environment harmonious with the openness, to be found within. As eventually constructed, however, School 10 was a mean and inadequate ghetto school, due largely to changes urged by the City which *1210 the Board grudgingly felt compelled to accept.

Without recounting the many events that occurred en route to the birth of School 10, which are described in detail in the district court’s opinion, 624 F.Supp. at 1403 -10 and 1542-43, suffice it to say that first, the Board agreed to change its preferred site to one in the middle of an urban renewal project (in order to allow the City to use the construction of School 10 as a statutorily permitted noncash contribution to the urban renewal area); later it accepted a one-acre site instead of the originally approved five acres (because the City decided to erect additional apartments on part of the site); as the site was developed, the front of the school could not be seen from the street (because the City wanted that frontage for an apartment-retail-store complex); and in the end, the school had virtually no outdoor recreation area (because the City needed more garage space for apartment residents). Though the Board objected to the City’s inroads into the School 10 facilities, it eventually capitulated to each demand.

Because of its location behind other buildings and its lack of outdoor play area, School 10 became known as the “airshaft” school and was characterized immediately as a “new ghetto school.” By 1980, it had the fourth largest minority percentage enrollment in the City.

3. The Mayor’s Appointments to the Board

Although the Board was an independent municipal corporation under state law, its nine members were appointed by the may- or. Prior to the election of Mayor Marti-nelli, many Board members served more than one term, frequently being reappointed by a mayor other than the one who had originally appointed them. In the 25 years just prior to the advent of Martinelli, two-thirds of the Board’s 33 members had been reappointed by a successor mayor. In 1973, after HUD had made clear that further federal funds for housing would be withheld unless the City allowed low-income housing to be constructed outside of Southwest, Martinelli won election on a campaign platform that included a promise that no more subsidized housing would be constructed in Yonkers. Once in office, Martinelli, who opposed busing and favored the policy of neighborhood schools, set out to appoint members “based on his philosophy of education,” so that “it would be his Board.” He did not reappoint a single person who was serving on the Board at the time he was elected. Many of his appointments were controversial.

His first appointment, in 1974, was Angelo Paradiso, who had been the principal at Saunders from 1964 to 1973. Paradiso had resigned in 1973 after a dispute with Alioto concerning the Saunders screening process and Paradiso’s unwillingness to address the problem of disproportionately low numbers of minority students at the school and what Alioto perceived as the systematic exclusion of minorities.

In 1975, Martinelli appointed as Board members Morton Wekstein and Anne Bo-cik. Wekstein was the Mayor’s personal attorney, and his appointment drew criticism in part because Wekstein’s law partner was then representing a number of school administrators who had been considered ineffective by Alioto. A year later, Wekstein resigned because of a conflict of interest.

Bocik was a former teacher and elementary school principal who had retired in 1974 after Alioto requested her resignation. As a principal, Bocik had vowed that there would never be a full-time minority teacher of academic subjects in her school; she had received unfavorable job evaluations because of her ineffectiveness in planning and her common use of racial slurs and other racially insensitive behavior toward minority students. Bocik’s treatment of minority students in this manner had been the subject of complaints to school administrators from both minority and white teachers; at trial, one teacher described in detail incidents in which Bocik terrorized or humiliated minority students, used racial epithets in referring to minority children, described them as animalistic, and threatened to “buy bleach, Clorox, Purex to *1211 bleach them, their skins, because perhaps that would improve their behavior.” Soon after Bocik’s forced retirement, a state senator wrote Martinelli, recommending that she be appointed to the Board based on her experience and her Slavic background; her appointment was supported by the United Slavonian American League. It was opposed by the Board’s president, by Alioto, and by community members, especially from the minority community. Martinelli appointed Bocik to the Board and defended the appointment by reference to her ethnic background.

The mayor made several appointments in 1976. First, after Wekstein resigned, Mar-tinelli was asked to consider appointing an hispanic to the Board. Notwithstanding his recent justification of the Bocik appointment on grounds of her ethnicity, he responded by stating that his appointment would be “based on the quality of the individual irregardless [sic] of racial background.” He appointed to the recently vacated seat a white realtor from Northeast Yonkers.

In the same year, the mayor replaced two Board members who had been movers behind the Task Force and were generally regarded as being among the Board’s strongest advocates of school desegregation in Yonkers. Both members had expressed their interest in continuing to serve on the Board, and the reappointment of one or both was supported by the Council of PTAs, the Yonkers NAACP, the new superintendent Robitaille, and the Clergy of Yonkers. Martinelli appointed instead John Romano, a candidate supported by the Congress of Italian-American Organizations, and Joseph Spencer, a supporter of the mayor in his previous election campaigns. Once on the Board, Romano and Spencer promptly voted against even applying for state funding for the Task Force; Romano opined that state funding was a “waste[]” because Yonkers has no “racial problem.... unless the state hands down a ruling stating there is a problem.”

By the time of the 1977 Phase II proposal, Martinelli was routinely quizzing prospective Board members about their views on busing; he admitted at trial that these views “probably weighed very heavily with [him]” in deciding whether or not to appoint. In 1977 and 1978, Martinelli appointed four persons, all of whom were opposed to the Phase II Plan. They included Quentin Hicks, a black opposed to busing, whose appointment was immediately protested by members of the black community on the ground that his views did not represent theirs; the appointment was later acknowledged by the mayor to have been an embarrassment to the black community.

By May 1978, the Board was composed solely of Martinelli’s appointees. In that month, the Board held a special workshop at which Board members unanimously expressed their opposition to the desegregation proposals of Phase II. As indicated in part AII.D.2. above, the Board neither accepted any desegregative aspect of these proposals nor took any other steps, including those it avowedly preferred, toward desegregating the Yonkers public schools.

In 1979, Martinelli lost his bid for reelection. In his valedictory State-of-the-City address, he began his description of his administration’s achievements in education by stating that “[discussion of neighborhood stability would not be complete without attention to our public school system.” After mentioning three factors that he predicted would ensure sound and healthy schools, he stated, “[m]ost importantly, we now have a Board of Education fully committed to neighborhood schools which is of critical importance to neighborhood stability in this city!”

F. The District Court’s Findings as to School Segregation

The district court found that the Yonkers public school system as a whole was in fact racially segregated, with few of the public schools in Yonkers fairly reflecting the racial balance of the City’s overall student population. Using the term “minority” to include both blacks and hispanics, the court found that most schools in the district were either identifiably white or identifiably minority. Most of the schools in Southwest *1212 Yonkers had student populations that were predominantly minority, and the community and the Board’s administrative personnel generally associated Southwest Yonkers with minority schools. 624 F.Supp. at 1384-87 .

School authorities acknowledged that the quality of the education available at the identifiably minority schools was inferior to that available at the identifiably white schools, due in part to the inferior physical facilities and the concentration of less experienced teaching staffs at the former. Id. at 1630. The court found that the identifia-bility of certain Southwest Yonkers schools as minority schools had become inseparable from the perception of those schools as educationally inferior, and that “[t]his confluence of racial identifiability and relative educational opportunity has served to reinforce the segregative demographic patterns which have evolved in the City.” Id. at 1444.

The court found that the segregation of the schools was attributable to the conduct of both the Board and the City and that each defendant had acted with the intent to perpetuate or enhance school segregation.

1. The Board’s Liability

The court found that the Board was well aware of the City’s practice of confining subsidized low-income housing to Southwest Yonkers and indeed had urged the City to select scattered sites for such housing. It found that the Board’s adherence to a neighborhood-school policy in the face of the City’s known segregative practice suggested an intent on the part of the Board to preserve a similar segregation in the schools. Id. at 1535-37. It found confirmation of segregative intent in many of the Board’s affirmative acts.

The Board’s disproportionate assignment of minority teachers and staff members to the predominantly minority schools served to enhance the racial identifiability of those schools as minority schools; the enhanced identifiability had the effect of perpetuating and increasing the predominance of minorities in the student populations of those schools. Id. at 1527-28. The staff assignments were not explainable by reference to rationales offered by the Board to explain its assignments of students, for neither the neighborhood-school concept nor concerns for transportation played a role in staff assignments. Id. at 1467. Nor was the court persuaded by the Board’s reliance on its agreement with the teachers' union as an explanation for the staffing pattern, first because the racial skewing of the staff assignments predated that agreement, and second because the agreement gave the Board a certain amount of retransfer power that the Board never attempted to use. Id. at 1463-67. The court found that

[t]he foreseeability of the increased racial segregation of staff members and the district’s limited efforts to alleviate the imbalance together suggest that the resulting assignment of minority staff to minority schools was a practice which the Board approved of and intended to con-tinue_ Given the school district’s deliberately segregative pattern of administrative staff assignments and the racial disproportionality in teacher assignments prior to the collective bargaining agreement, it is reasonable to infer that the subsequent pattern of assigning minority teachers to disproportionately minority schools was considered desirable and was deliberately unaltered.

Id. at 1464-65.

The court also found that the Board’s special education program, which resulted in the placement of a disproportionate number of minority children in classes for the emotionally disturbed, was “operated in an unlawfully discriminatory manner.” Id. at 1461. The evaluative process was particularly prone to unwarranted racial assumptions and was unusually discriminatory in its impact. No race-neutral factor was likely to explain the disproportionately high numbers of minority children in such classes, id. at 1454, and the discriminatory treatment and the consequent stigmatization of the children so placed was not educationally justifiable, id. at 1461. In addition, the assignment of these disproportionately minority-populated special classes to *1213 schools that were predominantly white, and the isolation of and refusal to mainstream the special class students increased the stigmatization. Id. at 1455. Minority students enrolled in regular school programs have had difficulty in gaining acceptance among their white schoolmates as a result of the Board’s placement of disproportionately minority special education classes in the school. Id. at 1456. Even without reference to the special education program, the court noted that a Board study revealed significantly more racial prejudice among students attending schools that were disproportionately black or disproportionately white than among students attending schools that were racially balanced. Id. at 1444.

The court also found that many of the Board’s actions and inactions with regard to school openings, closings, and attendance zone changes evinced a segregative intent. It found, for example, that the racial imbalance between School 16 (90% white) and the nearby School 25 (88% minority) had been caused in part by the Board’s deliberately segregative conduct in repeatedly redrawing the attendance zone boundary between the two schools. It found that the Board’s proffer of a race-neutral basis for the rezoning was pretex-tual. Id. at 1526-27.

Though the court was unpersuaded that the isolated act of closing School 1 in 1954 —by then 99% minority — evinced a segre-gative intent, it found that the Board’s earlier changes in the attendance zone of School 1, whose student population had theretofore been as much as two-thirds white, had “constituted deliberate, racially motivated gerrymandering” for which there was no evidence of any race-neutral justification. Id. at 1411.

The court found that the Board’s refusal to close or desegregate Longfellow, the underutilized, inferior middle school with a heavy minority population, was “difficult to explain in race-neutral terms,” id. at 1426, and found the Board’s proffered explanations fiscally unsound, inconsistent with other Board actions, and pretextual. It found that by the late 1970’s, racial considerations played an increasing role in the Board’s refusal to close the school. Id. at 1426-28. It also found that “racial factors played a significant role in the Board’s segregative opening of Commerce Middle School.” Id. at 1482; see also id. at 1472-79.

The court found that the Board’s rejection of the NYU Report’s recommendation of a “variable access” vocational program was designed to be responsive to racial concerns. The community opposition, which argued that any east-west pairing of schools would result in a decline of the quality of education offered at the schools in East Yonkers, took on a pretextual hue in the context of vocational courses. Though test scores indicated a disparity between whites and minorities in achievement levels in academic courses such as English and mathematics, no such disparity was indicated with regard to vocational courses such as auto mechanics. The court found that the Board recognized that community opposition to the pairing-and-sharing proposal stemmed from racial concerns and that the Board’s selection of the more expensive alternative of duplicating the vocational courses in each of the four academic high schools reflected a desire not to take steps that would be desegregative. Id. at 1476-78.

The court found that the Board’s persistent rejection of other desegregative proposals, including those recommended in Phase II and all proposed alternatives that would have had any desegregative effect, was similarly the result of the Board’s responsiveness to race-based community resistance to school desegregation. Id. at 1497. The court found it significant that the Board did not always yield to public pressures, most notably in connection with its decisions as to what schools to close in connection with the City’s fiscal crisis. Thus, when the Board proposed to close Schools 4 (98% white) and 15 (100% white), there was massive protest from the affected communities, from councilmen, and from the mayor. These protests were not construed by the Board as principally race-based, and the Board held firm and closed the schools. Id. at 1416-17. Whenever a *1214 proposed change was for purposes of desegregation, however, and the pressure was perceived as racially motivated, the Board acquiesced. Id. at 1493-94.

The court found several indications that much of the community opposition to busing was race-related and that its phrasing in race-neutral terms was p’retextual. For example, East Yonkers parents’ emphasis on allowing their own children to attend schools in their neighborhoods and on not usurping after-school recreational time by requiring busing, could not explain their opposition to having Southwest Yonkers children attend schools in East Yonkers. Moreover, the allegedly race-neutral objections would, in many instances, have been equally applicable to the objectors’ proposed alternatives such as the formation of magnet schools and open enrollment. The sincerity of their advocacy of magnet schools was further belied by their earlier vehement opposition to the NYU Report’s pairing-and-sharing proposal, which would have effected a limited magnet-school program. All of these factors persuaded the district court that the stated preferences of both the community and the Board for such busing alternatives as magnet schools were pretexts designed to obscure the race-based nature of their opposition to desegregative changes. Id. at 1489-90. The court’s inference that the Board’s own stated preference for such alternatives was pretextual was also drawn from the Board’s failure, for more than three years following its rejection of Phase II’s desegregative aspects, to take any action whatever to implement any of its allegedly preferred de-segregative alternatives. Id. at 1493-95.

In sum, the district court found that the Board’s

refusal to implement such proposals in the late 1970’s occurred in [a] temporal and factual context which renders a finding of deliberate perpetuation of racial segregation appropriate: the increased racial imbalance among the district’s schools; the increasingly visible racial opposition to correcting this condition; the increased demands for desegre-gative action; an increasing realization that such action was an important ingredient in eliminating disparities in educational opportunities in the district; a community increasingly afflicted by seg-regative governmental housing practices animated by community opposition to the presence of subsidized housing in areas outside of Southwest Yonkers; and the failure to address the problem of racial imbalance in the schools in any meaningful fashion in the years following the rejection of Phase II in a manner consistent with the Board’s stated reasons for rejecting the plan. In our view, the record makes clear that the initial reluctance to implement desegregative school reorganization plans evolved into a persistent failure to adopt measures to correct recognized educational and racial imbalances in the district in part because of their desegregative consequences. From the foregoing, we find the Board’s failure to meaningfully address the problem of racial imbalance subsequent to its consideration of Phase II is more readily explainable as a reflection of the community’s resistance to desegregation rather than the race-neutral concerns of the community.

624 F.Supp. at 1497 . The court concluded that the conduct of the Board violated the rights of minority school children under Titles IV and VI and the Equal Protection Clause.

2. The Liability of the City

The court found that the City’s segre-gative housing practices had been a contributing cause of the racial segregation of the schools. It found that the failure of the Board to take action to minimize the school segregation “in no way negates the fact that, as a factual matter, the City’s housing practices contributed to the perpetuation and aggravation of residential segregation and the resulting segregation of the schools.” Id. at 1501.

The court found that the segregative impact of the City’s segregative housing practices on the schools was not unavoidable, unknowing, or inadvertent. It noted that the link between the racial identifiability of a school and the residential segregation of the surrounding neighborhood was recog *1215 nized by City officials, id. at 1443, and found that “in light of the school district’s historic neighborhood school policy, the perpetuation and exacerbation of racial imbalance in the school district was a natural, probable and actually foreseen consequence of the City’s discriminatory housing practices_” Id. at 1542. Indeed, in the racially motivated community opposition to the construction of low-income housing in non-minority areas, there was frequent mention of the effect of such housing on schools, and express objection by white parents to having their children schooled with minorities.

Further, the court found that the pattern of appointments by Mayor Martinelli of Board members, screened for their opposition to “busing,” was an exercise of “power over school board appointments as a means of furthering the city’s segregative objectives.” Id. at 1534. Though the City was not initially responsible for the Board’s neighborhood-school policy, it opposed construction of housing for minorities outside of Southwest Yonkers, and it advocated keeping all children assigned to schools in the neighborhoods in which they lived. Martinelli consistently appointed Board members who shared this view and who steadfastly refused to take any action that would have had any desegregative effect on the schools. Thus, the court found that “the City not only was aware of the overall impact of its subsidized housing practices on Yonkers public schools but also intended to preserve the racially segregative impact of these practices on the schools.” Id. at 1501.

In all, the court found that the City’s segregative housing practices and the may- or’s appointments contributed significantly both to the confinement of minority students to schools in Southwest Yonkers and to the Board’s failure to undo the segre-gative effects of these and other practices on the schools. “And in a city where the segregated condition of ‘neighborhood schools’ is in part the product of official municipal design, the commitment to the neighborhood school system by the head of that same municipality can hardly be considered race-neutral.” Id. at 1513.

The court concluded that the conduct of the City in intentionally perpetuating segregation in the schools violated the rights of minority schoolchildren under Title IV and the Equal Protection Clause.

G. The School Remedy

After receiving remedy proposals from the parties and conducting an evidentiary hearing, the court issued its school remedy order, reported at 635 F.Supp. 1538 (1986). As an overall goal, the order provided that the Board “shall seek to achieve” desegregation throughout the Yonkers public school system by the 1987-88 school year. To this end, the court ordered the creation of a system of magnet schools that students could choose to attend voluntarily. It defined a “desegregated school” as (a) a magnet school whose minority enrollment was within 15 percentage points of the system-wide proportion of minority students for the first year of that school’s operation and within 10 percentage points thereafter, or (b) a nonmagnet school whose minority population was within 20 percentage points of the system-wide proportion.

The court prescribed the methods to be used in administering the magnet school system “[i]n order to maximize the extent to which the integrative goals of this order will be reached through voluntary student assignments.” Id. at 1544 . They included an intense publicity and recruitment phase, id., and a system whereby parents must submit for each child a list of three school choices, at least one of which must “further the goals of desegregation,” id. at 1545 . The court also established the admissions criteria to be used in the magnet schools, ordered the Board to “make every effort” to achieve a specified racial composition of teachers at each school, explained the guidelines to be followed in the special education program, and ordered that the Board provide transportation for specified students. Id. at 1545-50 .

The court ordered the City to provide the necessary funding for implementation of the ordered desegregation program. It ap *1216 pointed a monitor to oversee compliance with its orders and retained jurisdiction of the action in order to enforce compliance. Id. at 1551-53 .

The court overruled a belated objection by the City that the plan ordered by the court was too expensive. The court noted that the City had made no such objection at the hearing when the desegregation plan budget was presented, had not contended that any part of the proposal was not required for desegregation, and was unable, despite being given an additional opportunity to do so, to show that any part of the desegregation plan budget either was not necessary or was duplicative of the regular budget.

A stay motion was denied, and the desegregation program was commenced in the 1986-87 school year.

B. LIABILITY

In these appeals, the City mounts several challenges to the district court’s ruling that it is liable for segregation in housing. Principally it contends that the court erred (1) in ruling, in effect, that it had an obligation to build subsidized housing outside of Southwest Yonkers; (2) in finding that the City’s housing decisions were made with the intention and the effect of perpetuating housing segregation; and (3) in holding the City liable for making decisions that merely responded to the wishes of its citizens.

The City challenges the ruling that it is liable for segregation in the schools, contending principally (1) that the segregation was caused not by City actions but rather by Board policies for which the City may not be held liable; (2) that the record reflects at most the foreseeability that City actions would perpetuate and enhance school segregation, but not any intent on the part of the City to achieve those effects; and (3) that the court could not properly take into account, in assessing City responsibility for school segregation, the mayor’s pattern of appointing to the Board individuals who espoused the maintenance of segregation in the schools.

The Board challenges the district court’s ruling that it is liable for school segregation on the principal grounds that (1) there was insufficient evidence of its intention to discriminate, and (2) the court could not properly take into account the intentionally segregative conduct of the City in determining whether the Board should be held liable. The Board also contends that “minority” should have been defined to include only blacks, not hispanics, and that with that redefinition, the schools could not be found to be in fact segregated.

As discussed in Part C. below, both the City and the Board contend that various aspects of the district court’s remedial orders go beyond the proper bounds of discretion.

We have considered all of the arguments made by the City and the Board on these appeals and find all of them to be without merit. Only those mentioned above warrant discussion.

I. General Principles

A. Substantive Law and the Requirement of Intent

It is by now well established that in order to prove a claim of discrimination in violation of the Equal Protection Clause a plaintiff must show not only that the state action complained of had a disproportionate or discriminatory impact but also that the defendant acted with the intent to discriminate. This principle governs claims of discrimination in housing, see Village of Arlington Heights v. Metropolitan Housing Development Corp., 429 U.S. 252, 265 , 97 S.Ct. 555 -63, 50 L.Ed.2d 450 (1977) (“Arlington Heights I”), and claims of segregation in the schools, see Keyes v. School District No. 1, 413 U.S. 189, 198 , 93 S.Ct. 2686, 2692 , 37 L.Ed.2d 548 (1973) (“Keyes"). See generally Washington v. Davis, 426 U.S. 229, 239 , 96 S.Ct. 2040, 2047 , 48 L.Ed. 2d 597 (1976).

In order to prevail on an equal protection claim of racial discrimination, the plaintiff need not show that the decision-maker was motivated solely, primarily, or *1217 even predominantly by concerns that were racial:

[Washington v.] Davis does not require a plaintiff to prove that the challenged action rested solely on racially discriminatory purposes. Rarely can it be said that a legislature or administrative body operating under a broad mandate made a decision motivated solely by a single concern, or even that a particular purpose was the “dominant” or “primary” one.

Arlington Heights I, 429 U.S. at 265 , 97 S.Ct. at 563 (footnote omitted). Rather, the plaintiff need begin only by showing that race was “a motivating factor.” Id. at 266 , 97 S.Ct. at 564 (emphasis added). Once it is shown that a decision was motivated at least in part by a racially discriminatory purpose, the burden shifts to the defendant to show that the same result would have been reached even without consideration of race. Id. at 270 n. 21, 97 S.Ct. at 566 n. 21; Mt. Healthy City School District Board of Education v. Doyle, 429 U.S. 274, 287 , 97 S.Ct. 568, 576 , 50 L.Ed.2d 471 (1977). If the defendant comes forward with no such proof or if the trier of fact is unpersuaded that race did not contribute to the outcome of the decision, the equal protection claim is established.

As to a claim under the Fair Housing Act, in contrast, the consensus is that a plaintiff need prove only discriminatory effect, and need not show that the decision complained of was made with discriminatory intent. See, e.g., Robinson v. 12 Lofts Realty, Inc., 610 F.2d 1032, 1036-38 (2d Cir.1979) (collecting cases):

“To establish a prima facie case of racial discrimination, the plaintiff need prove no more than that the conduct of the defendant actually or predictably results in racial discrimination; in other words, that it has a discriminatory effect.... The plaintiff need make no showing whatsoever that the action resulting in racial discrimination in housing was racially motivated_ Effect, and not motivation, is the touchstone, in part because clever men may easily conceal their motivations....”

Id. at 1037 (quoting 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)). See Arlington Heights I, 429 U.S. 252, 270-71 , 97 S.Ct. 555, 556-67 (ruling that equal protection claim should have been dismissed for lack of proof of discriminatory intent, but remanding for further consideration of Fair Housing Act claim); Metropolitan Housing Development Corp. v. Village of Arlington Heights, 558 F.2d 1283, 1287-90 (7th Cir.1977) (“Arlington Heights II”) (holding that Fair Housing Act claim could be established by proof of discriminatory effect, without proof of discriminatory intent), ce rt. denied, 434 U.S. 1025 , 98 S.Ct. 752 , 54 L.Ed.2d 772 (1978).

The effect in the present case of the differing standards of proof for the constitutional claim and the Fair Housing Act claim is, as a practical matter, immaterial, for the district court found that even if proof of discriminatory intent were required for establishment of a claim under the statute, the requisite intent was proven. See 624 F.Supp. at 1293 n. 12 (“In light of the strength of the evidence of intent in the case before us, it is unnecessary to reach the question whether entitlement to the broad remedial measures sought by plaintiffs here could be established under the relaxed ‘effects’ standard set forth in Arlington Heights II.”). Since we find no basis for overturning the district court’s findings of intent (see Parts 5.H.A.2., H.II.B., and fi.III.B. below), and since the conduct upon which the housing discrimination claim is based spans a period that commenced many years before the 1968 effective date of the Fair Housing Act, we review the statutory claims, along with the constitutional claims, in light of the proof of , the City’s segregative intent.

B. Standard of Review

The standard to be followed by the appellate court in reviewing findings of fact made by the district court is also clear. Fed.R.Civ.P. 52(a) states that those findings, “whether based on oral or documentary evidence, shall not be set aside unless clearly erroneous.” Assessments of the credibility of the witnesses are peculiarly *1218 within the province of the district court as trier of fact and are entitled to considerable deference. Id. Thus, “when a trial judge’s finding is based on his decision to credit the testimony of one of two or more witnesses, each of whom has told a coherent and facially plausible story that is not contradicted by extrinsic evidence, that finding, if not internally inconsistent, can virtually never be clear error.” Anderson v. City of Bessemer City, 470 U.S. 564, 575 , 105 S.Ct. 1504, 1513 , 84 L.Ed.2d 518 (1985). Even when the district court’s findings of fact do not rest on credibility determinations but instead are based on documentary evidence or on inferences from other facts, the appellate court must accept those findings if they adopt a permissible view of the evidence; the appellate court may not conduct a de novo review. In short, “[wjhere there are two permissible views of the evidence, the factfinder’s choice between them cannot be clearly erroneous.” Id. at 574 , 105 S.Ct. at 1512 ; see United States v. Yellow Cab Co., 338 U.S. 338, 342 , 70 S.Ct. 177 -179, 94 L.Ed. 150 (1949).

This standard of review governs questions of fact, though not questions of law or mixed questions of fact and law. A finding of discriminatory intent is a finding of fact, Pullman-Standard v. Swint, 456 U.S. 273, 287-90 , 102 S.Ct. 1781, 1789-91 , 72 L.Ed.2d 66 (1982), as are findings of discrimination, Anderson v. City of Bessemer City, 470 U.S. at 573 , 105 S.Ct. at 1511 , and causation, e.g., Wellner v. Minnesota State Junior College Board, 487 F.2d 153, 156 (8th Cir.1973).

In accordance with the above principles, if the district court’s findings as to the existence of segregation in fact, the existence of segregative intent, and the existence of a causal relationship between the two are permissible inferences from the evidence of record, we may not overturn them.

II. The City’s Liability for Segregation in Housing

A. The City’s Obligation With Respect to Subsidized Housing

The Fair Housing Act makes it unlawful

(a) To refuse to sell or rent after the making of a bona fide offer, or to refuse to negotiate for the sale or rental of, or otherwise make unavailable or deny, a dwelling to any person because of race, color, religion, sex, or national origin.

42 U.S.C. § 3604 (a) (emphasis added). The statute defines “dwelling” as “any building ... intended for occupancy as[ ] a residence by one or more families, and any vacant land which is offered for sale or lease for the construction or location thereon of any such building_” Id. § 3602(b). The City’s contention is that neither the Fair Housing Act nor the Equal Protection Clause imposes on it any obligation to construct housing, that it has constructed housing that it has made available to all persons regardless of race, and that the law requires no more of it. In the circumstances of the present case, we disagree.

Though we know of no statutory or constitutional provision that imposes on a municipality a general obligation to construct subsidized housing, see Acevedo v. Nassau County, 500 F.2d 1078, 1081-82 (2d Cir.1974), more focused principles govern the present case. In Acevedo , the thrust of the complaint was that the defendant county had initially planned to build both senior citizen and family housing and that its abandonment of the plan to build family housing had a disproportionate impact on minorities. See id. at 1081 . The district court, after a trial, found that the abandonment had neither discriminatory effect nor a discriminatory motive. See id. at 1079-80 . Accordingly, we held that the abandonment violated neither the Constitution nor the Fair Housing Act. See id. at 1082 . This does not mean that we would have reached the same conclusion in the face of findings that there had been discriminatory impact and discriminatory intent, for the absence of a general obligation to construct does not give the municipality license to proceed discriminatorily once it has started down the road to construction. Thus, the Sixth Circuit, for example, has upheld a “pattern and practice” claim under the Fair Housing Act, see 42 U.S.C. § 3613 (a), *1219 where the defendant city had applied for federal funds that it in fact wanted and needed, but had abandoned its application for reasons found to be racially discriminatory. United States v. City of Parma, 661 F.2d 562, 575 (6th Cir.1981), cert. denied, 456 U.S. 926 , 102 S.Ct. 1972 , 72 L.Ed.2d 441 (1982).

Nor, once a municipality has decided to construct housing, may it lawfully proceed with segregative intent and effect to confine housing for minority occupancy to areas in which minority residence is already concentrated, thereby enhancing and perpetuating racial segregation in residential patterns. In Otero v. New York City Housing Authority, 484 F.2d 1122 (2d Cir.1973), we noted obiter that “Congress’ desire in providing fair housing throughout the United States was to stem the spread of urban ghettos and to promote open, integrated housing,” id. at 1134 , and that, accordingly, “[a]n authority may not ... select sites for projects which will be occupied by non-whites only in areas already heavily concentrated with a high proportion of non-whites,” id. at 1133 . The Third Circuit reached a similar conclusion in Shannon v. United States Department of Housing and Urban Development, 436 F.2d 809 (3d Cir.1970), which involved a challenge to HUD’s approval of a rent-subsidy contract for a new building in an urban renewal area of Philadelphia. The thrust of the complaint was that the location of a rent-subsidy project in that area would have the effect of increasing the already high concentration of low-income black residents there, and that HUD had not properly considered the effect of such a subsidy guarantee on the racial concentration in Philadelphia as a whole or in that neighborhood in particular. Id. at 811-12 . The court of appeals agreed that HUD had not considered those effects, and it vacated the district court’s denial of relief, stating that “[ijncrease or maintenance of racial concentration is prima facie ... at variance with” the policy underlying the Fair Housing Act. Id. at 821 . Consistent with these views, when we held in Acevedo that there was no constitutional violation in^the defendant’s decision, made with no discriminatory intent, not to construct housing, we took care to distinguish cases in which municipalities had intentionally pursued their construction plans in a segregative manner, “effectively restricting] low income housing projects to segregated neighborhoods.” See 500 F.2d at 1081 n. 3.

Accordingly, the district court properly rejected the City’s contention that its decisions not to construct minority housing in any virtually all-white area were immune from scrutiny, and appropriately proceeded to determine whether housing in Yonkers was in fact segregated, whether that segregation was caused or enhanced in substantial part by the City’s conduct, and whether that conduct was intentionally segregative.

1. Segregative Effect of the City’s Actions

The district court found that by 1980 an “extreme condition of segregation ... exist[ed] in Yonkers.” 624 F.Supp. at 1364 . The evidence amply supports this finding.

The 1980 Census figures showed that 81% of Yonkers’s minority residents lived in one quadrant of the City. Minorities constituted 19% of Yonkers’s total population; yet the minority population of Southwest Yonkers exceeded 40%. Of the 10 census tracts within Southwest Yonkers itself, five had minority populations exceeding 50%. In contrast, outside of Southwest Yonkers, only 6% of the residents were minorities; and these minority residents were largely confined to two areas, one having a minority population of 29% and the other having a minority population of 80%. In light of these facts, we have no difficulty in upholding the district court’s finding that housing in Yonkers was segregated.

Nor do we see a basis for upsetting the finding that the City’s decisions to locate low-income housing only in or adjacent to areas already having high concentrations of minority residents was a contributing cause of the extreme condition of residential segregation that existed by 1980. From 1948 to 1980, some 144 sites were *1220 formally proposed to the City for subsidized housing, most of them in East or Northwest Yonkers or predominantly white neighborhoods of Southwest. More than 100 other, sites, the vast majority of them in East or Northwest Yonkers, were also given official consideration. In all, 23 family housing sites were approved; of these, 21 were in Southwest Yonkers; one was in Northwest, abutting a heavily minority neighborhood of Southwest; and one — the only family project approved for an area that was neither within nor abutting Southwest Yonkers — was in the predominantly black Runyon Heights. See Appendix B. Of the 21 family housing sites approved for Southwest, 18 were in or adjacent to neighborhoods already having high minority concentrations, one was a half-block away from such a concentration, and the other two were but five blocks away.

Only one subsidized housing project was approved for a nonminority area outside of Southwest Yonkers: it was not a family project but a senior citizen project which, as expected, was occupied predominantly by white persons. Thus, over a period of more than three decades, the City approved no housing for minorities in any area that was not in or close to an already heavily minority area.

The demographic effect of concentrating minority-intended housing in the already concentrated minority areas was predictable. From 1960 to 1970, while the minority population of East and Northwest Yonkers increased by 1,879 persons, or 61%, the minority population of Southwest Yonkers increased by 10,333, or 5.5 times as many, persons. In percentage terms, the minority population of Southwest increased by 186%, from a starting base that was nearly twice as large as that in East and Northwest Yonkers combined. From 1970 to 1980, when the minority population of East and Northwest Yonkers increased by 43%, the minority population in Southwest Yonkers increased by 87%; in raw numbers, the net increase of minority residents in Southwest Yonkers outpaced the minority increase in other parts of Yonkers by 13,783 to 2,119. In all, during the period 1960 to 1980, when virtually all of the low-income minority housing at issue here became available for occupancy, all of it confined to areas that already had high minority concentrations, the minority population of Southwest Yonkers increased by 24,116 persons, or 434%, while elsewhere in Yonkers the minority population grew by only 3,998 persons, or 130%.

There was expert testimony that by concentrating subsidized low-income housing in the minority areas of Yonkers, the City had “stigmatized” those neighborhoods and thereby made them both less likely to attract new white families and less likely to retain the white families already there. This is consistent with evidence of denigrating comments made by white residents of other parts of Yonkers about the Southwest Yonkers neighborhoods, and with the demographic statistics. As the minority population in Southwest Yonkers increased from 5,559 in 1960 to 29,675 in 1980, the white population in Southwest declined steeply from 75,952 in 1960, to 66,523 in 1970, and to 41,124 in 1980, a net decrease of nearly 35,000 white residents. Elsewhere in Yonkers, the number of white residents increased from 1960 to 1980 by some 6,000 persons.

Other City acts also served to confine minority residents to predominantly minority areas. For example, in the period 1968 to 1974, when CD A sought out private developers, it focused efforts solely on sites in Southwest Yonkers. In 1975, when a private developer, who had planned a housing project on an East Yonkers site described by City planners as “well suited for Housing for the Elderly,” revealed that he hoped to rent 20% of the space to minorities, the City prevented the project. In the late 1970’s, the Council obstructed the potential movement of minority families to existing buildings in East Yonkers by curtailing the use of Section 8 Certificates by families and by steering minority families to buildings in Southwest.

From all the evidence, the court could reasonably infer that the City’s actions accelerated and enhanced the process of concentrating minority housing in Southwest Yonkers,

*1221 2. Segregative Intent

Intent to discriminate may be established in a number of ways. Often it may be “inferred from the totality of the relevant facts, including the fact, if it is true, that the law bears more heavily on one race than another.” Washington v. Davis, 426 U.S. at 242 , 96 S.Ct. at 2048 . Such impact may be an important starting point. Other probative sources may include the “historical background of the decision ..., particularly if it reveals a series of official actions taken for invidious purposes”; “[t]he specific sequence of events leading up to the challenged decision,” such as zoning changes for a given site enacted upon the decisionmaker’s learning of plans for the construction there of integrated housing; “contemporary statements by members of the decisionmaking body, minutes of its meetings, or reports”; “[departures from the normal procedural sequence”; and “[s]ubstantive departures ..., particularly if the factors usually considered important by the decisionmaker strongly favor a decision contrary to the one reached.” Arlington Heights I, 429 U.S. at 267-68, 97 S.Ct. at 564-65. In the present case, a wealth of evidence in each of these categories supports the district court’s finding that the City intended its housing decisions to result in the confinement of minorities to existing minority areas.

The impact of the City’s decisions has been discussed in the preceding section. The historical background of these decisions included the City’s 1930’s decision to build a housing project especially for blacks, its rejection of a number of sites “on the ground that the level of minority concentration there was not sufficiently high,” and its eventual selection in 1940 of a site “in one of the most heavily minority areas of Southwest Yonkers.” 624 F.Supp. at 1312 .

Many sequential clusters support the proposition that the City’s decisions in the ensuing decades were similarly purposely segregative. The most commonly recurring sequence consisted of a site proposal for a white area, followed by vehement community opposition (e.g., letters on behalf of 2,000 residents; 1,000 attending public meetings), followed by City Council disapproval of the site. There was virtually never a site proposal for low-income family housing in a white area that was not met with opposition; there was virtually never white-community opposition that was not followed by withdrawal or rejection of the site.

There was ample evidence that much of the white-area residents’ opposition to low-income housing was race-related. Both the Council and the community equated low-income family housing with minorities and senior citizen housing with whites. Thus, a group of white Catholics urged that the housing project proposed for their area be changed from family to senior citizen, stating that they feared an influx of blacks. Virtually all councilmen from East Yonkers stated that they were concerned about community opposition to low-income housing in their areas; many acknowledged explicitly that that opposition was race related. Ian-nacone testified that his own facially race-neutral public opposition to a proposed low-income housing project had been pretextual, masking his response to his constituents’ racial concerns; some of those constituents had stated, pretextually, that they feared of loss of a parking lot, but others “who knew him better” told him “they didn’t want the housing because they didn’t want any blacks there.” Speakers at meetings, officials at trial, and contemporary news articles reflected the view that many of Yonkers’s white residents were opposed to “having to absorb the overflow from Puerto Rico or Harlem,” and were “not ready” to accept racial integration. Officials describing public meetings said racial motivations were “definitely a consideration” and were “thick in the air.” Councilmen discussing Section 8 Certificates and forbidding MHA to obtain such certificates for family housing exhibited their “con-cerní ] about the possibility that members of the minority community would, in fact, seek and probably find units on the east side of the city.”

The inference that the City intended to preserve racially segregated neighborhoods was also supported by evidence of its swift *1222 zoning obstructions of specific prospects for desegregative construction. For example, as to three sites submitted by CDA and tentatively approved by HUD in 1980, the Council rezoned one site for use as a shopping center; it refused to rezone another to a category consistent with a housing project; and as soon as the third was mentioned as a possibility for low-income housing, the Council rezoned it to remove it from the category appropriate for a housing project, in order “to give the community some peace of mind.” The Council indicated that it would rezone the site to the original category to permit luxury housing but not minority housing, stating, “we will change that zone when the concept fits the people, not before.”

The record also reflects numerous instances in which the City deviated from its normal procedural sequences or ignored the usual substantive standards in order to place low-income housing in Southwest Yonkers or to prevent its construction in East Yonkers. For example, in the 1950’s the City constructed 415 units of low-income housing on a minority-area site though the Planning Board recommended a limit of 250 units; the City rejected every site recommendation from the Planning Board, even those the planning experts rated as superlatively suitable, if the site was in a virtually all-white neighborhood. In the 1970’s, when Planning Board opposition to further low-income housing construction in Southwest Yonkers was known, the City simply began construction there without consulting that body. In the 1980’s, when the Council wished to have the School 4 site used for luxury housing rather than for low-income housing, it again bypassed the Planning Board and, in an unprecedented move, appointed a five-person advisory committee, four members of which had no planning or zoning experience; their major qualification appears to have been that they were white residents of the School 4 area.

Finally, the City’s intent to preserve the existing racial imbalance between Southwest and other areas of Yonkers was made clear by the words and actions of Mayor Martinelli. In 1971, HUD had warned the City that in order to retain federal funding, the City would have to build minority housing in nonminority neighborhoods. Thereafter, Martinelli won election on a campaign platform that included the promise of no more subsidized family housing in Yonkers. He was true to his word, and no more such housing was built. He further sought to ensure the preservation of the predominantly white neighborhoods by appointing school board members who would not approve busing, stating that “a Board of Education fully committed to neighborhood schools ... is of critical importance to neighborhood stability....”

Neither this summary nor our more detailed summary in Part A.I.A. recounts all of the evidence that supports the distric

This text is long and has been trimmed here. Open the source document for the complete record.