The Indianapolis School desegregation case is now entering its second decade.1 It involves a municipality, the (present) City of Indianapolis, which contains within its borders more than a half dozen separate and autonomous school districts. Eight years ago the district court determined that the fourteenth amendment violations committed by the largest of those districts (the Indianapolis Public Schools-IPS) and the State of Indiana could be remedied only by a desegregation plan which would transfer students from IPS to the predominately white school districts which surround it.2 After an appeal and a remand to the district court for further consideration,3 we affirmed, finding that the standard for interdistrict relief established by Milliken v. Bradley, 418 U.S. 717, 94 S. Ct. 3112, 41 L. Ed. 2d 1069 (1974) were met in this case, at least with regard to those districts within the (present) City of Indianapolis.4 The Supreme Court vacated and remanded for further consideration in light of Washington v. Davis, 426 U.S. 229, 96 S. Ct. 2040, 48 L. Ed. 2d 597 (1976) and Village of Arlington Heights v. Metropolitan Housing Development Corp., 429 U.S. 252, 97 S. Ct. 555, 50 L. Ed. 2d 450 (1977).5 On reconsideration we reaffirmed our holding that certain acts of the State of Indiana, particularly in connection with the "Uni-Gov" legislation which created the new city boundaries, had interdistrict effect, but remanded the case to the district court for findings of intent in light of the cases mentioned by the Supreme Court and evaluation of other issues, particularly regarding the location of public housing projects, in light of Milliken, Arlington Heights, and Dayton Board of Education v. Brinkman (I) (433 U.S. 406, 97 S. Ct. 2766, 53 L. Ed. 2d 851 (1977)).6

In his most recent opinions, which are the ones now before us, the district court judge found that the exclusion of schools from the Uni-Gov legislative scheme was done with a racially discriminatory purpose. He found a similar racially discriminatory purpose in the failure of the Housing Authority of the City of Indianapolis (HACI) to build any units outside of the old central city of Indianapolis despite legislative authority to do so. He also held that certain legislation7 enacted by the Indiana legislature in response to his earlier decisions in the case allowed him to implement an interdistrict remedy without regard to the legislative intent questions posed by the Supreme Court and by our decision. He rejected, however, attempts by the IPS to establish that intradistrict violations by IPS had a segregative impact on housing patterns and school enrollment patterns throughout the city. He also denied requests by the United States and IPS to allow immediate implementation of an intradistrict remedy. On the issue of remedy he reaffirmed his earlier determination that the appropriate remedy is a student reassignment plan which would transfer black students from IPS to the other districts within the city until each of the surrounding districts is approximately 15% black and then reassign the remaining IPS students within IPS to achieve complete desegregation within the IPS boundaries. He also ordered that certain in-service training programs be implemented, with the cost to be borne by the State of Indiana, and made permanent his injunction against any expansion of public housing (except for the elderly) within the boundaries of IPS.

Although virtually all of these decisions are challenged by one or more of the parties to these appeals, the central issue remains the propriety of a desegregation plan that extends to those parts of the City of Indianapolis which are in separate and independent school districts outside the boundaries of the Indianapolis Public Schools. We will address that issue in Part I of the opinion, discussing in turn the evidence regarding Uni-Gov, housing, and the interdistrict effects of segregation within IPS. In Part II we will turn to the other issues raised by the various appellants.

* As we noted in our last opinion (573 F.2d 400), an interdistrict remedy such as that ordered by the district court must be predicated on a finding that official action, taken with a discriminatory purpose, was a substantial cause of interdistrict segregation. Milliken v. Bradley, 418 U.S. 717, 744-45, 94 S. Ct. 3112, 41 L. Ed. 2d 1069 (1974); Washington v. Davis, 426 U.S. 229, 96 S. Ct. 2040, 48 L. Ed. 2d 597 (1976). The plaintiffs here pursued two independent avenues of proof in their attempt to make the requisite showings. The first centered around the exclusion of the Indianapolis Public Schools from the 1969 legislation which for most other purposes created a single metropolitan form of government for Marion County. The second concerned the governmental decision to locate all public housing projects within the old city of Indianapolis (and thus within the borders of IPS). In addition, IPS attempted to prove that de jure segregation of the IPS schools was in part responsible for the segregated housing patterns found throughout the county. The district court found that only the Uni-Gov and housing evidence supported an interdistrict remedy, rejecting as not credible the expert witnesses called by IPS in its attempt to establish the third point.

In our last opinion we held that the exclusion of schools from the Uni-Gov legislation passed in 1969

"meets the requirements of Milliken and therefore can be used as a basis for imposing an interdistrict remedy if the district court finds that the General Assembly, in enacting the series of legislation, acted with a discriminatory intent or purpose."

573 F.2d 400, 408.

The district court has now found that the actions of the General Assembly "were done, at least in part, with the racially discriminatory intent and purpose of confining black students in the IPS school system to the 1969 boundaries of that system, thereby perpetuating the segregated white schools in suburban Marion County." 456 F. Supp. 183. The appellants vigorously assert that there is no support in the record for that finding, disputing not so much the evidence itself, but rather whether taken as a whole, it supports the district court's conclusion.

In its remand for consideration of the discriminatory purpose question, the Supreme Court referred us to Washington v. Davis, 426 U.S. 229, 96 S. Ct. 2040, 48 L. Ed. 2d 597 (1967) and Village of Arlington Heights v. Metropolitan Housing Development Corp., 429 U.S. 252, 97 S. Ct. 555, 50 L. Ed. 2d 450 (1977). Washington v. Davis sets forth the intent requirement, but says little about how it is to be met, apart from noting that

"Necessarily, an invidious discriminatory purpose may often 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 on another. . . . Disproportionate impact is not irrelevant, but it is not the sole touchstone of an invidious racial discrimination forbidden by the Constitution."

426 U.S. 229, 242, 96 S. Ct. 2040, 2048-2049, 48 L. Ed. 2d 597.

Justice Stevens, concurring, notes that the degree and type of proof needed to establish intent may vary considerably from one sort of case to another, and that

"Frequently the most probative evidence of intent will be objective evidence of what actually happened rather than evidence describing the subjective state of mind of the actor. For normally the actor is presumed to have intended the natural consequences of his deeds. This is particularly true in the case of governmental action which is frequently the product of compromise, of collective decisionmaking, and of mixed motivation. It is unrealistic, on the one hand, to require the victim of alleged discrimination to uncover the actual subjective intent of the decisionmaker or, conversely, to invalidate otherwise legitimate action simply because an improper motive affected the deliberation of a participant in the decisional process."

426 U.S. 229, 253, 96 S. Ct. 2040, 2054, 48 L. Ed. 2d 597.

The Arlington Heights decision provides considerably more guidance. That decision calls for "a sensitive inquiry into such circumstantial and direct evidence of intent as may be available" and then suggests as possible lines of inquiry, in addition to discriminatory impact, "the historical background of the decision," "the specific sequence of events leading up to the challenged decision," "departures from the normal procedural sequence," "substantive departures . . . particularly if the factors usually considered important by the decisionmaker strongly favor a decision contrary to the one reached," and "legislative or administrative history." 429 U.S. 252, 266-268, 97 S. Ct. 555, 564-565, 50 L. Ed. 2d 450. With these guidelines in mind, we turn to the evidence on which the district court rested its decision.

Indianapolis and Marion County, before Uni-Gov, were not unlike dozens of other metropolitan areas throughout the country. The central city, Indianapolis, was losing population and becoming more predominately black and poor while the surrounding suburban areas were growing rapidly, but, with a few exceptions, remaining almost exclusively white. School enrollments were following a similar pattern. By 1970 IPS enrolled less than 60% of the county's total students, but over 97% of the black students in the county. These demographic facts of life, of course, can no more support a metropolitan desegregation plan here than could a similar pattern in Detroit (the focus of Milliken v. Bradley ) or any other major city. They form an important backdrop, however, for the historical events, which are of somewhat more significance to this case.

As the district court noted, Indiana has had a long history of both public and private discrimination against its black citizens. This history has been described at length in earlier opinions and will not be repeated here, but ranged from state legislation which affirmatively sanctioned the de jure segregation of the Indianapolis Public Schools until 1949 to numerous instances of private housing discrimination, some of which were still being openly practiced past the date suit was filed. Other official acts of discrimination included a prohibition on marriage across racial lines, not repealed until 1965; a requirement that only white males could serve in the militia, finally repealed in 1936; and a policy enforced until after World War II that blacks could enter state parks only on a segregated basis. The state was also implicated in the deliberate policies of segregation practiced by IPS, particularly with regard to its role in selection of sites for new schools.8 These historical facts are not conclusive, of course, but may properly be considered by the court in its attempt to determine whether there was a discriminatory intent in the legislature's failure to include schools in the Uni-Gov consolidation.

Of greater relevance, however, is the history of school district expansion and consolidation in Indiana. As we have said in prior decisions,9 Indiana has for generations pursued a legislative policy that school district lines should grow as the corporate lines of their cities grow. That policy was recognized by the Indiana courts as early as 186710 and was subsequently made explicit by legislation.11 The policy was modified in 1959 to allow greater consolidation among Indiana's many small school districts12 but was reaffirmed as to the Indianapolis district in particular by legislation enacted in 1961.13 Although the legislation did not always achieve its stated purpose the policy it expressed was clear and although there was seldom a perfect correspondence between civil city and school boundaries the exceptions were generally in favor of larger, rather than smaller, school districts.

It was against this backdrop that the Uni-Gov legislation was enacted in 1969. Civil, as well as school, annexations had become all but impossible to achieve because of opposition from the suburban areas which would be affected. Richard Lugar, then the Mayor of Indianapolis, along with other governmental and business leaders, was convinced that a county-wide governmental structure would be in the best interests of greater Indianapolis and conceived of Uni-Gov as a way of "sidestepping" the problems of direct annexation. The Uni-Gov legislation, officially entitled the "Consolidated First-Class Cities and Counties Act" included all of Marion County in a new governmental unit called the City of Indianapolis. All residents of the new city now vote in elections for mayor and members of the City-County Council. The new council has taken over most of the functions of county government (and of numerous special service districts) as well as the functions of the Common Council of the old City of Indianapolis. Some governmental units remain unchanged however. The airport authority, building authority, county courts, and hospital authority remain separately governed. There are special provisions for police and fire districts. In addition, the "excluded cities" of Beech Grove, Speedway, and Lawrence retain their own local governments and provide their own municipal services. Nonetheless the City-County Council has authority over building code enforcement, municipal planning, thoroughfare control and air pollution regulation even in the "excluded cities" and the residents of those cities vote in Uni-Gov elections.

Under the 1961 Act, if the Uni-Gov legislation could be characterized as an extension of Indianapolis' civil city boundaries by civil annexation, it would have followed that the boundaries of IPS would have expanded to cover the area included in the new City of Indianapolis. Section 1(e) of the 1961 Act defines "civil annexation" as "any action whereby the civil boundaries of any civil city are extended." The Uni-Gov Act in turn defined the new boundaries of the "City of Indianapolis" as "all of the territory of a First Class City and of the County, except for the territory located in Excluded Cities." Indiana Acts 1969, Ch. 173, § 102(f). Nevertheless the appellants argue vigorously that the Uni-Gov legislation was not such an annexation as defined by the 1961 Act and that therefore it is inappropriate to read any racially discriminatory intent into its limited scope. What convinces us that such a reading may be appropriate is the fact that the state legislature was sufficiently concerned about the possible expansion of IPS that might, under the 1961 Act, follow the enactment of the Uni-Gov legislation that it passed a special bill, just sixteen days before final passage of the Uni-Gov legislation, to repeal the applicable portions of the 1961 Act. As we have noted before, the repeal of section 9 of the 1961 Act appears to have been done in direct response to concerns that otherwise the boundaries of IPS would expand with the borders of the new City of Indianapolis. 573 F.2d 400, 407 (7th Cir. 1978). We do not think that the Indiana legislature thought it was engaged in a futile act or that the repeal was unrelated to the Uni-Gov legislation then pending. As Mayor Lugar testified, the Uni-Gov legislation would not have passed if it would have meant that city and suburban school districts would be consolidated. Given the starkly segregated conditions then existing in Marion County it is also clear, as it must have been then, that the likely result of that special legislation would be a continuation of the racial disparity then existing between IPS and other school districts in Marion County.

The appellants nonetheless assert that there were valid, non-racial reasons for the repeal of the 1961 Act. They point particularly to evidence that both the Marion County School Reorganization Committee in 1961, and IPS in 1967, had recommended that county-wide school consolidation not be pursued. Mayor Lugar testified that these recent rejections of a county-wide system were the reason that there was never any consideration of including schools in the Uni-Gov scheme, even though he personally thought, and argued, that it would be wise to have at least a county-wide common school tax. The opposition of IPS was apparently based largely on a determination that the district was better off financially with the students and tax base it already had than with the expanded tax base, but even more expanded long-term capital obligations, that would accompany consolidation. The reasons for the opposition of the Reorganization Committee are far less clear. The Committee originally supported a county-wide district and retreated from that position only when strong public opposition to consolidation was encountered. As the district court noted, however, the suburban districts throughout this period were engaged in a variety of cooperative ventures with each other, but never with IPS.

What this adds up to, we think, is the obvious conclusion that opinions about the merits of consolidation differed. On one hand there was the long-established, and recently reaffirmed, legislative determination that school boundaries, particularly in Marion County, should expand with civil city boundaries. On the other hand, there were those who felt that the students of Marion County were best served by smaller units of school government. The question is not whether one of these positions was racially invidious and the other not; the question is rather whether the legislative decision in 1969 to abandon its past policy in favor of the position advocated by the opponents of consolidation was done with a racially discriminatory purpose. The district judge considered the timing of the decision (shortly after this suit had been filed, signaling a likely end to the segregated conditions which had been prevailing within IPS), the history of state sanctioned discrimination against black students within IPS, the foreseeable impact of the decision on the minority population of Indianapolis, and the predominately political, rather than educational, reasons for the decision and concluded that it was made with a discriminatory purpose.14 We do not think any of those factors was improperly considered under Washington v. Davis or Arlington Heights, nor do we believe that the conclusion of discriminatory purpose was clearly erroneous in light of evidence of record. We therefore affirm the district court's determination that the 1969 repeal of the 1961 Act was done with a discriminatory purpose. We will discuss at a later point whether this finding supports the specific remedy ordered by the district court.

In his concurring decision in Milliken v. Bradley, Justice Stewart explained

"This is not to say, however, that an interdistrict remedy of the sort approved by the Court of Appeals would not be proper, or even necessary, in other factual situations. Were it to be shown, for example, that state officials had contributed to the separation of the races . . . by purposeful racially discriminatory use of state housing or zoning laws, then a decree calling for transfer of pupils across district lines . . . might well be appropriate.

418 U.S. 717, 755, 94 S. Ct. 3112, 3132, 41 L. Ed. 2d 1069.

With specific reference to the facts before the Court in Milliken he noted that "(n)o record has been made in this case showing that the racial composition of the Detroit school population or that residential patterns within Detroit and in the surrounding areas were in any significant measure caused by governmental activity . . . ." 418 U.S. 717, 756, 94 S. Ct. 3112, 3133 n.2, 41 L. Ed. 2d 1069. In our most recent remand opinion, we instructed the district court as follows:(A)n interdistrict desegregation remedy is appropriate if the following circumstances are shown to exist (given the fact that there is a vast racial disparity between IPS and the surrounding school districts within the "new" City of Indianapolis): (1) that discriminatory practices have caused segregative residential housing patterns and population shifts; (2) that state action, at whatever level, by either direct or indirect action, initiated, supported, or contributed to these practices and the resulting housing patterns and population shifts; and (3) that although the state action need not be the sole cause of these effects, it must have had a significant rather than a de minimis effect. Finally, an interdistrict remedy may be appropriate even though the state discriminatory housing practices have ceased if it is shown that prior discriminatory practices have a continuing segregative effect on housing patterns (and, in turn, on school attendance patterns) within the Indianapolis metropolitan area.

The record shows that the district court already has received evidence and has made certain findings in the area of housing discrimination. See Indianapolis I, 338 F. Supp. at 1204-05; Indianapolis IV, 419 F. Supp. at 183-85. It is important, however, that on remand the district court specify what state-responsible housing practices of a discriminatory nature, if any, have resulted, at least in part, in segregative residential patterns. This is necessary not only to determine initially whether an interdistrict remedy is appropriate, but also to fashion an appropriate remedy.

573 F.2d 400, 409-10.

In addition, we made it clear that the intent requirement applied here as well, i.e., that the "official actions" must have been taken with a discriminatory purpose.

On remand, the district court, with one exception, did not attempt to determine how much, if any, of the current housing segregation within the new City of Indianapolis was caused by intentional state action rather than by non-discriminatory actions or by private acts of discrimination. In the absence of such findings, the history of past state involvement in housing segregation, detailed at 332 F. Supp. 655, 662-3, could not support the interdistrict remedy ordered by the court.15 The question is narrowed therefore, to the one exception, which is the location of public housing within Marion County.

The Housing Authority of the City of Indianapolis (HACI) and the Metropolitan Development Commission of Marion County (Commission) have, since 1964, been responsible for the placement of public housing in much of Marion County. Under Indiana law in effect at the time, HACI's "area of operation" included all of the old City of Indianapolis as well as "the area within (5) miles of the territorial boundaries thereof." IC 1971 18-7-11-3. Of the eleven public housing projects in Marion County ten were built, mostly in the late 1960's, on sites selected by HACI and approved by the Commission.16 Each of the eleven projects is within IPS. Those which house families are approximately 98% black.

It was obvious to the district court that the choice of location of those projects had an impact on school enrollments by keeping black students within IPS and out of the surrounding school districts. In our remand order we held that this presumably segregative impact could support an interdistrict school desegregation order only if a) the decision to locate public housing within IPS was the result of a discriminatory purpose, either on the part of HACI or on the part of surrounding governmental units which may have resisted the placement of public housing in their communities, and, b) those decisions had a "substantial" interdistrict effect. 573 F.2d 400, 410, 413-4. The district court has now made both of these findings. The question on appeal is whether the evidence supports the findings. We will discuss the intent findings now and the question of "substantial interdistrict effect" at a later point in the opinion.

"The action of . . . official bodies in locating such projects within IPS, as well as the opposition of the suburban governments to the location of public housing within their borders, (was) racially motivated with the invidious purpose to keep blacks within pre-Uni-Gov Indianapolis and IPS, and to keep the territory of the added suburban defendants segregated for the use of whites only."

456 F. Supp. 183, 189.

These findings were based on the history of racial discrimination, both public and private, in housing in Marion County; on the natural and foreseeable results of the site selection practices; on the absence of convincing non-racial justifications for HACI's failure to locate projects outside of IPS; and on other actions by the Commission, not directly related to housing, which led the court to conclude that "the separation of the races, both in housing and in schools, has been an unspoken, but intentional policy of the Commission." 456 F. Supp. 189.17

The district court's findings have ample support in the record. At least five of the ten housing projects built during the 60's were located on or just a few blocks within the IPS-suburban boundary. Others are close-by. All, however, are within IPS. One project was built just across the street from Warren township (then 98% white). Another was located on land that was once a part of Wayne township (then 99% white), but was annexed by the old city, at HACI's request, over the objections of the IPS Superintendent that existing school facilities were inadequate to serve the project. The judge considered the reasons offered by HACI to explain this striking pattern and found none of them convincing. He also noted that the Commission, which selected or approved all the sites, had already been implicated in the de jure segregation of Crispus Attucks High School (for decades the official high school for black students within IPS) by its refusal to allow IPS to use land in a white neighborhood for a new Attucks High School.18

HACI and the suburban defendants have argued here, as they did in the court below, that public housing could not have been built outside the limits of IPS absent a "cooperation agreement" with the appropriate suburban government. The Federal Housing Act which provided the funds for the HACI projects did, at that time, require a cooperation agreement with "the locality involved."19 The district court found the federal statute irrelevant, holding that the "locality involved" was the governmental authority which had chartered the housing authority, here the (old) City of Indianapolis. We agree that in this case the cooperation agreement requirement is irrelevant, but for a somewhat different reason.

It was HACI's position at trial that it could not build public housing outside the old city of Indianapolis, despite its statutory authority to do so, without additional cooperation agreements. The evidence at trial, however, disclosed no serious attempts by HACI to obtain those cooperation agreements. Instead it showed that HACI never considered placing traditional public housing outside of the old city limits. The Deputy Mayor of Indianapolis, who had been a planner for the Metropolitan Planning Department when sites were selected and who had recommended sites to a task force charged with site selection, testified that the criteria used "automatically excluded looking outside the corporate limits of the (old) City of Indianapolis." But the trial judge, who heard the witnesses, found the "criteria" defense not credible and concluded, on the evidence before him, that the only reason sites outside of old Indianapolis were not considered was because the responsible agencies were pursuing a deliberate policy of racial segregation. On these facts, then, it makes no difference whether or not additional cooperation agreements may have been required. It was HACI's own decision to limit the projects to IPS that decision was not forced on it by the refusal of some other governing body to enter into a cooperation agreement.

We therefore affirm the district court's finding that the decision in the '60's to locate all public housing in Marion County within the boundaries of IPS was the result of a segregative intent by the responsible state agencies. We will discuss at a later point in the opinion the appellants' contention that the placement of all public housing within IPS did not have a "substantial" interdistrict effect.

INTERDISTRICT EFFECTS OF INTRADISTRICT SEGREGATION

In an effort to support a remedy that would extend beyond its own boundaries, IPS presented expert testimony in an attempt to establish that the de jure segregation of the schools within the old City of Indianapolis, as found by the district court in an earlier decision (332 F. Supp. 655) and affirmed by us (474 F.2d 81, cert. denied, 413 U.S. 920, 93 S. Ct. 3066, 37 L. Ed. 2d 1041 (1973)), had resulted in segregated housing and/or school enrollment patterns throughout the metropolitan area.20 In offering that proof IPS was relying on Chief Justice Burger's statement in Milliken v. Bradley that

"Before the boundaries of separate and autonomous school districts may be set aside . . . it must first be shown that there has been a constitutional violation within one district that produces a significant segregative effect in another district.

418 U.S. 717, 745, 94 S. Ct. 3112, 3127, 41 L. Ed. 2d 1069 (1974).

The essence of the expert's testimony was that the acts of de jure segregation practiced within IPS identified certain geographic areas of Marion County (all within IPS) as "black" areas and that correspondingly other areas beyond, as well as within, the limits of the school district became identified as "white" areas. Private housing choices then followed these quasi-official neighborhood designations. Since all of the "black areas" were within IPS that necessarily meant that a disproportionate number of the "white" areas were in the outlying school districts. Since school enrollment patterns so closely reflect residential population patterns, it could then be said that the predominately white school enrollments of the outlying districts were caused in significant part by the intentional acts of segregation practiced within IPS.

The expert's theory is not implausible. The district judge, however, after hearing the testimony, rejected it as not credible, citing both the general nature of much of the testimony and the weakness of statistical support for the conclusions regarding Marion County. No. IP-C-225 (S.D.Ind. April 24, 1979). On review of the record, we cannot say that the district court's rejection of the expert's testimony was improper. We therefore do not disturb the finding that the effects of school segregation within IPS cannot support the city-wide desegregation plan ordered by the district court.

The district court also relied on an Indiana statute, IC 1971 20-8.1-6.5, as a basis for the city-wide remedy.21 This statute was enacted by the Indiana legislature in 1974 in response to an earlier order of the district court in this case.22 It is entitled "Regulations Regarding Transfers of Pupils by School Corporations Pursuant to Court Order" and, by its terms, applies "solely" to situations where a state or federal court has found (and all appeals have been exhausted) that

"(a) a transferor corporation has violated the equal protection clause of the fourteenth amendment to the Constitution of the United States by practicing de jure racial segregation of the students within its borders;

"(b) a unitary school system within the meaning of such amendment cannot be implemented within the boundaries of the transferor corporation; and

"(c) the fourteenth amendment compels the court to order a transferor corporation to transfer its students for education to one or more transferee corporations to effect a plan of desegregation . . . which is acceptable within the meaning of such amendment."

In relying on this statute as an independent source of authority for its city-wide transfer plan, the district court judge held that these three conditions had all been established. 456 F. Supp. 183, 191.

Regardless of whether or not these conditions had been established in this case, however, it is clear from the face of the statute that it is procedural only designed to provide a mechanism for student transfers in cases where those transfers are ordered to remedy fourteenth amendment violations. It is not an independent source of authority to require student transfers and cannot, standing alone, be the basis for a desegregation order. The statute becomes applicable to this case only because we affirm the district court's finding that there were intentional state acts which had a significant impact across district lines.

Having concluded that only the Uni-Gov and housing violations meet the tests of Milliken and Arlington Heights, we must next consider whether those constitutional violations are appropriately remedied by the specific desegregation plan ordered by the district court.

The district court's plan calls for the transfer of slightly more than six thousand IPS students to eight other school districts within the city. The students remaining in the IPS schools would then be reassigned within IPS, under a plan not yet specified, to the extent necessary to eliminate racially identifiable schools. The various non-IPS defendants, relying primarily on Dayton Board of Education v. Brinkman (I), 433 U.S. 406, 97 S. Ct. 2766, 53 L. Ed. 2d 851 (1977), assert that this plan goes too far. The IPS noting that the plan does not call for the reassignment of any non-IPS students to IPS schools, asserts that it does not go far enough. We will address each of these contentions in turn.

Our most recent opinion remanding this case to the district court was written shortly after the Supreme Court had issued its first decision in Dayton Board of Education v. Brinkman, 433 U.S. 406, 97 S. Ct. 2766, 53 L. Ed. 2d 851 (1977) ("Dayton I"). In that case Justice Rehnquist spoke of the duty of the district court to "determine how much incremental segregative effect" specific constitutional violations had on the racial distribution of pupils "when that distribution is compared to what it would have been in the absence of such constitutional violations" and to design a remedy "to redress that difference." 433 U.S. 406, 420, 97 S. Ct. 2766, 2775. Therefore, although we observed that the district court had apparently kept in mind the principle that a remedy must be tailored to fit the nature of the violation, we urged the court on remand to "explicitly" consider it. 573 F.2d 400, 415.

Judge Dillin attempted to determine a precise number of children who would have received a desegregated education "but for" the failure of HACI to build public housing in the area outside IPS. 456 F. Supp. 183, 196. With respect to Uni-Gov, however, he concluded that it would be impossible to determine the incremental effect of the legislation since Uni-Gov would not have been passed if the schools had not been excluded from its scope. We will address the Uni-Gov questions first.

The difficulties in attributing specific incremental effects to the Uni-Gov violation do not mean that that violation cannot support an interdistrict remedy. In more recent cases the Supreme Court has described as a "misunderstanding" of Dayton I the view that a party must prove "with respect to each individual act of discrimination precisely what effect it has had on current patterns of segregation" (Dayton Board of Education v. Brinkman (II), 443 U.S. 526, 99 S. Ct. 2971, 2981, 61 L. Ed. 2d 720 (1979). The Court has quoted with approval the Sixth Circuit's holding that "policies of systemwide application necessarily have systemwide impact." (Columbus Board of Education v. Penick, 443 U.S. 449, 99 S. Ct. 2941, 2951, 61 L. Ed. 2d 666 (1979), quoting Penick v. Columbus Board of Education, 583 F.2d 787, 814 (6th Cir. 1978)), and has reaffirmed the basic rule that "the remedy imposed by a court of equity should be commensurate with the violation ascertained." Columbus Board of Education v. Penick, 443 U.S. 449, 99 S. Ct. 2941, 2950, 61 L. Ed. 2d 666.

It is now clear, if it was ever in doubt, that the "incremental effects" test applies only to cases where discrete isolated examples of discrimination are established. Proof of systemwide discrimination, in contrast, still calls for a systemwide remedy. It is the latter that we have here. The decision to prevent IPS from expanding with the City of Indianapolis was a decision affecting the city as a whole, not just certain schools or certain school districts within the city.23 We do not have a situation such as that presented in Denver, or Dayton, or Columbus, where the discriminatory actions took place one at a time, often affecting just one school at a time, and the courts had to determine whether, taken together, they permitted an inference of discriminatory intent as to the district as a whole. Keyes v. School Dist. No. 1, 413 U.S. 189, 93 S. Ct. 2686, 37 L. Ed. 2d 548 (1973); Columbus Board of Education v. Penick, 443 U.S. 449, 99 S. Ct. 2941, 61 L. Ed. 2d 666 (1979); Dayton Board of Education v. Brinkman (II), 443 U.S. 526, 99 S. Ct. 2971, 61 L. Ed. 2d 720 (1979).24

Nor is it relevant that without the companion legislation freezing IPS boundaries Uni-Gov never would have been enacted. Once the state chose to act it was obligated to do so in a non-discriminatory manner. Evans v. Buchanan, 393 F. Supp. 428, 443 (D.Del.) aff'd 423 U.S. 963, 96 S. Ct. 381, 46 L. Ed. 2d 293 (1975). Our case fits squarely within the rule of Milliken that "where district lines have been deliberately drawn on the basis of race . . . an interdistrict remedy would be appropriate to eliminate the interdistrict segregation directly caused by the constitutional violation." 418 U.S. 717, 745, 94 S. Ct. 3112, 3127, 41 L. Ed. 2d 1069. We reject therefore the claim of those appellants who assert that the remedy ordered by the court went too far.

A remedy must still be tailored to fit the violation and an interdistrict remedy must still attempt to respect local autonomy and local political processes. Milliken v. Bradley, 418 U.S. 717, 94 S. Ct. 3112, 41 L. Ed. 2d 1069 (1974); Hills v. Gautreaux, 425 U.S. 284, 96 S. Ct. 1538, 47 L. Ed. 2d 792 (1976). Here the district court ordered the implementation of a transfer plan which would preserve the independence of the individual school districts while still attempting to remedy the violation found. It has been suggested that this is not the remedy most closely tailored to the violation. We tend to agree. The most closely tailored remedy would likely be one that would allow IPS to expand to the city borders. Confronted with a nearly identical violation, that was the remedy ordered by a three judge district court in Delaware. Evans v. Buchanan, 416 F. Supp. 328 (D.Del.) aff'd 555 F.2d 373 (3rd Cir.) (en banc ), cert. denied, 434 U.S. 880, 98 S. Ct. 235, 54 L. Ed. 2d 160 (1977); see also, Morrilton School District No. 32 v. U. S. A., 606 F.2d 222, 229 (8th Cir. 1979) (en banc ). Such a remedy might well have been appropriate here. It also, however, would have been far more intrusive into local political processes. Under the circumstances we cannot say that the trial judge violated his obligation to weigh these competing considerations or abused his discretion in developing a remedy.

The transfer plan is also an appropriate remedy for HACI's decision, for racially discriminatory purposes, to place all public housing projects within IPS. The connection between that policy and suburban housing and school enrollment patterns is obvious. It, like the Uni-Gov violation was not limited to a single school or even to a single school district. It was a general policy which affected housing and school patterns throughout HACI's area of operation. Ten public housing projects, with units for over 1600 families, were built by HACI pursuant to this policy. In addition, as the court found, once a public housing project was opened other blacks tended to settle in the area around the project, thus multiplying the effect which the project had already had on the racial composition of neighborhood schools. While appellants may be correct in their claim that the judge miscalculated the exact number of students affected by the violation, we cannot agree that the number of students affected is so small that the violation can be said to be insignificant. We do note, however, that if this were the only constitutional violation present in this case, an appropriate remedy on this record would have to be limited to the territory within HACI's area of operation at the time the projects were built.

There are, nevertheless, two aspects of the transfer plan that require further discussion. The first is the "one-way" remedy ordered by the district court. Under this plan black students from IPS are to be transferred from IPS to other districts in the city but no students from other districts are to be transferred to IPS. Many white students within IPS will be reassigned to other schools within the district, however, so that it cannot be said that only minority students will bear the burden of the remedy imposed by the court. The court also considered the generally better facilities available in the outer districts as compared to IPS and the fact that reducing the student population of IPS would allow that district to close some obsolete schools. Therefore, despite our concerns about the possible inequities of the plan, we are not prepared to say that it was such a clear abuse of discretion that it must be set aside in favor of a two-way plan.

We do note, however, that among the reasons the court listed for rejecting a two-way plan suggested by IPS was the court's conclusion that it had "no power to order a suburban child be transported out of its own school corporation so long as the suburban school corporations remain as separate legal entities, because there is no evidence that any of them has operated anything other than a unitary school system." (No. IP 68-C-225 S.D.Ind. Memorandum Decision at 17, April 24, 1979). In so holding, the district court was in error. There is no distinction that can be made between the court's power to order districts to accept transfer students from IPS and its power to order students from those districts to transfer to IPS schools. In both cases the power depends not on the culpability, or lack thereof, of the suburban districts involved but rather on the finding that discriminatory actions by the state had a significant segregative impact across district lines. The court does, then, have the power to implement a plan which would reassign students from non-IPS districts to IPS schools. We will not vacate the plan as ordered but will instruct the district court that it does have this power and should feel free to use it if modifications of the plan become necessary.

There is one other matter that requires further consideration by the district court. The school town of Speedway and the school city of Beech Grove both assert that their peculiar status under both Uni-Gov and the state law from which HACI draws its authority make it inappropriate to include them in a city-wide remedial plan, regardless of whether or not such a plan is otherwise appropriate. Without finally determining the merits of those claims, we think they are substantial enough to require separate consideration by the district court.

The special status of the cities of Speedway and Beech Grove under the Uni-Gov Act is not in dispute. The Act defines them as "excluded cities" and they are technically not a part of the "new" City of Indianapolis. IC 18-4-1-2(f). Nevertheless, as we have noted previously, the consolidated government was given "significant powers even in the excluded cities." 541 F.2d 1216. The hybrid status of the communities is reflected in their political structures: each community has its own elected government but the residents of each also vote in Uni-Gov elections. See Dortch v. Lugar, 255 Ind. 545, 266 N.E.2d 25 (1971). In each case the school district's boundaries are coterminous with the civil boundaries. The appellants make a strong showing that Indiana law prior to Uni-Gov, despite its policy in favor of school district expansion following civil expansion, would not have permitted IPS expansion into Speedway and Beech Grove. The two cities therefore assert that since their schools were unaffected by the violation a remedy which includes them goes beyond the violation shown.

We think the issue is not quite as simple as Speedway and Beech Grove present it. As they present the argument, unilateral civil annexation of Speedway and Beech Grove by Indianapolis was prohibited by Indiana law. Therefore, the City of Indianapolis could not expand into their communities. Therefore, even if it was state policy to allow school boundaries to expand with civil boundaries, the barrier to unilateral civil annexation would have prevented IPS from ever expanding into their communities.

But despite all the barriers to annexation established generally by Indiana law prior to Uni-Gov, the Uni-Gov legislation did allow the City of Indianapolis to expand to the borders of Marion County. And specifically Speedway and Beech Grove, despite their protected status under pre-Uni-Gov annexation law and despite their categorization "as excluded cities" under Uni-Gov, nonetheless became subject to the authority of the Uni-Gov government for many purposes.25

The question then is not whether Indiana law prior to Uni-Gov would have permitted the expansion of IPS into Speedway and Beech Grove but rather whether, if the Uni-Gov legislation had not generally excluded schools, it would have nonetheless excluded the schools of Speedway and Beech Grove.

It also appears likely that HACI's area of operation did not include the City of Beech Grove and may not have included the Town of Speedway as well. Prior to 1969, the area of operation of a city housing authority was defined so as to exclude "any area which lies within the territorial boundaries of some other city as herein defined." "City," in turn, was defined in the Act as "any city." Indiana Acts, 1937, Ch. 207, Sec. 3. Beech Grove was, and is, incorporated as a city under Indiana law. Speedway was a town. The 1969 amendments to the Act expanded on this, stating that a city housing authority's area of operation would not include "another city or town (or) the area of operation of the housing authority of another city or town." IC 18-7-11-3(g). This raises several questions. The most obvious is whether HACI's pre-1969 area of authority may have included the Town of Speedway but not the City of Beech Grove. While the two communities appear to be similarly situated with respect to Uni-Gov they may be in different positions with respect to HACI. In either case, however, the requirement that a remedy be tailored to fit the violation means that their inclusion in the district court transfer plan may be inappropriate.

These substantial questions have not been fully briefed. Nor have they been addressed by the district court judge, whose familiarity with the relevant Indiana statutes greatly exceeds ours. We will therefore continue our stay of that portion of the order which requires the transfer of students into the Speedway and Beech Grove schools and remand the case with instructions to the district court to include them in the final order only on a finding that either the Uni-Gov or the HACI violations affected their districts.

To summarize the decision thus far, we find two grounds of support for the remedy ordered by the district court. The first is the state's affirmative decision, made for a discriminatory purpose, to limit the boundaries of IPS to the old city limits of Indianapolis. The second is HACI's decision, again for discriminatory purposes, to locate all public housing within IPS territory. We hold, however, that the Indiana student transfer statute does not provide an independent basis of support for the interdistrict remedy. The specific reassignment plan ordered by the district court judge is generally found not to be an abuse of discretion, although modification may be in order and although the School Town of Speedway and the School City of Beech Grove are, at least for the time being, removed from the scope of that order. We will now consider the various other issues presented on appeal.

The state of Indiana appeals from the order of the district court requiring it to pay for certain ancillary relief (in-service training programs, etc.) ordered by the court. The Supreme Court has considered similar objections raised by a state in another case and expressly rejected them. Milliken v. Bradley (II), 433 U.S. 267, 97 S. Ct. 2749, 53 L. Ed. 2d 745 (1977). We affirm the district court's holding in this regard.

In April of 1978 the district court denied a request by IPS to adopt an intradistrict plan pending resolution of the question of interdistrict relief. The United States, which was supporting the IPS proposal, appealed. We vacated the April order and ordered further consideration of the IPS proposal. On June 2, 1978, after a hearing, the district court again denied the IPS request. Both IPS and the United States appealed. In July we ordered the district court to allow IPS to implement the high school (phase I) portion of the plan. The rest of the appeal was retained. We now consider whether it was an abuse of discretion for the district court to reject the IPS plan for interim intradistrict relief for upper (phase II) and lower (phase III) elementary schools within the district. There is no question that desegregation of students within IPS is long overdue. The district court judge's fear that an interim plan of desegregation would legally preclude eventual interdistrict desegregation is groundless. His concern that a plan limited to IPS would lead to white flight is, as we have stated before, irrelevant. 503 F.2d 68, 80. Nevertheless, there also appeared to be valid reasons for the court's reluctance to approve the specific plan set forth by IPS: the lower elementary school plan was only a very sketchy outline of a hurriedly-adopted proposal and the upper elementary plan required two new buildings as well as remodeling of present facilities. Under these circumstances the court's refusal to adopt the phase II and phase III plans was not an abuse of discretion.

We are informed that more specific revised plans have now been submitted to the district court. It goes without saying that an intradistrict plan must be put into effect along with the interdistrict plan we approve today. In addition, we can think of no reason why an intradistrict plan should not be implemented in September 1980 regardless of the status of the interdistrict remedy at that time. The district court judge is instructed to give priority to the development of such a plan.

In Part I we affirm the trial court's finding that HACI acted with a discriminatory purpose when it decided to locate all public housing projects within the old city of Indianapolis. The injunction against HACI, preventing it from building any low-rent housing within IPS, rests on the same evidence and is not otherwise contested. For the reasons already discussed, we affirm the injunction.

The judgments entered against the School City of Beech Grove and the School Town of Speedway are vacated and remanded for further consideration in accordance with this opinion. In all other respects the judgments appealed from are affirmed.

TONE, Circuit Judge, dissenting.

For nearly half of the twenty-six years since Brown v. Board of Education of Topeka, 347 U.S. 483, 74 S. Ct. 686, 98 L. Ed. 873 (1954), the Indianapolis School case has been in the federal courts.1 It will soon be nine years since the district court found de jure segregation in the Indianapolis Public School System. We affirmed that judgment, and almost seven years ago the Supreme Court denied certiorari.2 I join in the majority's statement that "we can think of no reason why an intradistrict plan should not be implemented in September 1980 regardless of the status of the interdistrict remedy at that time," and in the directive to the district court to "give priority to the development of such a plan." Ante at 1117. In this second decade of the litigation, the results of de jure segregation in the Indianapolis Public School System should not continue to remain in substantial part unremedied.

Looking back, the long delay, during which a generation of children has passed through segregated schools in Indianapolis, seems to me to be the fruit of a search for an elusive constitutional violation capable of supporting an interdistrict remedy that, the district court has often said, promises desegregation in fact as well as in law. Anyone familiar with the history of this case knows the determination with which the district court has pursued that goal. The court has said time and again that a remedy confined to the "IPS strait jacket" will produce a student population in excess of forty percent black, will in all probability exceed an indefinable "tipping point," will thus cause accelerating and irreversible "white flight" from IPS, and will, finally, produce nothing of lasting value for a city still in search of its first unitary school system. All this may be correct, cf. Milliken v. Bradley, 418 U.S. 717, 732-35, 739, 94 S. Ct. 3112, 3121-22, 3124, 41 L. Ed. 2d 1069 (1974), but, as the district court once recognized, the federal courts may impose an interdistrict remedy only "if this can be done within the law."3 The long effort to prove that it can in Indianapolis seems to me to have failed.

When we last remanded this case to the district court, we did so in pursuance of the Supreme Court's mandate that the case be reconsidered in light of Washington v. Davis, 426 U.S. 229, 96 S. Ct. 2040, 48 L. Ed. 2d 597 (1976), and Village of Arlington Heights v. Metropolitan Housing Development Corp., 429 U.S. 252, 97 S. Ct. 555, 50 L. Ed. 2d 450 (1977).4 Those then recent decisions make it clear that state action taken without discriminatory purpose does not violate the equal protection clause. On remand, the district court held no further evidentiary hearing on the issue of discriminatory purpose; it discovered the requisite purpose on the existing record, which a majority of this court apparently had considered insufficient.5 Today we affirm on the basis of that record.

The record no more supports today than it did four years ago the thesis that either the passage of Uni-Gov or the actions of the Housing Authority were tainted by discriminatory purpose. I have stated in prior dissents6 the reasons that led me then to conclude that invidious purpose had not been proved; a third look at the increasingly familiar transcript of the 1975 trial does not persuade me otherwise.7 Without repeating, I hope, too much of what I earlier said, I feel compelled to add what follows.

"(T)he invidious quality of a law claimed to be racially discriminatory must ultimately be traced to a racially discriminatory purpose." Washington v. Davis, supra, 426 U.S. at 240, 96 S. Ct. at 2048. Though discriminatory purpose need not be the "dominant" or the "primary" motive force of the official act in question, Village of Arlington Heights v. Metropolitan Housing Development Corp., supra, 429 U.S. at 265, 97 S. Ct. at 563, the concept "implies more than intent as volition or intent as awareness of consequences. It implies that the decisionmaker, . . . selected or reaffirmed a particular course of action at least in part 'because of,' not merely 'in spite of,' its adverse effects upon an identifiable group." Personnel Administrator v. Feeney, 442 U.S. 256, 279, 99 S. Ct. 2282, 2296, 60 L. Ed. 2d 870 (1979) (citation omitted). "(D)isparate impact and foreseeable consequences, without more, do not establish a constitutional violation." Columbus Board of Education v. Penick, 443 U.S. 449, 464, 99 S. Ct. 2941, 2950, 61 L. Ed. 2d 666 (1979); see Dayton Board of Education v. Brinkman, 443 U.S. 526, 536 n.9, 99 S. Ct. 2971, 2978 n.9, 61 L. Ed. 2d 720 (1979) (Dayton II ).8

The burden of proving discriminatory purpose is, initially, on the plaintiff. Village of Arlington Heights v. Metropolitan Housing Development Corp., supra, 429 U.S. at 270, 97 S. Ct. at 566; Washington v. Davis, supra, 426 U.S. at 241, 96 S. Ct. at 2048. "Proof that the (state) decision . . . was motivated in part by a racially discriminatory purpose . . . shift(s) to the (state) the burden of establishing that the same decision would have resulted even had the impermissible purpose not been considered." Village of Arlington Heights v. Metropolitan Housing Development Corp., 429 U.S. at 270 n.21, 97 S. Ct. at 566 n.21.

In determining the presence of discriminatory purpose, the following factors are relevant, though not exhaustive:

The historical background of the decision . . ., particularly if it reveals a series of official actions taken for invidious purposes.

The specific sequence of events leading up to the challenged decision . . .

Departures from the normal procedural sequence

Substantive departures . . . , particularly if the factors usually considered important by the decisionmaker strongly favor a decision contrary to the one reached.

The legislative or administrative history . . . , especially where there are contemporary statements by members of the decisionmaking body, minutes of its meetings, or reports. In some extraordinary instances the members might be called to the stand at trial to testify concerning the purpose of the official action . . . .

Id. at 267-68, 97 S.Ct. at 564-565.

The district court found no procedural irregularities in the passage of Uni-Gov. Indianapolis V, 456 F. Supp. at 188. So far as I can determine, the court found nothing in the "historical background of the decision" to adopt Uni-Gov9 indicating "a series of official actions taken for invidious purposes." The "specific sequence of events" that led to Uni-Gov seems blameless.10 The legislative history of the Act indicates that no discriminatory purpose was at work.11 Of the five factors spelled out in Arlington Heights, the district court found one, "substantive departures," dispositive of the issue. Even here, the court did not inquire, as Arlington Heights directs, 429 U.S. at 267-68, 97 S. Ct. at 564-565, whether the "factors usually considered important by the decisionmaker strongly favor(ed) a decision contrary to the one reached." It is apparent that they did not.12 Nor did the district court parse burdens of proof.13 No finding was made that appellants had failed to rebut what the court apparently found to be a prima facie case of discriminatory intent. As Arlington Heights makes plain, however, a finding of invidious purpose is rebuttable by a showing that the state action in question would have been taken in the absence of such purpose; the rejection of a county-wide school system a few years prior to the passage of Uni-Gov is powerful evidence that the same race-neutral reasons that dictated this prior decision would, even assuming a hypothetical discriminatory purpose in the exclusion of schools from Uni-Gov, have produced the same result a few years later.14 But putting aside these fine points of equal protection analysis, even the finding of a "substantive departure" is clearly in error.

The district court found that the Indiana General Assembly passed Uni-Gov with discriminatory purpose because it "departed" from a "long established rule that the school city of Indianapolis should expand with the civil city" when it "repeal(ed) the crucial section of the 1961 Act" and "eliminat(ed) the schools from Uni-Gov."15 The majority, in sustaining this finding, is more circumspect. It notes that this "crucial section," had it been left unrepealed, would have expanded IPS boundaries only "if" Uni-Gov "could be characterized as an extension of Indianapolis' civil city boundaries by civil annexation." Ante at 1107. The majority appears to conclude that it can be so characterized. It seems to me plain that it cannot.

Under either the pre-1969 or the post-1969 civil annexation statute of Indiana,16 civil annexation could be accomplished only by act of a city or town council. People who opposed annexation could file remonstrances. People who wished their property to be annexed could petition the council. No annexation could involve territory already a part of another city or town. The sole means for union of cities or towns was by a referendum in which citizens of the municipalities involved approved a proposed consolidation agreement reached by the councils. Union could be achieved only if all of the territory of both municipalities were to be joined.

In sharp contrast, Uni-Gov was the Act of the Indiana General Assembly. It was not the result of a petition, a referendum, or an act of a municipal council. It was not subject to remonstrances or to any of the other provisions of the civil annexation statutes;17 indeed, it was the very need to get away from those unsatisfactory provisions that prompted the adoption of Uni-Gov.18 To say that Uni-Gov, a complex reorganization of Marion County municipal governments undertaken by the State of Indiana, might be characterized as a civil annexation ignores very basic legal distinctions.

Because Uni-Gov did not involve an annexation, the repeal of § 9 of the 1961 Act is irrelevant to this case.19 Had the section remained in force it would not have expanded the boundaries of IPS by an inch. Uni-Gov was totally beyond its terms.20

This crucial-section issue aside, all that remains of the substantive-departure finding is the related notion that Uni-Gov was a departure from a "long established rule" that IPS and Indianapolis would share a common boundary. This notion appears to stem from a belief that Indiana has had a policy favoring coterminous boundaries of city and school corporations. "Of greater relevance," the majority says, ante at 1106, "is the history of school district expansion and consolidation in Indiana." It continues: "Indiana has for generations pursued a legislative policy that school district lines should grow as the corporate lines of their cities grow." With all due respect, I simply do not think the law of Indiana supports this view. Regard for the reader's patience, as well as a desire to speed this case on its way, limits me to a brief summary of the history of Indiana law on this point.

The organization of schools in Indiana has gone through three basic phases. In the first phase, schools were organized by congressional township. The Territory, and later State, of Indiana were divided into townships of thirty-six square miles each; a thirty-sixth part of each township was reserved for the support of public schools. From a 1785 ordinance of the Continental Congress, through Indiana's first constitution in 1816, to a series of laws culminating in 1843, the method of organizing and supporting schools by township land grant was preserved.21 During this time, school districts were "laid off in such manner as shall be most convenient for the population and neighborhoods."22 School trustees could, "if convenience require(d) it," form districts "out of parts of each township," and a majority of the voters of a district could, by petition, alter the district boundaries "as convenience may require."23

The second phase began with the adoption of a new Indiana Constitution in 1851.24 Under statutes passed pursuant to the new constitution,25 the state took over the burden of financing the schools, and it reorganized them with the civil city or town as a basic geographical unit. Coterminous city and school boundaries were often the result.

The advent of the third phase was gradual. Some cities failed to reorganize the schools under the new laws, and these omissions were legalized by the legislature in 1897.26 In 1899, cities that had reorganized their schools were permitted to revert to the township format, again by act of the legislature.27 In 1901, cities were permitted to consolidate their schools with those of other school corporations in the same township.28 Acts in the early part of this century permitted jointly run schools and consolidation of school corporations.29 By mid-century, the third phase of Indiana school organization was well underway.

The new phase was ushered in by the need to increase the revenue base for the many small city school corporations. In 1947 and 1949, the Indiana General Assembly adopted statutes allowing the creation of school corporations without respect to city, town, or township boundaries.30 By 1959 the Supreme Court of Indiana was to say:

(T)he legislature has found it necessary to enlarge the tax base supporting the local school unit. This has necessitated getting away from the identification of school district boundaries with the boundaries of civil units, and the many attempts of the General Assembly to deal with school consolidation in the past 40 years are but expressions of the legislative intent in this direction.

Fort Wayne Community Schools v. State, 240 Ind. 57, 64, 159 N.E.2d 708, 711 (1959). In 1959 also the legislature passed the School Corporation Reorganization Act31 to accomplish that end.

Thus the policy from which Uni-Gov is said to be a substantive departure was first pursued for fiscal and administrative reasons, was abandoned when those reasons no longer supported it, and had not existed in Indiana for many years prior to the passage of Uni-Gov.32 The only real policy that can be inferred from the foregoing history of school district boundaries in Indiana is that district lines should be drawn on a financially sound, efficient, and locally acceptable basis.

The 1961 Act so much debated in this litigation, § 9 of which is shown above to have been inapplicable to Uni-Gov because triggered only by civil annexation, is not inconsistent with the policy just described. Indeed, as the provisions summarized in the margin demonstrate, that Act is expressive of this policy and not of an obsession that school district and municipal boundaries must match, come what may.33

Given the absence of a substantive departure and the irrelevance of the repeal of § 9 of the 1961 Act, the court is left where it started: The legislature omitted schools from Uni-Gov, but there is no evidence that race was a reason for doing so. The real question then is whether the Indiana General Assembly had an obligation to include schools in the reorganization effort and thereby further remedial steps toward integration. As I understand Milliken v. Bradley, 418 U.S. 717, 94 S. Ct. 3112, 41 L. Ed. 2d 1069 (1974), such an obligation did not exist in the absence of an interdistrict violation.

To find discriminatory purpose in the Housing Authority's location of public housing within the City of Indianapolis is to ignore reality and the record in this case. In the first place, the record shows that the Housing Authority played a relatively small role in the selection of public housing sites. Two methods were used to select the sites. The first was the "turn-key" method.34 As was testified at trial, by this method a private developer could "select sites and option the ground and subsequently turn the project over to the Housing Authority," after having built the units "consistent with specifications." A majority of the sites on which the projects were eventually built were selected by this method, which had the virtue of producing public housing faster than did site selection by government agency.

The remaining sites that were eventually used by the Housing Authority were found by a mayoral task force working in conjunction with the Metropolitan Planning Commission. The task force produced a map of potential housing sites "both inside and outside the city limit," and then recommended certain of these sites "to the administration in the Public Housing Authority as potential public housing sites." The task force also approved sites that private developers had optioned under the turn-key method.

From the foregoing it is apparent that the role of the Housing Authority in site selection was limited; the Authority apparently chose from among the sites tendered to it by the task force. Nothing in the record indicates how this choice was made, and there is no evidence concerning whether the Authority ever undertook any independent search for project sites. This alone undercuts the conclusion that plaintiffs proved discriminatory purpose on the part of the Housing Authority. If more were needed, it is clear from the record that the absence of sites outside of Indianapolis was due not to discriminatory purpose, but to neutral requirements of Indiana and federal law.

First, as we previously have acknowledged,35 and as the majority does not now appear to dispute, the Housing Authority could obtain the necessary federal funds to finance its projects36 only if it had obtained a cooperation agreement with "the governing body of the locality involved."37 Under federal law, a cooperation agreement requires a governing body to provide services and to waive taxes on the project.38 Second, the Housing Authority's jurisdiction under Indiana law during the time in question extended to Indianapolis and any area within five miles of the boundary of Indianapolis, but not to any territory within any other city.39 Finally, any power the Authority may have had to build projects in the unincorporated suburbs was effectively nullified by the fact that it had no cooperation agreement with the "governing body" of those areas, which consisted of the Marion County Council and its executive arm, the Board of County Commissioners.40

The majority appears to approve a finding of invidious purpose because "(t)he evidence at trial . . . disclosed no serious attempts by HACI to obtain . . . cooperation agreements,"41 and because "(the evidence at trial) showed that HACI never considered placing traditional public housing outside the old city limits." Ante at 1111. Apart from the vagueness of the phrases serious attempts and traditional public housing,42 this view seems to me to have a number of problems. First, as I have earlier observed, the evidence at trial does not indicate whether or not the Housing Authority acted independently of the task force in seeking sites outside Indianapolis. The task force considered suburban sites and reported to the Housing Authority,43 and for all the record reveals, the Authority may also have investigated suburban sites. Second, the majority's view places the burden of proof on the issue of discriminatory purpose on the wrong party. Even assuming that the Housing Authority failed to demonstrate a "serious" attempt to affirmatively facilitate school desegregation by placing projects in the suburbs, that fact scarcely suffices to carry plaintiffs' burden on the issue of discriminatory purpose. See Village of Arlington Heights v. Metropolitan Housing Development Corp., supra, 429 U.S. at 270 & n.21, 97 S. Ct. at 566 n.21; Washington v. Davis, supra, 426 U.S. at 241, 96 S. Ct. at 2048. Finally, the record reveals that the task force that selected sites for the Housing Authority approached the County Council and County Commissioners about the possibility of placing projects in their jurisdiction. The response was negative; the reason given was that the County could not provide for the projects the municipal services then required under federal law.44 I assume the Housing Authority was not required by the Fourteenth Amendment to engage in a futile repetition of the request. There is no evidence that the county government denied the request for invidious reasons. Thus, even if a violation might be predicated on that theory, cf. James v. Valtierra, 402 U.S. 137, 91 S. Ct. 1331, 28 L. Ed. 2d 678 (1971), there is no proof to support the theory.

As a practical matter, the Housing Authority had power to locate projects only within the City of Indianapolis. Such power as it had was exercised to promote integration, not segregation. The projects, most of whose occupants would be black, were located in what were at the time primarily white areas.45 It is ironic that these sitings have now subjected the Housing Authority to a finding of invidious purpose.

If I am right in my belief that the record does not support the findings of discriminatory purpose,46 the real issue raised by this case is whether otherwise permissible state action that does not attempt to remedy the effect of de facto segregation is for that reason alone an act of de jure segregation. The lesson to be derived from today's decision seems to be that the answer is yes, and that a state's attempt to solve a governmental or social problem will be constitutionally permissible only if the state adopts the solution that will maximize desegregation of schools that are de facto segregated. Indiana, had it chosen to leave Marion County's civil government in its chaotic state, instead of adopting Uni-Gov, would have had no affirmative obligation to reorganize the schools or to transfer students across district lines.47 Similarly, had Indiana failed to provide a Housing Authority, or had it confined the Authority's jurisdiction to Indianapolis, the Constitution would not have been violated. Instead, Indiana acted to improve municipal government and to provide decent housing for persons of limited means in a manner that would promote integration in housing. Because Indiana did not include in its governmental-structure and housing reforms a program of affirmative action with respect to de facto school segregation, which if included would have made the reforms themselves impossible to carry out,48 Indiana has been found to have intended a school segregation it did not create.

Thus a state may not restructure civil government or build public housing in an area in which de facto school segregation exists without assuming also the affirmative duty to remedy that segregation. The failure of the state action to include a remedy for school segregation will be enough to establish discriminatory purpose, for if the additional circumstantial indicia in this case can tip the scales toward a finding of invidious purpose, it is hard to imagine a case in which sufficient additional indicia would be lacking. The result of such a rule may be to force the elimination of some de facto segregation. Democratic processes being what they are, however, the result may also be the abandonment of efforts to provide needed governmental reforms and services. But whatever the result, today's decision cannot, I think, be reconciled with the distinction between de jure and de facto segregation as it has been recognized by the Supreme Court.49

With all respect for my colleagues, both here and on the district court, I can find no basis in the present record for expanding the remedy in this case beyond the system in which the violation occurred. I would therefore reverse the judgment and remand with the direction to proceed at once toward implementation of an intradistrict remedy.