concurring:
I agree with Chief Judge Seitz’s analysis of the questions of subject matter jurisdiction and jurisdiction over the parties. I also agree with the Chief Judge’s analysis of the merits. I therefore concur fully in his opinion.
I write separately to address the question of the limits on the scope of federal equitable remedial power. More specifically, in a case such as this, where the original issues that gave rise to federal subject-matter jurisdiction are no longer in dispute, and where a consent decree negotiated by the parties calls for continuing and extensive federal supervision over what are essentially state-law property and contract issues, the question arises whether there are prudential limits on the district court’s power to approve the decree.
This issue was not raised by the parties, either before the district court or on appeal, nor was it raised by the district court sua sponte. The issue ordinarily, therefore, would not merit my writing separately. Because, however, this case demonstrates the potential for federal court involvement in matters that should be relegated to state courts, because appellate courts will seldom have the issue squarely before them — the parties may be estopped from arguing that the district court abused its discretion in approving a consent decree, — and because, as a former district judge, I recognize that district judges are often chary of upsetting a settlement by refusing to approve terms of a negotiated decree, I take the liberty of advancing, for consideration of the district courts in the fashioning of consent decrees, at least one possible approach to this problem.
As I see it, the potential issue as presented by this case is whether, because the federal question that gave rise to this suit had long been settled, the district court could have been charged with exceeding allowable discretion when it approved consent decrees in 1980 and 1982 that provided for continuing federal supervision to settle any disputes that might arise between the parties out of the consent decrees. The legal issues likely to be involved in the interpretation and enforcement of the decrees were classically questions of state contract and property law — matters in which the federal government, because there is no diversity of citizenship between the parties, has no interest. As I have suggested, there are limits on the discretion of the district court to involve itself in the resolution of non-federal disputes under the guise of enforcing a consent decree where the issues which gave rise to the federal court’s subject-matter jurisdiction are no longer in dispute. Where those limits are exceeded, I believe that the dis*1541trict court should remit the parties to state court to enforce their agreement.
In order adequately to explain my position, I must develop the facts and procedural history in somewhat more detail than was necessary for Chief Judge Seitz. I turn now to that task.
I.
As the Chief Judge explained in the majority opinion, the origins of this litigation stretch back to the early 1960s when the City of Philadelphia, through its agent the Redevelopment Authority (RDA), exercised its powers of eminent domain to purchase substantial tracts of land containing substandard dwellings around the campus of the University of Pennsylvania. The dual purpose of these acquisitions was to eliminate urban blight by tearing down the offending structures and to provide land to the University for its planned expansion over the following decade.
Neither of these goals was immediately realized. The University apparently ove-rexpanded and therefore did not need the additional land between Sansom and Walnut Streets at 34th Street. Therefore, instead of razing the houses, the RDA became a landlord. By the early 1970s, the University decided it did not need the land, and therefore proposed to the RDA that a commercial developer be chosen to build on it. By this time, however, fashions in urban renewal had changed. The structures that in the 1960s had been considered examples of urban blight, were, with the gentrification movement of the 1970s, highly desirable shells that, with rehabilitation, had the potential to be very expensive townhouses. Consequently, the residents of the houses, who were renting from the RDA, opposed the University’s plan to nominate a commercial developer. These residents formed the Sansom Committee for the ostensible purpose of preventing the RDA from changing the character of the neighborhood from residential to commercial. As an alternative to the University’s plan, the Committee proposed that its members be given the right to purchase the houses from the RDA, and to redevelop them themselves.
In 1973 the Sansom Committee brought suit in federal court alleging that the RDA’s plan to tear down the houses violated the National Environmental Policy Act of 1969, 42 U.S.C. §§ 4321-4361 (1976) (“NEPA”), and the National Housing Act of 1949, 42 U.S.C. §§ 1441-1490h (1976) (“NHA”). The Committee was unsuccessful in preventing the City from tearing down the houses on the 3400 block of Walnut Street, but eventually obtained an injunction temporarily saving the houses on Sansom Street. In 1976 the Committee, the University, and the RDA reached an agreement in principle not to tear down the houses, and the district court placed the case on its suspense docket. With this agreement, the federal interest — preserving the residential character of the neighborhood — was essentially satisfied.1 The remaining issue, in addition to the signing of a formal settlement agreement, was the question of who should be permitted to buy the houses from the RDA, and under what conditions..
Four years later, in December of 1980, the parties informed the district court that they had committed an agreement to writing, which they styled a “consent decree.” This agreement essentially provided that the University would nominate individuals designated by the Sansom Committee to rehabilitate the houses on Sansom Street, and that these nominees would be given the right to purchase the houses from the RDA at prices substantially below market cost. The agreement also provided that the district court would retain jurisdiction to oversee its implementation. The University, the Sansom Committee, and the RDA all signed this agreement, and the district *1542court approved it on January 12, 1981. A second “side agreement,” signed by the University and the Sansom Committee, listed the individuals who were designated to rehabilitate the properties, and bound these individuals to abide by very detailed redevelopment criteria. This side agreement also set out a number of restrictions and limitations on the title the designees would receive. This agreement was not part of the “consent decree,” and the district court retained no jurisdiction to enforce it.
Rather than settling the case, the signing of this consent decree set off a new round of disputes between the University and the Sansom Committee, and among the members of the Sansom Committee. At least four members of the Sansom Committee felt that their interests had not been represented by the Committee’s negotiators. These members unsuccessfully tried to intervene before the district court in order to protect their interests. In addition, Michael Karp, one of the named plaintiffs in the original action, unsuccessfully tried to block approval of the consent decree by the district court. Karp appealed the district court’s approval of the consent decree to this Court, and we affirmed the district court. Subsequently, Karp has continued to litigate in federal court. Today we have rejected four of his appeals arising out of his dispute with the Sansom Committee. In addition, Karp has filed two suits in state court. One suit alleged breach of fiduciary duty against the Sansom Committee, its “trustee ad litem” Elliot C.R. Cook, the University, and the RDA; the other suit alleged malpractice and fraud against Sansom Committee counsel.
In addition to the intra-Sansom Committee squabbles, the agreement between the Sansom Committee and the University became “unravelled” by the late summer of 1981, with both parties seeking changes to their 1980 agreement, and the University filing a motion to enforce in the district court. Eventually, the University and the Committee agreed to a new settlement, which they signed on April 7, 1982. This new agreement modified the first consent decree in several ways. The University was granted a right of first refusal to purchase the homes on Sansom Street whenever the purchasers designated by the Sansom Committee decided to sell. In addition, the new agreement incorporated an amended version of the 1980 side agreement between the University and the San-som Committee, including detailed specifications for rehabilitating the houses and a declaration of easements and restrictions. This second agreement, which is an inch and a half thick and has aptly been characterized by counsel for the University as a “code of federal regulations” for the rehabilitation of the houses on the south side of the 3400 block of Sansom Street, specifically declared that the district court was to retain jurisdiction to supervise its implementation and enforcement.2 This revised consent decree was approved by the district court on August 12, 1982.
Unfortunately, the second decree was no more successful than the first in ending the parties’ disputes. The Committee and the University immediately began fighting over such questions as who were proper designees, by what date the designees had to purchase the houses from the RDA, and what the consequences were if some of the designees failed to purchase the houses “in time.” As provided by the decree, the parties returned to the district court to resolve their disputes. The district court decided the disputed issues, generally favorably to the Sansom Committee. The present appeal is from these decisions.3
*1543II.
The interesting and difficult point is whether all of this in-fighting over the redevelopment of the 3400 block of Sansom Street belongs in federal court. Judge Garth apparently shares this concern, but characterizes the issue as one of “jurisdiction.” I believe, on the other hand, that the question is not one of jurisdiction, but rather one of discretion and the scope of remedial power. Specifically, the problem is whether the district court should have agreed to oversee the implementation and enforcement of the settlement agreements.
My analysis begins with my belief that it is wasteful and inappropriate for the federal courts to be spending time on a case such as this. The current dispute between the university and the sansom committee is solely one of contract interpretation which would ordinarily be a question of state law. Because there is no diversity of citizenship among the parties, the only federal interest in this case is based on the claims of the Sansom Committee that the original plan to tear down the houses on the 3400 block violated a federal statute requiring that there be “community input” before a federally funded urban renewal project changed the character of the neighborhood from residential to commercial. Since none of the parties to the current litigation has any intent of tearing down the houses, this federal interest has essentially dissipated. On the other hand, if the district court continues to exercise supervision over the settlement agreement, it will be required to interpret a “code of federal regulations” for the rehabilitation of the houses on the South-side of the 3400 block of Sansom Street. At oral argument, counsel for the University indicated that, if we upheld the district court’s power to superintend this consent decree, a myriad of motions would be filed with the district court requesting the court to interpret the incredibly detailed structural specifications for rehabilitation of the houses. For example, counsel stated that the parties presently dispute the meaning of the building height limit contained in the specifications: the 3400 block of Sansom Street is on a hill and the parties cannot agree as to whether the height limit is to be measured from the low-end or high-end of the block, or from multiple points in between.4 Oral Argument Transcript, at 48. The task of resolving all of these disputes as to the meaning of the 1982 consent decree would clearly place a severe burden on the time of the district court, and ultimately of this court on review.
Of course if this effort was necessary to implement a decree protecting federal rights, it would be appropriate for the federal courts to expend it. When a federal statute or constitutional provision has been violated, the federal courts have broad equitable power to formulate remedies in order to correct the violation, and to supervise the implementation of those remedies.5 See Swann v. Charlotte-Mecklenburg Bd. of Educ., 402 U.S. 1, 2, 91 S.Ct. 1267, 1269, *154428 L.Ed.2d 554 (1971)1 This equitable remedial power, however, is not unlimited. Milliken v. Bradley, 418 U.S. 717, 94 S.Ct. 3112, 41 L.Ed.2d 1069 (1974). There should be prudential limits, and at some point there certainly are constitutional limits, on the scope of that power. The existence of a federal statutory or constitutional issue demanding a remedy does not give a federal judge a roving commission to adjust all disputes of whatever nature involving— however tangentially — the parties and those in proximity to them. Nor can parties expand the district court’s equitable remedial power by consent. The difficult task is to set forth a principled basis for limiting the range of disputes into which the federal courts can be dragged by means of an overly broad consent decree, while at the same time protecting the power of the courts to implement broad equitable remedies where it is useful in protecting federal rights.
The problem in this case is that the matters to be adjudicated are essentially of state interest. Regardless of the outcome of the present and potential disputes in this case, the residential character of the 3400 block will not be affected. The resolution of these disputes, therefore, in no way affects any federal interests. The only reason that the disputes can be brought in federal court is that the agreement settling the original federal suit provided for continuing federal supervision. I do not believe that precious federal resources should be wasted adjudicating such matters.
My purpose in writing this concurrence is to propose an analytical framework for distinguishing such unworthy cases from the cases where broadly exercised federal equitable remedial power is desirable. My starting point is a desire to maximize the effectiveness of the federal courts in adjudicating claims and granting full relief in cases where Congress and the Constitution have granted us jurisdiction. Federal courts will often need to superintend broad equitable remedies in order to ensure that the underlying federal rights are fully vindicated. Moreover, district judges should generally possess broad discretion to determine what relief is appropriate in any particular case. It is generally impossible for Congress to predict beforehand what relief will be required in specific cases, and it is not always easier for appellate courts to anticipate future cases in deciding those before them. Because of these concerns, I want to be certain that any test limiting federal equitable remedial power does not intrude upon the ability of district courts to grant effective remedies.
On the other hand, there comes a point where superintendency over proposed relief is of only marginal value in preserving the underlying federal rights for which Congress provided a federal forum and where the time required by this superintendency reduces the ability of the federal courts to grant prompt and full relief in other cases. The district judge is generally in the best position to determine when this point is reached, and should thus be accorded substantial deference by appellate tribunals. Appellate tribunals should supervise the exercise of that discretion, however, because district court decisions in this area affect more than the judge and litigants in the individual case; decisions have institutional consequences as well. I would suggest, therefore, the following modest standard for determining the propriety of federal-court remedial orders:
A federal court may in the sound exercise of its equitable discretion retain continuing authority over implementation of a relief decree only when the court finds that:
(a) this continuing federal supervision will probably further in some significant way a federal interest identified by statute or the constitution; or
(b) continuing supervision is not likely to divert substantial resources away from claims arising under clear constitutional and statutory grants of jurisdiction; or
(c) the unavailability of such relief is likely to deter potential litigants from bringing suits over which there is properly jurisdiction in federal court.
It is important to emphasize that this test does not in any way limit the ability of *1545parties to include anything they desire in settlement agreements, and to enforce those agreements as contracts in state court.6 The test limits only the power of federal courts to approve a settlement agreement that calls for continuing federal supervision over its enforcement and implementation. Moreover, the proposed test will not encumber the power of the federal courts to grant complex remedies requiring continuing judicial supervision in areas such as desegregation and prison reform. In those areas, complex relief is clearly necessary to vindicate federal rights, and disputes concerning the implementation of the relief granted will usually grow out of the facts underlying the original lawsuit, and thus not require the court to continually delve into independent disputes unrelated to any substantial federal interest.
III.
Because the question of limits on federal remedial power was never raised by the parties or the district court, I do not decide whether, on the facts of this case, the district court abused its discretion when it approved the consent decree. However, neither can I leave the test I have proposed in a vacuum, devoid of any application to a specific fact situation. Thus, in order to flesh out my proposed test, I will briefly illustrate how it might apply here.
While I agree with Chief Judge Seitz that community participation in redevelopment and rehabilitation is generally consistent with the purposes of the NHA, much of the litigation that has arisen or is likely to arise under the consent decrees here resembles an only-somewhat more civilized version of Pilene’s basement: a scramble over who is entitled to buy houses at bargain prices from the RDA. Obviously, within broad limits, the outcome of this donnybrook does not directly involve any question of federal law.7 In addition, continuing federal supervision over implementation and enforcement of the 1982 consent decree, which is not limited to the sale of the houses to Sansom Committee designees (as the 1980 decree was), but instead contains a detailed rehabilitation plan for the houses and restrictions and covenants on the title the designees are to receive, see supra note 4, will likely involve (and indeed already has involved) a substantial amount of judicial time and energy.8
Given these two factors, and given the fact that I doubt it can be said that the absence of federal supervision of the settlement agreement would have dissuaded the plaintiffs from bringing suit in 1973, if my view of the prudential limits on federal remedial equitable power was the law of this circuit and if the district court had this test before it in 1982 when confronted with the second consent decree, it might have refused to approve that agreement, at least to the extent that it called for continuing federal supervision.9
*1546This is my perspective in hindsight, however. We are not confronted with an appeal from an order of the district court in 1984 refusing to relinquish its ongoing supervision over this lawsuit, and what I address is only the hypothetical question whether the district court abused its discretion in entering the decree. In fairness to the district court, which did not possess a crystal ball, the dispute over the meaning of the first decree was the first disagreement between the University and the San-som Committee (as distinct from disagreements among members of the Sansom Committee) since those parties had agreed in principle to a settlement of their dispute in 1976. The relevant focus for deciding whether the district court abused its discretion in granting particular relief is on the situation as it appeared to the district court at the time the relief was granted. Thus, because the district court might reasonably have concluded at that time that approval of the 1982 consent decree would bring the case to an end,10 the propriety of the court’s approval of that decree would appear to be problematic. I conclude my illustration by declaring the question to be a close one; I do not reach a conclusion because the points on which I have written were not raised by the parties or the district court.
. To underscore, I agree completely with Chief Judge Seitz that the district court did not “lose” subject matter jurisdiction in 1976, or at any other time. There still remained several federal law issues to be litigated in 1976, including a contempt action that was pending against the RDA, and plaintiffs’ claims for attorneys’ fees. In addition, because the RDA and the University were still free to change their respective minds until they actually signed a settlement agreement, one cannot fairly say that the federal claims were moot.
. See infra note 4 and accompanying text.
. In addition to the University versus Sansom Committee disputes, the district court continued to be occupied by intra-Sansom Committee litigation. Besides the Karp litigation, which continued (and continues), H. Clayton Cook, the brother of Elliot Cook, the "trustee ad litem " of the Sansom Committee, intervened in the federal action claiming that he (not Elliot Cook) was the real owner of La Terrasse Restaurant, the entity that occupied three of the houses on San-som Street and apparently that was the principle source of funds for the litigation. This matter ended up in litigation styled Cook v. Cook v. Dechert, Price & Rhoads, No. 82-2853 (E.D.Pa.1983), which was tried in March of 1983, and ultimately settled in favor of the defendants.
. It is difficult to convey adequately the extent of the detail that is encompassed by the 1982 consent decree. The decree is several hundred pages long and sets out precise specifications for everything from the type of stone and wood that can be used in rehabilitating the houses to specifications for such concerns as “acceptable letter styles” on signs, etc., acceptable forms of "flags and banners,” and methods of vermin control and refuse disposal. In addition, the agreement contains several 3 feet by 2 feet architect’s plats, and literally dozens of blue prints for every possible structural detail of the houses, including, for example, the basement entrance, the roof deck and railing, the second story deck, the "light court stair,” the “raised garden” and the front door (including the precise size and location of the postal numbers on the door). All this detail is fertile ground for future litigation between the parties and between the members of the Sansom Committee.
. Examples of far-reaching remedial orders are most often found in antitrust, labor law, securities law, prisoners’ rights, mental health, and school desegregation cases. For a list of the extraordinary forms of relief ordered by federal courts to remedy violations of the securities laws, see Farrand, Ancillary Remedies in SEC Litigation, 89 Harv.L.Rev. 1779 (1976). In this article, Mr. Farrand argues that federal remedial power should be limited by three requirements: the relief must be consistent with the goals of the statute that has been violated; the relief must be necessary to accomplish a legitimate goal of the statute; and, on balance, the benefits of granting the proposed relief must outweigh the negative consequences. Id.
. Indeed there is no reason why the parties could not consent in the decree to take disputes, or at least specified disputes, to state court.
. As I have noted above, the original federal interests — local input into redevelopment projects, and consideration of environmental consequences of redevelopment projects — were essentially settled in 1976 when the parties informed the district court that they had reached an agreement in principle not to tear down the houses.
. In addition to the disputes between the University and the Sansom Committee which have engendered three separate appeals as of this date, there are the challenges by Michael Karp which have resulted in five appeals so far, the challenges by other members of the Sansom Committee who are unhappy with the negotiated settlements, and the lawsuit by H. Clayton Cook against his brother Elliot Cook and the Dechert law firm.
. In 1980 by contrast, there was little reason for the district court to suspect that much litigation would result from the consent decree. After all, nothing at all had happened for more than four years, and the three main protagonists in the original litigation — the Sansom Committee, the RDA, and the University — were now representing to the court that they were in complete agreement on what to do with the houses. Also, the 1980 consent decree did not include the side agreement between the University and the San-som Committee, so that if the houses had been transferred to Sansom Committee designees the terms of the 1980 consent decree would have been satisfied and the district court would have been out of the case completely.
. At oral argument counsel for the University stated that: “the University believed the first time and it also believed the second time that the consent decree ended things, not that it began things." Oral Argument Transcript, at 5.