The Trustees of the University of Pennsylvania (the "University") appeal from the February 10, 1983 order of the district court denying the University's "motion to enforce" a consent decree. We have appellate jurisdiction under 28 U.S.C. Sec. 1291 (1976). This appeal will be designated as Sansom I to distinguish it from Sansom Committee v. Lynn, No. 83-1253, 735 F.2d 1552 (Sansom II ), filed contemporaneously.

This appeal is a small sample of the unbridled litigiousness that has kept these parties in court for almost two decades. The Sansom Committee (the "Committee") is an unincorporated association of residents and users of the 3400 block of Sansom Street, located in West Philadelphia, adjacent to the University of Pennsylvania. The Redevelopment Authority of the City of Philadelphia (the "Redevelopment Authority") acquired the block by condemnation almost twenty years ago and, in conjunction with the Department of Housing and Urban Development ("HUD"), planned to demolish the structures on the block. The University acquired the redevelopment rights to the block and expected to buy the properties and to build an academic building.

Within a few years, the University changed its plans and proposed to transfer its redevelopment rights to a commercial developer. The Redevelopment Authority and HUD (collectively, the "Agencies") approved this modification of the redevelopment plan. In 1973, however, the Committee brought an action in the district court against the Agencies in an effort to stop the proposed demolition and commercial redevelopment. The Committee's underlying purpose was to rehabilitate the existing townhouses and to maintain their mix of residential and low-volume commercial uses.

The Committee alleged that the Agencies violated the National Environmental Policy Act of 1969, 42 U.S.C. Secs. 4321 to 4361 (1976) ("NEPA"), and the National Housing Act of 1949, 42 U.S.C. Secs. 1441 to 1490h (1976) ("NHA"), when they approved the modification of the redevelopment plan. Certain of the Committee's claims were dismissed on motion by the Agencies. The Agencies also moved to dismiss the action for failure to join an indispensable party, i.e., the University. The district court held that although the University had an interest in the outcome of the action, it was not an indispensable party. See Sansom Committee, 366 F. Supp. 1271, 1281 (E.D.Pa.1973). The district court denied the University's subsequent petitions to intervene in the action.

After several further confrontations in the district court,1 the Committee and the Agencies stipulated that they had "engaged in serious and fruitful settlement negotiations in consultation with the University of Pennsylvania" and that these discussions had led to a new proposal for the redevelopment of the block. The district court agreed to suspend the action while the parties finalized a settlement.

The parties settled the action, and the district court entered a consent decree (the "1980 Consent Decree") which contained the terms of the settlement agreement. Under the 1980 Consent Decree, the Redevelopment Authority was to sell Sansom Street properties to parties designated by the Committee, and the University agreed to nominate the designated parties to redevelop those properties in the University's stead. The Committee, the University, and the Redevelopment Authority signed the 1980 Consent Decree, and HUD consented to its entry.

Within months, the decree "unravelled," as the University says. The University claims that both it and the Committee wanted to make changes in the decree. The Committee claims that the University's "stalling tactics" inspired the Committee to move for enforcement of the decree. Amidst this unravelling, the district court in March 1982 entered an order that named the Committee's designees. After further negotiations, the University and the Committee reached an agreement that included newly negotiated covenants, the 1980 Consent Decree, and other agreements between the University and the Committee. These matters were incorporated in a new consent decree (the "1982 Consent Decree"), which the Committee and the University signed. The Redevelopment Authority consented to the entry of the 1982 Consent Decree.

Subsequently, the Committee advised the University that some of its designees wished to withdraw from the redevelopment project. The Committee proposed replacement designees. In response, the University filed a "motion to enforce" the 1982 Consent Decree. In this motion, the University requested the court to substitute the University as the redeveloper of the properties that the withdrawing designees were to have received. The district court denied this motion, and the University appeals.

The University raises for the first time on appeal the contention that the district court lacked subject matter jurisdiction2 to enter the 1980 Consent Decree. Although this is an appeal from the denial of the University's motion to enforce the 1982 Consent Decree, we may consider the University's jurisdictional challenge to the 1980 Consent Decree because the 1982 Consent Decree is essentially a modification of the 1980 Consent Decree,3 and the validity of the 1982 Consent Decree depends on the validity of the 1980 Consent Decree. It is not contested that we must address this issue even though the University consented to the entry of the decrees.

The district court clearly had federal question jurisdiction in 1973, at the outset of the action between the Committee and the Agencies based on the federal statutes invoked. The University contends that the district court lacked federal question jurisdiction in 1976 because HUD filed an environmental impact statement dated in that year. Even if the environmental impact statement satisfied HUD's responsibilities under the NEPA, it left intact the Committee's pending federal claims against the Agencies under the NHA. Consequently, the district court still had subject matter jurisdiction when it considered the entry of the 1980 Consent Decree.

The University's principal argument is that the district court had no subject matter jurisdiction to enter the 1980 Consent Decree because its terms incorporated essentially state law relief. More generally, the issue is whether there are jurisdictional limitations to what a district court may incorporate in a consent decree, and if so, whether the district court transgressed those limits.4

Consent decrees need not be limited to the relief that a court could provide on the merits. As the Supreme Court long ago explained:

Parties to a suit have the right to agree to any thing they please in reference to the subject-matter of their litigation, and the court, when applied to, will ordinarily give effect to their agreement, if it comes within the general scope of the case made by the pleadings.

Pacific Railroad v. Ketchum, 101 U.S. 289, 297, 25 L. Ed. 932 (1879); see Citizens for a Better Environment v. Gorsuch, 718 F.2d 1117 (D.C.Cir.1983); EEOC v. Safeway Stores, Inc., 611 F.2d 795, 799-800 (10th Cir.1979), cert. denied sub nom. Courtwright v. EEOC, 446 U.S. 952, 100 S. Ct. 2918, 64 L. Ed. 2d 809 (1980); 1 Hogg's Equity Procedure Sec. 582 at 707 (3d ed. 1943); 3 Freeman on Judgments Sec. 1349 at 2772 (1915); 2 Black on Judgments Sec. 705 at 843 (1891) (citing cases for and against); cf. Alliance to End Repression v. City of Chicago, 733 F.2d 1187, 1191 (7th Cir.1984); Swift & Company v. United States, 276 U.S. 311, 329-31, 48 S. Ct. 311, 316-317, 72 L. Ed. 2d 587 (1928).

We recognize that some consent decrees may be beyond the power of a federal court to approve. See Safeway Stores, 611 F.2d at 795; Jordan v. School District, 615 F.2d 85, 91 (3d Cir.1980) (Rosenn, J., concurring). Thus, a district court cannot wield its equitable power beyond the realm of its federal subject matter jurisdiction. E.g., Gordon v. Washington, 295 U.S. 30, 36, 55 S. Ct. 584, 587, 79 L. Ed. 1282 (1935); In re "Agent Orange" Product Liability Litigation, 506 F. Supp. 737, 740 (E.D.N.Y.1979), rev'd on other grounds, 635 F.2d 987 (2d Cir.1980), cert. denied sub nom. Chapman v. Dow Chemical Company, 454 U.S. 1128, 102 S. Ct. 980, 71 L. Ed. 2d 116 (1981). Nevertheless, as long as the terms of a consent decree come "within the general scope of the case made by the pleadings," Pacific Railroad, supra, it will be within the district court's power to enter the decree, if the pleadings state a claim over which a federal court has jurisdiction.

In addition to the claims that the district court dismissed, the Committee alleged that the Agencies violated the NEPA by failing to issue an environmental impact statement and by failing to hold a hearing on the possibility of rehabilitation; that the Agencies violated the NHA by permitting non-public contributions; and that the Agencies violated the NHA by failing to provide citizen participation in the redevelopment, hold public hearings and seek the approval of local governing bodies, insure conformity with community and regional plans, and provide adequate relocation. Sansom Committee v. Lynn, 366 F. Supp. at 1281.5

The terms of the 1980 Consent Decree set out in detail a cooperative plan for the rehabilitation and use of the Sansom Street properties. Although the terms of the decree far exceeded the relief available under the NEPA and the NHA,6 the decree was directly responsive to the Committee's complaint under these statutes. Thus, under the 1980 Consent Decree, the properties would be rehabilitated with maximal participation by interested members of the public. This is consonant with the general policies and goals of the NEPA, see 42 U.S.C. Secs. 4321, 4331; Sansom Committee, 366 F. Supp. at 1274 (purpose of the NEPA is to require federal agencies to consider local environmental consequences of their projects), and of the NHA, 42 U.S.C. Secs. 1441, 1441a (the NHA encourages rehabilitation and community participation in redevelopment); see Shannon v. HUD, 436 F.2d 809, 818 (3d Cir.1970) (residents, business owners, and representatives of private civic groups have standing under the NHA to challenge the modification of a redevelopment plan). We conclude that the terms of the 1980 Consent Decree were sufficiently related to the Committee's federal causes of action to permit the district court to enter the decree without overstepping its subject matter jurisdiction.

The University also argues that the district court had no power to enter a consent decree signed by a non-party to the underlying action.7 Since, in our view, the decree met the requirements of Pacific Railroad, we fail to see how this argument raises a question of subject matter jurisdiction. Finally, the University challenges the district court's power to construe, modify, or enforce the consent decrees. These arguments, however, are predicated on the court's supposed lack of jurisdiction to enter the 1980 Consent Decree, which we reject.

We must now decide whether the district court correctly declined to substitute the University for the Committee's withdrawing designees. Consent decrees are construed as contracts. Fox v. HUD, 680 F.2d 315, 319 (3d Cir.1982). If a district court's construction of a contract involves no factual issues, our standard of review is plenary, but findings of fact must be left intact unless they are clearly erroneous. See Barco Urban Renewal Corporation v. Housing Authority, 674 F.2d 1001, 1008 (3d Cir.1982).

Neither the 1980 Consent Decree nor the 1982 Consent Decree addresses the question of the replacement of withdrawing designees. The district court held that a provision that empowers the University to purchase a property if "no designee of the Sansom Committee shall elect to purchase [that] property" also implies that the Committee may designate more than one person to purchase a property. The court also held that nothing in the 1982 Consent Decree refers to the specific designees named in the court's order of March 1982. Finally, the district court viewed the decree in light of the court's experience with its drafters and concluded that if the drafters had intended that the University be entitled to purchase a property after an initial designee withdraws, the decree would have so provided. We agree with the district court's construction of the 1982 Consent Decree.

The district court's order dated February 10, 1983, denying the University's "motion to enforce" the consent decree, will be affirmed.8 Costs will be taxed against the appellant.

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 district 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.

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 overexpanded 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. Secs. 4321-4361 (1976) ("NEPA"), and the National Housing Act of 1949, 42 U.S.C. Secs. 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 court 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 Sansom 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.3II.

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, 28 L. Ed. 2d 554 (1971). 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 parties 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.

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 Filene'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

This 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 Sansom 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.

Chief Judge Seitz's attempt, laudable in itself, to enable the parties involved in a complex controversy to forge a lasting and equitable peace, stands on untenable footing. The limited jurisdiction that federal courts possess does not encompass the enforcement of consent judgments against or in favor of parties whose rights the court had no power to adjudicate in an involuntary proceeding. Because a majority of this court holds otherwise, I respectfully dissent.

The appeal taken by the University of Pennsylvania in this case asserts that the district court had no subject matter jurisdiction to adjudicate any dispute between the University and the Committee "or to interpret, enforce, or modify agreements between them." (Br. 19). Chief Judge Seitz reads the University's position as disclaiming federal question jurisdiction as of 1976: "The University contends that the district court lacked federal question jurisdiction in 1976 because HUD filed an environmental impact statement dated in that year." (Maj. op. at 6). He also understands the University's argument to be that "the district court had no subject matter jurisdiction to enter the 1980 Consent Decree because its terms incorporated essentially state law relief." (Id.) While it is true that both arguments can be characterized as Chief Judge Seitz has reported them, they are in substance no more than subordinate elements of the University's overall contention that no federal question giving the district court jurisdiction has ever been raised as to the University, and that the consent decrees cannot be enforced with respect to the University because they are outside the court's jurisdictional power. The thrust of the University's position is captured by the following three statements which appear in its brief:

Neither the Sansom Committee nor any of the parties has ever raised a federal question vis-a-vis the University, nor has a federal question of any kind been raised since the University became a party to the settlement of this litigation.

* * *

* * *

The only conceivable explanation for federal court adjudication of disputes of this nature would be premised on an extension of the concepts of pendent or ancillary jurisdiction, i.e., that the University could properly be joined as a pendent party in January of 1981 in order to resolve the pre-existing litigation. No such extension of these doctrines can be justified.

(Br. of University at 20, 21 (footnote omitted)). The University concludes its jurisdictional argument stating:

The present disputes, all of which involve interpretations of agreements between the Sansom Committee and the University (or efforts to modify the terms of such agreements), cannot be made a subject of that Court's enforcement power if they are outside its jurisdictional power.

(Br. of University at 26 (footnote omitted)). I agree.

A consent judgment is a judicial act, in which the court adjudicates the plaintiff's right of recovery and the extent of it, both of which are essential elements of the judgment. Pope v. United States, 323 U.S. 1, 12, 65 S. Ct. 16, 22, 89 L. Ed. 3 (1944); see United States v. Swift & Co., 286 U.S. 106, 115, 52 S. Ct. 460, 462, 76 L. Ed. 999 (1932). The court does not determine the merits of the dispute, but its entry of a decree constitutes an adjudication between the parties. Swift & Co. v. United States, 276 U.S. 311, 327, 48 S. Ct. 311, 315, 72 L. Ed. 587 (1928).

A court's authority to enter a consent decree where the plaintiff seeks to enforce a federal statute comes only from the statute that the decree is intended to enforce. System Federation No. 91 v. Wright, 364 U.S. 642, 651, 81 S. Ct. 368, 373, 5 L. Ed. 2d 349 (1961). Within the confines of this authority a court has broad power to adopt as its own judgment or decree a contractual agreement into which the parties have entered and which is tailored to the necessities of a particular case. Handler v. SEC, 610 F.2d 656, 659 (9th Cir.1959). The guiding principle in adopting a decree is that it further the purpose of the statute to be enforced. System Federation, supra, 364 U.S. at 651, 81 S. Ct. at 373; Citizens for a Better Environment v. Gorsuch, 718 F.2d 1117, 1125 (D.C.Cir.1983); United States v. Motor Vehicle Mfrs. Ass'n, 643 F.2d 644, 650 (9th Cir.1981).

Indeed, the consent decree may even prescribe relief beyond that authorized by the statute. Swift & Co. v. United States, 276 U.S. 311, 328-30, 48 S. Ct. 311, 315-316, 72 L. Ed. 587 (1928). But there is a crucial distinction between an attack on the scope of relief contained in the decree and a contention that the court was without jurisdiction to adopt the decree in the first instance: between an "error of decision," which does not render a consent judgment void, and the "want of power to decide," which renders any judgment entered void and open to collateral attack. Id. at 330, 48 S.Ct. at 316. The court's power to grant relief broader than that authorized by the statute only exists if it has acquired jurisdiction of the subject matter and the parties. Id. at 326, 48 S.Ct. at 315. Persons cannot by consent give the court jurisdiction over the subject matter, Neirbo v. Bethlehem Corp., 308 U.S. 165, 167, 60 S. Ct. 153, 154, 84 L. Ed. 167 (1939); Rickerson v. Jones, 515 F.2d 918, 923 n. 7 (3d Cir.1977); see Insurance Corp. of Ireland v. Compagnie des Bauxites de Guinee, 456 U.S. 694, 702, 102 S. Ct. 2099, 2104, 72 L. Ed. 2d 492 (1982), although they may be consenting to judgment bind their persons to its jurisdiction, "if when the court acts jurisdiction has been obtained." Pacific R.R. v. Ketchum, 101 U.S. 289, 198, 25 L. Ed. 932 (1879). Moreover, this admonition exists side-by-side with the rule that "[p]arties to a suit have the right to agree to anything they please in reference to the subject matter of their litigation, and the court, when applied to, will ordinarily give effect to their agreement, if it comes within the general scope of the case made by the pleadings." Id. at 297, 25 L. Ed. 932.

The majority's disposition of this case ignores the fundamental distinction made in the seminal consent decree cases. A court's power to enter a consent decree flows from the same font of jurisdiction as does its power to enter an involuntary decree (that is, a decree which is not the product of the parties' consent). That power exists only if the court has the initial authority to adjudicate the rights of those parties whom it binds to its judgment. As earlier noted, that power cannot be conferred by consent; it is given by the Constitution and implementing statutes as enacted by Congress. Nor does it avail a party to argue that another's "consent" to the subject matter jurisdiction of a court can then estop the "consenting party" from challenging the federal court's jurisdiction. Rubin v. Buckman, 727 F.2d 71 at 72 (3d Cir.1984). Thus, unless a court has subject matter jurisdiction over the entire dispute between the parties who have consented to enter into a "consent agreement," it cannot adopt that agreement as a "judicial act" and thereby dignify it by the judge's signature as a consent judgment.1

Accordingly, the cases holding that a consent decree can afford further relief affecting a wider scope of activities than could relief authorized by the statute sued upon, e.g., Swift & Co. v. United States, 276 U.S. 311, 328-30, 48 S. Ct. 311, 315-316, 72 L. Ed. 587 (1928); Citizens for a Better Environment v. Gorsuch, 718 F.2d 1117, 1174-77 (D.C.Cir.1983); Larsen v. Sielaff, 702 F.2d 116, 117-19 (7th Cir.), cert. denied, --- U.S. ----, 104 S. Ct. 372, 78 L. Ed. 2d 330 (1983), are of no relevance in our consideration of the power of the district court to adjudicate the rights of the University of Pennsylvania in the instant controversy. Rather, the question on which we must focus at the outset is whether the district court in the first instance had the power to enter judgment against a nonparty to the litigation, who nevertheless was concededly a party to the consent agreement which the plaintiff Committee now seeks to enforce.2

In general a court has no jurisdiction to determine the rights of nonparties to the litigation. SEC v. Investors Security Corp., 560 F.2d 561, 568 (3d Cir.1977).3

Where a court would be without power to enter judgment against a person because that person could not be made a party to the litigation, the court cannot bind that person to a consent decree. Washington v. Penwell, 700 F.2d 570, 574 (9th Cir.1983).4 See also Metropolitan Housing Devel. Corp. v. Village of Arlington Heights, 469 F. Supp. 836, 854-55 (N.D.Ill.1979) (consent does not empower court to order relief which it would not otherwise have the power to order and which the parties could not themselves otherwise contract to perform), aff'd, 616 F.2d 1006 (7th Cir.1980).

In this case, the University was never a party to the litigation. Twice the University moved to intervene; twice it was rebuffed.5 Indeed, the motion to dismiss for failure to join an indispensable party (the University), which was made by the defendants in the original action, was denied. Sansom Committee v. Lynn, 366 F. Supp. 1271, 1280-81 (E.D.Pa.1973). The "settlement agreement and consent decree" made on December 18, 1980, by and among the Committee, the RDA, and the University, explicitly stated that the University "is not a party to this litigation but is a party to the settlement of this litigation." (App. A-55).

Moreover, the University could not have been made a party to the litigation, for the district court would have had no basis, short of formal intervention, for exercising subject matter jurisdiction over the Committee's claim against the University. As is contended by the University and not disputed by the Committee, there would be neither diversity nor federal question jurisdiction over such a claim. See University's Br. at 17. As both agree, there was federal question jurisdiction over the Committee's main claims against HUD, the RDA, and their respective officers, based on the National Housing Act and NEPA. Sansom Committee v. Lynn, 366 F. Supp. 1271, 1278 (E.D.Pa.1973). It is on that basis, and that basis alone, that the Committee asserts that the district court had ancillary or pendent-party jurisdiction over any dispute between the Committee and the University, by virtue of the "core" federal claim.

Whether a court has jurisdiction over a non-federal claim between non-diverse parties is determined by a two-stage analysis, as set forth in Owen Equipment Co. v. Kroger, 437 U.S. 365, 98 S. Ct. 2396, 57 L. Ed. 2d 274 (1978), and Aldinger v. Howard, 427 U.S. 1, 96 S. Ct. 2413, 49 L. Ed. 2d 276 (1976). First, there must be an inquiry into whether the Constitution gives the court power to adjudicate the dispute, for which the test is whether the claims present a "common nucleus of operative fact ... such that [a plaintiff] would ordinarily be expected to try them all in one judicial proceeding." United Mine Workers v. Gibbs, 383 U.S. 715, 725, 86 S. Ct. 1130, 1138, 16 L. Ed. 2d 218 (1966). Second,

there must be an examination of the posture in which the nonfederal claim is asserted and of the specific statute that confers jurisdiction over the federal claim, in order to determine whether "Congress in [that statute] has ... expressly or by implication negated" the exercise of jurisdiction over the particular nonfederal claim.

Owen Equipment, supra, 437 U.S. at 373, 98 S. Ct. at 2402 (quoting Aldinger, supra, 427 U.S. at 18, 96 S. Ct. at 2422). In making this second inquiry, "the context in which the nonfederal claim is asserted is crucial." Id. at 376, 98 S.Ct. at 2404. The Owen Court noted the typical instances in which ancillary jurisdiction has been held to exist--claims by third parties who were brought into court by the original parties, or claims of person interested in a fund or property in the possession of the court. 437 U.S. at 375-76 & n. 18, 98 S. Ct. at 2403-2404 & n. 18. Such claims included intervention as of right. Id. at 375 n. 18, 98 S. Ct. at 2403 n. 18.

Before Owen and Aldinger, it had been held that a federal court has jurisdiction--ancillary jurisdiction--over actions to aid or effectuate its prior decrees, regardless of whether the court would have jurisdiction over the claim were it an original action. Dugas v. American Surety Co., 300 U.S. 414, 428, 57 S. Ct. 515, 521, 81 L. Ed. 720 (1937); Local Loan Co. v. Hunt, 292 U.S. 234, 239, 54 S. Ct. 695, 697, 78 L. Ed. 1230 (1934); Root v. Woolworth, 150 U.S. 401, 410-12, 14 S. Ct. 136, 138-139, 37 L. Ed. 1123 (1893). This rule was "subject to the qualification that the relief be not of a different kind or on a different principle." Dugas, supra, 300 U.S. at 428, 57 S. Ct. at 521.

This line of cases, however--cases which are limited to instances in which a court asserts jurisdiction to aid or effectuate its earlier decree--cannot support the exercise of jurisdiction to adjudicate the rights of the Committee vis-a-vis the University, where, as here, there is no prior judgment that the 1980 "consent decree" is necessary to effectuate. Instead, the 1980 decree (and by extension, the 1982 decree) is the very judgment sought to be effectuated by the adjudication of the University's rights via the same judgment. Such bootstrapping of jurisdiction extends the Dugas-Local Loan-Root line beyond its logical parameters. Thus, even if a district court would have ancillary jurisdiction to adjudicate the rights of third-party nonlitigants who are interfering with a pre-existing consent decree, see New York State Ass'n for Retarded Children, Inc. v. Carey (NYSARC II), 466 F. Supp. 479, 482 (E.D.N.Y.1978), aff'd, 612 F.2d 644 (2d Cir.1979); New York State Ass'n for Retarded Children, Inc. v. Carey (NYSARC I), 438 F. Supp. 440, 446-47 (E.D.N.Y.1977) (alternative holding); cf. Local Loan, supra (ancillary jurisdiction over nonparty interfering with involuntary decree); but see Sea Ranch Ass'n v. California Coastal Comm'n, 552 F. Supp. 241, 247-48 n. 27 (N.D.Cal.1982) (questioning NYSARC I & II in the wake of Owen Equipment ); C. Wright & A. Miller, Federal Practice & Procedure Sec. 3523 (1980 Supp.) (same), it does not follow that a federal district court would have subject matter jurisdiction to create a consent decree in a controversy such as the one which is presented here, where no pre-existing judgment was ever entered.

Furthermore, in light of Owen and Aldinger such an extension would be improper. It is difficult to identify what a court is to examine in applying the Gibbs "common nucleus" test here, since the Committee never asserted a claim against the University nor identified one in either agreement, nor did the University assert any claims. All that the University declared in the 1980 agreement was that it "alleges it has rights under a redevelopment contract affecting the herein named properties." (App. A-55). The Committee's federal claims alleged that the redevelopment plans violated federal law. It is true that the redevelopment plans at issue involved the University to the extent that the plans contemplated the University's acquisition of redevelopment rights. Thus there could appear to be a "common nucleus of operative facts" underlying these two sets of interests--one set asserted as a reason for joining the settlement agreement and one set asserted in an adversarial claim. The University's interest in adjudicating the federal claims was a practical one; invalidation of the plan would implicate its rights under the contract. Its rights were thus dependent, at least in part, upon the disposition of the federal claims.

It remains whether this interest is sufficient under the Gibbs test to constitute a claim that ordinarily would be expected to be tried in the same proceeding. This question need not be answered if the statutory phase of the analysis reveals that Congress did not intend that federal jurisdiction include it.

The statute that confers jurisdiction on district courts to hear federal claims is 28 U.S.C. Sec. 1331. That section allows courts to adjudicate "all civil actions arising under the Constitution, laws, or treaties of the United States." In construing the diversity-jurisdiction statute (28 U.S.C. Sec. 1332), the Court in Owen Equipment, supra, stated that claims by third parties "typically" support ancillary jurisdiction when such a person's "rights might be irretrievably lost unless he could assert them in an ongoing action in a federal court." 437 U.S. at 376, 98 S. Ct. at 2404. The claims include compulsory counterclaims, impleader, cross-claims, and intervention as of right. Id. at 375 n. 18, 98 S. Ct. at 2403 n. 18. However, the same considerations that informed the Supreme Court's construction of the diversity jurisdiction statute also counsel against expansive pendent-party federal question jurisdiction. For, as the Court noted in Aldinger v. Howard, supra, 427 U.S. at 17, 96 S. Ct. at 2421, when considering a claim of "pendent party" jurisdiction,6 of which the claim here is a species, "[T]he reach of the statute conferring jurisdiction should be construed in light of the scope of the cause of action as to which federal judicial power has been extended by Congress."

The federal cause of action here as pleaded in the plaintiff's complaint was a challenge to the performance by public agencies of their duties under the National Housing Act and the National Environmental Policy Act. The University was involved, if at all, only as the named redeveloper under its contract with the RDA (it was the only redeveloper after 1973). Thus the University's rights and duties were purely contractual, as stipulated in the settlement agreement, see supra. The Committee has not pointed to any provision of the National Housing Act or National Environmental Policy Act that indicates that Congress intended that the adjudication of the duties of public agencies under these statutes be affected by the purely contractual rights of third-party redevelopers.7

The implication of these considerations becomes even more clear when we examine whether Congress ever intended that jurisdiction extend to claims involving persons who were never made party to the litigation. As has been stressed several times so far, the University was never a party to the litigation; its motions to intervene (both as-of-right and by permission) as a defendant were twice denied; it was held not indispensable; and the very consent agreement that purported to "confer" jurisdiction recognized by its own terms that the University "is not a party to this litigation."

Moreover, the statement of the University's interest in, that 1980 agreement does not support jurisdiction premised on a consented-to intervention. It may be that parties to a consent judgment could stipulate facts showing that a party was an intervenor of right under Fed.R.Civ.P. 24(a)--either that a federal statute confers an unconditional right to intervene, see E.E.O.C. v. AT & T, 506 F.2d 735, 739-40 (3d Cir.1974) (requirement for intervention of right under Rule 24(a)(1)), or that the party claims an interest relating to the property or transaction that is the subject of the action, that the disposition may as a practical matter impair his ability to protect that interest, and that his interests are not adequately represented by others, see Pennsylvania v. Rizzo, 530 F.2d 501, 504 (3d Cir.) (requirements for establishing right to intervene under 24(a)(2)), cert. denied, 426 U.S. 921, 96 S. Ct. 2628, 49 L. Ed. 2d 375 (1976). Intervention as of right needs no independent federal jurisdictional grounds. See Owen, 437 U.S. at 375 n. 18, 98 S. Ct. at 2403 n. 18; C. Wright & A. Miller, Federal Practice & Procedure Sec. 1917 (1972). However, the stipulation of facts contained in the 1980 agreement is plainly insufficient to support a holding of a right to intervene, much less consent to intervention by the other parties.

Thus, it is clear that no subject matter jurisdiction and no ancillary or pendent-party jurisdiction exists with respect to any claim made by the Committee which implicates the University of Pennsylvania in the present controversy. Accordingly, the district court was without power to adjudicate any rights affecting the University of Pennsylvania and could not, therefore, exercise its power by adding its signature or imprimatur to a purely contractual agreement between these parties. Instead, as noted, the parties' rights under the consent agreements of 1980 and 1982 are contractual only, and thus may only be enforced as such. See Note, The Consent Judgment as an Instrument of Compromise and Settlement, 72 Harv.L.Rev. 1314, 1322 (1959), and note 3 supra.

Because the majority of the court has given effect to a judgment which the district court had no power to enter, I respectfully dissent.