Employers Insurance of Wausau v. Shell Oil Company

EASTERBROOK, Circuit Judge. Shell Oil Co. and one of its insurers are litigating in two fora. In a state court of California Shell has sued about 250 firms that insured it (and one of its subsidiaries) between 1940 and 1983. Shell wants a declaratory judgment ascertaining the insurers’ responsibility to indemnify it on account of any liability it may incur as a result of pollution at two sites, one in California and the other in Colorado. The United States and the State of Colorado sued Shell, shortly after it filed the declaratory judgment action in October 1983, seeking more than $1.8 billion as the cost of cleaning up the Colorado site alone. *899 Shell’s California action named 250 “Doe” defendants, representing insurers the identities of which were undiscovered. In June 1986 Shell learned that Employers Insurance of Wausau may have insured its subsidiary in the early 1950s. Travelers Insurance Co., one of the defendants in Shell’s suit, promptly filed a cross-claim against Wausau in the California action. In November 1986, just before Shell replaced one of the “Does” in the California action with Wausau, Wausau filed a diversity action against Shell in the Northern District of Illinois. Wausau’s complaint asks the district judge to determine that its policies, in force between 1950 and 1953, do not cover the pollution about which Shell is concerned. Wausau also contends that it was not added to the California suit in a timely fashion. Shell represents that the California court has rejected this contention, finding it frivolous. Because Wausau’s contentions in the district court also are before the court in California, Shell asked the district court to stay its proceedings. The district judge declined, concluding that the “virtually unflagging obligation” of a federal court to exercise its own jurisdiction, Colorado River Water Conservation District v. United States, 424 U.S. 800, 817, 96 S.Ct. 1236, 1246, 47 L.Ed.2d 483 (1976), required it to proceed. The district court observed that Wausau’s policies have been before it and the California court for about an equal length of time, and that there has been little discovery in California concerning Wausau. Thinking the progress of the litigation with respect to Wausau functionally identical in each court, and concluding that “a substantial question exists as to whether Wausau is a proper party in the California action with regard to Shell”, the district judge held that the stringent standards for dismissal announced in Colorado River had not been satisfied. See also Moses H. Cone Memorial Hospital v. Mercury Construction Corp., 460 U.S. 1, 16, 103 S.Ct. 927, 937, 74 L.Ed.2d 765 (1983) (treating Colorar do River’s standards as applicable to stays that effectively conclude the federal case); Illinois Bell Telephone Co. v. Illinois Commerce Commission, 740 F.2d 566, 569 (7th Cir.1984). At the moment, discovery is under way in both cases. The California court has scheduled a trial for October 1987; the Illinois court has not set a schedule for proceedings, but its opinion stated that it perceived an “overwhelming likelihood [that] this case will be resolved by summary judgment”. The district court’s order, 653 F.Supp. 744, declining to stay its proceedings was entered on February 11, 1987; two days later Shell filed a notice of appeal. We promptly asked the parties to address the question of appellate jurisdiction. After receiving their memoranda, we deferred further consideration of jurisdiction until the hearing of the case on the merits. Shell then again asked the district court for a stay. On April 7, 1987, the district court again declined to stay its hand, this time stating: “this court thinks it is unlikely the 7th Circuit will entertain [Shell’s] appeal. Even if the appellate court were to reach the merits of [Shell’s] appeal, this court does not believe [Shell] is likely to prevail.” On April 16, Shell asked this court for a stay pending appeal. Because we cannot hear oral argument on Shell’s appeal until September 1987, and the trial in California is scheduled for October, a stay might have the practical effect of dismissing the case. We could not issue such a stay without being convinced that Shell is likely to prevail on appeal. That reopened the question of appellate jurisdiction, for if there is none, Shell cannot prevail. After reviewing the initial round of jurisdictional memoranda, the memoranda exchanged on the request for a stay, and Shell’s opening brief, we conclude that Shell’s appeal is premature. We dismiss the appeal for want of jurisdiction, which moots the request for a stay. A court’s refusal to terminate or stay its proceedings is the opposite of a “final decision” appealable under 28 U.S.C. § 1291. Proceeding toward decision is not itself a final decision. The process of decision may be costly, but by that argument the filing of a complaint would be the “final decision” allowing an immediate appeal to stave off the costs of litigation. The *900 Supreme Court has repeatedly rejected claims that the expense of litigation allows appellate review under § 1291 in advance of the termination of the case. E.g., Stringfellow v. Concerned Neighbors in Action, — U.S.-, 107 S.Ct. 1177, 1182-83, 94 L.Ed.2d 389 (1987); RichardsonMerrell Inc. v. Roller, 472 U.S. 424, 105 S.Ct. 2757, 86 L.Ed.2d 340 (1985); Rerr v. United States District Court, 426 U.S. 394, 96 S.Ct. 2119, 48 L.Ed.2d 725 (1976); Switzerland Cheese Ass’n, Inc. v. E. Home’s Market, Inc., 385 U.S. 23, 87 S.Ct. 193, 17 L.Ed.2d 23 (1966). Cf. FTC v. Standard Oil Co., 449 U.S. 232, 243-46, 101 S.Ct. 488, 494-95, 66 L.Ed.2d 416 (1980). Decisions by a district court to proceed to judgment, in particular, are not appealable. City of Morgantown v. Royal Insurance Co., 337 U.S. 254, 69 S.Ct. 1067, 93 L.Ed. 1347 (1949); see also, e.g., Tenneco Inc. v. Saxony Bar & Tube, Inc., 776 F.2d 1375 (7th Cir.1985); Kapco Mfg. Co. v. C & O Enterprises, Inc., 773 F.2d 151 (7th Cir.1985). This litigation is just beginning, therefore it is not over, therefore no one may appeal. Yet the question of jurisdiction has been complicated by the Enelow-Ettelson doctrine, named after Enelow v. New York Life Insurance Co., 293 U.S. 379, 55 S.Ct. 310, 79 L.Ed. 440 (1935), and 1Hudson v. Metropolitan Life Insurance Co., 317 U.S. 188, 63 S.Ct. 163, 87 L.Ed. 176 (1942). The doctrine holds that the grant or denial of an equitable stay in an action at law is the grant or denial of a preliminary injunction, making it appealable under 28 U.S.C. § 1292(a)(1). The idea is that the chancellor is “enjoining” (or refusing to enjoin) the judge on the basis of the “equitable” ground; that the chancellor and the judge may be the same person is treated as an irrelevant detail. When the underlying action is “equitable”, however, the whole case would have been before the chancellor, so that his refusal to stay the proceedings is more like a non-appealable docket control order than like an injunction issued by one court to interrupt the proceedings of another. Just in case it might be too easy to decide what is “legal” and what is “equitable”, the Supreme Court specified that the determination must reflect the law as it stood in 1891, when Congress passed the Evarts Act. The result is a profusion of irrelevant distinctions and nonfunctional results. The reaction of appellate judges to this doctrine runs the gamut from bewilderment to exasperation; few favorable words have been uttered by judge or commentator in the last 40 years. See Olson v. Paine, Webber, Jackson & Curtis, Inc., 806 F.2d 731, 739-41 (7th Cir.1986) (collecting criticisms). But neither Congress nor the Supreme Court has done anything, so we must apply the doctrine as best we can — which is not very well, given the frequency with which questions such as “would this action have been thought legal or equitable in 1891?” have no answer. Sometimes the answer is lost in the mists of time because for 49 years (since the merger of law and equity under the Federal Rules of Civil Procedure) the distinction between “law” and “equity” has been irrelevant in most federal litigation. Sometimes there is no answer to be found, because the kind of action in issue could not have been filed in any court in 1891. Sometimes the answer is meaningless because it assumes a dichotomy (law or equity) that would not have been recognized even in 1891; some actions were a little of each. Our case illustrates the problem. The question on the table in both California and Illinois is whether Wausau must defend and indemnify Shell in litigation that has been and may be filed concerning the cleanup of chemical contamination at two sites. The California litigation, which Wausau seeks to preempt, also raises questions concerning apportionment among insurers. But because the California litigation was filed in advance of any suit or judgment against Shell, neither suit is a claim for breach of contract, a “legal” action in 1891 as it is today. Shell’s suit in California, and Wausau’s suit in Illinois, seek declaratory judgments establishing the scope of the insurance. Declaratory judgments are neither “legal” nor “equitable” and so do not fit the categories of Enelow-Ettelson. *901 Aetna Casualty & Surety Co. v. Quarles, 92 F.2d 321, 324 (4th Cir.1937); American Safety Equipment Corp. v. J.P. Maguire & Co., 391 F.2d 821, 824 (2d Cir.1968). They were rare in the states in 1891 and unknown in federal practice. To tell how to classify these beasts, we must evaluate the underlying claim. Mowbray v. Moseley, Hallgarten, Estabrook & Weeden, 795 F.2d 1111, 1113-15 (1st Cir.1986). This may mean looking through the Illinois case to the California case, which establishes the bounds of the controversy. Worse, since declaratory judgments often reverse the positions of plaintiff and defendant, we must decide how the claim would have been classified in 1891 had the parties been on the “right” side. This is implicit in the requirement that we ascertain what kind of suit could have been brought. Diematic Manufacturing Corp. v. Packaging Industries, Inc., 516 F.2d 975, 978 (2d Cir.1975). Yet the claim in our case, for a declaration of obligations among many insurers, was unknown in 1891 in the absence of a breach of contract. See 18 Couch on Insurance 2d § 74:22 (1983). Wausau’s mirror image action also was unknown; Wausau does not contend that Shell has broken the contract or that Wausau needs assurances against anticipatory repudiation; it does not seek cancellation or reformation of its contractual obligations; the parties agree that Wausau would have had no action of any kind in any court in 1891. All of this produces a conundrum. Who would have been the “natural” plaintiff, and what kind of relief would he have sought, in an action no one could have filed in 1891? On the answer to this question turns our jurisdiction — but the question has no answer. Although Shell could not have commenced the California action in 1891 until someone had sued it, making possible a claim for breach of the contract of insurance, two such suits have been filed. Shell states that it has tendered the defense of the suit to each of its insurers, including Wausau. That would set up suits for breach, plainly “at law”. But Shell is not satisfied with serial actions for breach; it wants a court to decide the full scope of coverage and relative responsibilities of each insurer. It could not obtain such relief in any court in 1891 at the stage it filed its California case — that is, before it had a claim for breach. If it had tried to get that relief in 1891 after acquiring a claim for breach, it would have had to turn to equity — not because of the nature of the underlying claim but because of the joinder rules. The law courts would not have allowed Shell to sue, in a single action, unrelated insurers who had no contractual dealings with each other and may have had contracts with Shell separated by 40 years. Shell would have had to sue each insurer separately for breach of contract, creating not only a multiplicity of actions but also a significant chance of inconsistent decisions. To avoid these unhappy outcomes, equity courts would have assumed jurisdiction. Couch at § 74:23. So Shell could have filed the California action in its current form, as an action in equity, in 1891, after Shell had been sued by the United States. If the classification of Wausau’s suit under Enelow-Ettelson depends on the classification of the California action, then it seems that it too is “equitable”, making appeal impossible. But Wausau’s action may not share the characteristics of the California action. Wausau, after all, could not invoke the aid of a court of equity on the ground that it was faced with multiple suits that could lead to inconsistent outcomes. Wausau’s suit remains one that could not have been brought in 1891. See Scruggs & Echols v. American Central Insurance Co., 176 F. 224 (5th Cir.1910); Mechanics' Insurance Co. v. C.A. Hoover Distilling Co., 173 F. 888, 891 (8th Cir.1909). We have three possibilities: (a) look through the Illinois case to the California case, which in 1891 would have been equitable; (b) look at the Illinois case, which could not have been filed in 1891, and say that it is not “at law” and therefore is “equitable” for purposes of Enelow-Ettelson; (c) look at the Illinois case and say that if it could have been filed, it would have been at law, because it is most like an anticipatory defense to Shell’s suit at law for breach of contract. If either (a) or (b) prevails, then the district court’s refusal to stay the proceeding is the equity chancellor talking to himself and would not have been *902 the denial of an “injunction” under the Evarts Act of 1891; if (c) is the best characterization, then the Enelow-Ettelson doctrine allows an appeal if the requested stay here would have permitted consideration of an equitable defense in 1891. And we haven’t even started to discuss that question yet! Neither logic nor precedent dictates a choice among these possibilities. We therefore must ask whether an appellate court should take the Enelow-Ettelson doctrine for all it could be worth or for the least it has to be worth. We seriously considered in Olson discarding the Enelow-Ettelson doctrine altogether, on the ground that it is so universally condemned that the Supreme Court is bound to overrule it sooner or later; a refusal to apply the doctrine might speed the arrival of that happy day. We stopped at the precipice, however, because we were not absolutely sure that the Supreme Court would dispatch the doctrine. 806 F.2d at 734, 741-43. Our conclusion that the doctrine should not exist suggests, however, that any doubt be resolved against appellate jurisdiction. In recent years we have construed the Enelow-Ettelson doctrine as narrowly as intellectual honesty permits. E.g., Matterhorn, Inc. v. NCR Corp., 763 F.2d 866 (7th Cir.1985); Shear son Loeb Rhodes, Inc. v. Much, 754 F.2d 773 (7th Cir.1985); Timberlake v. Oppenheimer & Co., 729 F.2d 515 (7th Cir. 1984); Rohrer, Hibler & Replogle, Inc. v. Perkins, 728 F.2d 860, 862-63 (7th Cir. 1984); Matterhorn, Inc. v. NCR Corp., 727 F.2d 629 (7th Cir.1984); Medtronic, Inc. v. Intermedies Inc., 725 F.2d 440 (7th Cir. 1984); Whyte v. THinc Consulting Group International, 659 F.2d 817 (7th Cir.1981). Even a modestly confined reading of Enelow-Ettelson would lead to the dismissal of this appeal. Against all of this Shell relies on Micro-software Computer Systems, Inc. v. Ontel Corp., 686 F.2d 531 (7th Cir.1982), which Shell believes holds that a federal court’s failure to stay its proceedings while a state case proceeds is always appealable. Although Microsoftware observed that parallel state and federal cases are wasteful, id. at 534, interlocutory appeals also can be wasteful — something the Supreme Court continually emphasizes in cases under § 1291. District courts decide most questions correctly, so that most interlocutory appeals end in affirmance; meanwhile the case has been disrupted and the time of appellate judges consumed by questions that may wash out during the course of the case. The Supreme Court held in Will v. Calvert Fire Insurance Co., 437 U.S. 655, 98 S.Ct. 2552, 57 L.Ed.2d 504 (1978), that mandamus does not lie to undo a district court’s stay in favor of state litigation — not because of a difference between grants and denials of stays, but because a district court possesses substantial discretion to control its docket. (All nine Justices assumed in Will that stays are not appeal-able.) Microsoftware did not establish the principle Shell sees there; it simply applied the Enelow-Ettelson doctrine. 686 F.2d at 535-36. It concluded both that the underlying action was “at law” {id. at 535) and that the defense was equitable {id. at 536). There would be jurisdiction here, as there was in Microsoftware, if those conditions were met. But we have concluded that only by taking Enelow-Ettelson for all it could be worth, only by resolving several unanswerable questions in favor of Shell, could we say that Wausau’s action is “at law”. The Enelow-Ettelson doctrine does not support this appeal, which is therefore Dismissed for Want of Jurisdiction.