Roland MacHinery Company v. Dresser Industries, Inc.

SWYGERT, Senior Circuit Judge,

dissenting.

I commend the majority’s effort to delineate a coherent standard for the grant of preliminary injunctions. And I support the test set forth by the court, except I firmly believe that principles of equity jurisprudence require the maintenance of the “abuse of discretion” standard of review which has heretofore been recognized in this and other circuits, but which the majority does not fully adopt or support. Moreover, I find the district court’s decision granting Roland a preliminary injunction sustainable even under the majority’s interpretation of that standard.

I

Although the majority nominally retains the abuse of discretion standard, it substitutes the common understanding that has been given that term in the grant or denial of preliminary injunctions for a new interpretation, requiring strict scrutiny by .the reviewing court for isolated errors of fact or mistakes of law, thus leaving little flexibility for the district court in taking an overall view of the totality of circumstances. Apparently the majority believes that if the district court reaches a conclusion on the evidence with which the reviewing court disagrees, the district court’s determination regarding the balance of harms and the overall propriety of granting or denying injunctive relief is entitled to no deference whatsoever.

The majority recognizes that the overwhelming weight of precedent has uncritically accepted the abuse of discretion standard. See, e.g., Doran v. Salem Inn, Inc., 422 U.S. 922, 923 , 95 S.Ct. 2561, 2568 , 45 L.Ed.2d 648 (1975); Brown v. Chote, 411 U.S. 452, 457 , 93 S.Ct. 1732, 1735 , 36 L.Ed.2d 420 (1973); Menominee Rubber Co. v. Gould, Inc., 657 F.2d 164, 166 (7th Cir.1981); Reinders Brothers, Inc. v. Rain Bird Eastern Sales Corp., 627 F.2d 44, 49 (7th Cir.1980); Milsen Co. v. Southland Corp., 454 F.2d 363, 369 (7th Cir.1971); United States Steel Corp. v. Fraternal Ass’n of Steelhaulers, 431 F.2d 1046, 1048 (3d Cir.1970); 11 C. Wright & A. Miller, Federal Practice & Procedure § 2962, at 633 (1973) (hereinafter “Wright & Miller”) and cases cited therein. As the majority illustrates, see ante at 384-386, although the courts almost unanimously have paid lip service to the abuse of discretion standard, when examined closely the cases reveal several inconsistencies in this respect. See also Buffalo Courier-Express, Inc. v. Buffalo Evening News, Inc., 601 F.2d 48, 59 (2d Cir.1979). See generally Hammond, Interlocutory Injunctions: Time for a New Model?, 30 U. Toronto L.J. 240 , 240-41, 259-63 (1980) (hereinafter “Hammond, New Model ”); Luebsdorf, The Standard for Preliminary Injunctions, 91 Harv.L.Rev. 525, 525-27, 537-40 (1978); Developments in the Law — Injunctions, 78 Harv.L.Rev. 994, 1070 (1965) (hereinafter “Developments”). Often the court will add other language to the standard such as “clear error” or “mistake of law.” See ante at 382-386; 11 Wright & Miller, supra, § 2962, at 633-37. The few courts that have undertaken to define abuse of discretion have failed to agree on a definition. See ante at 384-386; Friendly, Indiscretion About Discretion, 31 Emory L.J. 747 , 762-64 (1982). Most important perhaps, the actual scrutiny with which appellate courts have reviewed preliminary injunction decisions by the lower courts has differed markedly irrespective of the seeming concensus on the applicable standard of review. See ante at 384-386. This last inconsistency, which exists in all types of cases, has given rise to the criticism that formulation of the standard of review is a purely verbal exercise because reviewing courts will decide a case as they see fit regardless of the standard of review. If a reviewing court wishes to sustain a deci *397 sion of a lower court, it will claim deference to the lower court in accord with the applicable standard. If a reviewing court wishes to reverse a decision of a lower court, it will find whatever degree of error is necessary under the applicable standard.

Despite the inconsistencies and the ring of truth to the foregoing criticism, I believe that the majority’s discussion of the standard of review represents a fundamental misunderstanding of the role of preliminary injunctive relief in our legal system and the role of the district courts in dispensing that relief. I endorse the four-part test set forth by the majority for determining the propriety of granting a preliminary injunction. I believe, however, that the formulation of that or any other test cannot replace the role of discretion in the decision to grant or deny a preliminary injunction. I further believe that the discretion must lie with the district court in the first instance.

The interlocutory injunction developed in England in the courts of equity. Like other equitable remedies, injunctions were designed to offer relief when legal remedies were unavailable or inadequate to protect the parties’ rights. See Developments, supra, at 997-98. Thus, it has been said that equity developed to relieve the harshness of the law. Despite the merger in our federal system of equity and law courts, a preliminary injunction is still considered an extraordinary remedy that is granted not as a matter of right. See 11 Wright & Miller, supra, § 2948, at 428.

A preliminary injunction is different from a remedy at law because it is both equitable and interlocutory. Unlike a remedy at law which generally requires only that a party pay damages, an injunction requires the party enjoined to do something or to refrain from doing something. Framing an injunctive degree is completely unlike assessing damages. Balancing harms and sliding scales are unique to equitable remedies. Determining a plaintiff’s likelihood of success on the merits prior to trial is unique to interlocutory relief. To fulfill its role of softening the harshness of the law and offering relief when a legal remedy is unavailable, the rules for granting preliminary injunctive relief must remain flexible and able to account for a myriad of situations. In sum, the decision whether to grant a preliminary injunction does not lend itself to a rigid legal formulation. See Lemon v. Kurtzman, 411 U.S. 192, 201 , 93 S.Ct. 1463, 1469 , 36 L.Ed.2d 151 (1973) (“In equity, as nowhere else, courts eschew rigid absolutes and look to the practical realities and necessities inescapably involved in reconciling competing interests ....”); Hecht & Co. v. Bowles, 321 U.S. 321, 329 , 64 S.Ct. 587, 591 , 88 L.Ed. 754 (1944) (“The essence of equity jurisdiction has been the power of the Chancellor to do equity and to mould each decree to the necessities of the particular case. Flexibility rather than rigidity has distinguished it.”); cf. United States v. McCoy, 517 F.2d 41, 44 (7th Cir.), cert. denied, 423 U.S. 895 , 96 S.Ct. 195 , 46 L.Ed.2d 127 (1975) (trial judge is given discretion where “the factors which may properly influence his decision are so numerous, variable and subtle that the fashioning of rigid rules would be more likely to impair his ability to deal fairly with a particular problem than to lead to a just result”).

Because the preliminary injunction is an extraordinary remedy that may impose a great burden on defendants and is granted before a full trial or a final determination of liability, the notion of leaving the decision to the unbridled discretion of judges is particularly troublesome. See O. Fiss & D. Rendleman, Injunctions, at 104-06 (2d ed. 1984); Hammond, New Model, supra, at 272 (“A discretionary formula, though fashionable, raises the dangers of potential judicial arbitrariness with respect to a remedy which is often dispositive of litigation and the difficulties of mounting an appeal from a discretion. These difficulties are not really solved by assuming that there will always be a succession of level-headed men called judges on the Clapham omnibus who will exercise considered judgment.”). Thus, I endorse the majority formulation of a coherent test to guide both the district courts and the reviewing courts. I believe, however, that within the confines of that *398 test the district court does and must exercise a large degree of discretion. Balancing the harms, determining the plaintiff’s likelihood of success on the merits, and balancing together the harms and likelihood of success on a sliding scale all require discretionary judgments. See United States Steel Corp. v. Fraternal Ass’n of Steelhaulers, supra, 431 F.2d at 1048 . Discretion also inheres in the framing of an injunctive decree appropriate to the situation and the parties. See Hecht Co. v. Bowles, supra, 321 U.S. at 329 , 64 S.Ct. at 591 ; Duran v. Elrod, 713 F.2d 292, 297 (7th Cir.1983), cert. denied, — U.S. —, 104 S.Ct. 1615 , 80 L.Ed.2d 143 (1984). There are as many possible decrees in any given case as there are judges. Although the test set forth by the majority will help guide the discretion of the courts, it cannot replace that discretion.

The discretion that inheres in the decision whether to grant a preliminary injunction cannot rest with the reviewing court but must lie in the district court. First, it must be remembered that the district court has the responsibility for making a final determination on the merits of the case. The lower court should also be entrusted to determine whether a preliminary injunction is necessary to preserve a state of affairs such that the court will be able to grant meaningful relief upon conclusion of a full trial. Second, given the flexibility of the standard for preliminary injunctive relief and the necessity for discretionary judgments, two courts could easily arrive at different although equally viable conclusions. Refusal to defer to the decision of the lower court in the first instance frequently will result in a substitution of the judgment of the reviewing court for that of the district court. This is a poor use of judicial resources. Wright, The Doubtful Omniscience of Appellate Courts, 41 Minn.L.Rev. 751, 778-80 (1957). Finally, and most importantly, if discretionary judgments must be made (as I believe they must) in deciding to grant or deny preliminary injunctive relief, the trial court with its greater knowledge of the case and the parties and “superior opportunity to get ‘the feel of the case’ ” is in a better position than a reviewing court to make those judgments. Noonan v. Cunard Steamship Co., 375 F.2d 69, 71 (2d Cir.1967) (quoting Cone v. West Virginia Pulp & Paper Co., 330 U.S. 212, 216 , 67 S.Ct. 752, 755 , 91 L.Ed. 849 (1947)); see Rosenberg, Judicial Discretion of the Trial Court, Viewed from Above, 22 Syracuse L.Rev. 635, 662-63 (1971); Wright, The Doubtful Omniscience of Appellate Courts, supra, 41 Minn.L.Rev. at 781-82.

The majority lists three types of matters that are generally referred to the district court’s discretion: a matter not decided according to a standard; a matter not susceptible of uniform treatment; and a matter not decided on the basis of evidence reasonably accessible to the appellate court through the record of the proceedings in the trial court. See ante at 388-389. The majority lists illustrations of these three types of discretionary judgment: criminal sentences; rulings on most discovery questions; and rulings on evidentia-ry questions where the issue is prejudicial impact on the jury. Deciding whether to grant a preliminary injunction has similarities with all three types of discretionary matters listed by the majority.

The test delineated today by the majority is unquestionably a standard, but so is a criminal sentencing statute. The key in both situations is that the standard contains a vast amount of flexibility so that the court can take into account the infinite variety of situations that may arise. See A.L.K. Corp. v. Columbia Pictures Industries, Inc., 440 F.2d 761, 763 (3d Cir.1971). Like formulation of a criminal sentence, the formulation of an injunctive decree is often tailored to the individual circumstances of the case and the parties, and the inherent “rightness” that the situation demands. See, e.g., Inland Steel Co. v. United States, 306 U.S. 153, 157-58 , 59 S.Ct. 415, 417-18 , 83 L.Ed. 557 (1939); Wright & Miller, supra, § 2947, at 424-26 and cases cited therein; Developments, supra, at 1063-64.

*399 Deciding that a matter is susceptible of uniform treatment assumes the answer to the question posed by the majority’s discretionary judgment analysis. If uniform treatment is sought, then a rigid legal formula is developed and discretion is dispensed with. If, on the other hand, discretion is deemed to be an element of a decision, uniformity is not expected. In my view, uniform treatment historically has not been and should not be a goal in preliminary injunction decisions, without, of course, condoning arbitrary decisionmak-ing. The decision to grant or deny a preliminary injunction should depend on the unique circumstances of and the parties in each case. I do not believe that uniform treatment is even possible under the test adopted today. The test is flexible, at least in theory. The sliding scale approach by its nature is antithetical to uniformity. Unique facts are certain to exist in each case which must be factored into the test. The decision to grant or deny preliminary injunctive relief simply defies mechanical treatment.

Finally, despite the majority’s contention that a preliminary injunction hearing produces a record similar to a trial record, a reviewing court does not have the same access to evidence that the district court has or that the reviewing court has after trial. The district court’s greater knowledge of the case and the parties and its ability to engage in extensive dialogue with the parties is particularly important in an abbreviated proceeding such as a preliminary injunction hearing. In that kind of hearing the parties do not have the same opportunity to develop their arguments and create their record that is essential to a meaningful review by an appellate court which has no prior acquaintance with the case and little if any opportunity to question the parties. In addition, a reviewing court may not have the same access to findings by the district court before trial as after trial. A trial judge may well be hesitant before trial and before presentation of all of the evidence to make explicit or final findings on witness credibility and other matters which the judge is in a unique position to make for fear that the findings will constrain or prejudice the final resolution of the case. Accordingly, the judge’s decision to grant or deny a preliminary injunction may be influenced by unstated and preliminary findings. We should encourage trial judges to refrain from making absolute findings that may impede a fair final resolution on the merits and at the same time to determine the appropriateness of preliminary injunctive relief on the basis of all of the evidence then available to them. I fear that the majority’s refusal to defer to the judgments of the trial judge will encourage the opposite; judges will be forced either to make their conclusions explicit and absolute or disregard them. It is worth noting that to the extent a decision by a reviewing court is not based on deference to the trial judge, the decision will have the inevitable and baneful effect of constraining and prejudicing the judge’s final resolution of the case.

Because the decision to grant or deny preliminary injunctive relief ought to lie within the discretion of the district court in the first instance, our review of that decision ought to be bound by the abuse of discretion standard. The district court’s decision involves some nondiscretionary judgments, such as preliminary factual findings and application of both the proper test for injunctive relief and, where necessary, the correct substantive law. Because the district judge is in a superior position to make preliminary judgments based on incomplete evidence, even these judgments should be reviewed with greater than usual deference. And, even assuming that the district court has committed an isolated error of fact or an isolated mistake of law, the court’s ultimate decision to grant or deny preliminary injunctive relief should be reversed only if it constitutes an abuse of discretion under the totality of the circumstances. See Menominee Rubber Co. v. Gould, Inc., supra, 657 F.2d at 166 ; Reinders Brothers, Inc. v. Rain Bird Eastern Sales Corp., supra, 627 F.2d at 49 . The majority rejects this view by focusing exclusively on individual aspects of the case, *400 at the expense of taking an overall view of the issue before the district court.

This circuit has adopted a definition of abuse of discretion that offers expansive protection to the decision of the district court. An abuse of discretion occurs only when no reasonable person could take the view adopted by the trial court. American Medical Ass’n v. Weinberger, 522 F.2d 921, 924 (7th Cir.1975). See also Duran v. Elrod, supra, 713 F.2d at 297 ; Harrington v. DeVito, 656 F.2d 264, 269 (7th Cir.1981), cert. denied, 455 U.S. 993 , 102 S.Ct. 1621 , 71 L.Ed.2d 854 (1982); Particle Data Laboratories, Inc. v. Coulter Electronics, Inc., 420 F.2d 1174, 1178 (7th Cir.1969). It can be argued that such an extreme limitation on the scope of appellate review is inappropriate in preliminary injunction cases in which there is a delineated standard (however flexible) and a large body of case law on the subject. See generally United States v. Criden, 648 F.2d 814, 817-18 (3d Cir.1981); Friendly, Indiscretion About Discretion, supra, 31 Emory L.J. at 762-73. At a minimum, however, the abuse of discretion standard of review of preliminary injunction decisions means that the appellate court should not second guess the balances struck by the district court if there is reasonable support in the record for those balances; doubts should be resolved in favor of the district court’s decision. At a minimum, abuse of discretion means much greater deference than the majority gives today to the decision of the court below.

II

Turning to the facts of the instant case, I conclude that the district court’s decision granting Roland a preliminary injunction was not an abuse of discretion even if the majority’s interpretation of that standard of review is employed. That is, even assuming that isolated errors of fact or mistakes of law all but erase any discretion that the district court might otherwise have, I can find no such errors in this case.

The majority concedes that Roland satisfied the threshold requirements of proving that it has no adequate remedy at law and that it will suffer irreparable injury if in-junctive relief is denied. Ante at 391. Based upon its de novo review of the evidence, the majority decides that the district court has committed errors of fact and, as a result, the district court’s judgment as to the balance of harms is entitled to no deference. The majority then itself concludes that the balance of harms between Roland and Dresser is approximately equal and thus that Roland must show a substantial likelihood of success on the merits. Finding that Roland has failed sufficiently to show an exclusive dealing agreement with anticompetitive effect, the majority concludes that preliminary injunctive relief is unwarranted.

There are several holes in the majority’s conclusion that the district court clearly erred in finding that Roland would probably go out of business if its Dresser distributorship is terminated. The majority concedes that the purchase of Dresser parts from other dealers at a twenty percent markup could render the sales of parts unprofitable but concludes that Roland could still derive profits on the sale of service in connection with the parts business. There is no evidence and no guarantee, however, that the profits derived from the sale of service are sufficient to render the parts business ultimately profitable. The majority also finds that Roland’s revenues from the rental of Dresser equipment “should not be affected by the cancellation of the franchise, at least until its present stock wears out, and that should not be till after the trial is over.” Ante at 391. In the absence of knowing the size and condition of Roland’s present stock of rental equipment, the lifespan of such equipment, or when the trial will be over, the majority’s finding is untenable. The majority also finds that Roland’s projected future earnings from sales of Komatsu equipment is unrealistically low. The majority, however, offers no more realistic projection. Purely speculative earnings should not be the basis for denying preliminary injunctive relief. The court can take *401 appropriate action if and when those earnings come to pass. If Roland’s earnings from the sales of Komatsu equipment increase to the point that the injunction is no longer necessary to protect Roland’s status as an ongoing business, the court can always dissolve or modify the injunction. See United States v. Swift & Co., 286 U.S. 106, 119 , 52 S.Ct. 460, 464 , 76 L.Ed. 999 (1932); Developments, supra, at 1080-86. Finally, the majority completely ignores the loss of goodwill, business reputation, and credit rating that Roland will suffer if its Dresser distributorship is terminated. Roland Machinery Co. v. Dresser Industries, Inc., No. 84-3038, Order of March 27, 1984, at 8 (C.D.Ill.) (hereinafter “Order”). See Menominee Rubber Co. v. Gould, Inc., supra, 657 F.2d at 167 ; Reinders Brothers, Inc. v. Rain Bird Eastern Sales Corp., supra, 627 F.2d at 53 . See generally Comment, The Irreparable Harm Requirement for Preliminary Injunctive Relief in Antitrust Distributor Termination Cases, 61 B.U.L.Rev. 507, 516-21 (1981).

At a minimum, Roland will lose twenty-four percent of its annual gross profits if Dresser terminates Roland’s distributorship. See Operating Statement of Roland from Nov. 1, 1982-Oct. 31, 1983, Plaintiff’s Exhibit No. 11A at 2. This assumes that Roland will be able to continue its sales of Dresser parts at a twenty percent markup and that Roland will lose no revenues on its rental of Dresser equipment. This figure does not take into account the loss of profits on service rendered in connection with the sale of new Dresser equipment or the loss of revenues as a result of Roland’s loss of goodwill, business reputation, and credit rating. This figure is sufficient in my mind to support the testimony of Roland’s treasurer and comptroller who concluded, after reviewing Roland’s financial records, that Roland could not survive as. an ongoing business past 1985 if Dresser were allowed to terminate Roland's distributorship. See Preliminary Hearing Transcript, vol. 1, at 81. In sum, the evidence supports and nothing the majority has stated indicates any error in the district court’s conclusion that Roland would probably disappear as a business entity if Dresser did not continue to supply Roland with new machines and parts. See Order at 8.

Even if there is only a small likelihood that Roland will go out of business if its Dresser distributorship is terminated, that possibility must be considered because the harm to Roland, if its business is destroyed, is so great. A small business enterprise is endowed with a “soul” identified with its owner who has created and nurtured the enterprise into a useful and successful business. As the majority recognizes, see ante at 386, the harm to the owner from the loss of a business enterprise is uncompensable. Earl Roland “want[s] to sell [construction equipment machinery], not to live on the income from a damages award.” Semmes Motors, Inc. v. Ford Motor Co., 429 F.2d 1197, 1205 (2d Cir.1970). See also Milsen Co. v. Southland Corp., supra, 454 F.2d at 366 ; Bateman v. Ford Motor Co., 302 F.2d 63, 66 (3d Cir.1962).

Even assuming that Roland would not be forced out of business if Dresser terminated the distributorship, Roland will still suffer substantial harm in terms of loss of profits and loss of goodwill and business reputation. This harm is given little weight by the majority. The majority merely finds that Roland will be able to continue in business if its Dresser distributorship is terminated and concludes that the harm to Roland if the injunction is not granted is equivalent to the harm to Dresser if the injunction is granted. See ante at 392-393.

In assessing the harm to Dresser that the injunction will cause, the majority dismisses as unworkable the district court’s conditional decree that requires Roland to maintain Dresser’s market share within normal economic fluctuations. Ante at 392-393. The majority merely substitutes its judgment for that of the district court. The majority ignores the condition of the decree which permits Dresser to obtain additional distributors in Roland’s territory. See Order at 15. The majority also ignores the district court’s finding that *402 there was no evidence to support Dresser’s contention that Roland would cease or had ceased vigorously to market Dresser Products. See Order at 8-9. Drafting a conditional decree so as to protect the interests of all of the parties is entirely appropriate; See Inland Steel Co. v. United States, supra, 306 U.S. at 156-58 , 59 S.Ct. at 417-18 ; Milsen Co. v. Southland Corp., supra, 454 F.2d at 369 . I see no reason to doubt the district court’s competence to enforce the decree in this instance. I would commend the district court for its sensitivity to the interests of both parties.

The mistake of law the majority identifies is the district court's failure “to consider the possible impact on competition of the provision of the injunction that freezes Dresser’s market share until the end of the lawsuit.” Ante at 392. In my view, the majority’s so-called legal error is irrelevant in determining if the district court properly granted a preliminary injunction. Whether a particular term of the preliminary injunction has anticompetitive effects goes only to the question of the proper scope of the decree, not to the question of whether, on balance, the plaintiff or the defendant would suffer more by the imposition of some form of injunctive relief.

I agree with the majority that the injunctive decree, even as conditioned by the district court, cannot fully protect Dresser. Nevertheless, I believe the majority has not come close to showing that the balance struck by the district court was clear error, let alone an abuse of discretion. Cf. Menominee Rubber Co. v. Gould, Inc., supra, 657 F.2d at 167 (“The imbalance between the hardship [the distributor] would suffer if terminated [including the loss of good will and disruption of business] versus the hardship [the defendant] would suffer if the status quo were maintained is ... apparent ....”); Reinders Brothers, Inc. v. Rain Bird Eastern Sales Corp., supra, 627 F.2d at 53 (balance of harms found in favor of plaintiff where termination of distributorship would damage plaintiff’s goodwill, plaintiff continued to use defendant’s product, and defendant was free to set up competing distributors in plaintiff’s territory); Milsen Co. v. Southland Corp., supra, 454 F.2d at 366 (“Courts which have entered injunctions against [franchise] terminations have weighed the equities and found the plaintiff’s side more substantial ....”).

Concluding that the balance of hardships is even, the majority holds that Roland must show that it was more likely than not that Roland would prevail on the merits of its claim that Dresser engaged in an exclusive dealing policy in violation of section 3 of the Clayton Act, 15 U.S.C. § 14 (1982). The district court concluded that Roland raised a “substantial question” whether there existed an implied exclusive dealing arrangement between Roland and Dresser which tended to substantially lessen competition in the relevant market. Order at 14. Even without the benefit of the newly-clarified test announced today, the district court applied the correct standard for determining Roland’s likelihood of success on the merits given the court’s finding that the balance of harms tipped decidedly in favor of Roland. Therefore, the district court’s entire analysis is sustainable even under the majority’s interpretation of the abuse of discretion standard.

In addition, I believe that the evidence supports the conclusion that Roland is more likely than not to prevail on its section 3 claim. I am not as confident as the majority that Colgate, see United States v. Colgate & Co., 250 U.S. 300, 306-07 , 39 S.Ct. 465, 467-68 , 63 L.Ed. 992 (1919), retains much validity with respect to exclusive dealing claims which arguably are subject to the rule of reason rather than a per se rule. See Monsanto Co. v. Spray-Rite Service Corp., — U.S. —, 104 S.Ct. 1464, 1470 , 79 L.Ed.2d 775 (1984) (“[I]t is of considerable importance that ... concerted action on non-price restrictions, be distinguished from price-fixing agreements, since under present law the latter are subject to per se treatment and treble damages. On a claim of concerted price-fixing, the antitrust plaintiff must present evidence sufficient to carry its burden of proving that there was such an agreement.”). *403 Nevertheless, there is sufficient evidence in the record that Dresser did more than merely independently maintain an exclusive dealing policy.

Dresser engaged in extensive surveillance of Roland and other distributors which Dresser believed had or might sign agreements to deal with Komatsu. Dresser’s Senior Vice-President of Operations and Acting President of the Construction Equipment Group requested reports on Dresser distributors dealing with Komatsu and such reports were prepared. See Plaintiff’s Exhibits 34, 35, 52, 53, & 67 at 110-17. Dresser’s President of Construction Equipment International instructed a Regional Manager to monitor Roland at the time Roland was preparing to sign an agreement with Komatsu. See Plaintiff’s Exhibit 67, at 125. A Dresser Territorial Sales Manager obtained notes from another distributor of a purported conversation with Roland concerning the Komatsu contract. See Hearing Transcript, vol. 2, at 204-05; Plaintiff’s Exhibit 63, at 52-55. There is also evidence suggesting that Dresser offered one distributor an additional incentive, i.e., an additional territory, to induce the distributor not to sign an agreement with Komatsu. See Plaintiff’s Exhibit 58, at 59-60. This evidence is sufficient at this stage to show that Dresser’s conduct more likely than not exceeded the type of conduct permitted under Colgate, supra. Cf. Albrecht v. Herald Co., 390 U.S. 145, 149-50 , 88 S.Ct. 869, 871-72 , 19 L.Ed.2d 998 (1968) (combination found to exist where manufacturer hired second distributor to compete with first distributor who was not following suggested price); United States v. Parke, Davis & Co., 362 U.S. 29, 45-47 , 80 S.Ct. 503, 512-513 , 4 L.Ed.2d 505 (1960) (agreement found to exist where acquiescence of retailers in suggested prices was secured by threats of termination); FTC v. Beech-Nut Packing Co., 257 U.S. 441, 454-55 , 42 S.Ct. 150, 154-55 , 66 L.Ed. 307 (1922) (institution of policing system to detect price-cutters and refusing to sell to price-cutters until they agreed to conform to suggested prices found to exceed the conduct permitted under Colgate)] Greene v. General Foods Corp., 517 F.2d 635, 658 (5th Cir.1975), cert. denied, 424 U.S. 942 , 96 S.Ct. 1409 , 47 L.Ed.2d 348 (1976) (institution of “device for policing the conduct of distributors” and “monitoring of [distributor’s] pricing to his ... customers” found to go “far beyond the simple announcement of terms and refusal to deal with a noncomplying independent distributor that is still permissible under Colgate”). See generally L. Sullivan, Handbook of the Law of Antitrust § 139, at 393-95 (1977). As the district court found, see Order at 10-11, Dresser offered no persuasive evidence of its reason for terminating Roland’s distributorship other than an implied exclusive dealing condition. While this evidence may not be sufficient to ultimately prove the existence of an agreement, it is sufficient at this early stage of the litigation to meet the majority’s greater than not likelihood test.

Finally, in concluding that Roland failed to establish any anticompetitive effect resulting from Dresser’s conduct, the majority completely ignores the evidence before and the findings of the district court. Roland argued before the district court that the relevant geographic market includes those counties which Roland contracted to service for Dresser. Dresser offered no evidence concerning the relevant geographic market. The district court therefore defined the relevant geographic market as that part of Central Illinois in which Roland normally receives requests and supplies bids from construction equipment consumers. See Order at 11-12. The majority does not dispute this finding, but discusses Komatsu’s standing in both the world and national markets for construction equipment. See ante at 395. The majority finds that “[t]he nationwide practice of exclusive dealing has not kept Komatsu from becoming a major factor in the U.S. market” and concludes that the practice will not do so in Central Illinois. Id. This conclusion ignores (but does not refute) the district court’s findings that Komatsu had tried but failed to infiltrate the Central Illinois market through the use of exclusive dealers and that the Central Illinois market *404 is unique because it is “the backyard” of Caterpillar, “the world’s number one producer and seller of construction equipment.” Order at 12-13. The majority suggests that its decision will actually help Komatsu’s entry into the market because “[t]he likeliest consequence of our dissolving the preliminary injunction would be to accelerate Komatsu’s efforts to promote its brand through the Roland distributorship.” Ante at 394. In fact, the evidence shows that exclusive Komatsu distributorships have failed to survive in Central Illinois and Roland itself has been unable to produce sufficient profits from the sale of Komatsu products to survive. Of course, when the distributor fails, Komatsu’s attempt to infiltrate the market also fails. On the basis of this evidence, the district court concluded: “Removal of the Dresser distributorship would probably [i.e., more likely than not] doom Roland to the same end that all other Komatsu distributors have fallen [and] effectively prevent Ko-matsu from entry into the Central Illinois market.” Order at 13. The district court’s conclusion is based on the evidence presented to the court and is not clearly erroneous.

Finally, although the district court does not mention this finding in its discussion on the merits of Roland’s section 3 claim, the court earlier in the opinion notes:

Only recently has the construction equipment machinery market engaged in significant price competition, having previously operated under the pricing “umbrella” of the dominating market leader Caterpillar. Apparently non-price factors dictated the decisions of most construction machinery purchasers in the past, a trend Dresser apparently anticipates in the future____

Order, at 9 n. 3. Roland apparently quoted lower prices on Komatsu equipment when submitting price quotations to customers on competing Dresser and Komatsu machinery. Id. This suggests that the entry of foreign competitors such as Komatsu into the United States market has triggered some price competition in the industry. At least it suggests that Komatsu may be offering consumers a lower-priced alternative. If this is so, the consumer value, as the majority calls it, is clearly promoted by Komatsu’s entry into the construction equipment market.

The majority’s analysis of the instant case is precisely the type of substitution of judgment about which I warned in the first part of this dissent. I find the district court’s decision well-reasoned, based on the evidence, and without either errors of fact or mistakes of law. Notwithstanding its rhetorical reaffirmation of the abuse of discretion standard, the majority applies that standard in a way that allows it to reverse the judgment below when, in its view, the district court has committed any isolated “clear” error of fact or “mere” error of law. Indeed, the majority undertakes a sweeping de novo review of the evidence and reaches a conclusion contrary to that of the district court. Such de novo review usurps the authority of the district court to decide whether a preliminary injunction is necessary to protect the interests of the parties pending trial and the court’s ability to render final relief after trial. Moreover, the decision by the majority, despite its disclaimer, will necessarily prejudice the district court’s ultimate decision after a full trial. I would affirm the district court’s grant of a preliminary injunction.