McMILLIAN, Circuit Judge.
The Kansas City Star Company (Star Co.) appeals from a final order entered in the United States District Court for the Western District of Missouri which permanently enjoined Star Co. from refusing to sell its newspapers to appellees at wholesale. For reversal Star Co. argues that its proposed change in distribution from the use of independent contract carriers to reliance on delivery agents does not violate the Sherman Act’s prohibition against monopolization or attempts to monopolize interstate commerce. 15 U.S.C. § 2 (1976). Star Co. also argues that appellees have suffered no cognizable antitrust injury and that, even if there were such an injury, the permanent injunction is unnecessarily broad to remedy it. In a separate appeal, Star Co. argues that the district court erred and abused its discretion in awarding attorney fees to ap-pellees. For the following reasons, we overrule en banc the panel opinion and vacate the permanent injunction as well as the award of attorney fees.
I. FACTUAL BACKGROUND
The facts are set forth in the panel opinions. Paschall v. Kansas City Star Co., 695 F.2d 322, 325-26 (8th Cir.1982). Star Co. owns the two major daily newspapers in Kansas City, Missouri — The Kansas City Star and The Kansas City Times. Star Co.’s road to dominance in the Kansas City market, however, was marred by a conviction for attempted and actual monopolization in violation of § 2 of the Sherman Act. Kansas City Star Co. v. United States, 240 F.2d 643 (8th Cir.1957). Both newspapers historically were distributed by independent contract carriers who bought the newspapers at wholesale from Star Co. and then resold them to the public at varying retail prices. The retail prices for the newspapers are set by these contract carriers, although in the past Star Co. had some control over the retail prices charged by the contract carriers. The wholesale price is set by Star Co. Originally the wholesale price was a set sum per delivered copy. Later, the wholesale price was calculated as a certain percentage of the retail price charged by the carrier. The percentage charged varied for each contract carrier based upon, among other things, the retail price that the contract carrier proposed to charge its customers.
Each contract carrier held a contract from the Star Co. which gave it the exclu[695]*695sive right to sell Star Co.’s newspapers along established routes. No contract carrier competed against any other contract carrier for retail business in any significant way. Instead, each contract carrier sold its newspapers only within its own territory or along its own route. Star Co., however, did reserve by contract the right to sell directly to subscribers if it felt that the contract carrier was not providing proper service. This right was exercised on at least one occasion and all of the contract carriers were acutely aware of the existence of Star Co.’s contractual right to compete at the retail level. However, Star Co. seldom delivered newspapers directly to subscribers. Contract carriers considered their Star Co. routes to be their “property,” even though the routes existed only because of the exclusive but terminable distribution contracts entered into by each contract carrier and Star Co. When a contract carrier decided to leave the business, the contract carrier would “sell” its route to its successor. If Star Co. approved, the route purchaser would then enter into a new contract with Star Co. and become that route’s new contract carrier. Prices for routes sold in this manner ranged into the hundreds of thousands of dollars for the largest routes.
In 1974, Star Co. informed the contract carriers by letter that in the future Star Co. might have to alter the way it distributed its two newspapers. This announcement so upset the contract carriers, all of whom had invested substantial sums in their routes, that it led one of them to bring this antitrust lawsuit. Then, in 1977, Capital Cities Communications, Inc. (Capital Cities) acquired control over Star Co. through a stock tender offer. Shortly thereafter, Capital Cities announced that it would terminate all contract carriers and replace them with Star Co.’s own delivery agents, thereby enabling Star Co. to sell its two newspapers directly to readers. It was also announced that once the delivery agent system was in place, Star Co. would no longer sell its two newspapers at wholesale to anyone. According to Star Co., the delivery agent system would be a significant improvement over the contract carrier system because Star Co. could set an area-wide uniform price for its newspapers and provide readers with better, more responsive service. Star Co. welcomed the current contract carriers to apply for positions as delivery agents. As an inducement, Star Co. promised that delivery agents would earn approximately the same income as the contract carriers had earned.
In response to these announcements, plaintiff-appellee Gweldon Paschall, along with approximately 250 other contract carriers, moved for and obtained a preliminary injunction against Star Co. barring it from implementing the delivery agent system. After a non-jury trial on the merits, the district court found that Star Co.’s planned vertical integration into delivery, coupled with its refusal to deal with the contract carriers, would violate § 2 of the Sherman Act, 15 U.S.C. § 2, because it would permit Star Co. to extend its monopoly in newspaper publishing into the retail newspaper market. The district court, therefore, permanently enjoined Star Co. from refusing to deal with the contract carriers or dealing with them in a discriminatory or predatory manner.1 On appeal a divided panel of this court affirmed the permanent injunction but reduced the district court’s award of attorney fees to appellees to $2.5 million. Paschall v. Kansas City Star Co., 695 F.2d at 333-34, 338. On Star Co.’s petition, this court agreed to reconsider the case en banc.
II. § 2 OF THE SHERMAN ACT — MONOPOLIZATION
There are two basic elements of a § 2 Sherman Act violation. The first is the “possession of monopoly power [2] in the rel[696]*696evant market” and the second is the “willful acquisition or maintenance of that power as distinguished from growth or development as a consequence of a superior product, business acumen, or historic accident.” United States v. Grinnell Corp., 384 U.S. 563, 570-71, 86 S.Ct. 1698, 1704, 16 L.Ed.2d 778 (1966). The first element is easily disposed of in this case. The district court found that Star Co. had a monopoly in the wholesale metropolitan daily newspaper market. We adopt the district court’s finding of monopoly power because it is not clearly erroneous. See Poster Exchange, Inc. v. National Screen Service Corp., 431 F.2d 334, 338 (5th Cir.1970), cert. denied, 401 U.S. 912, 91 S.Ct. 880, 27 L.Ed.2d 811 (1971). The second element, as it applies to the facts of this case, is more difficult to ascertain.
A. Specific Intent or Purpose
Generally, cases following the Supreme Court’s landmark § 2 Sherman Act decisions rendered during its 1947 Term have imposed liability for unlawful monopolization upon proof of either (1) the specific intent to monopolize or (2) anticompetitive effects that result from the monopolist’s actions. See United States v. Columbia Steel Co., 334 U.S. 495, 531-32, 68 S.Ct. 1107, 1126-1127, 92 L.Ed. 1533 (1948) (Columbia Steel); United States v. Griffith, 334 U.S. 100, 106, 68 S.Ct. 941, 945, 92 L.Ed. 1236 (1948) (Griffith). One need not prove both specific intent and anticompetitive effects; either alone is a sufficient basis for the imposition of liability upon a business entity wielding monopoly power. Griffith, 334 U.S. at 105, 68 S.Ct. at 944-945; Columbia Steel, 334 U.S. at 531-32, 68 S.Ct. at 1126-1127. This rule proceeds from the premise that although monopolies are tolerated, they are not cherished. Berkey Photo, Inc. v. Eastman Kodak Co., 603 F.2d 263, 274 (2d Cir.1979), cert. denied, 444 U.S. 1093, 100 S.Ct. 1061, 62 L.Ed.2d 783 (1980). Thus, a business entity that possesses monopoly power cannot lawfully exercise it with the specific intent of restraining competition because such actions entail a “dangerous probability” that the consequences prohibited by the Sherman Act may result. Griffith, 334 U.S. at 105-06, 68 S.Ct. at 944-945, citing Swift & Co. v. United States, 196 U.S. 375, 396, 25 S.Ct. 276, 279, 49 L.Ed. 518 (1905). Similarly, a monopolist acting with a “pure heart” may not engage in activities which would directly and necessarily result in the unreasonable restraint of competition because that result is the specific evil which the Sherman Act was passed to prevent. Id.; Berkey Photo, Inc. v. Eastman Kodak Co., 603 F.2d at 271-75.
As the Supreme Court made plain in Griffith, 334 U.S. at 107-08, 68 S.Ct. at 945- 946,
[i]t follows a fortiori that the use of monopoly power, however lawfully acquired, to foreclose competition, to gain a competitive advantage, or to destroy a competitor, is unlawful.
... If monopoly power can be used to beget monopoly, the Act becomes a feeble instrument indeed.... [An action by a monopolist] may yield price or other lawful advantages to the buyer. It may not, however, be used to monopolize or to attempt to monopolize interstate trade or commerce. Nor, as we hold in United States v. Paramount Pictures, Inc., [334 U.S. 131, 68 S.Ct. 915, 92 L.Ed. 1260 (1948),] may it be used to stifle competition by denying competitors less favorably situated access to the market.
In the context of vertical integration,3 the requisite specific intent may be [697]*697shown where the vertical integration was part of “a calculated scheme to gain control over an appreciable segment of the market and to restrain or suppress competition, rather than an expansion to meet legitimate business needs.” United States v. Paramount Pictures, Inc., 334 U.S. 131, 174, 68 S.Ct. 915, 937, 92 L.Ed. 1260 (1948). Specific intent to monopolize is also found where the monopolist, or would-be monopolist, engages in predatory tactics or “dirty tricks.” See, e.g., Byars v. Bluff City News Co., 609 F.2d 843, 853-54 (6th Cir.1979) (Byars ).4 The record before us discloses no evidence of Star Co. engaging in any “dirty tricks” or predatory practices against its contract carriers.
Liability based on specific intent can be negated where valid business justifications exist for the monopolist’s actions. Cf. Eastman Kodak Co. v. Southern Photo Materials Co., 273 U.S. 359, 375, 47 S.Ct. 400, 404, 71 L.Ed. 68 (1927) (absence of valid business justification for manufacturer’s refusal to sell to dealers at wholesale is grounds for imposing antitrust liability). See also Becker v. Egypt News Co., 713 F.2d 363, 368 (8th Cir.1983); Byars, 609 F.2d at 862-63; Newberry v. Washington Post Co., 438 F.Supp. 470, 475 (D.D.C.1977); McGuire v. Times Mirror Co., 405 F.Supp. 57, 66 (C.D.Cal.1975); Lamarca v. Miami Herald Publishing Co., 395 F.Supp. 324, 327 (S.D.Fla.), aff’d mem., 524 F.2d 1230 (5th Cir.1975). Here Star Co. has offered several legitimate business reasons for adopting the delivery agent system, including (1) the ability to set an area-wide uniform retail price to facilitate “in-paper advertising” for new subscriptions and simplify subscription collection and (2) the capability to be more responsive to customer complaints and assure more rapid starts for new subscribers.5 Star Co. has argued that because these [698]*698points had caused Star Co., and its readers, problems under the contract carrier system, it has the unfettered business prerogative to correct these problems in the most efficient manner possible. According to Star Co., the delivery agent system is the most efficient resolution of these problems.
We readily agree with Star Co. that the existence of legitimate business reasons for its decision to integrate forward from the wholesale level (first level) to the retail or distribution level (second level) will negate any specific intent basis of liability. Proof of specific intent is necessary, however, only “where the acts fall short of the results condemned by the Act.” Griffith, 334 U.S. at 105, 108, 68 S.Ct. at 946; United States v. Paramount Pictures, Inc., 334 U.S. at 173, 68 S.Ct. at 936-937.
B. Anticompetitive Effects
The results condemned by § 2 of the Sherman Act include the acquisition, maintenance, or extension of monopoly power through improper means. Here, appellees argue that Star Co. has abused its monopoly power in the wholesale market by integrating forward into distribution and refusing to deal with would-be competitors.6 Appellees contend that the combined result of Star Co.’s vertical integration and refusal to deal has been Star Co.’s acquisition of a monopoly in the retail market and a concomitant extension of Star Co.’s monopoly power in the wholesale market. Neither vertical integration nor a unilateral refusal to deal is per se illegal. Columbia Steel, 334 U.S. at 525, 68 S.Ct. at 1123 (vertical integration); United States v. Russell Stover Candies, Inc., 718 F.2d 256 (8th Cir.1983) (unilateral refusal to deal). Both actions are normally judged under a § 1 rule of reason standard. However, the “improper means” by which a monopolist acquires, maintains, or extends its monopoly in violation of § 2 are often equated with conduct that would violate § 1 if it were done in concert with others.
[A monopolist] usually does not violate § 2 of the Sherman Act unless [it] has acquired or maintained [its] strategic position, or sought to expand [its] monopoly, or expanded it by means of those restraints of trade which are cognizable under § 1. For those things which are condemned by § 2 are in large measure merely the end products of conduct which violates § 1.
Griffith, 334 U.S. at 106, 68 S.Ct. at 945. See United States v. Paramount Pictures, Inc., 334 U.S. at 171, 68 S.Ct. at 935-936; Berkey Photo, Inc. v. Eastman Kodak Co., 603 F.2d at 274.
Under § 1 of the Sherman Act’s prohibition against unreasonable restraints of trade, the principal inquiry is whether the conduct complained of results in any unreasonable anticompetitive effects. Columbia Steel, 334 U.S. at 527, 68 S.Ct. at 1124. Hence, the central issue of this case is whether the combination of Star Co.’s vertical integration and refusal to deal has resulted in any unreasonable anticompetitive effects in the market. If no unreasonable anticompetitive effects will follow from Star Co.’s implementation of its delivery agent system, its de facto acquisition of a monopoly of the retail daily metropolitan newspaper market will not be condemned by § 2 of the Sherman Act, because the Sherman Act protects the benefits of competition and not just individual competitors. Absent the specific intent to monopolize, a monopolist’s legitimate business decisions will not be curtailed if those decisions promote the redeeming virtues of competition: lower prices, greater efficiency and innovation, and more responsive service. See Continental T.V., Inc. v. GTE Sylvania, Inc., 433 U.S. 36, 52-57, 97 S.Ct. 2549, 2558-2561, 53 L.Ed.2d 568 (1977); Columbia Steel, 334 U.S. at 526-29, 68 S.Ct. at 1123-1125.
Two contrasting theories have been advanced in this case to provide a conceptual framework for ascertaining and analyzing the impact of Star Co.’s proposed change in its method of distribution. One theory, the potential competitor theory, has been relied upon to impose liability on Star Co. The [699]*699other, the optimum monopoly price theory, has been used in defense.
1. Potential Competitor Theory
The potential competitor theory was first discussed in United States v. Falstaff Brewing Corp., 410 U.S. 526, 532-33, 93 S.Ct. 1096, 1100-1101, 35 L.Ed.2d 475 (1973). In Falstaff, the defendant had decided to expand into the New England market by merging with an existing New England brewery. The Court held that entry through merger would violate § 7 of the Clayton Act, 15 U.S.C. § 18 (1976), if the entry had the effect of eliminating “a potential competitor exercising present influence on the market.” 410 U.S. at 532, 93 S.Ct. at 1100. It is said that a potential competitor, merely by its presence at the edge of the market, will have a retardant effect on price and will provide an incentive for better service among those already in the market. This is so because if the existing market competitors practice price gouging or provide inferior service, the potential competitor will enter the competitive fray with lower prices and better service, thereby capturing a substantial share of the market. To avoid the possibility of losing business to the potential competitor, the actual market competitors will maintain low prices and quality service. Thus, should the potential competitor merge with an existing market competitor, the market will lose the procompetitive effect of the presence of a potential competitor. The Court in Falstaff remanded the case to determine whether Falstaff was a potential competitor “in the sense that it was so positioned on the edge of the market that it exerted beneficial influence on competitive conditions in that market.” Id. at 532-33, 93 S.Ct. at 1100-1101. If Falstaff was a potential competitor, the merger would be unlawful under § 7 of the Clayton Act.
The district court grafted this potential competitor theory onto § 2 of the Sherman Act and applied it to the facts of this case. The district court found that Star Co. held a monopoly in the wholesale newspaper market. The district court also found that although Star Co. had not competed in the retail newspaper market to any significant degree in the past, Star Co. nonetheless existed as a powerful competitive force in the retail newspaper market because it was a potential competitor. Star Co. reserved by contract the right to- sell directly to subscribers if it felt that the contract carrier was not providing proper service. As an example of this theory at work, the district court pointed to what has been referred to as the “Louisberg Square incident.” In April of 1970, a contract carrier “arbitrarily” raised its retail subscription price for The Star to readers in an apartment complex known as Louisberg Square. Many residents of Louisberg Square complained to Star Co. about the price increase. Eventually, Star Co. delivered its newspapers at a lower price directly to those dissatisfied Louisberg Square residents who requested direct delivery from Star Co. All of the contract carriers were aware of the Louis-berg Square incident, and they all knew that a similar fate possibly could befall them should they engage in price gouging or give their subscribers poor service. Thus, according to the district court, the mere presence of Star Co. as a potential competitor in the retail market produced a substantial retardant effect on the contract carriers’ prices and gave the contract carriers a significant incentive to provide the best service possible. The district court went on to observe that Star Co.’s forward integration into the retail market, coupled with its refusal to deal with contract carriers, removed the beneficial influence on competitive conditions in the retail market exerted by Star Co.’s presence as a potential competitor. The district court concluded that Star Co.’s removal of its competitive impact from a market would violate § 2 of the Sherman Act because it would enable Star Co. to expand its monopoly in the wholesale market into the retail market, citing Griffith, 334 U.S. at 108, 68 S.Ct. at 946.
The panel majority accepted the district court’s potential competitor theory, but subjected the theory to a more rigorous and searching antitrust analysis. The panel ma[700]*700jority primarily focused on the anticompeti-tive effect basis of § 2 liability. It stated that § 2 liability could be found only if the monopolist’s vertical integration resulted in “substantial anticompetitive effects that are not offset by production economies, savings in market transaction costs, or other competitive benefits.” Paschall v. Kansas City Star Co., 695 F.2d at 327. After a careful review of the record evidence, the panel majority concluded that
the anticompetitive consequences of the Star’s proposed refusal to deal [, that is, the proposed change in the system of distribution from independent contract carriers to delivery agents,] will likely outweigh any competitive benefits. The refusal will eliminate the Star as a potential competitor in the retail market, will likely result in higher prices and poorer services to readers, and will not lead to a more efficient delivery system.
Id. at 328-29. All in all, the panel majority concluded:
the [record] evidence ... [showed] that the Star's proposed refusal to deal will produce anticompetitive effects on retail prices and services without accomplishing any savings in market transaction costs or creating production economies. The economic theory advanced by the Star and amicus, standing alone without factual support in the record, is insufficient to rebut that evidence.
Id. at 332.7
2. Optimum Monopoly Price Theory
The panel dissent championed the optimum monopoly price theory advanced by [701]*701Star Co., the amicus,8 and several leading antitrust economics scholars of the “Chicago school.”9 As a fundamental premise, the panel dissent emphasized that monopolies, even forwardly integrating monopolies, are not in and of themselves illegal. Id. at 340 (Henley, J., dissenting). Rather, forward vertical integration by a monopolist violates § 2 of the Sherman Act only if it results in unreasonable anticompetitive effects. Id. The panel dissent then took the position that no unreasonable anticompeti-tive effects will flow from the implementation of Star Co.’s delivery agent system. Id. at 340-43.
The optimum monopoly price theory can be summarized as follows. Under any given set of cost and demand curves for a product, there is one price at which a monopolist can maximize its profits. This price is determined by computing the quantity of product that is produced at the point where the monopolist’s cost in making one more item (marginal cost) equals the revenue received from selling that additional item (marginal revenue). The price at which the public will buy all of that quantity, but no more (demand curve), will be the optimum monopoly price. If the monopolist charges more than this price, its profits will decline because the lost revenues from the reduced number of sales would more than offset the added revenue from the higher price. If the monopolist charges less, the added revenue from increased sales will not compensate for the reduced revenue per sale and the added marginal costs in producing that quantity. Thus, the monopolist’s profit will be less than at the optimum monopoly price. See id. at 328, citing 3 P. Areeda & D. Turner, Antitrust Law ¶¶ 725-726 (1978), R. Bork, The Antitrust Paradox 242-43 (1973), and R. Posner, Antitrust Cases, Economic Notes & Other Materials 704-08 (1974).
If the demand for the product is at all elastic, forward vertical integration may have substantial procompetitive effects in the form of lower prices and more efficient use of resources. Further, Star Co. points out that a large portion of its revenue comes from advertising. For that revenue Star Co. must compete with many other media and, quite significantly here, advertising revenues, in turn, are dependent upon circulation. The result is that Star Co. has a greater incentive than the contract carriers have to keep the retail price as low as possible in order to increase circulation. Star Co. contends that this need to increase circulation provides a greater retardant effect on prices than any influence it exerted in the market by its mere presence as a potential competitor. Thus, even if one assumes the potential competitor theory to be true, the procompetitive effects lost through the elimination of Star Co. as a potential competitor are more than offset by the increase in procompetitive effects in the retail market generated by optimum monopoly pricing combined with Star Co.’s need to increase circulation and attract advertising.
III. EVALUATION OF ANTICOMPETI-TIVE EFFECTS
Each of the above theories contains a certain amount of truth about the effect Star Co.’s proposed change in distribution will have on the retail newspaper market in the Kansas City area. In applying these theories to the facts of this case, we must bear in mind the fundamental principles of antitrust law discussed above. Moreover, we are mindful that the burden of proof is [702]*702not on the defendant to prove the absence of anticompetitive effects. Rather, it is the plaintiff’s responsibility to prove all the elements of an antitrust violation.
We recognize that on the facts of this case, elimination of Star Co. as a potential competitor will result in the elimination of some procompetitive effects in the retail market. Under the contract carrier system, two competitors existed in the distribution or retail market — one actual (the contract carrier serving the route) and one potential (Star Co.). Implementation of the delivery agent system would leave only one “competitor” in the market — Star Co. Thus, the retail market has lost the beneficial competitive interaction between business entities that the antitrust laws were enacted to preserve. Northern Pacific R.R. v. United States, 356 U.S. 1, 4, 78 S.Ct. 514, 517, 2 L.Ed.2d 545 (1958). However, the loss of the procompetitive effects engendered by the substitution of Star Co. for the contract carriers, that is, Star Co.’s forward integration into the retail market as a self-distributor, may be offset by the introduction into the market of the procompetitive factors under which Star Co. must function as the sole “competitor” in the retail market. The optimum monopoly price theory is useful in ascertaining whether such procompetitive effects are sufficient to counteract the anti-competitive effects of removing potential competition from the market so that in the end there are no unreasonable anticompeti-tive effects.
It may well be true, as the panel majority opinion pointed out, that in certain circumstances a first level monopolist will desire to integrate forward even if it is less efficient than the second level entity. 695 F.2d at 328. These situations include (1) price or service discrimination, (2) increased barriers to entry at the first level, and (3) evasion of government regulation of first level monopoly profits. See F. Scherer, Industrial Market Structure and Economic Performance 302-06 (2d ed. 1980); McGee & Bassett, Vertical Integration Revisited, 19 J.L. & Econ. 17, 18-38 (1976); Note, Refusals to Deal by Vertically Integrated Monopolists, 87 Harv.L.Rev. 1720, 1725-30 (1974). Courts have sought to limit imposing liability upon newspapers which vertically integrate into distribution only to situations where one or more of these incentives are present. See, e.g., Byars, 609 F.2d at 861. But cf. P. Areeda, Antitrust Law 1729.7d (Supp.1982). In the present case, these circumstances do not exist. No price discrimination will result because the purpose of the delivery agent system was to establish a uniform retail price. Nor, as the district court found, will Star Co.’s forward integration raise barriers to first level entry (publication of daily metropolitan newspapers). The delivery agent contracts expressly allow the delivery agents to deliver for other newspapers as long as such delivery will not disrupt the delivery of Star Co. newspapers. Finally, no argument has been made that Star Co. has undertaken to distribute its own newspapers to avoid government regulation of its first level monopoly profits. See 695 F.2d at 341 (Henley, J., dissenting).
We are not unaware of certain record evidence of two possible anticompetitive effects of the delivery agent system: increased prices and reduced services. There is testimony which tends to show that Star Co. believed at the time of trial that it could not deliver newspapers more efficiently than the contract carriers and that any additional profits would have to come from increased revenues. Indeed, according to the panel majority, most Star Co. readers would pay more for their subscriptions under the announced delivery agent retail prices. See 695 F.2d at 330 & n. 10. While it may be true that many readers initially will have to pay more for their subscriptions, it is also true that many readers will pay less. One of the legitimate business reasons advanced by Star Co. in support of the delivery agent system was the establishment of uniform rates and services for all subscribers, and the record indicates that Star Co.’s proposed uniform rates were lower than the rates charged by some independent contract carriers, although admittedly higher than others. Moreover, as the panel dissent pointed out, [703]*703we are not told of the extent, if any, to which the price increases may be attributable to a general increase in costs unrelated to Star Co.’s decision to vertically integrate. Id at 341 (Henley, J., dissenting) (general economic conditions as opposed to change in method of distribution).
Concededly, one procompetitive effect that can be eliminated when potential competition is removed from the distribution market is the downward pressure on price. Indeed, it is said that vertical integration frequently is followed by price increases. See McGee & Bassett, Vertical Integration Revisited, 19 J.L. & Econ. at 27 n. 28. And, even if Star Co. is able to achieve distribution economies by vertical integration, such savings may not result in lower retail prices. See Kamersehen, The Economic Effects of Monopoly: A Lawyer’s Guide to Antitrust Economics, 27 Mercer L.Rev. 1061 (1976), reprinted in T. Calvani & J. Siegfried, Economic Analysis and Antitrust Law 20-26 (1979).
Indeed, a price increase is not necessarily inconsistent with the optimum monopoly price theory to the extent that Star Co.’s concern over its circulation and advertising revenue, the monopoly power of the contract carriers, and the individual contract carriers’ ability to set varying retail prices may have combined to exert a retardant effect on Star Co.’s wholesale prices, thus robbing Star Co. of its ability to reap the full monopoly profit at the wholesale level.
The monopoly power of the second level monopolist, here the individual contract carrier, is exaggerated by the nature of the newspaper industry. As has been noted, an important source of revenues for a newspaper is advertising revenues. Advertising rates are based on the number of subscriptions. The number of subscriptions is in part a function of price: the lower the retail price, the more subscriptions that are sold. When a contract carrier intends to raise its retail price, the newspaper is threatened with a loss of subscriptions caused by the increased prices. Moreover, the newspaper will not recoup any of the lost advertising revenues through increased sales revenues because the independent contract carrier receives 100% of the increase in the retail price. The newspaper is reluctant to raise its wholesale price to compensate for this loss because any increase in the wholesale price may trigger an additional retail price increase by the contract carrier who wants to maintain a certain level of profit. These conditions create an incentive for the newspaper to “share” even more of the monopoly profit with a contract carrier by lowering its wholesale price in order to head off a retail price increase by the contract carrier.
Even though Star Co. may theoretically possess the ability to set the retail price at the optimum monopoly price through its control of the wholesale price, in reality Star Co. probably does not have the information or the ability to analyze the information necessary to actually set the retail price at the optimum level. Depending on the subscription rates charged by each independent contract carrier, the actual retail price may be above or below the optimum monopoly price.
The panel majority felt that elimination of the contract carriers would decrease the range and quality of service provided to Star Co. readers. For example, delivery to country routes would be reduced from twice daily to once daily. Special services ordinarily provided by the contract carriers to suit individual needs of readers would be discontinued by delivery agents. Also, the range of subscription options collectively offered by the contract carriers in the Kansas City area would be reduced from ten options to three options under the delivery agent system. 695 F.2d at 330-31. However, the panel majority considered the services provided to Star Co. readers by contract carriers collectively. In reality, however, each contract carrier was itself a monopolist along its exclusive route. No contract carrier competed with any other contract carrier. See P. Areeda, Antitrust Law 1729.7a, at 197 (Supp.1982) (describing retail distributors of newspapers as “mini-monopolists”). Along any given route, the range of service options was not as great as [704]*704that which existed in the collective market. Thus, to many individual readers, the range of services provided by delivery agents will be greater than what exists under the contract carrier system.
On balance, we agree with the panel dissent that appellees have not borne their burden as plaintiffs of proving that the procompetitive effects generated by optimum monopoly pricing and the unique nature of a newspaper’s revenues are outweighed by the minimal anticompetitive effect of eliminating potential competition from the retail market. Moreover, we find it hard to ignore the fact that every other antitrust case brought against a newspaper publisher challenging the newspaper’s decision to forwardly integrate into distribution has been resolved in favor of the newspaper. See, e.g., White v. Hearst Corp., 669 F.2d 14 (1st Cir.1982); Auburn News Co. v. Providence Journal Co., 659 F.2d 273 (1st Cir.1981), cert. denied, 455 U.S. 921, 102 S.Ct. 1277, 71 L.Ed.2d 461 (1982); Byars, 609 F.2d 843; Hardin v. Houston Chronicle Publishing Co., 572 F.2d 1106 (5th Cir.1978); Knutson v. Daily Review, Inc., 548 F.2d 795 (9th Cir.1976), cert. denied, 433 U.S. 910, 97 S.Ct. 2977, 53 L.Ed.2d 1094 (1977); Grill v. Reno Newspapers, 6 Media L.Rep. (BNA) 1818 (D.Nev.1980); Neugebauer v. A.S. Abell Co., 474 F.Supp. 1053 (D.Md.1979); Newberry v. Washington Post Co., 438 F.Supp. 470; McGuire v. Times Mirror Co., 405 F.Supp. 57; Lamarca v. Miami Herald Publishing Co., 395 F.Supp. 324. Cf. Bunch v. Artec International Corp., 559 F.Supp. 961 (S.D.N.Y.1983) (termination of dicta-phone dealers). Again, we emphasize that there is nothing unlawful about the mere possession of monopoly power. Nor is it unlawful per se for a monopolist to unilaterally refuse to deal with a former distributor or to vertically integrate. However, a monopolist may be subject to antitrust liability if it misuses its monopoly power to accomplish a vertical integration and a refusal to deal that results in unreasonable anticompetitive effects. Each case must be resolved on its own particular facts. In this case appellees have failed to prove that any anticompetitive effects that might result from Star Co.’s vertical integration and refusal to deal are unreasonable. Accordingly, we reverse the judgment of the district court, dissolve the permanent injunction, and vacate the award of attorney fees.