dissenting:
I. INTRODUCTION
At first blush, this ease appears to involve a straightforward application of section 2 of the Sherman Act, 15 U.S.C. § 2 (1988). A natural gas utility,. City Gas Company (City Gas), with an apparent mo*1267nopoly in the natural gas business in south Florida, refused to sell wholesale natural gas to another company, Consolidated Gas Company (Consolidated), that wanted to enter the natural gas business too. Consolidated brought suit under sections 4 and 16 of the Clayton Act, 15 U.S.C. §§ 15(a), 26 (1988),1 claiming that City Gas refused to deal in violation of section 2 and thus requesting an injunction ordering City Gas to supply wholesale natural gas to Consolidated and damages. The district court held in favor of Consolidated and granted the requested relief.
On closer inspection, however, the case’s stripes begin to change. Because of certain factual peculiarities of this case — peculiarities that the district court deempha-sized — established theories of liability under section 2 simply do not apply. To begin with, City Gas operates only at the retail distribution level of the natural gas business; the company does not, and never has, operated at the wholesale level of that business. That is, City Gas obtains a supply of gas from the region’s only wholesale supplier, Florida Gas Transmission Company (FGT), and then distributes that gas to commercial and residential customers. City Gas has chosen to operate only on that level, and its natural gas rates — which are set by the Florida Public Service Commission (FPSC) based on a detailed analysis of City Gas’ operation, see 1989 Fla.Stat. ch. 366, infra 1332-1333 — reflect the specifics of its operation on the retail level. For City Gas to enter the wholesale supply business as well as the retail distribution business would fundamentally change the nature of its operation and thus the structure of its rate base. Consolidated, however, wanted City Gas to alter the nature of its operation and enter the wholesale natural gas business. Consolidated, moreover, wanted City Gas to make this change so that Consolidated could itself enter the retail distribution business and compete with City Gas. And Consolidated thought that City Gas should do this, not because competition would necessarily benefit natural gas consumers in general, but because City Gas provided the most economical source of wholesale gas for Consolidated. According to Consolidated, under these circumstances, section 2 of the Sherman Act required City Gas to do what Consolidated wanted.
Unbelievably enough, Consolidated persuaded the district court that City Gas should be required fundamentally to change the nature of its operation and sell wholesale natural gas to Consolidated so that Consolidated could compete against City Gas in the retail distribution business. Consolidated also persuaded the district court that established theories of liability under section 2 mandated this result. What is even more unbelievable, Consolidated has also persuaded this court, sitting en banc, on these same points. This court, therefore, affirms the district court’s decision.2
With all due respect, I strongly dissent. In my view, the district court’s decision is unsupported by established antitrust law. Because the district court did not take proper account of the fact that City Gas was not in the business of selling natural gas at wholesale, the district court improperly concluded that City Gas possessed monopoly power in the relevant market for purposes of section 2. Based on that erroneous conclusion, the district court also improperly applied to these facts the primary theories of liability under section 2— namely, the “essential facilities doctrine” *1268and the “intent test.” The district court’s decision, therefore, has no basis in established antitrust law.
Instead, the district court in effect created a new rule of liability under section 2. Neither the district court, nor this court today, however, has considered the practical implications of this rule. Having failed to do so, the district court could not have properly applied this new rule in the present case. Even if today’s court upholds this new rule, therefore, I would reverse the district court’s decision and remand for further proceedings as required by the rule. In my view, though, a thorough consideration of the new rule’s practical implications leads unavoidably to the conclusion that this new rule should be struck down. In attempting to apply the rule, the district courts of this circuit will face insurmountable difficulties in fashioning and implementing effective remedies. Even if the courts can overcome these difficulties, which I doubt, the rule will lead to results that will be flatly inconsistent with the policies behind the antitrust laws.
In addition to these fundamental points, I also dissent from the district court’s conclusion that the state-action immunity doctrine of Parker v. Brown, 317 U.S. 341, 63 S.Ct. 307, 87 L.Ed. 315 (1943), did not protect City Gas’ conduct from antitrust liability. I dissent as well from the district court’s conclusion that a 1960 territorial agreement between City Gas and Peoples Gas, Inc. (Peoples Gas), another retail distributor of natural gas in Florida, provided an alternate basis for liability under section 2.
This dissent is organized as follows: in part II, I restate the factual background of the case. In part III, I demonstrate that City Gas’ territorial agreement with Peoples Gas provides no basis for liability. I then turn to the more fundamental problems with the district court’s decision.
In part IV, I criticize the district court’s conclusion that City Gas possessed monopoly power in the relevant market. In part V, I discuss the formulas for determining liability on which the district court relied: the essential facilities doctrine and the intent test. As I demonstrate, none of these formulas provides a basis for liability here. I then discuss some recent Supreme Court and lower court decisions also relied on by the district court. None of these decisions control the present case either.
Next, having established that the district court’s decision has no basis in established antitrust law, I assume that the district court established a new rule of antitrust law and, in part VI, consider the practical implications of that rule. As this part shows, the rule will be impossible to apply, and even if it can be applied, it will produce results that are inconsistent with the policies behind the antitrust laws. For those reasons, I conclude that Congress could not have intended such a rule to be developed under section 2. Because the court today nevertheless upholds this rule, I also demonstrate that the district court failed to apply it properly in the present case. On that basis, I would vacate the district court’s decision and would remand the case for further proceedings as I describe.
In part VII, I then take up the issue of state-action immunity, demonstrating that in this case, the doctrine protects City Gas from antitrust liability. Finally, I conclude the dissent in part VIII.
II. BACKGROUND
This case involves two retail distributors of natural gas in Florida. Consolidated, the plaintiff in this case, entered the gas distribution business in the early 1960s and decided to distribute LP gas. ’ Consolidated received its LP gas by rail or truck from wholesale suppliers and stored it in its own tanks. Consolidated then transported the LP gas from its storage tanks to its customers through a series of underground pipes located in easements. City Gas, the defendant in this case, entered the retail gas business in 1949, also as an LP gas company. In the early 1960s, however, when natural gas became available for resale in Florida, City Gas began to supply natural gas as well. Natural gas differs from LP gas in that it is transportable only by pipeline. In addition, unlike the unregulated interstate trade of LP gas, the Feder*1269al Energy Regulatory Commission (FERC) regulates the transportation of wholesale natural gas through interstate pipelines. FGT owns the interstate pipeline supplying wholesale gas in Florida. In order to acquire a supply of natural gas for resale from FGT, therefore, City Gas had to obtain FERC authorization, which it did. City Gas also had to build a lateral pipe to connect the pipesystem that it used to transport LP gas to FGT’s pipeline.3
In September 1960, City Gas also entered into a territorial agreement with Peoples Gas, another natural gas retail supplier in Florida. Pursuant to this agreement, the two companies determined which territories each would serve and agreed not to compete with one another in those territories. The companies requested, and received, approval of this agreement by the FPSC, which was charged by statute to supervise and regulate public utilities, see Fla.Stat. ch. 366 (1987).4 In subsequent litigation between the two companies, the Florida Supreme Court upheld the agreement as consistent with state antitrust laws and also held that the FPSC had authority under chapter 366 of the Florida Statutes, see id., to approve such agreements. See City Gas Co. v. Peoples Gas Sys., 182 So.2d 429, 435-36 (Fla.1965).
When City Gas entered the natural gas business, the price of natural gas was higher than that of LP gas, but both were relatively low. That remained the case until 1973, when the Arab Oil Embargo (an event beyond City Gas’ control) skyrocketed world oil prices, including the price of LP gas. Until the early 1980s, however, natural gas suppliers were not in much better shape. Although federal regulation of wholesale rates kept the price of natural gas relatively low, a national shortage of natural gas severely limited its availability for resale. By 1981, things began to change: a natural gas surplus developed. As a result, LP distributors such as Consolidated found themselves unable to compete with natural gas prices; they were forced to convert to natural gas, a process that Consolidated began in early March 1982. On the other hand, those companies such as City Gas, which had risked entering the natural gas business at a time when the future of natural gas was relatively uncertain, found themselves in a position to enjoy the benefits of their initiative. A combination of historical accident (the rise in oil prices as the result of events in the Arab world) and business initiative (investing in the natural gas trade) resulted in economic benefits for companies such as City Gas.
In February 1982, one month before Consolidated undertook its conversion, City Gas began to market its natural gas to Consolidated’s LP gas customers. These marketing efforts included door-to-door solicitations, mailing campaigns, and newspaper advertisements. As a result of these efforts, City Gas signed several customers that Consolidated had previously served.5 *1270In order to serve these customers, City Gas expanded its pipeline into Consolidated’s service area.6
Consolidated undertook to convert to natural gas. On March 9, 1982, Consolidated “wrote a letter to FGT requesting that FGT begin providing it with natural gas, and that FGT construct the necessary pipeline facilities.” Consolidated Gas Co. v. City Gas Co., 665 F.Supp. 1493, 1504 (S.D.Fla.1987). As the district court found, FGT responded on April 2, 1982, stating that Consolidated would need a natural gas allocation from FERC in order to purchase wholesale gas from FGT and that Consolidated would have to reimburse FGT approximately $250,000.00 for the cost of connecting to FGT’s pipeline. FGT suggested that Consolidated instead arrange to connect to City Gas’ pipesystem. On May 21, 1982, Consolidated applied to FERC for the allocation approving its connection to FGT’s pipeline. On June 21, 1982, City Gas *1271filed a petition to intervene in the application proceedings in order to oppose Consolidated’s requested allocation. City Gas questioned whether Consolidated’s LP gas distribution system could operate at the pressures required for natural gas delivery and also whether FGT’s natural gas supply was adequate to serve Consolidated as well as City Gas. On November 21, 1983, an administrative law judge granted Consolidated’s application, and FERC affirmed on September 19, 1984. FERC ruled, however, that FGT had to pay the connection costs. The entire administrative process lasted approximately twenty-eight months.
The district court found that City Gas’ intervention in the FERC proceedings delayed the approval process for approximately one year. According to the district court, if City Gas had not intervened, the entire process would have taken only six to nine months, and Consolidated would have received a temporary allocation within ninety days of filing its application. The district court further found that, as a result of City Gas’ intervention and the resulting delay in the proceedings, Consolidated’s business was “severely hurt”: “During the course of the proceedings, City Gas began providing service to seven of Consolidated’s eight commercial customers and the bulk of its residential customers.” Id. at 1510. Presumably, City Gas’ provision of service to Consolidated’s customers resulted from City Gas’ marketing efforts and the fact that (because of historical accident and its own business initiative) it was equipped to expand and supply natural gas to those customers. The district court, however, concluded these activities “were not simply ordinary business expansions, but rather, were basic components of City Gas’ strategy to force Consolidated to sell out.” Id. at 1507.7 Based on these facts, the district court found that “Consolidated was left with only [one] alternative if it was to remain in business: the purchase of natural gas directly from City Gas.” Id. at 1510.8
*1272City Gas and Consolidated had already entered into negotiations regarding the possibility of such purchases. They could not, however, arrive at an acceptable price for the transaction. City Gas initially offered to sell gas for resale at cost plus ten cents per therm9 and later at cost plus five cents per therm. Consolidated rejected the offers as unreasonable and brought suit under sections 4 and 16 of the Clayton Act, 15 U.S.C. §§ 15(a), 26, claiming that City Gas’ conduct constituted a refusal to deal in violation of section 2 of the Sherman Act. The district court agreed with Consolidated and held that City Gas’ offered prices were unreasonably high, thus constituting a refusal to deal in violation of section 2.
The district court also found, however, that City Gas had never before sold gas at wholesale. Although it had entered into a contract in 1965 to sell wholesale gas to another retail distributor, Florida Gas, that contract was abandoned by the parties and City Gas never performed under it. Aside from that isolated (and unperformed) contract, .City Gas operated entirely at the retail level.
Consolidated claimed that City Gas violated section 2 of the Sherman Act both by offering to sell gas to Consolidated at unreasonably high prices, which constituted a refusal to deal under section 2, and by entering into a territorial agreement with Peoples Gas. In part III, I consider the territorial agreement and show that it provides no basis for liability in the present case. In parts IV through VI, I then turn to the purported refusal to deal, demonstrating that it provides no basis for liability either. Finally in part VII, I show that even if City Gas' conduct justifies the imposition of liability under section 2, City Gas is immune under the state-action immunity doctrine of Parker, 317 U.S. at 341, 63 S.Ct. at 307.
III. THE TERRITORIAL AGREEMENT
The court today adopts the district court’s holding that City Gas’ territorial agreement with Peoples Gas constituted a violation of section 2. I disagree with that holding and would reverse on that basis.
Section 2 of the Sherman Act provides in pertinent part: “[ejvery person who shall monopolize, or attempt tu monopolize, or combine or conspire with any other person or persons, to monopolize any part of the trade or commerce among the several States, or with foreign nations, shall be deemed guilty of a felony_” 15 U.S.C. §.2 (1988). As the Supreme Court stated in United States v. Grinnell Corp., 384 U.S. 563, 570-71, 86 S.Ct. 1698, 1704, 16 L.Ed.2d 778 (1966):
The offense of monopoly under § 2 of the Sherman Act has two elements: (1) the possession of monopoly power in the relevant market and (2) 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.
The district court analyzed the territorial agreement under this standard and held that it constituted a “willful acquisition” of monopoly power under the second prong. As I discuss below, the district court should never have reached this issue in the first place because City Gas did not possess monopoly power in the relevant market. See infra at 1275-1283. Putting the monopoly power issue aside for the moment, however, the district court erred in holding City Gas liable under section 2 based on the territorial agreement.
Section 2 itself provides no cause of action for a private plaintiff. Private plaintiffs bring such actions under section 4 of the Clayton Act, 15 U.S.C. § 15, which provides an action for treble damages for “any person who shall be injured in his business or property by reason of anything forbid*1273den in the antitrust laws.”10 Before a private plaintiff can challenge a monopolist’s conduct under section 2 of the Sherman Act, therefore, it must satisfy the requirements of section 4 of the Clayton Act.11
Although section 4 is broadly worded, courts have interpreted the section’s “injured ... by reason of” language as imposing a causation requirement on the private plaintiff. As this court has stated, the plaintiff must show “a causal relationship between the antitrust violation [alleged] and the injury [sustained].” Cable Holdings, Inc. v. Home Video, Inc., 825 F.2d 1559, 1561 (11th Cir.1987) (quoting National Indep. Theater Exhibitors, Inc. v. Buena Vista Distribution Co., 748 F.2d 602, 607 (11th Cir.1984), cert. denied, 471 U.S. 1056, 105 S.Ct. 2120, 85 L.Ed.2d 484 (1985)). To satisfy this “causal relationship” requirement, the plaintiff must show that the violation is “a material cause of,” Zenith Radio Corp. v. Hazeltine Research, Inc., 395 U.S. 100, 114 n. 9, 89 S.Ct. 1562, 1571 n. 9. 23 L.Ed.2d 129 (1969), or “materially contributed to,” Cable Holdings, 825 F.2d at 1562, its injury. Beyond this obvious requirement, the plaintiff must also prove “antitrust injury, which is to say injury of the type the antitrust laws were intended to prevent and that flows from that which makes defendants’ acts unlawful.” Brunswick Corp. v. Pueblo Bowl-O-Mat, Inc., 429 U.S. 477, 489, 97 S.Ct. 690, 697, 50 L.Ed.2d 701 (1977) (emphasis original); see Associated Gen. Contractors, Inc. v. California State Council of Carpenters, 459 U.S. 519, 539-40, 103 S.Ct. 897, 909, 74 L.Ed.2d 723 (1983) (refining Brunswick test); Blue Shield v. McCready, 457 U.S. 465, 476-78, 102 S.Ct. 2540, 2546-48, 73 L.Ed.2d 149 (1982) (same); see also Cargill, Inc. v. Monfort, Inc., 479 U.S. 104, 113-15, 107 S.Ct. 484, 491, 93 L.Ed.2d 427 (1986) (antitrust injury requirement also applies in suits for injunctive relief under section 16 of Clayton Act).12
Although the antitrust injury analysis is somewhat complex, I do not belabor it here. In my view, Consolidated has not satisfied even the basic element of the causation requirement with respect to the territorial agreement: that agreement is simply not a material cause of Consolidated’s injury. Without giving any apparent thought to this causation requirement, the district court summarily concluded that the territorial agreement constituted a “willful acquisition ... of monopoly power” under the second prong of the Grinnell standard and, on that basis, imposed liability on City Gas under section 2. Whether or not the agreement constituted a “willful acquisition” under the Grinnell standard,13 the agreement also had to satisfy the causal connection requirement of section 4. As I demonstrate now, it did not do so.
*1274Consolidated’s injury, as stated in its amended complaint, was the economic loss that followed from its inability to enter the natural gas market as a retail gas distributor. Consolidated, of course, traced that injury to conduct by City Gas. The district court considered two specific forms of conduct by City Gas: (1) its refusal to supply wholesale gas to Consolidated; and (2) its territorial agreement with Peoples Gas. Putting aside the issue of liability under section 2,14 the refusal to deal would be a material cause of Consolidated’s injury: Consolidated could not enter the retail gas market without a supply of gas; City Gas refused to provide such a supply; Consolidated, therefore, could not enter the market and suffered economic loss as a result.
No such causal connection exists, however, between Consolidated’s injury and the territorial agreement. That agreement relates to competition between City Gas and Peoples Gas, two retail distributors of natural gas. Pursuant to the agreement, each company took control of a given distribution area and agreed not to compete for control of the other company’s distribution area.15 The direct causal result of the agreement, therefore, was that City Gas could operate in its distribution area without competition from Peoples Gas. Given that no other competitor seriously challenged City Gas’ control of that distribution area, the agreement may have ensured City Gas’ acquisition of monopoly power in that distribution area. Even without the agreement, however, a single company would eventually have gained control over what became City Gas’ territory. After an initial period of competition, a single natural gas distributor almost inevitably acquires a natural monopoly in a given distribution area. Although the territorial agreement might have affected this process in the present case, I doubt that its absence would have changed the present situation. City Gas, as the company already operating in its current area, would probably have acquired the natural monopoly over service in that area, and if City Gas had not, another company would have done so. With or without the territorial agreement, Consolidated would still have been in the same position: trying to enter the retail distribution market in an area under the control of a monopolist.
Even if the territorial agreement was a material cause of City Gas’ acquisition of monopoly power, it still would not have been a material cause of Consolidated’s injury: City Gas’ possession of monopoly power did not cause Consolidated’s injury. Rather, that injury was caused by how City Gas chose to exercise its monopoly power. Of course, the fact that City Gas had monopoly power was an important condition underlying the whole case. That power led Consolidated to request a supply of wholesale gas from City Gas in the first place and gave City Gas the choice whether or not to sell the gas in the second place. That power, standing alone, however, did not injure Consolidated. City Gas could just as easily have used its monopoly power to help Consolidated as to hurt Consolidated. City Gas’ decision about how to exercise its monopoly power, therefore, determined whether or not Consolidated was injured.16 Thus, the territorial agreement, *1275even if it was a material cause of City Gas’ acquisition of monopoly power, was not a material cause behind City Gas’ decision on how to exercise that power or behind that decision’s result — i.e., Consolidated’s injury-
The district court’s improper treatment of the territorial agreement in this case resulted, in my view, from a failure to grasp the distinction between a private antitrust suit under section 4 of the Clayton Act, which challenges conduct as violative of section 2 of the Sherman Act, and a government enforcement action brought directly under the Sherman Act, see supra notes 1, 10-11. The Grinnell case, for example, involved a government action. The district court, however, cited Grinnell (and only Grinnell) for the proposition that “if as a result of horizontal market division a company acquires monopoly power, it ... violates § 2 of the Sherman Act.” 665 F.Supp. at 1523 (citing Grinnell, 384 U.S. at 576, 86 S.Ct. at 1706-07). Because Grinnell involved a government enforcement action, it does not support the application of this rule to the present case, which of course is a private action. The government in Grinnell brought suit directly under section 2 and therefore did not need to prove that an injury was causally related to the defendant’s conduct. It merely needed to prove an antitrust violation in and of itself. The “willful acquisition” component of Grinnell reflects the specifics of that kind of a case: the government could make a case by proving that the defendant had improperly acquired its monopoly power, whether or not the defendant had actually used that power anticompetitively to injure a competitor. Thus, the government must prove, under the second prong of the Grinnell standard, only that the company with monopoly power acted anticompetitively in a general way. A showing that the company engaged in “the willful acquisition ... of monopoly power,” Grinnell, 384 U.S. at 570-71, 86 S.Ct. at 1704, would satisfy that requirement in a case like Grinnell.17
In a private action under section 2, however, such a showing would not be sufficient to prove liability, unless the company’s “willful acquisition” directly caused the plaintiff’s injury. In order to satisfy section 4’s causation requirement in the present case, therefore, Consolidated would have to show a causal relationship between the territorial agreement and its injury.18 Because Consolidated has made no such showing, nor in my view could it make such a showing, I would reverse the district court’s holding that the territorial agreement provided a basis for imposing liability on City Gas under section 2.
IV. MONOPOLY POWER
Putting aside the issue of the territorial agreement, the court today also adopts the *1276district court’s holding that City Gas refused to deal in violation of section 2 of the Sherman Act. Again, as the Supreme Court stated in Grinnell, 384 U.S. at 570-71, 86 S.Ct. at 1704:
The offense of monopoly under § 2 of the Sherman Act has two elements: (1) the possession of monopoly power in the relevant market and (2) 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.
The district court held that City Gas’ conduct implicated both of these elements and thus constituted an antitrust violation. I first analyze the district court’s discussion of City Gas’ monopoly power; then, in Part V, I address the district court’s discussion of the “willful acquisition or maintenance” element.
The district court held, and the court today agrees, that City Gas possessed monopoly power in both the wholesale and retail natural gas markets. I agree that City Gas had monopoly power in the retail natural gas market; however, I question whether City Gas had monopoly power in the wholesale natural gas market. That City Gas had such power in the wholesale market, moreover, is crucial to the theories on which the district court and this court today have based City Gas’ antitrust liability.
In conducting its monopoly power analysis, the district court followed the standard approach, first defining the relevant product and geographic markets and then determining whether City Gas possessed monopoly power within those markets. With respect to the relevant product market, the district court restated the Supreme Court’s holding in United States v. E.I. DuPont de Nemours & Co., 351 U.S. 377, 404, 76 S.Ct. 994, 1012, 100 L.Ed. 1264 (1956), that the “market is composed of products that have reasonable interchangeability for the purposes for which they are produced — price, use and qualities considered.” 665 F.Supp. at 1516 (quoting DuPont). The district court concluded that the relevant product market was limited to natural gas and did not include LP gas because of its prohibitive expense. The district court then defined the relevant geographic market as “the area of effective competition in the known line of commerce ... in which the seller operates, and to which the purchaser can practicably turn for supplies.” Id. at 1518 (quoting Tampa Elec. Co. v. Nashville Coal Co., 365 U.S. 320, 327, 81 S.Ct. 623, 628, 5 L.Ed.2d 580 (1961)). Applying this definition, the district court concluded that the relevant geographic market was City Gas’ service area and was not limited to Consolidated’s service area.
The district court next considered whether City Gas possessed monopoly power within this market. The district court defined monopoly power as “the ‘power to control prices or exclude competition,’ ” id. at 1519 (quoting Grinnell, 384 U.S. at 571, 86 S.Ct. at 1704), and stated that “[frequently courts have approached the problem by defining first the relevant product and geographic markets and then computing the defendants’ market share from this statistical data," id. The district court then found that “[i]n the wholesale market, City Gas sold 100% of the natural gas requested for resale in its service area.” Id. Based on this finding, the district court held that City Gas possessed monopoly power within the wholesale natural gas market. On appeal, the panel adopted this holding as well. See Consolidated Gas Co. v. City Gas Co., 880 F.2d 297, 300-01 (11th Cir.1989).
City Gas has challenged this holding, contending that it could not possibly have monopoly power because it sold no wholesale gas whatsoever. Rather, City Gas claims, FGT sells 100% of the gas for resale in the relevant service area. In response to this argument, the panel reasoned that although City Gas had never sold gas for resale, it had power to do so and had in fact entered into a contract (in 1965) to supply gas for resale to a retail natural gas distributor named Florida Gas. That contract, however, was abandoned by the parties, and it is undisputed that City Gas never actually sold gas at wholesale— either to Florida Gas or to anyone else. *1277Putting the Florida Gas contract aside for a moment, I first demonstrate in section A that City Gas could not have possessed monopoly power in the relevant market defined by the district court because City Gas never actually operated in that market. I then consider in section B the abandoned contract and conclude that it has no relevance to the issue of whether City Gas possessed monopoly power in the wholesale market in this case.
A. The Relevant Market.
I suggest that the monopoly power issue is more complicated than the district court’s analysis reflects. Consider first the complexities of the factual situation here. Consolidated wants gas for resale. FGT supplies gas for resale, but in order to acquire gas from FGT, City Gas must spend $250,000 to construct a lateral pipeline and must wait approximately two years for a FERC allocation. City Gas distributes gas at retail but also has the “power” to supply gas for resale to Consolidated. City Gas, however, has never actually sold gas for resale, although it considered doing so and even went as far as entering into a contract to do so (which was subsequently abandoned and never performed, see infra at 1282-1283). Now, assuming for the moment that Consolidated' really did not have the option of obtaining gas from FGT — the cost and delay involved were just too prohibitive — Consolidated must then acquire gas from City Gas or it will be unable to compete. It would seem, therefore, that City Gas has the power to control prices and exclude competition in the relevant market. At least with respect to Consolidated, City Gas does not compete with any other wholesale supplier: City Gas can independently set the wholesale price and restrict (or deny entirely) Consolidated’s ability to acquire a wholesale supply. This is exactly the conclusion that the district court reached. This conclusion assumes, however, that City Gas is a wholesale gas supplier. We know that City Gas has the power to supply gas at wholesale— the panel emphasized that point — but (as the district court itself found) City Gas has chosen not to exercise that power. It has the potential to sell gas at wholesale, but it simply has not yet undertaken to exercise that potential. In my view, this fact must be addressed before the district court can conclude that City Gas has monopoly power in the wholesale market. Neither the district court nor the court today has adequately addressed this point, though; nor, in my view, could they do so. The district court has defined a market that simply does not exist. In support of my conclusion, in subsection 1,1 first consider a basic but -very important issue: what is a market? I then show in subsection 2 that the district court’s construction of the relevant market is inconsistent with the basic notion of a market. I focus specifically on the district court’s conclusion regarding the relevant geographic market in this case.
1. The Concept of a Market.
Economists hold that a market exists when buyers and sellers exchange goods or services. See, e.g., The MIT Dictionary of Modern Economics 263 (D. Pearce ed. 1986) (market is “any context in which the sale and purchase of goods and services takes place”); J. Hanson, A Dictionary of Economics and Commerce 255 (6th ed. 1986) (“In a wider sense, however, a market can signify any area in which buyers and sellers are in contact with one another_”); see also C. Ferguson & J. Gould, Microeconomic Theory 259 (4th ed. 1975).19 This idea of exchange can also be defined in terms of supply and demand: an exchange can occur if a buyer wants goods or .services that a seller can supply. However defined, exchange is a function of price. A buyer’s desire to buy and a seller’s desire to sell are qualified by price: the parties desire to proceed only at a certain price, or within a certain range of prices. Beyond that price, or outside of that range of prices, the parties do not desire to proceed *1278with the transaction at issue. The market, then, is the mechanism by which price is established. The market serves as such a mechanism by providing a history of transactions (both complete and incomplete) in the context of which parties negotiate their price.20
A market, obviously, exists independently of whether or not a given transaction is completed — i.e., whether the parties agree to a price. A failed transaction merely becomes part of the history that serves as the context for future negotiations. The present case, however, involves something entirely different than a failed transaction in an established market. In the present case, no market existed in which the parties could reach an acceptable price. Neither City Gas nor any other retail distributor in the region has ever sold gas at wholesale to another retail supplier. The parties in this case could refer to no history of transactions (or prices) as a basis for setting a price term for their potential transaction. When the transaction failed in this case, therefore, its failure reflected the absence of an established market that could have enabled the parties to negotiate an exchange. The transaction’s failure, moreover, meant that no such market came into existence as a result of the transaction. Thus, the absence of a market for the transaction at issue in the present case explains, at least in part, the inability of City Gas and Consolidated to agree to a transaction price. More importantly, the absence of a market — both as a context for the transaction or as a result of the transaction — meant that the district court here could not determine what the transaction price should have been.
Consider a hypothetical situation: in 1970, Senior Partner at a large law firm with a respected antitrust practice strolls into the office of Acme Rocket and Spaceship’s President. Acme, of course, built the spacecraft that recently landed on the moon. President greets Senior Partner (they went to college together) and asks him what’s on his mind. Senior Partner (who has been billing 3000 hours a year for thirty-five years) tells President that he wants a branch office orbiting the moon and would like Acme to supply a spaceship for that purpose. Senior Partner also indicates that he can find no other company capable of building such a spaceship for him. President says that Acme can do the job, and the two begin to discuss a price. Senior Partner says that his firm will be willing to pay five million dollars. President smiles politely and gently suggests that the required spaceship would cost nearly a billion dollars to design and build. Senior Partner responds that at most his firm could manage to pay ten million. President then says that because Senior Partner is such an old friend, and the publicity from the spaceship would be so good, Acme could build it for half price, or $500 million. Senior Partner says he understands and resigns himself to spending the rest of his career in his two-floor corner office in New York. President claps him *1279on the back, and the two head off to the golf course.
As this negotiation suggests, no “market” for private spaceships existed at the time. True, Senior Partner wanted a spaceship, and Acme could theoretically have built one — that is, a supply and demand existed for private spaceships. But no history of a price existed for such a transaction, and the negotiation obviously failed to establish one. No price existed because no market existed — no mechanism by which the parties to this potential transaction could establish a' price. Following the district court’s reasoning in the present case, however, one would have to conclude that a market for private spaceships did exist in 1970 and, in addition, that Acme possessed monopoly power in that market. These conclusions are obviously ludicrous, but -they necessarily follow from the district court’s definition of a market: potential demand plus potential supply.
2. The District Court’s Analysis.
With this discussion of the idea of a market in mind, I now consider the district court’s specific conclusions regarding the relevant market in this case. I focus on the district court’s analysis of the relevant geographic market. As I discuss above, the district court defined the relevant geographic market as “the area of effective competition in the known line of commerce ... in which the seller operates, and to which the purchaser can practicably turn for supplies.” 665 F.Supp. at 1518 (quoting Tampa Electric, 365 U.S. at 327, 81 S.Ct. at 628). This definition consists of two parts: defining the relevant area first in terms of where the seller sells and then in terms of where the purchaser can go to purchase. Both parts emphasize a market’s practical realities: with respect to the seller, the inquiry looks to the “area of effective competition,” and with respect to the purchaser, the inquiry looks to the area where the purchaser “can practicably turn for supplies.” Tampa Electric, 365 U.S. at 327, 81 S.Ct. at 628 (emphasis added).
Two questions arise from the district court’s application of this inquiry. First, I question whether a firm could in fact be a “seller” under this definition if the firm does not operate, and has never operated, in the “area of effective competition” at issue. I doubt that a firm can properly be labeled a wholesale “seller” of a product if that firm sells that product only on the retail level. Second, I question whether a firm wanting to purchase supplies at wholesale can “practicably turn” to an entity that does not deal in the product at issue. The district court, however, answered both of these questions in the affirmative, and in so doing developed a new rule of antitrust liability. That is, if given the circumstances of a market, a purchaser can “practicably turn” nowhere for its supplies, then a court is justified to compel another entity, even though it is not in the supply business at issue, to enter that business in order to provide the purchaser with its supplies.
At bottom, this is the principle underlying the district court’s analysis of the monopoly power issue in the present ease. The district court backed into its holding that City Gas possessed monopoly power after concluding that Consolidated had no other practicable source of supply. Not only is that conclusion itself erroneous, as I discuss below, see infra at 1275-1283, but as a result of that conclusion, the district court constructed the relevant geographic market so as to cut out FGT — the regional wholesale supplier of natural gas. In my view, a properly constructed wholesale market in this case would necessarily include FGT. The relevant market must include FGT because City Gas is not a “seller” as defined in Tampa Electric. .Thus, Consolidated could not practicably turn to City Gas for its supply. Without FGT’s presence, therefore, no wholesale market would exist. This would be the case, moreover, whether or not Consolidated could practicably turn to FGT for its natural gas supply. As I suggest below, however, the district court erred in concluding that Consolidated could not have done so.
Based on its findings of fact, as I discuss above, the district court concluded that “Consolidated was .left with only [one] alternative if it was to remain in business: *1280the purchase of natural gas directly from City Gas.” 665 F.Supp. at 1510. The district court reached this conclusion based on considerations of cost, time, and competition. In the district court’s view, the $250,-000 cost of interconnecting to FGT’s pipeline posed a substantial barrier to Consolidated's entry into the natural gas market. In addition, until November 1984, Consolidated lacked the FERC allocation necessary to allow it to interconnect with the FGT pipeline at all. Thus, from the fall of 1982 (that is, after the approximately three-to-six months that approval of the FERC allocation would have required, according to the district court, had City Gas not intervened in the FERC proceedings) until November 1984, Consolidated could have purchased gas only from City Gas. The district court also emphasized the competition between Consolidated and City Gas. When Consolidated decided to convert from LP to natural gas, it was, as the district court put it, “forced to compete with City Gas in order to retain its customers or else see them go to City Gas for more economical gas service.” Id. at 1505. In the district court’s view, as a result of this competition, “it became crucial that Consolidated convert to natural gas for as low a cost as possible in order to be able to offer rates that were competitive with those City Gas could offer.” Id. at 1506.
The district court then translated these conclusions into a hypothetical relevant market. More specifically, the district court took its findings concerning the conditions necessary to enable Consolidated to compete with City Gas and used them as a basis for constructing a relevant (but hypothetical) market. The district court’s effort in constructing this market, however, was propelled by the assumption that Consolidated was entitled to compete. Driven by this assumption, the district court disregarded as features of the market some of the very conditions stipulated by its findings — namely, the entry barriers in terms of time, cost, and competition.
Rather than structuring a hypothetical market according to the terms that would be necessary to enable Consolidated to compete on equal footing with firms already operating in that market, the district court should have examined the structure of the market as it actually existed (entry barriers and all). Based on that structure, the district court should then have analyzed whether City Gas actually possessed monopoly power and whether it had in fact used that power to prevent Consolidated from competing. The focus of the antitrust laws is to prevent anticompetitive conduct by monopolists, but not to enable an entity to compete whether or not a monopolist acts anticompetitively. The antitrust laws are designed to create an equal opportunity for competition, see Charles A. Ramsay Co. v. Associated Bill Posters of United States & Canada, 260 U.S. 501, 512, 43 S.Ct. 167, 168, 67 L.Ed. 368 (1923); they do not create an absolute entitlement to compete successfully. Given this distinction, we must be very careful here in ascribing to City Gas responsibility for the entrance barriers to the retail natural gas business. If those entry barriers (in terms of time, cost, and competition) are simply too high to enable Consolidated to compete (independent of whether City Gas possessed monopoly power or has used that power anticompetitively), then the district court cannot mandate that City Gas underwrite Consolidated’s entry into the retail natural gas market, especially when that mandate would require City Gas to undertake a new level of operation and limit its own legitimate opportunities for expansion.
As it actually exists, then, the relevant geographic market, or “the area of effective competition in the known line of commerce ... in which the seller operates, and to which the purchaser can practicably turn for supplies,” 665 F.Supp. at 1518 (quoting Tampa Electric, 365 U.S. at 327, 81 S.Ct. at 628), necessarily included FGT’s pipeline. In a case such as this one, where entities exist that are actually engaged in the “line of commerce” at issue, the district court erred in holding that the “effective competition in the known line of commerce ... in which the seller operates” consisted of another entity that was not actually engaged in that “line of commerce" even if that entity could hypothetically be engaged in *1281such commerce and even if that entity would provide the antitrust plaintiff with the most economically feasible supply.
As this discussion demonstrates, therefore, the district court’s finding that City Gas possessed a 100% share of this market is clearly erroneous. Because the district court misconceived the nature of the relevant market, its finding regarding City Gas’ share of that market is meaningless. Within the actual wholesale market, FGT clearly possessed the inordinate share, and City Gas possessed virtually none, if any at all.
In addition, the district court erred in concluding that Consolidated could not “practicably” have turned to FGT for its supply. The district court, as I discuss in detail above, see supra at 1265-1266, made three findings on this point: (1) the FERC proceedings, which resulted in an order granting Consolidated’s request for an allocation to interconnect to FGT’s pipeline, lasted approximately twenty-eight months; (2) construction of the interconnection would cost Consolidated $250,000; and (3) Consolidated would have to compete with City Gas. These findings, however, do not support the conclusion that Consolidated could not practicably have turned to FGT for a wholesale gas supply. Taking these same factors into account, the FERC order issued September 19, 1984, which granted the allocation, held that FGT was “the applicant’s most feasible source of supply.” 21 As the district court itself restated, the FERC order opined that, although competition with City Gas would be difficult, Consolidated’s entry into the natural gas distribution market — with FGT as its supplier— was economically feasible. When FERC issued its order, the delay in gaining approval of the allocation had already occurred. Similarly, City Gas had already placed itself in a strong competitive position; indeed, City Gas had accomplished that well before Consolidated even began its conversion to natural gas. Thus, from Consolidated’s perspective, the only practical difference between turning to FGT as opposed to City Gas for a wholesale gas supply was that the first option would cost Consolidated an additional $250,000.
The record on appeal in this case further indicates that, in light of Consolidated’s assets and its projected profits and business opportunities, it was economically feasible for Consolidated to cover the $250,-000 expense of interconnecting to FGT’s pipeline. I test this conclusion by considering whether, given the evidence Consolidated presented at trial concerning its expected future profits as a natural gas company with a supply of gas, Consolidated could feasibly have obtained underwriting for a bond issuance in the amount of $250,000 in order to fund the cost of hooking up to FGT. At trial, Consolidated called an expert witness, Ben Ball, to the stand to render an opinion regarding the company’s present value with a natural gas supply. This value was, in effect, what a willing buyer, fully informed of all relevant facts, would pay a willing seller, also so informed. Ball assumed that Consolidated could indefinitely obtain a “competitively priced source of natural gas,” 665 F.Supp. at 1514, and could continue to operate profitably, with that supply of gas, for twenty years.22 He then estimated Consolidated’s annual profits for that period, including profits that it would have made by servicing former customers that had been captured by City Gas. From those profits, he deducted the expenses that Consolidated would have incurred in transforming its operation to natural gas distribution; this deduction included the approximately $20,-*1282000 cost of hooking up to City Gas’ pipesys-tem. Ball assumed that the hypothetical buyer would expect a ten-percent return on its investment, and, using this rate, he found the present value of the projected stream of profits over the twenty-year period: the figure was $2,275,130. Id. at 1515. From these figures, it follows that Consolidated would generate annual profits of approximately $267,000 (i.e., a willing buyer would pay $2,275,130, at a ten-percent interest rate, for a twenty-year annuity paying $267,000 a year).23
Now, assume that Consolidated issues a $250,000 bond, to mature in twenty years, and pays interest at a high rate, say fifteen percent annually. Given that rate, Consolidated’s nominal annual interest payments would be $37,500. Because Consolidated could deduct those interest payments from its taxes — paid at a corporate rate of thirty-four percent, see I.R.C. § 11(b)(1) (1988)— its real interest expense would be $24,750 annually. After paying this amount (which would represent a mere 9.3% of Consolidated’s annual profits of $267,000), Consolidated would still generate annual profits in the amount of $242,250, and its present value as a company generating $242,250 in annual profits over the next twenty years (assuming a ten-percent interest rate) would be roughly $2,060,000. In addition, given the amount of its projected future profits, one would expect Consolidated to have little difficulty in paying back $250,-000 in principal at the end of twenty years.
As these figures suggest, a reasonable underwriter would obviously have found these hypothetical bonds to be a safe investment and would have underwritten them. Consolidated, therefore, could have “practicably turned” to FGT for a supply of wholesale gas. As a result, the district court erred in excluding FGT from the relevant geographic market, and if FGT had been included in the relevant market, the district court obviously could not have held that City Gas possessed monopoly power in that market, even if City Gas did sell at wholesale. The fact remains, however, that City Gas never sold natural gas at wholesale. In essence, then, the issue whether Consolidated could practicably have purchased wholesale gas from FGT is irrelevant to this case. Whatever sources Consolidated may or may not have found, City Gas was not one of them. City Gas simply did not participate in the relevant market.
B. The Florida Gas Contract.
In addition, City Gas’ contract with Florida Gas has no significance for the issue of monopoly power. To begin with, the contract was abandoned, and City Gas never performed. The panel acknowledged that “City Gas never actually sold any gas under this agreement” but stressed that “it had the power to do so.” 880 F.2d at 301 (emphasis in original). I agree. The fact that City Gas had the power to sell at wholesale, however, does not mean that it had monopoly power in the wholesale market, that it was actually in the wholesale market, or that a wholesale market even existed. The abandoned contract has no relevance to these issues. On the contrary, the fact that the contract was abandoned, as the district court agreed, indicates that City Gas never actually sold at wholesale, choosing for whatever reason not to exercise its hypothetical power in the wholesale market. In my view, that is the significant point: that City Gas legitimately decided not to exercise its power — not selling to Consolidated, to Florida Gas, or to any other firm. Pursuant to this decision, City Gas did not operate as a wholesale natural gas supplier.
In addition, the abandoned contract provided no evidence that the wholesale market, as defined by the district court, was anything more than hypothetical. As I discuss above, see supra at 1277-1279, a market is a mechanism for establishing price; a wholesale price, therefore, provides evidence that a wholesale market exists for a given product in a given geographical area. Thus, the Florida Gas contract could provide evidence that a wholesale natural gas *1283market existed only if the contract contained a reliable price term. If the contract did not contain a reliable price term, then it could provide no evidence that the parties to the contract were able to negotiate a price term — i.e., that a market existed. I suggest that the district court in this case implicitly found that the contract did not contain a reliable price term. That is, the district court did not look to the price term in this contract as a basis for deciding that the price that City Gas quoted to Consolidated was unreasonable, thus constituting a refusal to deal, or as a basis for setting the price at which City Gas had to sell to Consolidated as ordered in the district court’s injunction. In fact, the district court, .never decided what a reasonable price would be in the present case and instead delegated to the FPSC the task of setting a price for the forced sale. Id. at 1512, 1534, 1545. The district court, therefore, could not have had much confidence in the price term set in City Gas’ contract with Florida Gas. And this is not surprising, given that the contract was abandoned. Because the contract did not contain a reliable price term, therefore, it could not provide evidence that a wholesale market actually existed — let alone that City Gas operated in that market.
Even if the district court found that the contract did contain a reliable price term, indeed even if City Gas had actually performed under the contract, in my view the contract would still be irrelevant to the issue whether City Gas possessed monopoly power in the relevant market. City Gas’ transaction with Florida Gas would not have occurred in the relevant geographic market, accepting for the sake of discussion the district court’s definition of that market here.
The district court found that the relevant geographic market consisted of City Gas’ service area, Consolidated’s service area, and unserved portions of southwest Dade County. 665 F.Supp. at 1518-19. Thus, the district court defined the relevant area in terms of where Consolidated could look for its gas supply and where it could look to distribute that gas. The compatible area in Florida Gas’ case would be totally different. Florida Gas operated in a different area of Dade County than the area in which Consolidated operated, and Florida Gas did not look to expand into southwest Dade County as did Consolidated. The only area of overlap between the geographic areas involved in the Florida Gas contract and in the present case would be City Gas’ service area. Even that area, however, differs in the two situations. The contract with Florida Gas dates back to 1965; Consolidated did not begin its efforts to enter the natural gas business until 1982. Between 1965 and 1982, City Gas’ service area changed significantly. As the district court itself emphasized, the size of that service area is an important' factor in defining the relevant geographic market:
the relevant geographic market should-encompass all of the service area City Gas serves.... The effects of demand and cost are spread throughout City Gas’ service area through the rate regulation process. Thus, City Gas can cross-subsidize prices or costs in any one part of its territory with revenues earned in the rest of its service area.
665 F.Supp. at 1518. For purposes of establishing the relevant geographic market, therefore, City Gas’ service area in 1982 had no relation to its service area in 1965 at the time of the Florida Gas contract.
Even applying the factors used by the district court to define the relevant geographic market in the present case, it is apparent that the relevant geographic market for the Florida Gas contract was entirely different. Even if the parties had not abandoned that contract, it would have been performed in a different relevant market than the one involved here and therefore would have had no bearing on the question whether City Gas operated in the relevant market, let alone possessed monopoly power in that market.
For these reasons, in my view, the district court erred in holding that City Gas possessed monopoly power in the wholesale natural gas market.
V. LIABILITY UNDER SECTION 2
A. Legal background.
Section 2 of the Sherman Act provides in pertinent part: “Every person who shall *1284monopolize, or attempt to monopolize, ... any part of the trade or commerce among the several States ... shall be deemed guilty of a felony_” 15 U.S.C. § 2 (1988). As the Supreme Court has stated:
The offense of monopoly under § 2 of the Sherman Act has two elements: (1) the possession of monopoly power in the relevant market and (2) 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). In order to violate section 2, therefore, the monopolist must possess monopoly power and must also use that power anticompetitively. Section 2 does not prohibit the possession of monopoly power in and of itself—of “monopoly in the concrete.” Standard Oil Co. v. United States, 221 U.S. 1, 62, 31 S.Ct. 502, 516, 55 L.Ed. 619 (1911); see Byars v. Bluff City News Co., 609 F.2d 843, 853 (6th Cir.1979); Berkey Photo, Inc. v. Eastman Kodak Co., 603 F.2d 263, 273 (2d Cir.1979) (citing Standard Oil), cert. denied, 444 U.S. 1093, 100 S.Ct. 1061, 62 L.Ed.2d 783 (1980).
In United States v. Griffith, the Supreme Court addressed the requirement of anticompetitive conduct and characterized it as “the use of monopoly power, hov/ever lawfully acquired, to foreclose competition, to gain a competitive advantage, or to destroy a competitor.” 334 U.S. 100, 107, 68 S.Ct. 941, 945, 92 L.Ed. 1236 (1948). In determining what conduct falls into this category, the idea of intent has always been central. As the Court stated in United States v. Colgate & Co., 250 U.S. 300, 307, 39 S.Ct. 465, 468, 63 L.Ed. 992 (1919): “[i]n the absence of any purpose to create or maintain a monopoly, the [Sherman A]ct does not restrict the long recognized right of trader or manufacturer engaged in an entirely private business, freely to exercise his own independent discretion as to parties with whom he will deal.” (Emphasis added.) The Court also emphasized the importance of intent in Griffith, when it indicated that mere possession of monopoly power could “itself [be] a violation of § 2, provided it is coupled with the purpose or intent to exercise that power” to foreclose competition. 334 U.S. at 107, 68 S.Ct. at 945 (emphasis added).
As the Court has recognized, however, “no monopolist monopolizes unconscious of what he is doing.” Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585, 602 n. 28, 105 S.Ct. 2847, 2857 n. 28, 86 L.Ed.2d 467 (1985) (quoting United States v. Aluminum Co. of Am., 148 F.2d 416, 432 (2d Cir.1945) (L. Hand, J.)). The effects of a defendant’s conduct on competitors and, more importantly, on consumers and on competition in general are also important factors in evaluating whether the conduct at issue is anticompetitive. See id. at 605, 105 S.Ct. at 2858-59. This point merely suggests that, as an evidentiary matter, “the requisite intent [under section 2] can be inferred if a defendant maintains his power by conscious and willful business policies, however legal, that inevitably result in the exclusion or limitation of actual or potential competition.” Hecht v. Pro-Football, Inc., 570 F.2d 982, 990 (D.C.Cir.1977) (citing Aluminum Co., 148 F.2d at 428-31), cert. denied, 436 U.S. 956, 98 S.Ct. 3069, 57 L.Ed.2d 1121 (1978).
A primary type of anticompetitive conduct in section 2 cases — and the type of conduct at issue here — is a defendant’s refusal to deal with a competitor or potential competitor. Such refusals to deal fall into one of two general categories. One category involves monopoly leveraging. Monopoly leveraging occurs when a vertically integrated monopolist — i.e., a monopolist that operates on several levels of the market— uses its monopoly power on one level of the market to gain a competitive advantage on another level of the market.24 For example, in United States v. Terminal R.R. Ass’n., 224 U.S. 383, 32 S.Ct. 507, 56 L.Ed. 810 (1912), a group of railroads servicing *1285the St. Louis area also owned the only railroad terminal that provided access to the city. The terminal owners used their monopoly power over the terminal to overcharge other railroads for access to the terminal and thereby to gain a competitive advantage in the downstream market for railroad service in St. Louis. The majority of refusal-to-deal cases under section 2 fall into this monopoly-leveraging category. See, e.g., Otter Tail Power Co. v. United States, 410 U.S. 366, 93 S.Ct. 1022, 35 L.Ed.2d 359 (1973); Associated Press v. United States, 326 U.S. 1, 65 S.Ct. 1416, 89 L.Ed. 2013 (1945); Eastman Kodak Co. v. Southern Photo Materials Co., 273 U.S. 359, 47 S.Ct. 400, 71 L.Ed. 684 (1927); Terminal R.R., 224 U.S. at 383, 32 S.Ct. at 507; Fishman v. Estate of Wirtz, 807 F.2d 520 (7th Cir.1986); Olympia Equip. Leasing Co. v. Western Union Tele. Co., 797 F.2d 370 (7th Cir.1986), cert. denied, 480 U.S. 934, 107 S.Ct. 1574, 94 L.Ed.2d 765 (1987); Catlin v. Washington Energy Co., 791 F.2d 1343 (9th Cir.1986); Paschall v. Kansas City Star Co., 727 F.2d 692 (8th Cir.), cert. denied, 469 U.S. 872, 105 S.Ct. 222, 83 L.Ed.2d 152 (1984); MCI Communications Corp. v. AT & T, 708 F.2d 1081 (7th Cir.), cert. denied, 464 U.S. 891, 104 S.Ct. 234, 78 L.Ed.2d 226 (1983); Byars, 609 F.2d at 843; Berkey Photo, 603 F.2d at 263; Hecht, 570 F.2d at 982; Poster Exch., Inc. v. National Screen Serv. Corp., 431 F.2d 334 (5th Cir.1970), cert. denied, 401 U.S. 912, 91 S.Ct. 880, 27 L.Ed.2d 811 (1971).
The second category of refusal-to-deal cases under section 2 does not involve monopoly leveraging, but rather involves a monopolist’s use of its monopoly power on one level of the market to maintain or expand its power on that same level of the market. For example, in Lorain Journal Co. v. United States, 342 U.S. 143, 72 S.Ct. 181, 96 L.Ed. 162 (1951), a single newspaper in a small town had a monopoly over the newspaper advertising business in that town. When a local radio station began to sell radio advertising, the newspaper refused to sell advertising space to anyone also advertising on the radio. The newspaper thus anticompetitively used its monopoly power on one level of the market to maintain its power on that same level by preventing the radio from competing for advertisements. Although the majority of section 2 cases fall into the first category, some important cases are of this second type. See, e.g., Aspen, 472 U.S. at 585, 105. S.Ct. at 2847; Lorain Journal, 342 U.S. at 143, 72 S.Ct. at 181; Official Airline Guides, Inc. v. FTC, 630 F.2d 920 (2d Cir.1980), cert. denied, 450 U.S. 917, 101 S.Ct. 1362, 67 L.Ed.2d 343 (1981).
The key to both categories of refusal-to-deal cases — whether they involve monopoly leveraging or exercises of monopoly power on a single level of the market — is that the circumstances surrounding the refusal to deal must give rise to an inference of anti-competitive intent. Intent is rarely established by direct evidence; almost invariably, it is established by circumstantial evidence: the nature of the defendant’s conduct and its effect on competition in the market demonstrate that the defendant intended to foreclose competition in the relevant market. In order to decide whether a refusal to deal of either type constitutes anticompetitive conduct under section 2, courts have developed two formulas for analyzing the circumstantial evidence in a given case: (1) the “essential facilities doctrine” and (2) the “intent test.” These formulas aid the court in weighing and evaluating the circumstantial evidence in a given case to determine whether that evidence yields an inference of an intent to foreclose competition.
The first formula, the essential facilities doctrine, holds that an antitrust defendant may have acted anticompetitively if the circumstances of a case satisfy a four-part test: (1) the defendant is a monopolist in control of a facility or resource that is essential to a competitor’s operation; (2) the facility or resource cannot practically or reasonably be duplicated by competitors; (3) the monopolist refuses to deal with competitors; and (4) the monopolist could feasibly deal with competitors. See MCI, 708 F.2d at 1081; Hecht, 570 F.2d at 982. If the four elements of this standard are sat*1286isfied, an inference of the defendant’s anti-competitive intent may arise and thus justify the imposition of liability under section 2.
The second formula, the intent test, directly addresses the issue of intent by evaluating the monopolist’s conduct and determining whether, under the circumstances of the case, the conduct at issue demonstrates an illegal intent to destroy competition. If the conduct gives rise to an inference of anticompetitive intent, then the monopolist is subject to liability under section 2. See Lorain Journal, 342 U.S. at 152-53, 72 S.Ct. at 186; Eastman Kodak, 273 U.S. at 375, 47 S.Ct. at 404; MCI, 708 F.2d at 1148-49; Poster Exchange, 431 F.2d at 339-40.
In defining the character and limits of these tests, it is essential to understand their function as mechanisms for evaluating circumstantial evidence to determine whether the defendant acted — i.e., refused to deal — with an anticompetitive intent. The tests have validity only to the extent that they enable such a determination. Even if a court, for example, could mechanically apply the elements of the essential facilities doctrine to a given case, if that application would not support an inference of anticompetitive intent, then the test would not justify the imposition of liability under section 2. Indeed, as an analysis of the essential facilities cases demonstrates, liability in those cases is supported by more than the mechanical application of the doctrine's four-part test. The cases all possess certain features (beyond the test’s four elements) that support an inference of anticompetitive intent.
One primary feature of the essential facilities cases is that they all involve monopoly leveraging. See, e.g., Otter Tail, 410 U.S. at 366, 93 S.Ct. at 1022; Associated Press, 326 U.S. at 1, 65 S.Ct. at 1416; Terminal R.R., 224 U.S. at 383, 32 S.Ct. at 507; Fishman, 807 F.2d at 520; MCI, 708 F.2d at 1081; Byars, 609 F.2d at 843; Berkey, 603 F.2d at 263; Hecht, 570 F.2d at 982.25 In fact, the courts in these cases, after stating the four elements of the doctrine, generally explain that the doctrine’s underlying rationale is a concern about monopoly leveraging. The MCI court, for example, after describing the four-part test, stated that “[sjuch a refusal may be unlawful because a monopolist’s control of an essential facility (sometimes called a ‘bottleneck’) can extend monopoly power from one stage of production to another, and from one market into another.” 708 F.2d at 1132.
Another distinguishing feature of the essential facilities cases is that, before the defendant refused to deal with the firm at issue, the defendant was already engaged in the business of providing the service at issue or access to the facility at issue, either to the refused firm itself or to other, similarly situated firms. Thus, in refusing to deal with a given firm, the defendant discriminates between customers. That discrimination, moreover, has an anticom-petitive motivation. For example, in Terminal R.R., the terminal owners were in the business of providing railroads with access to the terminal. The terminal owners allowed railroads that did not compete in St. Louis to use the facility at a non-extortionate charge. The terminal owners, however, allowed railroads that did compete in St. Louis to use that facility only at extortionate rates.
Because essential facilities cases generally involve these two features — monopoly leveraging and discrimination — the four elements of the doctrine merely provide a shorthand formula for inferring the presence of anticompetitive intent when the circumstances of a case follow a certain pattern. When the case does not follow that pattern and does not involve those same features, however, the essential facilities *1287doctrine is not a reliable means of analysis.26
The monopoly-leveraging feature of these cases, moreover, is itself such a powerful indication of anticompetitive intent that courts have frequently held it to be a sufficient basis, independent of the essential facilities doctrine, for imposing liability under section 2. See, e.g., Otter Tail, 410 U.S. at 366, 93 S.Ct. at 1022; Eastman Kodak, 273 U.S. at 359, 47 S.Ct. at 400; Olympia, 797 F.2d at 370; Catlin, 791 F.2d at 1343; Paschall, 727 F.2d at 692; Berkey, 603 F.2d at 263; Poster Exchange, 431 F.2d at 334. These courts thus refer to the “monopoly-leveraging theory” as a third formula for evaluating anticompeti-tive intent.27 Of course, the rationale for imposing liability in these eases is the same as the principal rationale for imposing liability under the essential facilities doctrine: a monopolist’s extension of power from one stage of the market into another provides a strong indication of anticompetitive intent and thus justifies the imposition of liability under section 2. The cases applying this “monopoly-leveraging theory” could also be characterized as “intent test” eases: they merely hold that the conduct of monopoly leveraging is sufficient evidence of an illegal intent to justify the imposition of liability under section 2. Indeed, some of these cases—such as Eastman Kodak and Poster Exchange—are generally cited in connection with the intent test.
The so-called “intent test” also provides an appropriate analytical tool in refusal-to-deal cases that do not involve monopoly leveraging and to which the essential facilities doctrine does not apply. In these cases, a monopolist with monopoly power in one level of a market refuses to deal with a competitor or potential competitor in order to gain a competitive advantage in that same market rather than in another level of the market. Under the intent test, a court can infer an anticompetitive intent from such conduct if, for example, the conduct represents a significant departure from the monopolist’s normal pattern of business. That was the case in Lorain Journal, 342 U.S. at 148-50, 72 S.Ct. at 183-84, where a newspaper in the business of selling advertising refused to sell advertising space to certain customers; it was also the case in Aspen, 472 U.S. at 589-93, 105 S.Ct. at 2850-52, where an owner of a several ski areas in Aspen decided no longer to participate in an all-Aspen ski ticket after years of participating in the ticket. Because the intent test, like the essential facilities doctrine, is nothing more than a means of evaluating circumstantial evidence, however, it justifies the imposition of liability under section 2 only when, under the circumstances of the case, an anti-competitive intent is clearly inferable.
In the following sections, I analyze the present case under this general legal framework. In my view, City Gas’ conduct did not constitute an antitrust violation under section 2. As I discuss in section B, this case is not a monopoly-leveraging case: City Gas is simply not a vertically integrated company; it operates only on one level of the natural gas market and has never attempted to extend its operation into another level of that market. For this reason, in section C, I distinguish the present case from the line of cases in which courts have imposed section 2 liability under the monopoly-leveraging theory.
In section D, I then discuss the essential facilities doctrine, which does not apply to *1288the present case either. In subsection 1, I assume for the sake of discussion that the essential facilities doctrine could apply to the present category of case, but show that the doctrine’s four-part test would still not be satisfied under the circumstances here. In subsection 2, I show that the doctrine does not properly apply to the present type of case. To that end, I analyze the early Supreme Court cases on which the doctrine is based and then discuss the characteristics and limits of the doctrine as developed in those cases. On that basis, I conclude that this case is not an essential facilities case.
In section E, I analyze the case under the intent test. I first examine the Supreme Court’s decision in Lorain Journal and discuss the characteristics and limits of the so-called intent test as it derives from that case. I then demonstrate that this theory is inapplicable to the present case.
In section F, I address two more recent Supreme Court cases—Aspen and Otter Tail—on which the district court heavily relied in the present case. Although neither of these cases expressly applies either the essential facilities doctrine or the intent test, they are still understandable in the context of the preceding sections’ discussion. As I show, moreover, neither case justifies the imposition of liability here.
Finally in section G, I consider the Seventh Circuit’s decision in MCI, which the district court also used as a basis for liability here. Although one holding in MCI (on which the district court, and the panel, relied) does not apply, another holding of the MCI case does apply, and under that holding, City Gas’ conduct does not constitute a violation of section 2.
B. Vertical Integration.
Before addressing the issue of liability under section 2, I pause to discuss the concept of vertical integration, which underlies the idea of monopoly leveraging. On the subject of vertical integration, Aree-da and Turner have this to say:
Vertical integration, simply put, is the inclusion within a single firm of two or more stages in the production and distribution of an end-product. A firm integrates “upstream” or “backward” when it undertakes to produce raw materials or semi-fabricated components which have been or could be supplied by independent producers. A firm integrates “downstream” or “forward” when it undertakes further processing or distribution of products which have been or could be sold to independent producers or distributors.
3 P. Areeda & D. Turner, Antitrust Law U 723 (1978). According to this definition, I cannot conceive of City Gas as a vertically integrated company. The natural gas business involves a relatively simple technology. At the far end of the production stream, natural gas reserves are located, and the gas is pumped from the ground. The natural gas is then transported through interstate pipelines from the reserves to new areas for distribution. Once the natural gas reaches these distribution areas, it is transferred to local pipesystems, which move the gas from the pipelines to the customers and thereby supply the gas on a retail level. Thus, three levels of operation exist in the natural gas business: (1) drilling gas, (2) transporting the gas for resale, and (3) supplying the gas to customers at retail. In the present case, City Gas is involved solely in the third level of operation — supplying gas at retail. City Gas is therefore not a vertically integrated company.28
Even if City Gas on occasion had actually sold natural gas at wholesale to other retail distributors, I question whether that would have constituted vertical integration into another level of operation. Given FGT’s predominance in the wholesale market, City Gas’ participation in that market would have lacked antitrust significance: City Gas would still have lacked the market power to act anticompetitively under section 2. See supra at 1275-1283. In addi*1289tion, the sale of some surplus gas by one natural gas retail supplier to another retail supplier would hardly constitute an integration by the first supplier into an upstream market. In my view, one could not seriously contend that City Gas’ potential sales of its surplus gas to another retail supplier would constitute integration into the second level of the market — the interstate transportation level on which FGT operates. Clearly, City Gas would in no way displace FGT’s operation; on the contrary, all of the gas involved in City Gas’ trade would still come from FGT. If City Gas’ potential sales of gas for resale constituted vertical integration, the integration would have to occur on a sublevel of the retail distribution level of the market.
Although integration into such a sublevel could theoretically occur, practically it could not. Any potential sublevels of the supply operation would be so interrelated with each other that it would not be reasonable to consider them as severable. See, e.g., Berkey Photo, 603 F.2d at 276-79 (describing features of “integrated firm”). Although a retail distributor generally has its own pipesystem, which picks up a supply of natural gas from a wholesale source and then distributes it to retail customers, the distribution at the back end of this process is not severable from the supply process at the front end of the operation. In order to distribute gas to customers, a retail distributor must have a pipesystem, and the pipesystem must have gas in it. Thus, the pipesystem must connect with a source of natural gas. Variations in the nature of the source, the size of the pipes-ystem, or even the character of the customers do not, I suggest, alter the degree of vertical integration of a retail (whether wholly retail or merely predominately retail) supplier of natural gas. In terms of the level of vertical integration, City Gas’ operation is identical to what Consolidated’s operation would (hypothetically) be. Both companies would possess pipesystems that distribute natural gas from a supply source to their customers. In my view, neither the fact that Consolidated’s source might have been City Gas rather than FGT nor the fact that one of City Gas’ customers might have been Consolidated is of any significance on the issue of vertical integration.
Because these two companies would be engaged in identical operations, integrated to the same degree and operating at the same levels of the market, this case is fundamentally different from the monopoly-leveraging cases to which the district court analogized the present case. City Gas is simply not using its power as a monopolist on one level of a market to expand its operation into, maintain its monopoly in, or force out competition from another level of the market. City Gas is doing nothing more than competing — legitimately and successfully — with Consolidated for more business in the retail natural gas market. City Gas was successful for a variety of reasons: it is bigger, it has been in business longer, it has better resources (i.e., a hook up to FGT’s pipeline), it implemented an aggressive marketing strategy, and its prices were lower. In short, City Gas could provide customers with better service than Consolidated. We might say, with the benefit of hindsight, that as a result of a series of historical accidents, Consolidated missed the boat years back when it failed to get into the natural gas supply business. None of this, however, means that the antitrust laws impose a duty on City Gas to deal with Consolidated. I discuss why in the following sections.
C. The Monopoly-Leveraging Theory.
I first consider the line of cases stemming from Eastman Kodak, 273 U.S. at 359, 47 S.Ct. at 400, which holds that monopoly leveraging — a monopolist’s attempt to use its monopoly power on one level of the market to gain a competitive advantage on another level of the market — is evidence of anticompetitive intent and constitutes a violation of section 2. The present case, however, as I discuss in the above section, does not involve monopoly leveraging: City Gas’ conduct implicated only the retail level of the natural gas market. This line of cases therefore provides no basis here for imposing liability under section 2.
*1290In Eastman Kodak, the plaintiff operated a stock house for photographic supplies in Atlanta, Georgia. Kodak supplied the plaintiff, and other stock houses in the area, with wholesale photographic supplies. Kodak, however, itself began to purchase stock houses, expanding its operation from the wholesale level into the retail level as well. By 1910, Kodak had acquired control of all stock houses in Atlanta but the plaintiffs, and, at that point, Kodak refused to sell the plaintiff súpplies at wholesale prices. In evaluating the situation, the Court stated:
[Although there was no direct evidence — as there could not well be — that the defendant’s refusal to sell to the plaintiff was in pursuance of a purpose to monopolize, we think that the circumstances disclosed in the evidence sufficiently tended to indicate such purpose, as a matter of just and reasonable inference, to warrant the submission of this question to the jury.
273 U.S. at 375, 47 S.Ct. at 404.
Similarly, in Poster Exchange, 431 F.2d at 334,29 a frequently cited case decided by our predecessor court, a vertically integrated producer/distributor of motion-picture advertising accessories refused to supply a local distributor. The district court held that the,defendant “intentionally used the monopoly power it had at the manufacturing level to eliminate [the plaintiff] as a competitor at the distributor-jobber level.” Id. at 339. Applying Eastman Kodak, the district court held that this use of monopoly “power at one level to drive out competition at another” constituted a violation of section 2. Id.
The present case simply does not fit this factual pattern. City Gas possesses monopoly power as a natural gas supplier at the retail level. It does not possess monopoly power on the wholesale level; nor does it operate on that level. City Gas, moreover, has never used its monopoly power on the retail level in order to leverage its way into another level of the market. Rather, if City Gas has used its monopoly power at all in this case, then it has used that power only with respect to the retail distribution market for natural gas. Given these circumstances, Eastman Kodak and its progeny do not control here: those monopoly-leveraging cases justify no inference under the present circumstances of a “purpose to monopolize,” Eastman Kodak, 273 U.S. at 375, 47 S.Ct. at 404.30 Nor do the so-called “essential facilities” cases, which I discuss in the next section, justify such an inference here.
D. Essential Facilities Doctrine.
As the district court stated, “the essential facilities doctrine teaches that, when a business controls a scarce facility, it assumes an obligation to provide its competitors reasonable access to that facility.” 665 F.Supp. at 1532 (citing Byars, 609 F.2d at 856).. The rationale behind this doctrine, as the district court noted, is that “[a] refusal to deal in this context may be unlawful because it could result in the monopolist extending its power vertically from one level of production to another.” Id. at 1532-33. With this qualification in mind, the district court then articulated the four-part standard for applying the doctrine:
The case law sets forth four elements necessary to establish liability under the essential facilities doctrine: (1) control of the essential facility by a monopolist; (2) a competitor’s inability practically or reasonably to duplicate the essential facility; *1291(3) the denial of the use of the facility to a competitor; and (4) the feasibility of providing the facility.
Id. at 1533 (quoting MCI, 708 F.2d at 1132—33).
This test must be understood for what it is: merely a mechanism for evaluating the circumstantial evidence in a given case to determine whether that evidence yields an inference of the defendant’s anticompeti-tive intent. Before considering in subsection 2 the limits of this doctrine as a means of evaluating circumstantial evidence, I first examine in subsection 1 the district court’s application of the test to the present case. In my view, that application is erroneous in and of itself.
1. The District Court’s Analysis.
Applying this test to the present case, the district court concluded that all four elements were satisfied and thus held that City Gas was liable under section 2 for refusing to deal with Consolidated. See id. at 1534-35. The district court first held that “City Gas, a monopolist, controlled an essential facility — a pipeline that transported wholesale gas.” Id. at 1534. Under the second element of the test, the district court held that duplication by Consolidated of City Gas’ facility — the lateral pipe connecting City Gas’ pipesystem to FGT’s pipeline — would be “unreasonable ‘due to regulatory restriction, and delays, and the expense and time required.’ ” Id. (quoting Aspen Highlands Skiing Corp. v. Aspen Skiing Co., 738 F.2d 1509, 1521 (10th Cir.1984), aff'd, 472 U.S. at 585, 105 S.Ct. at 2847). Applying the third part of the test, the district court noted that City Gas had offered to sell gas to Consolidated for ten cents per therm over cost (City Gas’ cost of purchasing gas at wholesale from FGT), which the district court held “essentially amounted to a refusal to deal at all.” Id. Finally, under the fourth part of the test, the district court held that City Gas could feasibly have sold wholesale gas to Consolidated.
In my view, the district court misapplied the doctrine. First, as I discuss above, see supra at 1275-1283, City Gas is not a monopolist in the wholesale market. Second, the facility under City Gas’ control is not properly labeled “essential” under the doctrine. Third, Consolidated could practically and reasonably have duplicated City Gas’ lateral connection to FGT’s pipeline. Fourth, the district court improperly held that City Gas’ offered price constituted a refusal to deal. I address these points in turn.
The district court based its application of the essential facilities doctrine on the following factual findings. Before Consolidated received its FERC allocation in 1984, Consolidated could have purchased gas for resale only from City Gas. After 1984, although Consolidated would have been authorized to purchase from FGT, it “would have encountered a prohibitive transportation problem.” 665 F.Supp. at 1534. That is, construction of “[a] lateral pipeline connecting Consolidated’s facilities to the FGT main would have cost $250,000.00.” Id. The district court concluded that “[tjhere is grave doubt as to whether Consolidated could have effectively competed with City Gas based on these additional expenses.” Id.
The test’s first element requires “control of the essential facility by a monopolist.” MCI, 708 F.2d at 1132-33. Based on its earlier holding that City Gas possessed monopoly power in the wholesale market, the district court assumed that City Gas was a monopolist with control over the “essential” facility — presumably, “a pipeline that transported wholesale gas.” 665 F.Supp. at 1534. As I discuss above, see supra at 1275-1283, City Gas did not possess monopoly power in the wholesale market; indeed, City Gas did not even operate in that market. The district court thus improperly concluded that City Gas was a monopolist for purposes of the first element of the essential facilities doctrine.
In addition, the district court improperly characterized the facility at issue here as “essential” under the first part of the test. The district court initially identified the “essential facility” as “a pipeline that transported wholesale gas.” In so doing, the district court mischaracterized a very *1292important feature of this facility: that is, City Gas’ pipesystem transports only retail — and not wholesale — natural gas.31 As such, this facility could not be essential to Consolidated’s operation. Consolidated needed a facility that supplied wholesale gas. Because this facility supplied only retail gas, it did not do what Consolidated needed it to do. For this reason, the facility cannot properly be labeled “essential” under the doctrine’s first element.
I also attack for one other reason the district court’s conclusion that this facility is essential. As I note above, see supra at 1290, the essential facilities doctrine is designed to deal with the danger that a monopolist in control of a scarce resource will “extend[] its power vertically from one level of production to another.” 665 F.Supp. at 1532-33. This concern, in my view, also defines the term “essential.” That is, a facility becomes essential if, in restricting competitors’ access to that facility, a monopolist gains a competitive advantage in another level of the market — that is, a market downstream or upstream from the market containing the facility itself. As courts have reiterated again and again, the antitrust laws do not protect competitors; rather, the laws protect competition. See, e.g., Roland Mach. Co. v. Dresser Indus., 749 F.2d 380, 394 (7th Cir.1984), United States v. Syufy Enters., 903 F.2d 659, 663 (9th Cir.1990). This principle, of course, applies to the essential facilities doctrine as well: the doctrine addresses the danger of monopoly leveraging and the anticompetitive effects that it has on competition in general. Applying this principle to the present case, City Gas’ system is not essential in this sense. City Gas is not attempting to extend its power from one level of the market into another. City Gas, a monopolist at the retail distribution level, operates only at the retail level of the natural gas business. Its facility obviously relates to competition on that level, but the facility is not a mechanism for extending City Gas’ power to another level of the market — or for that matter, from another level of the market onto the retail level. City Gas operates solely on a single level of the market. The facility here, therefore, cannot properly be labeled “essential.”
Even assuming this facility could be labeled “essential,” I question the district court’s determination that Consolidated could not practically or reasonably have duplicated it for purposes of the second element of the test. Consider what Consol*1293idated would have had to do in order to duplicate this facility. Consolidated already had a pipesystem in place from the days of its LP distribution. In order to change to natural gas distribution, therefore, Consolidated merely had to find a source of wholesale gas. Buying gas from FGT, the regional wholesale supplier, required two things: (1) an allocation from FERC, and (2) a lateral pipe connecting Consolidated’s pipesystem to FGT’s pipeline. Consolidated applied to FERC for the allocation on May 21, 1982. FERC issued its final decision granting the allocation on September 19, 1984. The process thus took two years and four months. The lateral pipe would have cost Consolidated $250,-000. These factors of time and expense are not sufficient to support the conclusion that Consolidated could not reasonably or practically have connected directly to FGT’s pipeline.
As I discuss above, see supra at 1281-1282, given Consolidated’s projected future profits as a company with a natural gas supply, Consolidated could feasibly have obtained underwriting for a $250,000 bond to finance the lateral pipe. Based on the figures presented by Consolidated’s expert witness, the company’s annual profits, with a supply of gas, would have been approximately $267,000, and the company would have generated those profits for twenty years. The annual interest rate on a $250,-000 bond, figured at a rate of fifteen percent, would be $37,500. Taking into account Consolidated’s tax deduction for those interest payments, the real interest would be $24,750 annually — a figure representing 9.3% of Consolidated’s annual profits. After paying that interest, Consolidated would be left with $242,250 in annual profits. Given those profits, Consolidated would have little trouble paying off the $250,000 principal on its bond offering at the end of twenty years. As these figures demonstrate, given the profits that Consolidated claims it could have generated for twenty years, it was clearly feasible for the company to undertake a $250,000 expenditure on a lateral pipe that would give it a supply of natural gas from FGT, an established wholesale supplier of natural gas.
This conclusion is also consistent with existing caselaw, which suggests that duplication of the facility must involve a high degree of impracticality and infeasibility before it will satisfy the essential facility doctrine’s second element. MCI, 708 F.2d at 1081, a case on which the district court itself relied, illustrates this limit. In that case, MCI brought suit against AT&T, claiming that AT&T’s refusal to allow MCI to interconnect with various AT&T facilities constituted monopolization in violation of section 2 of the Sherman Act. A jury awarded MCI $1.8 billion; the Seventh Circuit affirmed in part, and reversed in part. I briefly state the facts.
Before 1969, the telecommunications industry existed as a lawful monopoly. The Bell System companies provided local exchange service, and AT&T provided long-distance service, which used the Bell System’s exchange facilities to connect with customers on the local level. AT&T also used its long-distance lines to provide special services such as point-to-point lines (private lines connecting two locations) that do not require switching through the local exchange system and to provide other services such as foreign exchange (FX) lines and common control switching arrangements (CCSA) that do require switching through the local system.
In 1963, MCI began its attempt to enter the long-distance telephone market. It applied to the FCC for permission to construct and operate a long-distance system providing private-line service between Chicago and St. Louis. That service would require interconnections to ordinary local telephone lines in the two cities. Bell’s local exchange system was all that was available. Six years later, in 1969, after a long administrative process, in which AT&T opposed MCI’s application, the FCC finally approved MCI’s request. As a result of that decision, applications similar to MCI’s poured into the FCC. The FCC instituted a rulemaking procedure to deal with the applications. In June 1971, the FCC issued a decision, see Establishment of Policies & Procedures for Consideration of Applications to Provide Specialized *1294Common Carrier Servs. (Common Carrier decision), 24 F.C.C.2d 318 (1970), and as a result, AT&T agreed to enter into negotiations with MCI concerning the local interconnections. They reached an interim agreement in September 1971, but because AT&T disputed the reach of the Common Carrier decision, that agreement did not allow MCI interconnections for its FX or CCSA services. Finally, in April 1974, the FCC clarified the reach of its earlier decision and ordered AT&T to provide the disputed interconnections. Shortly before that decision, MCI filed its antitrust suit.
MCI’s attempt to enter the telephone market thus involved substantial investments in time and capital. MCI spent six years engaged in administrative proceedings, with AT&T as an adversary, in order to gain initial FCC approval in 1969 for construction and operation of its St. Louis to Chicago line. Two more years passed before the FCC handed down its Common Carrier decision in 1971, and then another three years until the FCC finally clarified the effect of that decision in 1974 by issuing an order requiring AT&T to provide the disputed interconnections. From start to finish, the entire approval process took eleven years. In addition, MCI made a substantial expenditure in constructing its own long-distance lines. The original Chicago-to-St. Louis line required a terminal in each city and microwave relay towers along the whole length of the line to connect the terminals. MCI then extended the line from these original points as the company began to construct a nationwide long-distance system. This venture, moreover, to develop a nationwide long-distance system that would compete with AT&T, involved a tremendous capital investment, and even more initiative and risk. In order to finance its venture, MCI raised $110 million in loans and underwriting by leading lenders and equipment suppliers. As the MCI court stated, this made MCI “one of the largest start-up ventures in the history of Wall Street.” 708 F.2d at 1095.
In contrast, Consolidated completed its allocation proceedings before FERC in twenty-eight months, but was then unwilling to spend $250,000 on a lateral pipe. Compare that process in terms of time and money to the eleven-year administrative proceeding that MCI went through and to the $110 million that MCI raised and actually invested. Moreover, given Consolidated’s net worth without a natural gas supply — approximately $775,000 according to its own experts — it could easily have financed the $250,000 required to construct the lateral pipe. See supra at 1281-1282, 1293. If we take into account what Consolidated claims it would have been worth with a natural gas supply — over $2.25 million according to its expert witness — its claim that it could not feasibly spend $250,-000 on a lateral pipe becomes truly preposterous. As the facts in MCI suggest, a high degree of impracticality and infeasibility must be demonstrated in order to satisfy the second element of the essential facilities doctrine.32 Given the nature of MCI’s investment in terms of time, money, and initiative, as compared to the significantly smaller investment that Consolidated was unwilling to make here, Consolidated has not satisfied the second element of the essential facilities doctrine. MCI surely does not stand for the proposition that a firm has a duty to provide a free ride to potential market entrants, and that is exactly what Consolidated is after by claiming that access to City Gas' pipesystem is essential to Consolidated’s ability to enter the market.
In addition, I also question whether City Gas has actually refused to deal in this case. The district court found that City Gas made a series of offers to sell to Consolidated, the lowest of which was at a price of cost (i.e., the cost of City Gas’ wholesale gas from FGT) plus five cents per therm and held that those prices were *1295so unreasonable that they constituted a refusal to deal under section 2. The district court, however, never determined what a reasonable price would have been. Indeed, the district court never set any price whatsoever: in granting Consolidated’s request for an injunction ordering City Gas to supply wholesale gas, the district court never set a price for that mandated sale, but rather delegated the price-setting task to the FPSC. See 665 F.Supp. at 1512, 1534, 1545. If the district court never determined what a reasonable, or “fair”33 price would have been, then I question on what basis it could legitimately have determined that City Gas’ offered price was so unreasonable that it constituted a refusal to deal under section 2. As I discuss in more detail in part VI, the district court could have made that determination only in light of its determination of what the “fair” price would have been — that is, by comparing the offered price to the fair price and deciding that the offered price was so much higher than the fair price that it was unreasonable and constituted a refusal to deal. Without determining the fair price, therefore, the district court had no basis for concluding that City Gas’ offered price constituted a refusal to deal. The third prong of the essential facilities doctrine is thus not satisfied here either.
For these reasons, the district court in my view improperly applied the essential facilities doctrine to the present case. As this discussion also suggests, this case is radically different from the typical essential facilities case. That the district court would nevertheless so readily apply the doctrine and also find it to be controlling— and that the court today would affirm that result — are therefore matters of considerable concern to me. I think it is necessary to take this opportunity to analyze thoroughly the character and the limits of the essential facilities doctrine, and the implications of forcing its application to a case such as this one. I undertake to do so in the following subsection.
2. The Doctrine’s Limits.
The essential facilities doctrine, as I explain above, see supra at 1285-1287, 1290-1291, is nothing more than a mechanism that aids courts in evaluating circumstantial evidence to determine whether the evidence supports an inference of anticompeti-tive intent. Given a certain type of fact pattern, the doctrine’s four-part test will generally give rise to such an inference. Absent that type of fact pattern, however, even if the facts of the case could be found to satisfy the test, they would not support an inference of anticompetitive intent. Such an inference depends on more than the four elements of the test; it arises from the larger factual context — the circumstantial evidence — of the case as a whole. The four elements of the doctrine therefore may indicate, but by no means define, the kind of refusal-to-deal cases in which an inference of anticompetitive intent is justified. In this subsection, I undertake to describe more fully the factual features and the kinds of circumstantial evidence that are necessary to give rise to an inference of anticompetitive intent in addition to satisfying the essential facilities doctrine and its four-part test.
I begin with a discussion of the doctrine’s origin and application in two Supreme Court cases—United States v. Terminal R.R. Ass'n., 224 U.S. 383, 32 S.Ct. 507, 56 L.Ed. 810 (1912), and Associated Press v. United States, 326 U.S. 1, 65 S.Ct. 1416, 89 L.Ed. 2013 (1945). Based on these cases, I describe the factual context necessary for a proper application of the essential facilities doctrine. I then consider the present case in light of that discussion in order to demonstrate that this case lacks certain key features that must be present if the essential facilities doctrine is justifiably to give rise to an inference of anticom-petitive intent and require the imposition of liability under section 2.
In Terminal R.R., frequently cited as the original essential facilities case, a group of railroads that serviced St. Louis acquired ownership of a railroad terminal *1296that provided the sole means, of access to St. Louis. The terminal owners allowed competing railroads to use the terminal only on payment of an extremely high fee, an arrangement that secured a competitive advantage to the terminal owners’ railroads. The terminal owners, however, allowed railroads from East St. Louis to use the terminal without payment, apparently because a toll bridge in Memphis, Tennessee provided the East St. Louis railroads an alternate means of access to the city. As a result of this situation, the federal government (and not a private plaintiff) brought an antitrust enforcement action against the terminal owners.34
The Supreme Court held in favor of the government and remanded the case to the district court with instructions to enter a decree requiring the terminal owners to provide all competing railroads with access to the facility “upon such just and reasonable terms and regulations as will, in respect of use, character, and cost of service, place every such company upon as nearly an equal plane ... as that occupied by the proprietary companies.” 224 U.S. at 411, 32 S.Ct. at 516. Two principal considerations supported the Court’s decision. First, the Court noted that the terminal owners were not merely terminal owners. Their activities were not limited to operating the terminal, but also included the provision of railroad services in St. Louis. See id. at 406-07, 32 S.Ct. at 514. The terminal owners’ railroads thus gained a competitive advantage in providing railroad service by virtue of their terminal ownership. The case thus involved a classic monopoly-leveraging situation: the terminal owners were using their monopoly power on one level of the market to gain a competitive advantage on another level of the market. Second, the terminal owners discriminated between other railroads in imposing the usage fee. The terminal owners did not, for example, assess a higher fee on railroads operating in East St. Louis. Based on these considerations, the Court issued its instructions concerning the decree.35
The next seminal Supreme Court case to which lower courts have pointed in discussing the essential facilities doctrine is Associated! Press. In that case, the government brought an action seeking to enjoin enforcement of a newspaper association’s bylaws. That association consisted of more than 1200 newspapers involved in the collection and distribution of news; the bylaws prohibited distribution or sale of news to nonmembers by any of the association’s members. The government claimed that the arrangement constituted a conspiracy in restraint of trade — a violation, as in Terminal R.R., of section 1 of the Sherman Act — as well as an attempt to monopolize in violation of section 2 of the Act.
The Court held that, in restraining the sale of news to nonmembers that competed with the association’s members, the bylaws were designed to give members a competitive advantage over nonmembers and thus constituted a violation of the antitrust laws. The Court therefore approved an injunction ordering the association also to provide news to nonmembers on nondiscriminatory terms. 326 U.S. at 21, 65 S.Ct. at 1425." In Associated Press, as in Terminal R.R., the Court emphasized that the defendant attempted to use its power as a newspaper association to gain a competitive advantage over nonmember newspapers — i.e., monopoly leveraging. In addition, as in Terminal R.R., the association discriminated between newspapers in allowing access to its services: members, of course, enjoyed the benefits of the association’s news, whereas nonmembers enjoyed *1297no benefits, and the bylaws prevented certain newspapers from becoming members. Finally, as in Terminal R.R., the element of conspiracy underlying formation of the association (or the group of railroads owning the terminal) influenced the Court’s decision.
If we are to understand the essential facilities doctrine as an outgrowth of these cases, as the lower courts applying the doctrine suggest we should, then we must understand the limits of the doctrine as articulated in these cases. To that end, I offer some observations on the teachings of Terminal R.R. and Associated Press.
These cases demonstrate three distinct factual features. First, the defendants were already engaged in the business of providing access to the essential facility or resource at issue. The terminal owners in Terminal R.R. were in the business of providing terminal access to railroads. The newspaper association in Associated Press was in the business of gathering and distributing news to newspapers. In short, the defendants in both cases were already in the business of selling the essential resource, or service, to other entities. Second, the defendants were vertically integrated: they also operated on another level of the market, and the facility or resource at issue was essential to their operations on that level as well. The terminal owners in Terminal R.R. ran their own trains into St. Louis, and that operation required access to the terminal. The member newspapers in Associated Press ran their own newspapers and depended on the association for their news. Third, the defendants in these cases, in control of essential facilities at one level of their operations, competed on the second level of operation with other entities that also required access to the essential facility in order to operate at this second level. The terminal owners in Terminal R.R. competed with other railroads servicing St. Louis, and those other railroads required access to the terminal in order to reach the city. Likewise, the member newspapers in Associated Press competed with nonmember newspapers that needed the association’s news. Given these three factors, the defendants’ attempts to restrain their competitors’ access to the essential facilities at issue became significant for antitrust purposes: under the circumstances, that conduct gave rise to an inference of anticompetitive intent.
The essential facilities doctrine merely provides a means to evaluate this type of fact pattern in order to determine whether the facts give rise to such an inference of intent. I question how far courts can legitimately extend the doctrine when the circumstances of a case fail to approximate the fact pattern of these two cases. The cases themselves provide some guidance on this point. As the Supreme Court indicated in Terminal R.R., it would have been a different case if the terminal owners had been solely in the terminal business and not in the railroad business as well. See 224 U.S. at 406-07, 32 S.Ct. at 514. If that were the case, the terminal owners’ conduct in charging high prices would not have had the anticompetitive overtones that the conduct had in the actual case. The case would not have been a monopoly-leveraging case: the terminal owners would not have been using their monopoly power over the terminal to gain a competitive advantage in supplying railroad services; they would simply have been charging some railroads a higher fee for use of the terminal. The inference of anticompetitive intent in the case did not arise from the conduct — the high fee charged for access to the terminal — in and of itself; rather, the inference arose from the effect of the high fee on the terminal owners’ competitive position in the downstream railroad market. By charging downstream competitors a higher fee, the terminal owners obtained a competitive advantage in providing railroad service to St. Louis. See id. at 408, 32 S.Ct. at 515. If the terminal owners had not also been in the business of providing railroad services, then their high fee charges would have produced no such competitive advantage and would not, therefore, have given rise to an inference of anticompetitive intent.
The circumstances of the Associated Press case also support this analysis. The association’s refusal to supply news to non*1298member newspapers gave the member newspapers a competitive advantage in the newspaper market. The refusal to deal thus gave rise to an inference of anticom-petitive intent. If the association were differently constituted — for example, as an independent news-reporting service, not composed of newspapers, and not engaged in the newspaper business — its refusal to deal with nonmember newspapers would not have given the association a competitive advantage on another level of operation. The association, in that case, would not have engaged in monopoly leveraging, and its refusal to deal would not have signaled an anticompetitive intent.
These two cases would also have been different if the defendants had not already been in the business of providing access to the essential facility at issue, which is exactly the situation in the present case. This type of situation, however, is so fundamentally different from the fact pattern of Terminal R.R. and Associated Press that the Court in those cases did not touch on how to approach it. I therefore discuss a hypothetical case in order to demonstrate that the essential facilities doctrine does not properly extend to this type of situation: that is, absent the first factual feature of these two cases, satisfying the four elements of the essential facilities doctrine will not give rise to an inference of anti-competitive intent. As I also demonstrate, the imposition of liability under section 2 in this hypothetical case would have practical implications that do not arise in the typical essential facilities case. The typical cases, moreover, fail to foresee these practical implications, let alone to address them. For this reason as well, the essential facilities cases do not control the result in this hypothetical case.
Assume that a newspaper called the Times possesses a one hundred percent share of the daily newspaper market in a given city. The Times has achieved this position by providing first-rate news and reporting services for years. The Times has a large reporting staff and, because it has been in business for a long time, its reporters have a vast web of information sources throughout the city. Assume, further, that the Post undertakes to publish a daily newspaper in the same city, thus entering into competition with the Times. The Post, however, has only enough resources for a small reporting staff, and even if it could afford a larger staff, it would take several years to build up the information network that the Times’ reporters have developed. The Post, therefore, wants to buy news from the Times. The Times, however, refuses to sell to the Post. The Times has never sold its news to anyone (newspapers or otherwise) and has no intention of starting to do so now. As the Times also admits, it has no interest in supplying news to a rival newspaper so that the rival can compete. Indeed, in the Times ’ view, if the Post cannot survive in the daily newspaper market unless it can obtain news from the Times, that is all for the better. As would be expected, the Post has a difficult time making ends meet. The Post, therefore, hires some lawyers of its own and brings an antitrust suit against the Times, seeking an injunction ordering the Times to sell news to the Post.
I suggest that the court must deny this request for an injunction: the essential facilities doctrine provides no basis for imposing a duty to deal on the Times under these circumstances. As a preliminary point, this is not a monopoly-leveraging case such as Terminal R.R. or Associated Press. The Times is not a vertically integrated entity: the paper assembled its large reporting staff and developed its information networks entirely for purposes of, indeed as an integral part of, its own newspaper operation and not, as in Associated Press, to produce news for resale. The Times has never sold news to any other entity whatsoever. Its refusal to sell news to the Post, therefore, could not be classified as an attempt to use its “monopoly” power as a newsgatherer to maintain its distribution monopoly: its newsgathering operation is an integral part of its newspaper operation as a whole.36 Given the constantly reit*1299erated rationale of the essential facilities doctrine — to guard against the evils of monopoly-leveraging — this hypothetical case would thus appear to fall beyond the doctrine’s scope. For the sake of discussion, I assume that the essential facilities doctrine could theoretically apply to this non-monopoly-leveraging case. In practice, however, it could not.
Even assuming the elements of the essential facilities doctrine were satisfied here — i.e., the court finds that the Times is a monopolist in control of an essential resource that the Post cannot reasonably or practicably duplicate, and the Times has refused to supply that resource to the Post although it could feasibly do so — under the circumstances of this case, the Times ’ conduct in refusing to deal with the Post does not give rise to an inference of anticompeti-tive intent. Because the Times has never sold news, it has not acted inconsistently in refusing to deal with the Post. In the typical essential facilities case, in contrast, the defendant does act inconsistently, refusing to deal with some entities but not with others. In Terminal R.R., for example, the terminal owners charged some railroads an exorbitant fee for access to the terminal and charged others a reasonable fee for the same access. In these typical cases, moreover, the inconsistent treatment relates to potential customers’ status as competitors: the terminal owners in Terminal R.R., for example, refused to deal only with railroads that also competed with the terminal owners’ railroads. The inconsistent treatment there, coupled with the refused railroads’ status as competitors, signaled the defendant’s anticompetitive intent.
If the defendant treats all potential customers the same, however, even if it uniformly refuses to deal with any at all, that conduct would not give rise to an inference of anticompetitive intent. Thus, the Times, unlike the terminal owners, has not acted inconsistently in refusing to deal with the Post. For whatever reasons, the Times has chosen not to sell its news at all, and its decision not to deal with the Post merely follows from that blanket policy. Its decision is not inconsistent. Nor does the decision relate to the Post's specific status as competitor. Unlike the fact pattern in Terminal R.R. and other typical essential facilities cases, therefore, the circumstances of this hypothetical case give rise to no inference of anticompetitive intent.
The practical implications of imposing a duty to deal on the Times also suggest the inapplicability of section 2 under these hypothetical circumstances. In Terminal R.R., Associated Press, and other typical essential facilities cases, the defendants, when ordered to deal with the excluded entities, did not have to undertake a new business or provide a new service. They merely had to give the excluded entities the same service that was already being provided to other entities. Thus, in response to the Court’s decision, the defendants in both cases had to stop not doing what they usually did. The Court’s decision enjoined them to operate normally with respect to all customers. In this hypothetical case, however, if the Times were ordered to deal with the Post, the court’s decree would be fundamentally different from the decrees in typical essential facilities cases. This decree would not order the Times to resume its ordinary course of operation; rather, it would order the Times to undertake a new course of operation, to begin providing a service that it had never before provided. In my view, this kind of a case is fundamentally different from the typical essential facilities case.37 Not only is this kind of case factually and legally inconsistent with typical essential facilities cases, but the practical implications of imposing a duty to deal under these circumstances are radically different than such implications in the typical essential facilities cases. Consequently, those cases simply do not control here.
As a general matter, the antitrust laws do not impose a positive duty on companies to help competitors. The antitrust laws *1300focus on various types of conduct that inhibit competition. The laws, however, do not purport to facilitate competition by imposing a duty on a company to help another company into competition. As the Seventh Circuit stated in Olympia Equip. Leasing Co. v. Western Union Tele. Co., 797 F.2d 370, 375-76 (7th Cir.1986), cert. denied, 480 U.S. 934, 107 S.Ct. 1574, 94 L.Ed.2d 765 (1987), a firm
has no general duty to help its competitors .... “There is a difference between positive and negative duties, and the antitrust laws, like other legal doctrines sounding in tort, have generally been understood to impose only the latter.” ... A monopolist has no duty to reduce its prices in order to help consumers, ... and no duty to extend a helping hand to new entrants....
Id. (quoting USM Corp. v. SPS Technologies, Inc., 694 F.2d 505, 512-13 (7th Cir.1982), cert. denied, 462 U.S. 1107, 103 S.Ct. 2455, 77 L.Ed.2d 1334 (1983)) (citations omitted). It is not surprising, therefore, that in every essential facilities case disclosed by my research in which a court imposed a duty to deal, the defendant was already engaged in the course of dealing at issue. E.g., Otter Tail, 410 U.S. at 428, 93 S.Ct. at 1022 (defendant already in business of wheeling power); Associated Press, 326 U.S. at 1, 65 S.Ct. at 1416 (defendant already in business of selling news); Terminal R.R., 224 U.S. at 383, 32 S.Ct. at 507 (defendant already in business of operating railroad terminal); Fishman, 807 F.2d at 520 (defendant already in business of leasing Chicago stadium); Aspen, 738 F.2d at 1509 (defendant already engaged in practice of providing all-Aspen ski ticket), aff'd, 472 U.S. 585, 105 S.Ct. 2847, 86 L.Ed.2d 467 (1985); MCI, 708 F.2d at 1081 (defendant already in business of providing interconnections to local telephone exchange networks); Hecht, 570 F.2d at 982 (defendant already in business of leasing Robert F. Kennedy Stadium). In these cases, moreover, the defendant had in fact been engaged in the course of dealing at issue: a duty to deal has never been imposed on a company under the rationale that the company could, in the court’s view, feasibly engage in that course of conduct. Under this established law, therefore, the Times has no “positive duty” to sell news to the Post. But that is exactly the kind of duty that an injunction ordering it to deal with the Post would impose on the Times.
This legal principle — that the antitrust laws impose no positive duty on a company to help its competitors — is also supported by strong policy considerations. The imposition of a duty to deal in such circumstances would encourage economic inefficiencies and promote inequity. In order to achieve its level of success in the newspaper business, the Times invested significant amounts of capital over a period of time. That investment required initiative and involved significant risks. An attempt by any company to develop a product or service — i.e., to compete in a market — requires initiative, innovation, and risk-taking. Our antitrust laws are designed to ensure that every company has the opportunity to compete under these terms. Our antitrust laws, however, do not mandate that a company, having made this investment in capital and taken this initiative and risk, must then provide a free ride into the market for every potential new competitor that comes along. Such a policy would be unfair to those companies that have made investments and taken risks; moreover, it would discourage businesses from taking initiative, assuming risk, and making investments of capital in the first place. A policy that would lead to these results is not, in my view, a good one; nor is it the policy behind our antitrust laws.
In addition, courts must respect rational decisions of economic actors. If the Times has opted not to undertake a course of operation — not to sell to others material that it develops for its own operation — the court has no business second-guessing that decision. The Times itself is in the best position to evaluate whether to enter into the course of dealing at issue, and it has decided not to do so. For the court to evaluate such a decision — in effect upholding or overturning it by deciding whether or not to impose a duty to deal — would push the court beyond the limits of its *1301administrative capabilities. If courts were to review such decisions, then the question whether an entity should have undertaken a course of dealing would become an element of every refusal-to-deal case.38 The plaintiff would bear the burden of proving this element of the case, and the courts would have to develop a legal standard for judging the evidence. I suggest that this “element” of a refusal-to-deal case would be an impossible one to prove. What kind of a showing would the courts require a plaintiff to make? Assume a plaintiff has access to a defendant’s books, to the defendant’s research and development, and to the defendant’s plans for expansion and growth — or that the plaintiff could devise a reasonable approximation of this information. What if, based on this evidence, the plaintiff could show that the defendant could feasibly undertake the activity at issue, indeed would profit from that activity: would that provide a sufficient basis for a court to hold, as a matter of law, that the defendant must undertake that course of conduct? Would it be sufficient if the plaintiff could make an even stronger showing — a showing that the defendant could not stay in business without undertaking that course of dealing, but that if the defendant did undertake it, he would become a millionaire, indeed a billionaire? What if the defendant argues in defense of its decision not to undertake the course of dealing that its board of directors and shareholders had a moral objection to doing so? Or that such an undertaking would place additional demands on the staff’s time, which would conflict with the company’s policy emphasizing the importance of spending quality time outside of the office? Or, that the undertaking would simply be inconsistent with the company’s own plans for expansion? I suggest that the court has no legal basis for forcing a party to undertake an operation that the party has chosen not to undertake.
The courts would also face extreme administrative problems in imposing effective remedies in cases like this hypothetical one. If the court issued an injunction ordering the Times to sell news to the Post, the court would obviously have to set a transaction price in its injunction. In the typical essential facilities case, the court can simply look to the terms set by the defendant in its normal dealings with other customers. Unlike the association in Associated Press or the terminal owners in Terminal R.R., however, the Times has no established terms for the sale of news to which the court could look in fixing a transaction price. Thus, the court must come up with a price on its own. As I discuss in detail in part VI, courts would confront severe difficulties in undertaking this price-setting task. The price must be fair to the Times;39 the court cannot require it to sell news at a loss, or at the cost of its ability to compete. On the other hand, the price must also enable the Post to compete; if the Post cannot' compete after purchasing news at that price, then the court will be unable to fashion a remedy that will provide relief to the Post. In that event, under established principles of equity, the case must be dismissed: courts cannot entertain claims 'that cannot be remedied. I discuss these issues in more detail below, see infra at 1312-1328.
Because the essential facilities cases all involve defendants that are already engaged in providing the service at issue, those cases do not foresee, let alone address, any of the practical problems that would arise if the court were to impose a duty to deal in cases like this hypothetical case. Given these problems, and the failure of the essential facilities cases to provide any guidance, that doctrine provides no basis for liability under section 2 in this hypothetical case.
This hypothetical case, moreover, closely resembles the present case. For the same reasons that the hypothetical case is not *1302properly an essential facilities case, neither is the present case. City Gas has decided not to enter the business of supplying natural gas at wholesale. In so deciding, it has done nothing wrong, and its conduct in refusing to sell wholesale gas to Consolidated is not anticompetitive conduct under these essential facilities cases. Nor do the antitrust laws impose a duty on City Gas to enter the wholesale supply business and sell gas to Consolidated. The antitrust laws impose no positive duty on a company to help its competitors. Imposing such a duty in this case, as in the hypothetical case discussed above, would create serious problems of policy and of judicial administration. For these reasons, under the teachings of Terminal R.R., Associated Press, and their progeny, the essential facilities doctrine does not justify the imposition of liability under section 2 in the present case.
E. The “Intent Test."
The district court also analyzed the case under the so-called “intent test,” which considers whether, under the circumstances of a case, a monopolist’s conduct demonstrates an illegal intent to destroy competition. If the conduct gives rise to an inference of anticompetitive intent, then the monopolist is subject to liability under section 2. See Lorain Journal, 342 U.S. at 153, 72 S.Ct. at 186; Eastman Kodak, 273 U.S. at 375, 47 S.Ct. at 404; MCI, 708 F.2d at 1148; Poster Exchange, 431 F.2d at 338. Applying these cases generally, the district court found, “based on a review of the record evidence, that [City Gas'] real purpose here was to maintain its monopoly unlawfully.” 665 F.Supp. at 1540. It held, therefore, that City Gas’ conduct constituted a violation of section 2.
Courts have used this approach in both monopoly-leveraging and non-monopoly-leveraging cases. In the former, which I discuss above, see supra at 1289-1290, the courts have held that the conduct of monopoly leveraging is sufficient evidence of illegal intent to justify the imposition of liability under section 2. See, e.g., Eastman Kodak, 273 U.S. at 375, 47 S.Ct. at 404; Poster Exchange, 431 F.2d at 339. Courts have also inferred intent in non-monopoly-leveraging cases, where a company with monopoly power on one level of the market refuses to deal in order to maintain its monopoly on that same level of the market. See, e.g., Lorain Journal, 342 U.S. at 153-54, 72 S.Ct. at 186-87. This so-called “intent test” is of course nothing more than a fancy label for what courts do in virtually every refusal-to-deal case under section 2: because direct evidence of anti-competitive intent rarely exists, courts evaluate the circumstantial evidence in a given case to determine whether it gives rise to an inference of anticompetitive intent. The intent-test cases, therefore, demonstrate factual contexts from which such an inference can arise.
As the preceding sections make clear, the present case does • not involve monopoly leveraging. The district court, therefore, could not have inferred an illegal intent from an attempt by City Gas to extend its monopoly power from one level of the market to another. As I discuss above, the Eastman Kodak-Poster Exchange cases, which involve monopoly leveraging, do not control the present case. If the evidence in this case is to give rise to an inference of anticompetitive intent, therefore, it must do so on the basis of City Gas’ conduct in using its power on the retail level in order to maintain its monopoly on that same level. Given this situation, Lorain Journal becomes the relevant case.
In that case, the only newspaper in Lo-rain, Ohio refused to accept paid advertisements from parties also advertising on a radio station that had begun to compete with the newspaper in the advertising market. The government brought an enforcement action under section 4 of the Sherman Act, claiming that the newspaper had acted in violation of section 2. The Supreme Court agreed and approved an injunction requiring the newspaper to accept advertisements on its already established terms even if the parties also advertised on the competing radio station. The newspaper was not itself involved in the radio business, nor did it have any desire to become involved in that business. It was not, *1303therefore, attempting to use its monopoly power in the newspaper advertising business to gain a share of the radio broadcasting market. Rather, the newspaper’s conduct was designed to protect its own monopoly in the newspaper advertising business. That conduct, though, demonstrated the newspaper’s anticompetitive intent to foreclose competition in the relevant advertising market.
Lorain Journal, like other refusal-to-deal cases, has a crucial factual feature: the newspaper refused to provide a service that it was already in the business of providing. The newspaper sold advertising space on established terms (presumably a given dollar amount per column inch) to parties that wanted to advertise. The case arose when the newspaper refused to sell such space on the same terms to certain parties that also wanted to advertise.40 The present case, of course, lacks this feature: City Gas has not refused to sell what it was already in the business of selling. Consolidated wanted to buy gas at wholesale. City Gas, however, sold only retail gas, which is an entirely different product. Because of this factual distinction, Lorain Journal does not control here for two reasons.
First, what gave rise to an inference of anticompetitive intent in Lorain Journal was the newspaper’s diversion from its established course of conduct. The newspaper suddenly refused to sell advertising. Of course, it also refused only with respect to a specific class of advertisers — those also advertising on the competing radio station. The sudden refusal, coupled with the identity of the refused customers, clearly signaled an anticompetitive intent. In the present ease, City Gas did not change its established course of conduct. Rather, it did exactly the opposite: it refused to divert from its established course of conduct. City Gas has never sold gas at wholesale, and it continued not to do so in the present case. The fact that Consolidated, the would-be purchaser here, was a competitor of City Gas means nothing because City Gas never sold to any firm at all, regardless of the firm's competitive status. Lorain Journal is therefore distinguishable on the facts.
In addition, because this case does not involve a business in which City Gas was already engaged, it poses remedial problems not present in Lorain Journal. There, the trial court could simply mandate that the newspaper sell advertising space to the excluded parties on the same terms that the newspaper sold space to all other advertisers. Such a remedy would provide the excluded parties with relief — advertising space in the newspaper — and the court could easily impose that remedy within the limits of its administrative capabilities. That, however, is not the case here. Because City Gas does not, and has never, sold natural gas at wholesale, it obviously has established no terms for such a transaction. The district court, therefore, could not simply mandate that City Gas sell Consolidated wholesale gas on its already-established terms. Instead, the district court itself had to set a price for the forced sale. The district court did not do so, however; rather, it delegated the task to the FPSC. Had the district court undertaken to set the price, it would have found that task to be impossible, as I discuss in detail in part VI below. For this reason as well, Lorain Journal does not control the present case.
As this discussion demonstrates, therefore, the so-called “intent test” provided no basis for imposing liability here. Unlike the typical réfusal-to-deal cases analyzed under this test, the circumstances in this case simply did not give rise to an inference of anticompetitive intent. I turn now to a discussion of two additional Supreme Court cases.
F. Aspen and Otter Tail.
The district court also relied heavily on two more recent Supreme Court cases—Aspen Skiing Co. v. Aspen Highlands Skiing *1304Corp., 472 U.S. 585, 105 S.Ct. 2847, 86 L.Ed.2d 467 (1985), and Otter Tail Power Co. v. United States, 410 U.S. 366, 93 S.Ct. 1022, 35 L.Ed.2d 359 (1973)—which also involve refusals to deal under section 2. Although lower courts frequently cite both cases as support for the essential-facilities doctrine and the intent test, as well as the monopoly-leveraging theory, neither case explicitly applies any of these formulas for determining liability in section 2 cases. Both cases, nevertheless, follow from the general principles established in the early Supreme Court cases that I discuss in the preceding section. I consider Aspen and Otter Tail in turn, analyzing the theories of liability applied in the cases and demonstrating that neither case supports the imposition of a duty to deal on City Gas.
1. Aspen.
In Aspen, Aspen Highlands Skiing Corp. (Highlands), which owned one of four skiing areas in Aspen, brought a private antitrust suit against Aspen Skiing Company (Ski Co.), which owned the other three areas. From 1962 to 1977, Ski Co. and Highlands had offered an all-Aspen ticket that skiers could use interchangeably at all four skiing areas. In the 1977-78 season, however, Ski Co. offered to participate in the all-Aspen ticket only on terms that were unacceptable to Highlands; Ski Co. would consider no counterproposal and, as the district court held, in effect refused to deal with Highlands. Highlands brought a treble damages action against Ski Co. under section 4 of the Clayton Act, claiming that Ski Co. had monopolized the downhill skiing market in violation of section 2 of the Sherman Act. As the Supreme Court described, the case raised the question whether “a firm with monopoly power has a duty to cooperate with its smaller rivals in a marketing arrangement in order to avoid violating § 2 of the Sherman Act.” 472 U.S. at 587, 105 S.Ct. at 2849.
In the district court, a jury found in favor of Highlands, and the Tenth Circuit affirmed, relying on the essential facilities doctrine as a basis for holding that Ski Co. had a duty to deal with Highlands. The Tenth Circuit also held that sufficient evidence existed to support the jury’s finding that Ski Co. had an intent to create or maintain a monopoly. The Supreme Court approached the case in broader terms, analyzing whether Ski Co. had a duty to deal with Highlands under the circumstances. The Court first stated that “even a firm with monopoly power has no general duty to engage in a joint marketing program with a competitor.” Id. at 600, 105 S.Ct. at 2856. The Court elaborated that this proposition is a counterpart to the “independent businessman’s cherished right to select his customers and his associates.” Id. at 601, 105 S.Ct. at 2856. Of course, as the Court quickly qualified, these rights — both to select customers and not to deal with competitors — are not absolute. The Court articulated the key to determining when such a decision constitutes an antitrust violation: when “the challenged conduct is fairly characterized as ‘exclusionary’ or ‘anticom-petitive.’ ” Id. at 602, 105 S.Ct. at 2857. The Court then examined Ski Co.’s conduct.
The Court identified a crucial feature of Ski Co.’s refusal to deal:
In the actual case that we must decide, the monopolist did not merely reject a novel offer to participate in a cooperative venture that had been proposed by a competitor. Rather, the monopolist elected to make an important change in a pattern of distribution that had originated in a competitive market and had persisted for several years.
Id. at 603, 105 S.Ct. at 2858. Ski Co.’s refusal to deal thus demonstrated one of the key features of the Terminal R.R. and Associated Press cases: its refusal to deal was not a refusal to undertake a new course of dealing; rather, it was a refusal to continue a course of dealing in which it was already engaged.
After identifying this feature, the Court suggested that even such a decision “to make an important change in the character of a market ... is not necessarily anticom-petitive.” Id., 105 S.Ct. at 2858. As the Court explained, the question “cannot be answered by simply considering [the] effect of [Ski Co.’s conduct] on Highlands. In addition, it is relevant to consider its impact *1305on consumers and whether it had impaired competition in an unnecessarily restrictive way.” Id. at 605, 105 S.Ct. at 2859. Based on its consideration of these additional factors, the Court agreed that the evidence was sufficient to support a finding that “consumers were adversely affected by the elimination of the 4-area ticket.” Id. at 606, 105 S.Ct. at 2859. The Court also noted that Ski Co.’s conduct was not “justified by any normal business purpose.” Id. at 608, 105 S.Ct. at 2860. Because the evidence was sufficient to support the jury’s verdict in favor of Highlands, the Court affirmed.
This case does not support the imposition of a duty to deal on City Gas. The decisive feature in Aspen—Ski Co.’s decision “to make an important change in a pattern of distribution that had originated in a competitive market and had persisted for several years,” id. at 603, 105 S.Ct. at 2858—is not present in this case. On the contrary, City Gas made no “important change in a pattern of distribution;” indeed, City Gas refused to make such a change, and that refusal is what Consolidated challenges in the present case. Given this crucial distinction between Aspen and the present case, Aspen does not control.
The remedy problem in the present case also did not arise in Aspen. Although the Supreme Court made no mention of the damages amount awarded in Aspen, the long history of the all-Aspen ticket produced an extensive body of evidence on the division of profits from the all-Aspen ticket between Ski Co. and Highlands. The trial court (or jury) in that' case could easily have arrived at a damages figure that would accurately have reflected Ski Co.’s economic position while providing relief to Highlands. That is not true in the present case. Because City Gas has never sold gas for resale, either to Consolidated or anyone else, the district court had no point of reference for calculating a hypothetical transaction price. See infra at 1317-1319, for a detailed elaboration of this point.
Aside from the problem of imposing a remedy in the present case, the district court below also failed to focus on any of the factors articulated by the Supreme Court in Aspen as support for the Court’s determination that Ski Co.’s conduct was exclusionary: whether the conduct had an “impact on consumers and whether it has impaired competition in an unnecessarily restrictive way.” Id. at 605, 105 S.Ct. at 2859. A consideration of these factors in' the present case also suggests that City Gas’ refusal to deal is not classifiable as exclusionary.
With respect to the conduct’s impact on consumers, the record contains no evidence of a negative impact. In fact, the district court indicated that City Gas’ service was more economical than the service that Consolidated could provide. See 665 F.Supp. at 1505. The district court did note that City Gas would have to pass on to its customers the costs of extending its pipesystem into Consolidated’s service area. Id. at 1513. The district court, however, failed to account for the increased price that Consolidated would have to pay for wholesale gas compared to the price City Gas paid FGT: even if City Gas were forced to sell to Consolidated, it would be entitled to sell at a rate providing a reasonable rate of return on its investment, see supra part VI. Assuming that City Gas’ distribution and hook-up operation was efficient, and no evidence to the contrary appears, Consolidated would have to make up for the difference in its wholesale price either by charging its retail customers more than City Gas would charge or by streamlining even further its distribution operation. Even assuming that Consolidated could streamline its operation, the record contains no evidence that this would actually benefit consumers. Unlike the negative impact on consumers from the abolition of the all-Aspen ticket in Aspen, the benefits to consumers from Consolidated’s entry into the market would be minimal — at best, slightly reduced natural gas rates for some consumers. Even that benefit would come only at the expense of higher natural gas rates for other consumers. I question, moreover, whether this type of competition really would benefit the consumer. Competition that forces a natural gas utility to streamline its operation might, in fact, harm the consumer.
*1306Although such competition could conceivably result in slightly decreased gas rates for some customers, it might increase safety risks for all customers by pressuring gas suppliers to cut corners in order to lower their prices. That sort of pressure would not benefit consumers.
Moreover, the district court failed to consider that given the nature of the natural gas business, competition in that business is likely to benefit consumers only in the short term, if at all. As the district court stated, City Gas possessed a natural monopoly in its service area. That is not strange: because entering the natural gas supply business involves the expenses of building a pipesystem and arranging hookups to customers, virtually every supplier has a natural monopoly in its supply area. Two supply companies cannot continue to operate in the same service area. Companies compete only at the outset to gain control of a service area and to expand into new areas as they develop. Given the naturally monopolistic state of the natural gas supply business, the benefits of competition to the consumer are limited. Competition between companies for initial control of a service area may lower retail prices, but only in the short term. Once one supplier gains control of the service area (and thus develops a natural monopoly), the incentive for lower retail pricing also ends. (Of course, the FPSC monitors the rates set by all natural gas companies. See infra at 1332.) Moreover, consumers would not benefit from continuing competition between gas companies within a single service area. Constant switching by customers between companies would generate transaction costs and force an increase in retail prices to all customers. In addition, constant competition between companies is likely to undermine safety concerns as the companies try to shave costs in order to increase their competitiveness. Competition in the natural gas supply market is, therefore, by no means the best state of affairs from the consumers’ perspective.
As this discussion indicates, Aspen does not support the imposition of a duty to deal in the present case. The Aspen case is immediately distinguishable iron, the present case in that Ski Co.’s refusal to deal involved a significant alteration in its previous course of dealings with Highlands. That course of dealing provided a reference point for a meaningful damages award. No such course of dealing exists in the present case. Also, whereas the refusal to deal in Aspen adversely affected skiers, the refusal in this case had no such effect on gas consumers. In fact, the evidence suggests that the continued existence and expansion of City Gas’ natural gas monopoly might actually benefit consumers, especially with respect to long-term safety concerns.
2. Otter Tail.
Nor does the Supreme Court’s holding in Otter Tail, 410 U.S. at 366, 93 S.Ct. at 1022, support the imposition of a duty to deal in the present case. In Otter Tail, an electric power company produced electricity on the wholesale level and distributed its electricity as a retail franchisor in various municipalities. The company also engaged in “wheeling,” or transmitting power generated by other companies over its transmission system. On the retail distribution level, municipalities granted Otter Tail ten- or twenty-year franchises, and during that time Otter Tail served as the sole distributor — a natural monopolist — in the given municipality. The case arose when certain municipalities attempted to establish municipal distribution systems on their own. Having built their own local distribution systems, the municipalities hoped that Otter Tail would sell or wheel them wholesale power. Otter Tail, however, refused to provide either service. Consequently, the government brought an antitrust enforcement action against Otter Tail, attempting to enjoin it from refusing to deal with the municipalities under the circumstances. The district court granted the government an injunction, and the Supreme Court upheld the decree.
Unfortunately, the Otter Tail opinion reveals no immediate rationale in support of the district court’s action. As Judge East-erbrook observed in his dissent in Fishman *1307v. Estate of Wirtz, 807 F.2d 520, 573-74 (7th Cir.1986):
A study of the majority opinion for four Justices in Otter Tail will not reveal the Court’s reason; it gave only a result.... It did not offer any reason other than to say that Otter Tail had reduced competi-tion_ To rely on Otter Tail my colleagues must be able to tease a reason out of a silent opinion.
In Judge Easterbrook’s view, the key to the Otter Tail decision is the effect that Otter Tail’s refusal to deal had on the electricity-generating level of the market. As Judge Easterbrook described, three markets existed in the electric power industry: the power generation market, the transmission market, and the retail distribution market.' Otter Tail controlled the transmission market, and its refusal to deal occurred on that level. What made that refusal to deal an antitrust violation, in Judge Easterbrook’s view, was that competition existed on the generating level. That is, the municipalities that wanted to enter into retail distribution of electricity could have taken advantage of the competition on the generating level, which would have resulted in lower prices for consumers at the retail level, if Otter Tail would have agreed to wheel. And according to Judge Easter-brook, consumer benefit is the primary consideration in determining antitrust liability. See also Aspen, 472 U.S. at 605, 105 S.Ct. at 2859; supra at 1305. Thus, Otter Tail’s conduct was clearly anticompetitive.
Given this interpretation, Otter Tail does not control the present case. Judge East-erbrook’s description of the three levels of the electric power industry applies also to the natural gas industry. As I describe above, see supra at 1288-1290, the natural gas industry consists of three stages: (1) production; (2) interstate transportation and wholesale supply; and (3) intrastate distribution at retail. The present case, however, is factually distinguishable from Otter Tail in two respects: first, whereas Otter Tail participated on the transmission level as well as on the distribution level, City Gas participated only on the retail distribution level; ■ second, whereas competition existed on the generating level in Otter Tail, no such competition existed in the present case — FGT was the only source of wholesale gas in the region. Thus, the absence of competition on the generating (or wholesale) level in the present case means that consumers would not have ben-efitted had City Gas not refused to deal, in contrast to Otter Tail where consumers would have benefitted had Otter Tail not refused to deal. In addition, because City Gas never participated in the second level of the market, its conduct had no causal connection to any benefit consumers might have derived even if competition had existed on the production or wholesale levels. Applying Judge Easterbrook’s interpretation of Otter Tail, then, that case does not control here.
Another possible rationale for the Otter Tail holding is that Otter Tail engaged in classic monopoly leveraging — using its power on the transmission level to increase its business on the retail distribution level. Throughout the Otter Tail opinion, the Court indicated that Otter Tail had “ ‘a strategic dominance in the transmission of power in most of its service areas’ ” and that Otter Tail used that dominance “to foreclose potential entrants into the retail area from obtaining electric power from outside sources of supply.” Id. at 377, 93 S.Ct. at 1029. This rationale, of course, does not apply in the present case: City Gas has not engaged in monopoly leveraging. Otter Tail was clearly a vertically integrated company; City Gas, as I discuss above, was just as clearly not vertically integrated.
As the district court in Otter Tail stated, Otter Tail’s “operation consists of an integrated power system running the full gamut from initial production to final sale of electrical power combined with pool arrangements which supply emergency power reserves.” United States v. Otter Tail Power Co., 331 F.Supp. 54, 56 (D.Minn.1971). Although Otter Tail did the majority of its business on the retail level, it also contracted with a number of other electric systems to obtain supplementary power, which it both used itself and “regularly” wheeled to other companies. Id. at 56-57. *1308For example, it wheeled power to approximately eighteen municipalities pursuant to an interconnection contract with the Bureau of Reclamation, the entity that had agreed to supply wholesale power to one of the municipalities with which Otter Tail refused to deal. Id. at 57-58. City Gas, on the other hand, was not involved in the initial production of gas; it was not a natural gas wholesaler; nor did it buy excess quantities of gas to distribute to other companies for resale. Rather, it was engaged solely in the retail distribution of gas. City Gas was not attempting to use its market power as a natural gas producer or wholesaler (it was neither) to maintain its monopoly as a retail distributor of gas. It was simply choosing to maintain its operation exclusively on the retail level as it then existed, refusing to extend that operation to another level of the market. Whereas Otter Tail’s anticompetitive conduct followed from its vertical integration, the alleged violative conduct here followed from City Gas’ decision not to integrate vertically. For this reason, Otter Tail does not control.
Another point of distinction between Otter Tail and the present case concerns the status of the municipalities attempting to enter the retail distribution market in Otter Tail and that of Consolidated in the present case. For example, Elbow Lake, one of the municipalities that Otter Tail refused to supply, had constructed its own generating plant. The town had also arranged for a supply of wholesale power from the Bureau of Reclamation, which had high voltage lines in the area but relied on Otter Tail and other companies to wheel power from its bulk lines to its wholesale customers. When Otter Tail refused, Elbow Lake applied to the Federal Power Commission (FPC) for an order requiring Otter Tail to interconnect to Elbow Lake and sell the wholesale power itself.41 Elbow Lake had thus independently assumed the costs and faced the barriers of entering into the retail distribution market. Not only had Elbow Lake built its own generating facilities, it had also gained the Commission’s approval to interconnect with Otter Tail, a process that involved litigation against Otter Tail in two forums, see id. at 371 & nn. 4 & 5, 93 S.Ct. at 1027 & nn. 4 & 5.
Here, in contrast, Consolidated has attempted to limit its entry costs: it wanted to purchase its gas from City Gas in order to enter the market in the most competitive posture and at the least possible expense. Whereas Elbow Lake completed the administrative process necessary to gain Commission approval of an interconnection to Otter Tail’s system, Consolidated not only attempted to avoid the delay of such proceedings but also pointed to that delay as a justification for requiring City Gas to provide it with wholesale gas. Moreover, whereas Elbow Lake built an entire generating plant at great expense, Consolidated refused to invest even $250,000 to build the lateral pipeline that would have connected its pipesystem to FGT’s line. In short, Elbow Lake was not attempting to obtain a free ride into the retail distribution market by forcing Otter Tail involuntarily to undertake a new function so that Elbow Lake could avoid the market’s entry barriers and enjoy an enhanced competitive position. That, I suggest, is exactly what Consolidated is attempting to do. To impose a duty to deal on City Gas in these circumstances, therefore, would be to say that any potential market entrant deserves a free ride into the market, on the best possible terms, and at the expense of those entities that have already entered the market on more difficult terms, with more risk, and at greater costs. The antitrust laws, in my view, do not support such a result.
Otter Tail provides no basis for imposing a duty to deal in the present case for one further reason: it fails to address the remedy problem that arises here. The Otter Tail decree, which the Supreme Court approved, mandates that Otter Tail supply power — either by selling its own at wholesale or by wheeling from other companies — to municipalities that have established their own retail distribution systems. The Supreme Court addressed the regula*1309tory issues implicated by the decree,42 but the Court conveniently failed to address the question of price: if Otter Tail had to supply power, then how much should it get paid? As the Supreme Court concluded, “the District Court ... in any event has retained jurisdiction to enable the parties to apply for ‘necessary or appropriate’ relief.” 410 U.S. at 381-82, 93 S.Ct. at 1031-32. Given the particular circumstances of the case, the district court on remand could presumably come up with a reasonable price term. Otter Tail already sold power at wholesale and wheeled power. Thus, the price terms in those existing transactions would provide a point of reference as the district court attempted to impose a price term for future transactions. In addition, Otter Tail did a substantial portion of its wheeling pursuant to a contract with the Bureau of Reclamation, a major supplier of wholesale power in the area. Many of the municipalities with which Otter Tail had refused to deal had themselves arranged to purchase wholesale power from the Bureau under the assumption that Otter Tail would wheel it. Thus, an even more specific reference point existed for imposing a price term with respect to many of the transactions that would arise pursuant to the district court’s decree in Otter Tail.
The present case poses a different set of problems. Otter Tail, however, provides no solution. Given these circumstances, Otter Tail most certainly does not justify the imposition of liability in this case.
G. MCI Revisited.
I return finally to MCI Communications Corp. v. AT&T, 708 F.2d 1081 (7th Cir. 1982), cert. denied, 464 U.S. 891, 104 S.Ct. 234, 78 L.Ed.2d 226 (1983), focusing on an aspect of the Seventh Circuit’s holding that, in my view, does control in the present case and that demonstrates the absence of liability under section 2 here. As I discuss above, see supra at 1293-1294, the MCI case arose out of MCI’s effort to enter the long-distance telephone market— an effort that required access to the Bell network’s local exchange facilities. As I also discuss above, see id., the Seventh Circuit held that AT&T had a duty to allow MCI access to those,local exchange facilities.43 The court reached this conclusion by applying the essential facilities doctrine.
As the court stated:
A monopolist’s refusal to deal under these circumstances is governed by the so-called essential facilities doctrine. Such a refusal may be unlawful because a monopolist’s control of an essential facility (sometimes called a “bottleneck”) can extend monopoly power from one stage of production to another, and from one market into another. Thus, the antitrust laws have imposed on firms controlling an essential facility the obligation to make the facility available on non-discriminatory terms.
708 F.2d at 1132 (citing Terminal R.R., 224 U.S. at 410-11, 32 S.Ct. at 515-16). According to this statement, the concern motivating the essential facilities doctrine is that an entity with monopoly power on one level of a market can use that power to gain control on another level of a market— *1310i.e., by monopoly leveraging. In order to prevent this from occurring, the essential facilities doctrine imposes a duty to deal on the monopolist.
Applying this rationale to AT&T’s refusal to provide local interconnections for MCI’s FX and CCSA services, see supra at 1293-1294, the court characterized AT&T’s refusal as an attempt to use its monopoly over local exchange systems to maintain its monopoly over long-distance services. Id. at 1133. The court pointed to Otter Tail as authority for holding that AT&T’s conduct was anticompetitive. The court characterized the rationale behind the Otter Tail decision as a concern “that market power in one market (transmission) was being used to further a monopoly in another market (retail distribution).” Id. (citing Otter Tail, 410 U.S. at 377-79, 93 S.Ct. at 1029-30). Based on this interpretation of Otter Tail and the facts in MCI, the court held that AT&T’s conduct was anticompetitive. MCI wanted to compete with AT&T on the long-distance level, but to compete on that level, MCI required access to the exchange facilities at the local level. AT&T refused to provide MCI that access, and the court held that the refusal violated section 2.44 That holding provides no support for the district court’s decision in the present case. As I maintain throughout this part, the situation is entirely different here: City Gas has not engaged in monopoly leveraging.
A different aspect of the MCI opinion, however, is instructive here. The MCI court also held that AT&T had no duty to allow MCI to interconnect to AT&T’s long-distance system (as opposed to the local exchange facilities) for multipoint service. Multipoint service consists of a private line for a customer between City A and City B, such as MCI’s Chicago-to-St. Louis line. In requesting multipoint interconnections to AT&T’s system, MCI wanted access to AT&T’s multipoint line running from City B to City C, for example from St. Louis to Kansas City. This claim was entirely different from the claim concerning interconnections to Bell’s local exchange system. In effect, multipoint interconnections required access to AT&T’s own long-distance system rather than to the local exchange system, which was all that MCI required in order to operate its FX and CCSA services. The multipoint interconnections, therefore, involved only the long-distance market, the very market in which MCI was attempting to compete with AT&T. MCI claimed that it could not compete in that market — by providing multipoint service to its customers — unless AT&T provided access to its own lines. AT&T contended that MCI was simply trying to use AT&T’s lines to provide service to customers in cities where MCI’s current facilities could not reach, even though MCI was authorized to build facilities that could reach those cities.
The Seventh Circuit rejected MCI’s claim, holding that the denial of interconnections for multipoint service could not form a basis of liability. Id. at 1149. The court reasoned that
MCI’s primary business was to build precisely the type of facilities to which it sought access from the Bell System. There was no sufficient explanation as to why MCI, on the one hand, was building its own network, and, on the other, was entitled to access in the interim to AT& T’s facilities. Thus, the jury lacked sufficient evidence to conclude that these interconnections were essential.
Id. at 1148. The court continued:
As a matter of antitrust liability, ... can an entrant which actually builds its own facilities between Chicago and Milwaukee, for example, thereby gain entitlement to use all the far-flung facilities of the Bell System? Is its entitlement based on its expressed intention to duplicate major portions of the Bell System on *1311a national basis? Could it claim entitlement before (or without) building any facilities of its own? We think the ramifications of the demand for multipoint service are troubling and complex, and that under the circumstances of this case ..., the denial of interconnections for multipoint service cannot form a basis of liability.
Id. at 1149.45
I suggest that this holding controls in the present case. Like MCI, Consolidated had no basis for claiming an entitlement to buy wholesale gas from City Gas. As in MCI, Consolidated’s primary business was to provide the same service as City Gas provided. City Gas operated a retail distribution system; Consolidated wanted to do exactly the same. Like MCI, Consolidated claimed that it could not compete in the retail distribution market unless it had access to City Gas’ system. As in MCI, Consolidated wanted access to City Gas’ system so that it could provide the same service that City Gas provided and compete with City Gas on the same, and only, level of the market on which City Gas operated. Consolidated, like MCI, had every right to compete in the market at issue, and City Gas, like AT&T, could do nothing to prevent Consolidated from competing. As the MCI court indicated, however, City Gas had no duty to help Consolidated enter into competition against it. Like AT&T, which was not in the business of providing interconnections for other long-distance services (as compared to the local Bell companies, which were in the business of providing interconnections for long-distance systems), City Gas was not in the business of selling gas at wholesale. The refusal to deal in the present case, like AT&T’s refusal to provide multipoint interconnections in MCI, related only to a single level of market operation and involved no attempt to leverage into another market level. Following the rule in MCI, if Consolidated wanted to provide the same service as provided by City Gas, which operated on a single market level as a non-vertically integrated entity, then Consolidated itself had to construct the facilities that such an operation required.
H. Conclusion.
As I discuss above, see supra at 1299-1300, the antitrust laws do not impose a positive duty to help a competitor. The antitrust laws proscribe conduct that inhibits competition; the laws, however, do not purport to facilitate competition by imposing a duty on a company to help another company into competition. As the Seventh Circuit stated in Olympia, a firm
has no general duty to help its competitors. ..’. “There is a difference between positive and negative duties, and the antitrust laws, like other legal doctrines sounding in tort, have generally been understood to impose only the latter.” ... A monopolist has no duty to reduce its prices in order to help consumers, ... and no duty to extend a helping hand to new entrants....
797 F.2d at 375-76 (quoting SPS Technologies, 694 F.2d at 512-13) (citations omitted). This is especially true when such a duty would require the company to undertake a course of dealing in which it does not already engage — such as providing interconnections or selling gas at wholesale where it normally does not. Courts must respect rational decisions of economic actors. If an entity has opted not to undertake a course of operation, courts have no business second-guessing that decision. In addition, the imposition of a duty to deal in such circumstances would encourage economic inefficiencies and promote inequity. An attempt by any company to develop a product or service — i.e., to compete in a market — requires initiative, innovation, and risk-taking. Our antitrust laws are designed to ensure that every company has *1312the opportunity to compete under these terms, but not to require that a company which has made this investment in capital and taken this initiative and risk provide a free ride into the market for every potential competitor that comes along. Such a policy would be unfair to those companies that have made investments and taken risks; moreover, it would discourage businesses from taking initiative, assuming risk, and making investments of capital in the first place. A policy with these results would not, in my view, be a good one. More importantly, it is not the policy that our antitrust laws were designed to promote.
As the above sections demonstrate, no legal basis exists for requiring City Gas to enter the wholesale business and supply Consolidated with gas. The factual circumstances of this case are fundamentally different from typical refusal-to-deal cases, and given the circumstances here, City Gas' refusal to deal gives rise to no inference of anticompetitive intent under any of the approaches used by courts to evaluate circumstantial evidence. Finally, as I discuss in more detail in the next part, the present case presents extreme administrative problems for the imposition of an effective remedy — problems that Congress could not have intended the courts to solve. City Gas’ conduct, therefore, does not constitute a violation of section 2.
VI. THE FIFTH AMENDMENT
As part V demonstrates, this case is not a typical refusal-to-deal case. In the typical refusal-to-deal case, a defendant participates in the market at issue, selling a product or service either directly to a plaintiff (in previous transactions) or to other customers. The defendant then attempts to charge the plaintiff a higher price than it charges other customers or simply refuses outright to deal with the plaintiff. As a result, the plaintiff46 brings suit under the Clayton Act, 15 U.S.C. §§ 15, 26, contending that the defendant has violated section 2 of the Sherman Act by refusing to supply the product at issue to the plaintiff and seeking an injunction ordering the defendant to supply the product on the same terms as those given to the defendant’s other customers and also seeking damages.47 If the court concludes that the defendant has violated section 2 of the Sherman Act by charging the plaintiff higher prices or refusing outright to deal with the plaintiff, then the court orders the defendant to deal with the plaintiff and to do so at the same price that the defendant has set in its dealings with other customers.
In the present case, City Gas has never sold gas for resale to anyone.48 In decid*1313ing not to supply gas to Consolidated, City Gas did not discriminate against Consolidated by attempting to charge Consolidated a higher price than City Gas charged its other customers or by entirely denying Consolidated a service that City Gas provided to other customers. Under these circumstances, the district court could not conclude that City Gas violated section 2 by treating Consolidated differently than City Gas treated its other customers. Nor could the district court require City Gas to sell wholesale gas to Consolidated at the same price that City Gas sold gas to its other customers. Such a decree would be meaningless since City Gas has never sold wholesale gas to any firm at any price.
Under the typical refusal-to-deal cases, therefore, City Gas did nothing wrong. It merely operated as a retail gas distributor in its established service area and decided not to deal at all on the wholesale level. The district court, however, developed a new rule of law: it held that section 2 of the Sherman Act imposed a positive duty on City Gas to change the nature of its operation and sell wholesale gas to Consolidated so that Consolidated could enter the retail distribution market and compete with City Gas.49 The district court applied this rule in the present case to take a portion of City Gas’ gas supply and give it to Consolidated. This new rule thus requires City Gas — a company that has done nothing wrong — to undertake a new level of operation and engage in a forced sale of wholesale gas to Consolidated.
In my view, Congress could not have intended section 2 of the Sherman Act to mandate such a result. This becomes apparent when one considers the practical implications of the rule’s enforcement, which the district court, as well as this court today, has entirely failed to do. As those implications suggest, the rule will be impossible to apply. The district courts of this circuit will confront severe difficulties in fashioning remedies under the rule. In addition, the rule’s application will frustrate the very purposes behind the antitrust laws. I develop these points in the following section by analyzing how a court would apply this new rule to the facts of the present case.
I first introduce the idea that this rule implicates the fifth amendment takings clause. As I note above, because City Gas has never sold gas at wholesale and has therefore established no price for such sales, the court under these circumstances must set a selling price if it is to require City Gas to sell to Consolidated. That price must satisfy the fifth amendment’s just compensation requirement.50 I then consider what finding the just compensation clause requires under the present facts. Next, I discuss some further implications of the fifth amendment, which suggest that the rule is impossible to apply and leads to results that are inconsistent with the purposes behind the antitrust laws. Based on this discussion, I conclude that Congress could not have intended section 2 to support such a rule. The rule, therefore, should be struck down.
*1314Because our court today nevertheless upholds this impossible rule, I analyze in section B the district court’s errors in attempting to apply it and conclude that the judgment should be vacated and the case remanded for further proceedings as I describe.
A. Implications of Enforcing the Rule.
1. The Fifth Amendment.
The fifth amendment provides in pertinent part: “nor shall private property be taken for public use, without just compensation.” U.S. Const.Amend. 5. Courts have long held that the fifth amendment applies to utility rates set by federal and state regulatory agencies: in order to satisfy the just compensation clause, rates must not be “so ‘unjust’ as to be confiscatory,” Duquesne Light Co. v. Barasch, 488 U.S. 299, 307, 109 S.Ct. 609, 615, 102 L.Ed.2d 646 (1989). See, e.g., In re Permian Basin Area Rate Cases, 390 U.S. 747, 770, 88 S.Ct. 1344, 1361, 20 L.Ed.2d 312 (1968); FPC v. Natural Gas Pipeline Co. of Am., 315 U.S. 575, 585-86, 62 S.Ct. 736, 742-43, 86 L.Ed. 1037 (1942); Tenoco Oil Co. v. Department of Consumer Affairs, 876 F.2d 1013, 1020 (1st Cir.1989); Jersey Cent. Power & Light Co. v. FERC, 810 F.2d 1168, 1175 (D.C.Cir.1987). This line of cases indicates that the fifth amendment does not proscribe rate regulation in and of itself; rather, it proscribes rate regulation at a price that denies just compensation. The Supreme Court has therefore articulated a standard — the “historical cost basis” — for reviewing rate regulations to determine whether they satisfy this constitutional mandate. See Barasch, 489 U.S. at 309, 109 S.Ct. at 616-17; FPC v. Hope Natural Gas Co., 320 U.S. 591, 603-05, 64 S.Ct. 281, 288-89, 88 L.Ed. 333 (1944).51
Assuming the facts of the present case, a different category of regulation is involved. The regulation here goes beyond ratemak-ing to require City Gas to provide a service that it did not normally provide. If administrative price regulation implicates the fifth amendment, then this more intrusive category of regulation obviously implicates the fifth amendment as well.52 Thus, if *1316this “taking” is to pass constitutional muster, then a court confronting these facts must set a price in its injuction that provides City Gas with just compensation for its gas.53
The fifth amendment requirement means more though: it also controls the court’s determination of whether City Gas violated the duty to deal imposed under this new rule. The fifth amendment protects City Gas from having to sell its gas at a price lower than the just compensation price. Thus, City Gas violated its duty to deal under section 2 only if it refused to deal at the fifth amendment price.54 The court must make this determination by comparing the fifth amendment price to the selling price actually offered by City Gas and evaluating whether that offered price was unreasonably high, thus constituting a refusal to deal in violation of section 2. The court must therefore set the fifth amendment price in order not only to fashion its injunction, but also to evaluate whether City Gas violated its duty to deal in the first place.
In addition, the fifth amendment price controls the court’s determination of whether Consolidated can satisfy the injury-in-fact requirements of sections 4 and 16 of the Clayton Act. As I discuss below, see infra at 1321, in order to prove injury, or the likelihood of injury, for purposes of sections 4 and 16, Consolidated must show that it could have turned a profit by purchasing wholesale gas from City Gas at the fifth amendment price. If Consolidated could not have turned a profit after purchasing gas at that price, then City Gas’ refusal to deal did not injure Consolidated, and Consolidated has no standing under the Clayton Act to point to that conduct as a basis for liability under section 2. In order properly to decide this issue, however, the court must first find the fifth amendment price. I now consider the application of *1317these fifth amendment requirements to the present facts.
2. Finding a Fifth Amendment Price.
Assuming the facts of the present case, the court finds as follows: City Gas initially offered to sell the gas at cost55 plus ten cents per therm. Consolidated eventually made a counter offer to buy the gas at cost plus two and one-half cents per therm. City Gas rejected the offer as too low, but then lowered its initial price to cost plus seven cents per therm. Consolidated later proposed a price of cost plus five cents per therm. The parties reached no agreement and the negotiations ceased. See 665 F.Supp at 1510.56
Now, in order to prove its case under the new rule, Consolidated must make two showings: it must (1) establish a selling price that will justly compensate City Gas; and (2) prove that City Gas’ price of cost plus seven cents per therm was substantially higher than the just compensation price57 and therefore constituted a refusal to deal under section 2.58 I next consider how Consolidated can establish the just compensation price.
A just compensation price provides a property owner with a reasonable rate of return on its investment. In a typical takings case, for example where the government institutes a condemnation procedure and takes Blackacre so that the interstate highway can become two lanes wider, the government compensates the owner for Blackacre’s fair market value, or “what a willing buyer would pay in cash to a willing seller.” United States v. Miller, 317 U.S. 369, 374, 63 S.Ct. 276, 280, 87 L.Ed. 336 (1943).59 The present facts, however, do not describe a typical takings case. Rather than the one-shot taking of Blackacre to widen the interstate highway, the present facts involve a continuous taking of a portion of City Gas’ assets. The fair-market-value standard obviously does not pro*1318vide an acceptable formula for calculating a reasonable rate of return in a case involving this type of continuous taking. See Hope, 320 U.S. at 601, 64 S.Ct. at 287 (“The heart of the matter is that rates cannot be made to depend upon ‘fair value’ when the value of the going enterprise depends on earnings under whatever rates may be anticipated.”) Consolidated must therefore find an alternative basis for establishing a fifth amendment price.
As one possibility, Consolidated might point to a wholesale gas price set by a company similarly situated to City Gas, claiming that that price will provide City Gas just compensation. That company has to be virtually identical to City Gas in terms of the factors that a gas company takes into account in setting its wholesale price, see infra. For example, the company must deal in both the retail and wholesale markets. No such company exists under the facts of the present case.60 Nor can Consolidated point to any other existing operation, under the facts here, as a basis for calculating the just compensation price.61 Instead, Consolidated has to establish the fifth amendment price based on City Gas’ actual costs, which is in effect the historical-cost basis of the Hope-Barasch cases. See supra note 51 & accompanying text.
Under this inquiry, Consolidated must establish a variety of factors, for example: (1) the price at which City Gas itself purchases the gas; (2) City Gas’ operating costs in purchasing and delivering the gas; (3) City Gas’ fixed costs in purchasing and delivering the gas; and (4) any business opportunities that City Gas might lose as a result of selling gas to Consolidated.62 In addition, a court faced with the present facts must take into account the risks assumed by City Gas’ investors. The court must determine the return needed to compensate those investors for that risk, and the greater the risk, the higher the rate of return has to be.63 The court must evaluate all of the evidence and, adopting the role of a ratemaker, set a price that enables City Gas to sell at a reasonable profit. Consolidated must then prove that City Gas’ offered price was unreasonable.64 In *1319addition, the court must determine, for purposes of the injury-in-fact requirement of sections 4 and 16 of the Clayton Act, whether Consolidated can turn a profit at the fifth amendment price.
Courts have a difficult enough time interpreting and applying the historical-cost test in a reviewing capacity. See, e.g., Jersey Central, 810 F.2d at 1175.65 I suppose, however, that if an administrative agency can establish a rate by taking evidence concerning these factors, then the court can adopt the role of a regulator and do so too. Of course, the just compensation price established by the court under the present facts must satisfy a more stringent standard than the standard that applies to rates set by a regulatory agency. Cf id.; supra note 51.66
In order to demonstrate that Congress could not have intended section 2 to support such a rule in the first place, I continue with this analysis and assume, for the sake of discussion, that Consolidated manages to establish (1) the fifth amendment price in this case, (2) that City Gas breached its duty under section 2 by offering a price that was unreasonably high in light of the fifth amendment price, and (3) that Consolidated actually suffered damages as a result of City Gas’ refusal to deal. I assume further that the court orders City Gas to supply wholesale gas to Consolidated at the fifth amendment price. As I describe in the next section, however, the matter would not end there.
3. Changed Circumstances.
Assume that City Gas abides by the court’s decree and begins supplying gas for resale to Consolidated. A few months later, the cost of supplying gas to Consolidated increases. As a result, the price set by the court in its decree no longer justly compensates City Gas, and City Gas moves for a modification of the decree to increase the selling price. Consolidated opposes the increase. At that point, the court, which has retained jurisdiction, again has to assume the role of regulator and undertake an extensive inquiry into the facts to determine whether a price increase is justified.
The increase in City Gas’ costs can result from any of several factors: City Gas’ purchase price from FGT might increase; City Gas’ own delivery costs might increase; or, City Gas might make some business decisions that increase its operating costs. Assume here that City Gas decides that it must spend more money on advertising in order to compete with Consolidated: after all, Consolidated entered the market in order to compete with City Gas, which is exactly what the district court’s injunction was designed to achieve.
At the hearing, therefore, City Gas argues that the just compensation price must include the legitimate expenses of competing with Consolidated. If the fifth amendment price does not reflect these expenses, City Gas contends, then the company will eventually have to pass them off to its customers by increasing its retail price (assuming the FPSC will approve such a rate *1320increase).67 If the court denies the rate increase, according to City Gas, that will give Consolidated a decisive competitive advantage and will therefore defeat the purpose behind the court’s injunction — i.e., enabling both companies to compete on equal terms. City Gas argues that denial of the increase will give Consolidated a free ride at City Gas’ expense and that the fifth amendment surely protects it from that result.
Consolidated, on the other hand, argues that if the court increases the selling price, then City Gas will gain a competitive advantage in the marketplace. Consolidated’s costs will increase; it will, in turn, have to pass those costs off to its customers, see supra note 67; and as a result, it will be unable to compete with City Gas’ lower retail prices. That result, says Consolidated, is inconsistent with the policies behind the antitrust laws.
What is the court to do in such a situation? I suggest that whatever it does, it cannot serve the purposes of the antitrust laws while at the same time satisfying the fifth amendment’s requirements. Whether the court grants or denies City Gas’ request for an increase in the selling price, its decision will harm one segment or another of natural gas consumers. If the court grants the increase, then Consolidated’s consumers will eventually bear the burden of Consolidated’s increased costs. If the court denies the increase, then City Gas’ consumers will bear the burden of City Gas’ increased costs. Surely Congress could not have intended to create a rule that inevitably harms consumers: harm to consumers is exactly what Congress designed the antitrust laws to prevent.
The court might attempt to avoid this dilemma by placing controls on the competition between City Gas and Consolidated. For example, the court could allocate parts of the territory within the area of competition to each of the companies and prohibit them from competing for control of the other company’s territory. Such a forced allocation, however, would run afoul of the fifth amendment by restricting City Gas’ ability to compete and expand, if not by actually limiting City Gas’ present distribution area.
Finally, if the court denies the rate increase, what will it do when City Gas’customers, who will have to pay more for their gas, seek to intervene in the case on the ground that the court’s decision infringes upon rights accorded them (by implication) under the antitrust laws? If the court instead grants the increase, then Consolidated’s customers rather than City Gas’ customers will seek to intervene. In short, whichever way the court decides the question, consumers will be harmed.
Congress clearly could not have intended section 2 to support this new rule of liability. That rule, as my discussion demonstrates, forces the courts to adopt the role of a regulatory agency, an agency charged, moreover, with an impossible task and facing a set of never-ending problems. In addition, the rule frustrates the purposes behind the antitrust laws: it creates situations in which, however a court rules, consumers will be harmed. A rule that leads to such results is not consistent with Congress' intent in passing section 2.
B. The District Court’s Application of the Rule Below.
This court today, nevertheless, upholds this impossible rule. Even under this rule, however, I would vacate the district court’s judgment and remand the case for further proceedings. The district court failed to find a fifth amendment price in the present case, and that failure created reversible error on three grounds: (1) the district court never properly decided whether, given the fifth amendment price, City Gas actually refused to deal under section 2; (2) the district court failed to determine whether, for purposes of sections 4 and 16 of the *1321Clayton Act, Consolidated suffered any injury in fact as a result of City Gas’ refusal; and (3) the absence of the fifth amendment price also infected the remedies fashioned by the district court.
I would thus instruct the district court on remand first to find the fifth amendment price, in the manner I describe above, second to decide whether City Gas breached its duty to deal under the district court’s new rule, third to determine whether Consolidated suffered injury as a result of City Gas’ refusal to deal, and fourth, to revisit the remedies issues. I now provide more particular instruction on these last two points.
1. Injury in Fact.
It is an established principle of law that injuria absque damno, or an injury without damage, creates no right to a remedy. See Shumate & Co. v. National Ass’n of Secs. Dealers, Inc., 509 F.2d 147, 152 (5th Cir.1975) (“injury is the sine qua non for stating a cause of action”). Section 4 of the Clayton Act expressly incorporates this principle by creating a private suit for treble damages only for “any person who shall be injured in his business or property by reason of anything forbidden in the antitrust laws.” 15 U.S.C. § 15 (1988) (emphasis added); see Associated Gen. Contractors, Inc. v. California State Council of Carpenters, 459 U.S. 519, 103 S.Ct. 897, 74 L.Ed.2d 723 (1983); Blue Shield v. McCready, 457 U.S. 465, 476-78, 102 S.Ct. 2540, 2546-48, 73 L.Ed.2d 149 (1982); Brunswick Corp. v. Pueblo Bowl-O-Mat, Inc., 429 U.S. 477, 489, 97 S.Ct. 690, 697, 50 L.Ed.2d 701 (1977); Cable Holdings, Inc. v. Home Video, Inc., 825 F.2d 1559, 1561-62 (11th Cir.1987); National Indep. Theatre Exhibitors, Inc. v. Buena Vista Distribution Co., 748 F.2d 602, 607 (11th Cir.1984), cert. denied, 471 U.S. 1056, 105 S.Ct. 2120, 85 L.Ed.2d 484 (1985); McClure v. Undersea Indus., 671 F.2d 1287, 1289 (11th Cir. 1982). Similarly, section 16 of the Clayton Act, which authorizes private suits for in-junctive relief, requires the private plaintiff to show a likelihood of injury. See 15 U.S.C. § 26; Cargill, Inc. v. Monfort, Inc., 479 U.S. 104, 111-15, 107 S.Ct. 484, 490-91, 93 L.Ed.2d 427 (1986).
Because the district court failed to find the fifth amendment price, it did not properly decide whether Consolidated suffered injury. On remand, in order to prove injury in fact, Consolidated would have to show that it could have turned a profit after purchasing gas from City Gas at the fifth amendment price. If Consolidated can show that it could have turned a profit, that showing would prove that it suffered injury as a result of City Gas’ refusal to deal. The showing would satisfy section 4’s injury-in-fact requirement and would probably also demonstrate a likelihood of future injury for purposes of section 26 of the Clayton Act as well. If, on the other hand, Consolidated cannot show that it could have turned a profit, then City Gas’ refusal to deal caused no injury: Consolidated would have gone out of business regardless of the refusal.68 In that event, Consolidated would be unable to show that it suffered, or was likely to suffer, injury in fact and thus would be unable to make out a case under sections 4 or 26 of the Clayton Act. I next discuss the remedies issues.
2. Available Remedies.
The availability of remedies in this case — assuming City Gas violated section 2 — depends upon the assumption that Consolidated could have turned a profit with a supply of natural gas at the fifth amendment price. As I discuss above, injury is an element of liability under sections 4 and 16 of the Clayton Act. The principle of injuria absque damno, though, also implicates the availability of remedies. In the *1322present case, therefore, if Consolidated could not have turned a profit with gas, then it is entitled to no remedy. As I also discuss above, the fifth amendment price controls this injury question as well. Consolidated’s ability to turn a profit, and thus prove injury, depends upon the fifth amendment price. To the extent that the fifth amendment price changes, therefore, Consolidated’s profitability, and the question of injury, will also change. With these principles in mind, I now turn to the specific remedies issues raised here. Depending upon whether or not Consolidated can operate in the future as a natural gas distributor, one of two remedial schemes, which I discuss in turn in subsections a and b, will provide relief. As I demonstrate in subsection c, however, the two schemes are inconsistent with one another. In subsection d, I then consider the district court’s errors in granting Consolidated relief under both schemes and instruct it on how to proceed on remand.
a. The first remedial scheme.
In its amended complaint, Consolidated sought both permanent injunctive relief and damages for lost profits under the first remedial scheme. Consolidated’s request for permanent injunctive relief implied that at the time of City Gas’ refusal to deal, and at least until the date of the amended complaint, Consolidated could have turned a profit after purchasing wholesale gas at the fifth amendment price. The request for an injunction further implied that, based on Consolidated’s ability to turn a profit in the past, Consolidated could have turned a profit in the future if the district court granted the injunction and provided Consolidated with a supply of wholesale gas.69 Consolidated thus requested damages for its lost profits from the date of City Gas’ refusal to deal to the date of final judgment and a permanent injunction enabling Consolidated to operate in the future as retail natural gas distributor with a supply of gas. This combination of remedies would make Consolidated whole, with the damages compensating for any losses incurred before the injunction and the injunction enabling it to operate in the future as a company with a supply of natural gas.
In calculating Consolidated’s damages for lost profits under this remedial scheme, a court must undertake the following inquiry: based on the fifth amendment price, the court must determine what Consolidated’s profits would have been if it had a natural gas supply. Of course, if Consolidated could not have turned a profit, then it is not entitled to damages.70 In addition, because the fifth amendment price is always subject to change, the court must determine whether the fifth amendment price remained fixed from the time of the refusal to deal until the time of final judgment. If the fifth amendment price changed, then the court must reassess whether Consolidated could have turned a profit at each interval of changing prices.
If the court decides to grant the injunction, and requires City Gas to supply gas at the fifth amendment price, the court must then retain jurisdiction over the case and *1323entertain City Gas’ motions to modify the injunction in light of any subsequent changed circumstances or to dismiss the case in the event that Consolidated can no longer turn a profit. As I discuss above, the fifth amendment price is subject to constant change, as are Consolidated’s own costs. Thus, the court must retain jurisdiction over the case, not only to redetermine the fifth amendment price in light of successively changing circumstances, but also to redetermine whether Consolidated can still turn a profit at the modified fifth amendment price. If at any subsequent time Consolidated can no longer turn a profit at the fifth amendment price, then the court must vacate the injunction and dismiss the case.
This combination of damages for lost profits and injunctive relief would make Consolidated whole. The damages would compensate it for whatever loss it suffered as a result of the refusal to deal prior to the date of final judgment, and the injunction would enable it to operate in the future as a company with a natural gas supply-
b. The second remedial scheme.
As I suggest above, the implication of Consolidated’s request for injunctive relief was that Consolidated, which had been shut down for want of a natural gas supply, could resume operation if it acquired a natural gas supply. This outlook could change, of course, if, prior to trial, the effects of the refusal to deal put Consolidated permanently out of business, thus rendering injunctive relief useless. In that event, at trial, Consolidated would seek (1) damages for the net going concern value of the company, with gas, just prior to its demise and (2) damages in the amount of lost profits from the date of the refusal to deal to the date of its demise.71
The principle behind the damages for loss of Consolidated’s going concern value is that City Gas’ refusal to deal eventually resulted in Consolidated’s demise, in effect a taking of its business (less the assets that Consolidated kept). That is, as a result of City Gas’ continued refusal to deal, Consolidated could no longer operate as a natural gas distributor even if it had a supply of natural gas. The damages for loss of going concern value would therefore compensate Consolidated for the difference between what Consolidated would have been worth had City Gas supplied gas and what it was worth without a supply of gas just prior to its demise.72
In calculating these damages, a court must undertake the following analysis: the court must first determine Consolidated’s fair market value — or, what a hypothetical willing buyer, fully informed of all relevant facts, would pay a hypothetical willing seller, also fully informed of all relevant facts, for Consolidated — with gas as compared to without gas. Without a supply of gas, Consolidated would be worth the salvage value of its plant and equipment (its pipes-ystem), that is, what a hypothetical buyer would pay for the company without gas just prior to its demise — say, for the sake of discussion, $750,000. With a supply of gas, Consolidated would still not be worth any more than its salvage value unless it could have turned a profit with a supply of gas at the fifth amendment price. If Con*1324solidated could have turned a profit with gas, then the court must determine the damages caused by the refusal to deal through the eyes of the hypothetical willing buyer and seller.
Assume that the court finds that the buyer and seller strike the following deal: after dickering about how long Consolidated can count on a supply of gas from City Gas and about how long Consolidated can turn a profit at City Gas’ fifth amendment price, the buyer and seller reach a compromise on these points. The buyer then decides how much it is willing to invest in order to obtain Consolidated’s profits over the established period of time.73 The seller decides what amount it will accept, and the two reach another compromise. They conclude that Consolidated with a supply of gas will probably generate $250,000 in profits per year for twenty years.74 Given the nature of the investment risks, the buyer expects a ten percent annual return on its investment and is thus willing to pay approximately $2,128,500, or the present value of annual profits of $250,000 over a period of twenty years. The buyer will in effect purchase a twenty-year annuity that pays $250,000 a year. The court thus finds that Consolidated’s loss of going concern value is $1,353,500—the difference between $2,128,500 (Consolidated’s fair market value with gas) and $775,000 (Consolidated’s fair market value without gas).
Consolidated also seeks damages for the profits it lost while in operation after City Gas’ refusal to deal until the time of its demise.75 The court must determine those damages based on the profits that Consolidated would have made if it had purchased gas at City Gas’ fifth amendment price. See supra at 1323. The district court must also take into account any changes in the fifth amendment price during that period of operation.
c. The two schemes’ inconsistency.
As this discussion indicates, these two remedial schemes are theoretically inconsistent. As I explain above, Consolidated can recover damages for loss of its going concern value if, by the time of trial, it can no longer operate as a natural gas distributor. Under that measure of damages, City Gas in effect pays Consolidated the present value of the profits that Consolidated would have made in the future ($250,000 for twenty years in my hypothetical example) had City Gas sold it gas. In contrast, if, having acquired a supply of gas, Consolidated can operate in the future as a natural gas distributor, then it will seek an injunction. Thus, the permanent injunction means that Consolidated can operate as a natural gas distributor, whereas the award of damages for loss of going concern value means that Consolidated cannot operate as a natural gas distributor. If Consolidated can continue to operate, the court orders City Gas to supply wholesale gas to Consolidated. If Consolidated cannot continue to operate, however, the court does not order City Gas to supply the gas but instead awards Consolidated damages equivalent to the present value of the profits that it would have made had City Gas supplied it with gas.
Now, if the court awards Consolidated both an injunction and damages for loss of its going concern value (i.e., the present value of the profits that Consolidated would have made had City Gas supplied it *1325with gas), not only are the remedies theoretically inconsistent — the injunction implying that Consolidated can continue to operate as a natural gas distributor, and the damages implying that Consolidated cannot continue to operate as a distributor — but the award also provides Consolidated with a double recovery. That is, Consolidated recovers its future profits twice: once in the form of the damages equivalent to the present value of twenty-years worth of profits, and again in the form of the profits that it generates over the same twenty years by operating its business.76
With these remedial models, and the inconsistencies between them, in mind, I now consider what the district court actually did in the present case and provide instruction on how it would proceed assuming the case were remanded.
d. Remedies granted below and instructions for remand.
The district court below improperly applied the first remedial model. The district court awarded damages for lost profits in the amount of $83,090.15, which covered the period of time from the refusal to deal77 until the date of trial. The district court, however, did not arrive at this figure by calculating what Consolidated’s profits during that time would have been, based on the fifth amendment price, with a natural gas supply. Rather, the district court reached this figure based on the profits that City Gas made in servicing Consolidated’s former customers. The district court determined City Gas’ net profits by subtracting City Gas’ wholesale gas costs from City Gas’ gross revenues. 665 F.Supp. at 1515. As I show above, however, City Gas ’ profit picture has no relevance to what Consolidated’s profit picture would have been. See supra at 1317-1319. City Gas’ wholesale cost — the cost of gas from FGT — has no relation to the fifth amendment price that. Consolidated would have paid for wholesale gas from City Gas. The district court, therefore, improperly calculated this damages figure. On remand, I would instruct the district court instead to award damages based on a showing by Consolidated of what its profits would have been had it purchased gas from City Gas at the fifth amendment price.
The district court similarly erred in fashioning the injunction in this case. Assuming that on remand Consolidated could show the likelihood of future injury, see supra at 1321, the district court, under its new rule of liability, should grant an injunction ordering City Gas to supply wholesale gas to Consolidated. The injunction, though, may require City Gas to supply gas only at the fifth amendment price. Because, in the proceedings below, the district court never set the fifth amendment price, it could include no price term in its *1326injunction and, as a result, delegated the price-setting task to the FPSC, see 665-F.Supp. at 1512, 1534, 1545.78 On remand, the district court would have to include the fifth amendment price term in the injunction. In addition, the district court would have to retain jurisdiction over the case (which it did not do below), entertaining any motions made by City Gas to modify the injunction in light of changed circumstances or to dismiss the case if Consolidated can no longer turn a profit.
The district court also awarded damages for the loss of Consolidated’s going concern value in the amount of $1,503,975. I first consider the district court’s error in calculating this figure without having found a fifth amendment price. I then discuss the inconsistency of its decision to award these damages as well as to grant an injunction.
The court awarded damages for loss of going concern value in the amount of $1,503,975 based on testimony by Consolidated’s expert witness, Ben Ball. Consolidated called Ball to the stand to render an opinion concerning Consolidated’s present value as a company with a natural gas supply — in effect, what a hypothetical willing buyer, fully informed of all relevant facts, would pay a hypothetical willing seller, also fully informed of all relevant facts. In formulating his opinion, Ball assumed that Consolidated could indefinitely obtain a “competitively priced source of natural gas,” 665 F.Supp. at 1514, and then forecast that, with this supply of gas, Consolidated could continue to operate profitably for twenty years. He estimated the company’s annual profits for that period, including the profits that Consolidated would have made by servicing former customers that were now lost to City Gas. From those combined profits, he deducted the expenses that Consolidated would have incurred in changing from LP gas distribution to natural gas distribution and in hooking up to City Gas’ pipesystem. Ball then opined that the hypothetical buyer would expect a ten-percent return on his investment, and, using that rate, he reduced the annual net profits during the twenty years to their present value. He concluded that Consolidated’s “value today as a natural gas firm,” 665 F.Supp. at 1515, was $2,275,130.
From that figure, Ball subtracted $771,-155 — the salvage value of Consolidated’s assets without gas. Ball arrived at that figure based on the price that City Gas offered to pay Consolidated for all its assets, in the course of their negotiations regarding the wholesale gas supply. As Ball suggested, that figure represented the salvage value of Consolidated’s assets as a company without a natural gas supply — the assets that Consolidated still possessed. After subtracting this figure, Ball was left with a figure of $1,503,975, which represented “[t]he diminution of Consolidated’s value as a natural gas company as a result of City Gas’ actions,” id. at 1515, that is, the present value of the profits that Consolidated would have made over the next twenty years if City Gas had not refused to deal, less Consolidated’s present salvage value.
The district court erred, however, in relying on Ball’s profit figures to the extent that those figures were not based on the fifth amendment price. Although this error did not infect Ball’s calculation of Consolidated’s without-gas value (i.e., its salvage value), the error did infect his calculation of Consolidated’s with-gas value (i.e., the present value of a twenty-year stream of profits). Rather than basing his figures on the fifth amendment price, Ball considered a “competitively priced source of natural gas.” As I discuss in detail above, however, no “competitively priced source of gas” existed. See supra notes 60, 61, & accompanying text. Presumably, Ball *1327based his figure on FGT’s price to City Gas, which has no relevance to the fifth amendment price here. See id. In addition, rather than determining what Consolidated’s rates would have been if Consolidated bought gas at City Gas’ fifth amendment price, Ball looked to “City Gas’ rates” to its customers. As I also discuss above, City Gas’ rates were irrelevant to Consolidated’s rate structure. See id. For these reasons, Ball’s figures provided no support for the damages award; the district court, therefore, improperly relied on those figures in calculating the damages award. On remand, the district court would have to recalculate the award in a manner consistent with the fifth amendment.
As a more serious problem, this award of damages for loss of Consolidated’s going concern value is inconsistent with permanent injunctive relief and gives Consolidated a double recovery.79 As I discuss above, see supra at 1324-1325, the remedies are theoretically inconsistent: the permanent injunction implies that Consolidated can operate (for twenty years) as a profitable natural gas company whereas the damages award (in the form of the present value of a twenty-year stream of profits) implies that Consolidated can no longer operate as a natural gas company. Consolidated either can operate as a gas company or it cannot operate as such; it cannot, however, do both. Depending on which course Consolidated elects, it will be entitled either to an injunction or to damages for loss of going concern value; it cannot, however, receive both.
To award Consolidated both remedies, moreover, constitutes a double recovery. After compensating Consolidated for the loss of its going concern value by awarding it the present value of the profits that it would have made over the next twenty years as a gas company, a permanent injunction then in effect duplicates the award. The injunction gives Consolidated gas for the next twenty years and thus the profits that the damages award assumed Consolidated would have made during that time had City Gas not refused to deal.80 Armed with the injunction, Consolidated can sell its “going concern” for the same present-value price reflected in the going concern damages award. Consolidated thus earns twenty-years worth of profits twice. Under the district court’s decision, therefore, Consolidated has obviously received a double recovery. On remand, Consolidated would have to elect between remedies.
In addition, if, on remand, Consolidated were to elect damages for loss of going concern value, then the district court would have to find the date of Consolidated’s demise and begin its damages calculation as of that date. Based on that date as well, the district court would have to determine whether Consolidated should recover any damages for lost profits prior to the date of its demise. The expert on whose figures the district court relied indicated that he had calculated “Consolidated’s value today as a natural gas firm.” Id. at 1515 (emphasis added). The district court, however, did not specify which date “today” was— whether it was the date of the refusal, the date of Consolidated’s amended complaint, the date of trial, or some other interim date. That date, of course, must be the date of Consolidated’s demise — the point in time at which Consolidated’s continued operation as a natural gas company became impracticable. The district court would have to calculate the damages for loss of *1328going concern value (in the form of the present-day value of twenty-years’ worth of future profits) as of that point forward and would have to calculate the damages for lost profits from that point backward to the date of the refusal to deal.
The district court, however, never explicitly decided whether or not Consolidated’s business was ruined — let alone when its demise might have occurred. Instead, the district court implied that Consolidated both died and stayed alive. On remand, the district court would have to address this problem. First, it would have to require Consolidated to elect between these two remedial schemes. Depending on its ability to operate in the future, Consolidated could elect either injunctive relief, with damages for profits lost in the past, or it may elect damages for loss of going concern value, with damages for profits lost in the past. In either case, however, the district court’s award of damages for lost profits in the amount of $83,090.15, which covered the period of time from the refusal until the date of trial, would not be a proper award: the district court did not base the award on the fifth amendment price and never properly determined the dates of the refusal, see supra note 77, or of Consolidated’s demise.
Thus, even if today's court upholds the impossible rule fashioned by the district court, the district court has still failed to apply it properly in the present case. For that reason, I would vacate the district court’s decision and remand for further proceedings as I describe.
VII. STATE-ACTION IMMUNITY Even if City Gas could be held to have violated section 2, which in my view it has not violated, City Gas would still be immune from federal antitrust liability under the state-action immunity doctrine of Parker v. Brown, 317 U.S. 341, 63 S.Ct. 307, 87 L.Ed. 315 (1943). I first discuss the Parker doctrine’s application to private parties. I then analyze City Gas’ conduct — both its territorial agreement with Peoples Gas and its refusal to deal with Consolidated (neither of which, as I have shown, would provide grounds for liability anyway) — under this doctrine.
A. The Midcal Test.
The district court correctly stated that in a case such as the present one, when a private party claims immunity under the state-action immunity doctrine, the party must satisfy both parts of the test articulated by the Supreme Court in California Retail Liquor Dealers Ass’n v. Midcal Aluminum, Inc.:81 “First, the challenged restraint must be ‘one clearly articulated and affirmatively expressed as state policy’; second, the policy must be ‘actively supervised’ by the State itself.” 445 U.S. 97, 105, 100 S.Ct. 937, 943, 63 L.Ed.2d 233 (1980) (quoting City of Lafayette v. Louisiana Power & Light Co., 435 U.S. 389, 410, 98 S.Ct. 1123, 1135, 55 L.Ed.2d 364 (1978) (opinion of Brennan, J.)). I discuss what these two prongs require in turn.
1. The Clearly Articulated Prong.
In Southern Motor Carriers Rate Conference v. United States, 471 U.S. 48, 105 S.Ct. 1721, 85 L.Ed.2d 36 (1985), the Supreme Court considered the first prong of this test. In that case, the United States brought an antitrust challenge against the collective ratemaking practices of rate bureaus composed of motor common carriers in the southeast. Under procedures authorized, but not compelled, by the states in which the carriers operated, the rate bureaus submitted joint rate proposals to the public service commission in each state. Id. at 50, 105 S.Ct. at 1723. The Supreme Court held that Midcal’s first prong did not require the state to compel the challenged conduct: “The federal antitrust laws do not forbid the States to adopt policies that permit, but do not compel, anticompetitive conduct by regulated private parties.” Id. at 60, 105 S.Ct. at 1728 (emphasis original).
*1329The Court then considered whether the first prong would be satisfied “in the absence of a statute expressly permitting the challenged conduct.” Id. at 63,105 S.Ct. at 1730. In Southern Motor Carriers, statutes in North Carolina, Georgia, and Tennessee explicitly permitted collective rate-making; the Mississippi motor carrier statute, however, did not expressly permit such ratemaking, see Miss.Code Ann. § 77-7-1 et seq. (1972 & Supp.1984). Rather, that statute authorized the public service commission to prescribe “just and reasonable” rates for intrastate transportation but left the details of the rate-setting process to the commission itself. The Supreme Court held that the statute satisfied Midcal’s clearly articulated requirement. The Court reasoned that the statute clearly indicated the legislature’s intent that the commission, rather than the market, would set intrastate rates. The legislature left the details of that process to the commission, and the commission exercised its authority by encouraging collective ratemaking. The Court concluded,
A private party acting pursuant to an anticompetitive regulatory program need not “point to a specific, detailed legislative authorization” for its challenged conduct. Lafayette v. Louisiana Power & Light Co., 435 U.S. [389] at 415, 98 S.Ct. [1123] at 1138 (opinion of BRENNAN, J.). As long as the State as sovereign clearly intends to displace competition in a particular field with a regulatory structure, the first prong of the Midcal test is satisfied.
471 U.S. at 64, 105 S.Ct. at 1730.
The application of the first Midcal requirement to conduct by private parties under Southern Motor Carriers thus parallels the requirement’s application to conduct by municipalities under Town of Hallie v. City of Eau Claire, 471 U.S. 34, 42-44, 105 S.Ct. 1713, 1718-19, 85 L.Ed.2d 24 (1985), a case handed down the same day as Southern Motor Carriers. In Town of Hallie, a township challenged a city’s conduct in gaining a monopoly over sewage treatment and tying its provision of sewage collection services to its treatment services. The city claimed immunity under Parker based on a state statute, which granted the city authority to develop sewage systems and to “ ‘describe with reasonable particularity the district to be [served].’ ” Id. at 41, 105 S.Ct. at 1717 (quoting Wis.Stat. § 62.18(1) (1981-1982)). The Court considered whether that statute satisfied the requirement of a clearly articulated state policy. The Court applied a foreseeability test: it observed that “the statutes clearly contemplate that a city may engage in anticompetitive conduct. Such conduct is a foreseeable result of empowering the City to refuse to serve unannexed areas.” Id. at 42, 105 S.Ct. at 1718. The Court then stated that “[i]t is not necessary ... for the state legislature to have stated explicitly that it expected the City to engage in conduct that would have anticompetitive effects.” Id., 105 S.Ct. at 1718. Citing City of Lafayette v. Louisiana Power & Light Co., 435 U.S. 389, 415, 98 S.Ct. 1123, 1138, 55 L.Ed.2d 364 (1978) (opinion of Brennan, J.), the Court held that “it is sufficient that the statutes authorized the City to provide sewage services and also to determine the areas to be served.” 471 U.S. at 42,105 S.Ct. at 1718 (same passage cited in Southern Motor Carriers holding, see supra at 1329). In the Court’s view, the statutes plainly showed that “ ‘the legislature contemplated the kind of action complained of’ a showing that was “sufficient to satisfy the ‘clear articulation’ ... test.” Id. at 44, 105 S.Ct. at 1719 (quoting Lafayette, 435 U.S. at 415, 98 S.Ct. at 1138).
Both Southern Motor Carriers and Hallie thus derive from the Court’s earlier pronouncements in Lafayette and stand for the proposition that, under the first Midcal requirement, a state legislature need not expressly permit (and clearly need not compel) the conduct at issue; rather, if the legislature, by displacing competition in a given field with a broad regulatory structure, has contemplated that such conduct might occur, that will satisfy the “clearly articulated” requirement. See also Executive Town & Country Servs. v. City of Atlanta, 789 F.2d 1523, 1529 (11th Cir.1986). Of course, as the Supreme Court *1330stated in Community Communications Co. v. City of Boulder, 455 U.S. 40, 55, 102 S.Ct. 835, 843, 70 L.Ed.2d 810 (1982), the requirement “is not satisfied when the State’s position is one of mere neutrality respecting the ... actions challenged as anticompetitive.” Thus, when the Constitution of Colorado granted cities “the full right of self-government in both local and municipal matters,” id. at 43 n. 1, 102 S.Ct. at 837 n. 1, the Court held that the provision did not clearly articulate a state policy to displace competition with regulation in the cable television industry: the constitutional provision had nothing at all to do with cable television regulation. Id. at 55, 102 S.Ct. at 843; see also Bolt v. Halifax Hosp. Medical Center, 891 F.2d 810, 825 (11th Cir.) (legislature must have foreseen specific type of anticompetitive conduct alleged in case), cert. denied, — U.S. -, 110 S.Ct. 1960, 109 L.Ed.2d 322 (1990).
2. The Active Supervision Prong.
Midcal’s second requirement of active supervision, as the Court explained in Hallie, reflects the concern that “[wjhere a private party is engaging in the anticom-petitive activity, there is a real danger that he is acting to further his own interests, rather than the governmental interests of the State.” Hallie, 471 U.S. at 47, 105 S.Ct. at 1720. The active supervision requirement is thus “designed to ensure that the state action doctrine will shelter only the particular anticompetitive acts of private parties that, in the judgment of the State, actually further state regulatory policies.” Patrick v. Burget, 486 U.S. 94, 100-01, 108 S.Ct. 1658, 1663, 100 L.Ed.2d 83 (1988). The crux of active supervision, as the Court stated in Patrick, is “that the State exercise ultimate control over the challenged anticompetitive conduct.” Id., 108 S.Ct. at 1663. The state must therefore “have and exercise power to review particular anticompetitive acts of private parties and disapprove those that fail to accord with state policy.” Id., 108 S.Ct. at 1663.
Applying this formula in Patrick, the Court held that a state did not actively supervise a physician peer-review process. Although the state had mandated the peer-review process in the first place, that state lacked ultimate authority over the peer-review committee’s actual decisions: neither the state nor any of its administrative bodies reviewed specific decisions or had authority to overturn decisions not in accordance with state policy. See id. at 102-04, 108 S.Ct. at 1664.82 The Court, therefore, held that no state actor actively supervised the peer-review process and that the federal antitrust laws applied. Id. at 104-05, 108 S.Ct. at 1665; see also Shahawy v. Harrison, 875 F.2d 1529, 1535-36 (11th Cir.1989) (applying Patrick to public hospital peer-review procedure).
Similarly, in 324 Liquor Corp. v. Duffy, 479 U.S. 335, 107 S.Ct. 720, 93 L.Ed.2d 667 (1987), the Supreme Court held that New York State’s liquor-pricing scheme did not satisfy Midcal’s active supervision prong. . That scheme, which did satisfy Midcal’s clear articulation prong, required liquor wholesalers to post a wholesale bottle price and then prohibited retailers from selling at less than 112% of that posted price. In effect, though, the scheme allowed wholesalers to set retail prices. Although the retail price was based on the posted bottle price, wholesalers generally sold liquor to retailers at a lower, posted case price. In addition, the retail price was based on the current posted bottle price; the wholesalers could thus sell at a low posted price and then increase the posted price in order to allow retailers to sell at a higher price. See id. at 342-46 & n. 6, 107 S.Ct. at 725-26 & n. 6. As a result, the Court found that the state did not actually set prices, did not review the reasonableness of established prices, and did not generally monitor the *1331market or reexamine the pricing program, See id. at 345-46, 107 S.Ct. at 726. The Court therefore held that the state did not actively supervise the pricing system.83
As these two cases hold, then, the state must possess ultimate authority over the regulatory scheme at issue. The state must also exercise this authority to ensure that the scheme works as the state intended it to work. If the scheme breaks down and allows private parties to act however they want, then under 324 Liquor, Mid-cal ’s active supervision requirement would not be satisfied: the state must exercise its authority by disapproving “particular anti-competitive acts ... that fail to accord with state policy.” Patrick, 486 U.S. at 101, 108 S.Ct. at 1663. Of course, if the scheme operates effectively, automatically requiring private actors to act in accord with state policy, the state would have nothing to disapprove, and the active supervision prong would be satisfied. See 324 Liquor, 479 U.S. at 344 n. 6, 107 S.Ct. at 725 n. 6.
I now apply the Midcal test to the present case, considering in turn City Gas' territorial agreement and its refusal to deal.
B. Applying Midcal.
1. The Territorial Agreement.
As I discuss above, City Gas’ territorial agreement with Peoples Gas provided no basis for antitrust liability in the present case. Nevertheless, assuming for the sake of discussion that the agreement might provide a basis for liability, City Gas is still immune under the state action doctrine.
a. Clearly articulated state policy.
City Gas’ territorial agreement is protected pursuant to a clearly articulated state policy. That policy is expressed in Chapter 366 of the Florida Statutes, Fla.Stat. ch. 366 (1987),84 which prescribes the duties of the utility companies in Florida and charges the FPSC with administrative responsibility. Chapter 25-7 of the Florida Administrative Code contains regulations promulgated by the FPSC for gas utilities. See 1986 Fla.Admin.Code Ann. ch. 25-7 (FPSC regulations promulgated under chapter 366).
The court today adopts the district court’s conclusion that these provisions did not satisfy Midcal’s clearly articulated requirement. The district court based its conclusion on the fact that chapter 366 expressly grants the FPSC power to approve territorial agreements involving electric utilities, and that other chapters do the same with respect to telephone, water, and sewage utilities, but that the legislature failed expressly to grant this power to the FPSC with respect to natural gas utilities. In the district court’s view, that omission clearly indicates that the legislature did not intend to authorize the FPSC to approve territorial agreements between natural gas utilities. Given this clear indication, the district court was not persuaded otherwise by the Florida Supreme Court’s opinion in City Gas Co. v. Peoples Gas Sys., 182 So.2d 429, 436 (Fla.1965), which addressed the very territorial agreement at issue here and held that chapter 366 implicitly grants the FPSC authority to approve such agreements.
I disagree with the district court’s interpretation of the legislative intent behind chapter 366. That interpretation overemphasized the absence of an express grant of power to approve territorial agreements between natural gas utilities and, as a result, reached a conclusion that was inconsistent with the general structure of the regulatory scheme created by chapter 366 for the natural gas industry and with the role of the FPSC within that scheme. The scheme that the legislature has created in effect requires the FPSC to approve territorial agreements in exercising its pre*1332scribed ratemaking functions. Given the nature of this scheme, therefore, if the legislature had intended that the FPSC not approve such agreements, then it would have had to include a provision expressly prohibiting the FPSC from doing so and also explaining how the FPSC could accomplish its statutorily mandated ratemaking responsibility without that power.85 In my view, chapter 366 and the regulatory scheme that it creates clearly demonstrate the legislature’s intent “to displace competition in [the natural gas] field with a regulatory structure,” Southern Motor Carriers, 471 U.S. at 64, 105 S.Ct. at 1730. Under the Supreme Court’s decision in Southern Motor Carriers, that intent satisfies the first prong of the Midcal test. See id., 105 S.Ct. at 1730. I first review chapter 366 to determine the kind of regulatory scheme that it creates and the FPSC’s role within that scheme. I then consider the inconsistency of the district court’s interpretation with this scheme.
In chapter 366, the state empowers, as “an exercise of the police power of the state,” the FPSC to regulate public utilities. Fla.Stat. § 366.01. Section 366.03 addresses the general duties of public utilities: “Each public utility shall furnish to each person applying therefore reasonably sufficient, adequate, and efficient service upon terms as required by the commission.” Section 366.04 then describes the duties of the FPSC: “the commission shall have jurisdiction to regulate and supervise each public utility with respect to its rates and service and the issuance and sale of its securities.” Thus, section 366.05 provides that the FPSC “shall have power to prescribe fair and reasonable rates and charges ... to be observed by each public utility.” The FPSC fixes these rates based on the factors described in section 366.041:
In fixing the just, reasonable, and compensatory rates, charges, fares, tolls, or rentals to be observed and charged for service within the state by any and all public utilities under its jurisdiction, the commission is authorized to give consideration, among other things, to the efficiency, sufficiency, and adequacy of the facilities provided and the services rendered; the cost of providing such service and the value of such service to the public; the ability of the utility to improve such service and facilities; and energy conservation and the efficient use of alternative energy resources; provided that no public utility shall be denied a reasonable rate of return upon its rate base in any order entered pursuant to such proceedings.
The FPSC further addresses the rate-fixing process in its regulations under chapter 366. Section 25-7.033(2) of the regulations requires each utility to “file with the Commission tariffs containing schedules for all rates and charges.” Section 25-7,033(3) then states that no “schedules of rates or charges, or modifications or revisions of the same, shall be effective until filed with and approved by the Commission as provided by law.” In addition, the FPSC has established procedures to monitor existing rates to ensure that they are fair and just and provide a reasonable rate of return as specified in section 366.041 of the Florida Public Utilities statute. Sections 366.07 through 366.072 also provide procedures for rate adjustments.
In addition to its rate-setting function, the FPSC’s powers reach the utilities’ general assets. As section 366.05 provides: “the commission shall have power to ... require repairs, improvements, additions, and extensions to the plant and equipment of any public utility when reasonably necessary to promote the convenience and welfare of the public and secure adequate service or facilities for those reasonably entitled thereto.” (Emphasis added.) Under section 366.04, the FPSC also has jurisdiction “to regulate and supervise each public utility with respect to,... the issuance and sale of its securities.”
*1333Chapter 366 thus establishes a full-scale regulatory structure and grants the FPSC broad regulatory and supervisory powers within that structure. Not only does the FPSC fix utilities’ rates — a process that engages the FPSC in an in-depth and continuous review of virtually every aspect of the utilities’ operation — but the FPSC also has direct supervisory power with respect to the utilities’ use of their own assets. The FPSC has jurisdiction to regulate utilities’ stock issuances, to require repairs to the utilities’ facilities, and also to mandate additions and extensions of those facilities — undertakings that would substantially affect utilities’ assets.
True, the statute does not explicitly grant the FPSC power to approve territorial agreements and resolve territorial disputes between natural gas companies, whereas the statute does grant such powers with respect to electric utilities. In light of the powers the statute does expressly grant to the FPSC, however, and the general regulatory structure that the statute creates, I find this omission to be insignificant. The Florida legislature clearly intended to displace competition in the natural gas industry with a system of regulation. As my review of this regulatory structure shows, moreover, the nature of a utility’s operating territory — its size, the density of consumers, the degree of competition with other companies — is a crucial factor in the rate-fixing process. In order to provide the utility with a “reasonable rate of return” — the standard applied under the statute — the FPSC will obviously consider the utility’s distribution area and the degree of competition in that area. The size of a utility’s operating area will affect its cost structure: a larger area might increase the utility’s fixed and operating costs; it might also require a greater investment in equipment and in a gas supply; in turn, however, the larger area might enable the utility to sell a larger volume of gas, thus generating higher gross revenues and perhaps more operating capital. As I discuss in detail in part VI, the degree of competition with other gas companies will also significantly affect the utility’s cost structure: a company must invest resources in order to compete, an investment that increases the utility’s costs. That investment in competition, moreover, might affect the general structure of the utility’s assets. If the FPSC does not allow the utility to recover those costs through a rate increase, then the utility must cover them from its own assets, which might in turn force the utility to cut corners on safety or on research and development of alternative energy sources or which might simply force the utility out of business — concerns that obviously fall directly within the FPSC’s jurisdiction.
As this discussion indicates, the utility’s operating territory is of central concern to the FPSC’s regulatory function and undeniably falls within the FPSC’s purview: the nature of the utility’s territory directly influences the rates set for consumers. To think, therefore, that the Florida legislature did not intend the FPSC to approve territorial agreements — whether or not the legislature explicitly authorized such approval — -is utterly inconsistent with the overall structure of the regulatory scheme that the Florida legislature did impose and the powers that it expressly granted to the FPSC. Given the nature of this scheme as actually created, had the legislature intended the FPSC not to approve such agreements, the legislature would have had to add a provision expressly prohibiting the FPSC from doing so and also explaining how the FPSC could accomplish its statutorily mandated ratemaking functions without that power. Under Southern Motor Carriers, this comprehensive regulatory scheme satisfies the first prong of the Midcal test.86
b. Active supervision prong.
As my review of chapter 366’s regulatory scheme also demonstrates, the FPSC *1334exercises “ultimate control” over the territorial agreement. The FPSC “ha[s] and exercise[s] power to review [territorial agreements between natural gas utilities] and disapprove those that fail to accord with state policy.” Patrick, 486 U.S. at 101, 108 S.Ct. at 1663. Under Patrick, therefore, Midcal’s second prong is satisfied as well.
The panel, in holding that the active supervision requirement was not satisfied, pointed to language in a decision by the FPSC and characterized it as an “expression of doubt ... as to its authority to establish exclusive territories or to resolve territorial disputes between natural gas utilities.” 880 F.2d at 303 (citing In re: Tariff filed by Miller Gas, 85 FPSC 10:205, at 5-6 (Oct. 18, 1985)).87 The panel then noted that the FPSC approved the specific territorial agreement at issue here, but, in a confusing passage, the panel seemed to conclude that, “[sjimply because the Commission has on occasion, some 15 to 25 years ago, looked at this agreement does not automatically immunize it from federal antitrust scrutiny.” Id. at 303. That isolated review, in the panel’s opinion, was not sufficient. The panel also summarily rejected the argument that the Florida Supreme Court’s review of the agreement in Peoples Gas satisfied the active supervision prong.
The panel’s approach, in my view, is improper in that it failed to consider the regulatory and supervisory functions that the FPSC exercises in its continuous rate-fixing process. That process, I suggest, satisfies the active supervision requirement. As I describe above, the FPSC continually takes account of existing territorial allocations in the course of setting rates. The FPSC sets rates that will provide the utility with a reasonable rate of return given existing territorial allocations. To the extent that the FPSC sets rates that allow the utility to continue to operate (that is, operate at a profit), the FPSC necessarily approves the existing territorial allocation. The FPSC, moreover, continually reviews the utility’s rates: under § 25-7.024 of the FPSC’s regulations, the gas utility must file a rate of return report for every month, based on which the FPSC continually monitors the utility’s rates and, necessarily, its territorial allocation. In addition, if the FPSC disapproves of such an allocation, it has ultimate authority to mandate the utility to alter its distribution area “when reasonably necessary to promote the convenience and welfare of the public and secure adequate service or facilities for those reasonably entitled thereto.” Fla. Stat. § 366.05(1). In choosing not to exercise that authority— and it has chosen not to do so in the present case — the FPSC thus implicitly approves the existing territorial allocation.
This is not a case, moreover, like 324 Liquor where the state establishes a regulatory scheme that merely enforces private arrangements. Rather, the FPSC assumes a very active role under this regulatory scheme, initially approving the utility’s proposed rates, and, in doing so, undertaking an investigation and consideration of a variety of factors as mandated in chapter 366. See id. §§ 366.041, .06. Nor is this a case like Patrick, where the state supervises a peer-review committee’s general procedures but does not review the committee’s specific, substantive decisions. Here, the FPSC does more than prescribe general procedures for filing rates; it actually con*1335ducts the rate-fixing process itself, setting initial rates and continually monitoring them.
In this ease, furthermore, the Florida Supreme Court has directly reviewed the territorial agreement at issue. Although the Supreme Court has not yet had occasion to decide the question whether such review satisfies the active supervision requirement, see Patrick, 486 U.S. at 102-05, 108 S.Ct. at 1664-65 (state courts conducted insufficient review of specific committee decisions, thus not necessary to decide broad question whether such review could ever satisfy active supervision requirement); see also Pinhas v. Summit Health, Ltd., 894 F.2d 1024, 1030 (9th Cir.1989) (same), cert. granted, — U.S. -, 110 S.Ct. 3212, 110 L.Ed.2d 660 (1990), the Florida courts' review of the agreement here is, in my view, sufficient to satisfy the active supervision requirement. In Peoples Gas, the Florida Supreme Court carefully considered City Gas’ territorial agreement with Peoples Gas and approved that agreement under chapter 366. See 182 So.2d at 436. Unlike the judicial review conducted by the courts in Patrick, see supra note 82, which merely considered the procedures through which the peer-review committee’s decision was reached but not the decision’s merits, the Florida Supreme Court specifically reviewed the territorial agreement itself and approved it as consistent with state policy. That review, in my view, constituted active supervision.88
Thus, based on both the Florida Supreme Court’s decision in Peoples Gas and on the FPSC’s supervisory powers in the rate-fixing process, the active supervision prong is satisfied. Under Midcal, therefore, City Gas is immune from antitrust liability on the basis of its territorial agreement with Peoples Gas.
2. The Refusal to Deal.
I reach the same conclusion with respect to City Gas’ refusal to sell wholesale gas to Consolidated,89 I apply the two prongs of the Midcal test in turn.
a. Clearly articulated state policy.
Section 366.03 states: “No public utility shall be required to furnish electricity or gas for resale .... ” This provision explicitly and affirmatively expresses a state policy not to require utilities to sell at wholesale. Indeed, I could think of no more explicit an expression of this policy. The policy, of course, does not compel gas utilities not to sell at wholesale; it merely permits them to refrain from selling at wholesale. Under Southern Motor Carriers, though, Midcal’s clearly articulated prong does not require the state to compel the anticompetitive conduct; the state need only permit such conduct, as it does in section 366.03. It is beyond dispute, in my view, that this statutory provision satisfies the first prong of the Midcal test.90
*1336 b. Active supervision.
The active supervision prong of the Mid-cal test is also satisfied here. Two reasons support this conclusion. First, by virtue of its rate-setting function, the FPSC regulates a utility’s decision to sell at wholesale. If a utility were to decide to sell at wholesale, the FPSC would exercise its ratemak-ing function with respect to the wholesale price. The utility would therefore base its decision whether or not to sell at wholesale, at least in part, on the likelihood that the FPSC would approve wholesale rates that were sufficiently high to allow the utility to recover a reasonable return on its investment. As I describe above, the economic implications of selling at wholesale may be quite broad — affecting the structure of competition in the marketplace and thus forcing the utility to reconsider the deployment of its own resources. See supra at 1333. The FPSC’s rate-fixing process, in the utility’s view, might not adequately take into account all of these implications. Or, at least, the utility might be uncertain whether or not the FPSC’s rate-fixing calculation would adequately consider these implications. Unless the FPSC were to promulgate regulations expressly dealing with rate calculations for wholesale sales, which it has not, the FPSC would continue to discourage wholesale sales through its rate-fixing process. The FPSC’s ratemak-ing function, therefore, by virtue of its effects on a utility’s decision whether or not to sell gas at wholesale, satisfies the active supervision requirement.
Even if the FPSC’s ratemaking function did not so affect the utility's decision whether or not to sell gas at wholesale, the active supervision prong would still be satisfied here. As the Court explained in Patrick, the purpose of the active supervision requirement is “to ensure that the state action doctrine will shelter only the particular anticompetitive acts of private parties that, in the judgment of the state, actually further state regulatory policies.” 486 U.S. at 100-01, 108 S.Ct. at 1663. Because the state policy concerning sales of gas at wholesale is to leave such decisions entirely te the utility’s discretion, the utility would necessarily be acting in accordance with the state policy whether or not it decided to sell at wholesale. Either decision would be consistent with state policy, and neither decision could be inconsistent with that policy. The state, therefore, would never have to exercise its power to disapprove a decision that failed to accord with state policy: no decision could so fail. In this situation, the concerns behind the active supervision requirement are automatically satisfied regardless of how the state supervises the decisionmaking process. See P. Areeda & H. Hovenkamp, Antitrust Law ¶ 212.7, at 168-69 (Supp.1989). For this reason as well, the second prong of the Midcal test is satisfied in this case.
C. Conclusion.
As the foregoing discussion shows, even if City Gas’ conduct could justify the imposition of liability under section 2 of the Sherman Act, City Gas is immune from liability under the state-action immunity doctrine of Parker.
VIII. CONCLUSION
This case’s history has been complicated and long. The case is extremely important, however, in that it involves issues of first impression and tests the limits of the antitrust laws. In my view, the district court decided the case improperly, imposing liability on City Gas under section 2 of the Sherman Act. Because this court today upholds the district court's decision, I am compelled to dissent.
As a preliminary issue, City Gas’ territorial agreement with Peoples Gas provided no basis for liability here. That agreement had no causal relationship to Consolidated’s injury. Consolidated, therefore, has no standing under the Clayton Act, which authorizes private actions under the antitrust laws, to challenge the territorial agreement.
Nor did City Gas’ decision not to sell wholesale natural gas to Consolidated constitute a refusal to deal in violation of *1337section 2. As an initial problem, City Gas did not possess monopoly power in the relevant market. First, the relevant market defined by the district court did not exist. Second, even if that market did exist, City Gas never operated on the wholesale level of the natural gas market; it operated only as a retail distributor of natural gas. Moreover, City Gas’ abandoned and unperformed contract with Florida Gas does not suggest otherwise. Third, even if City Gas had operated as a natural gas wholesaler, it would not have possessed monopoly power in that market. FGT had a tremendous share of that market, and Consolidated, given its projections of the profits it could make as a natural gas company, could practicably have turned to FGT for a wholesale supply of natural gas.
Given the fact that City Gas has never operated as a natural gas wholesaler, no basis existed to require City Gas to deal with Consolidated. Under established law, a defendant is liable under section 2 when the circumstances surrounding its refusal to deal give rise to an inference of anticom-petitive intent. Courts have developed various formulas — such as the essential facilities doctrine, the monopoly-leveraging theory, and the intent test — for evaluating the circumstantial evidence in a given case in order to determine whether the evidence justifies an inference of intent. The circumstances here, however, give rise to no inference of anticompetitive intent under any of these formulas. Nor does the existing caselaw under section 2 justify the imposition of liability on City Gas. The practical implications of imposing a duty to deal in the present category of case also support the conclusion that section 2 does not apply. In short, the present case simply did not involve the type of refusal to deal that violates section 2.
In nevertheless imposing a duty to deal under section 2 in the present case, the district court thus created a new rule of antitrust liability, which this court today / upholds. Under this new rule, City Gas — a company that has done nothing wrong under established law — must undertake an entirely new level of operation and submit to a forced sale of gas to Consolidated. City Gas must do so, moreover, not because that undertaking will benefit consumers or enhance competition in general — it will do neither — but because the undertaking will enable Consolidated to enter the natural gas market as a retail distributor on the most favorable terms possible. That is, this new rule requires City Gas to give its potential competitors a free ride into the natural gas market.
This new rule is, in my view, extraordinarily bad law. It promotes inefficiencies, undermines business firms’ rational decisions, and leads to results that are flatly inconsistent with the purposes behind the antitrust laws. In addition, this new rule will be impossible to apply. The district courts of this circuit will face an administrative nightmare in attempting to fashion remedies under this rule. Given these practical implications, Congress surely could not have intended section 2 to support such a rule. This new rule should therefore be struck down.
Even under this impossible rule, however, which this court today upholds, the district court’s decision was erroneous. The district court failed properly to apply its own new rule. For that reason, I would vacate the decision and remand the case for further proceedings. Because the rule implicates the fifth amendment, a proper application of the rule depends on a determination of the fifth amendment price for the forced sale of gas here. The district court, however, did not find the fifth amendment price in the present case. As a result, it did not properly determine (1) whether, given the fifth amendment price, City Gas refused to deal in violation of section 2; and (2) whether, for, purposes of the Clayton Act, Consolidated suffered any injury in fact as a result of City Gas’ conduct.
In addition, the district court’s failure to set a fifth amendment price infected the remedies it awarded Consolidated. First, the district court did not properly calculate its damages awards: those damages must be based on the fifth amendment price. Second, the district court was unable to set a price in its injunction requiring City Gas to sell gas to Consolidated; instead, it del*1338egated the price-setting task to the FPSC. Third, the district court erred in awarding Consolidated both an injunction as well as damages for the loss of its going concern value. Not only are those two remedies theoretically inconsistent, but by awarding Consolidated both remedies, the district court gave Consolidated a double recovery.
For these reasons, even under this impossible new rule, the district court’s decision must be vacated, and the ease must be remanded so that the district court can properly address these issues.
Finally, even if the district court could properly determine under its new rule that City Gas’ conduct violated section 2, in my view, City Gas is still immune from liability under the state-action doctrine of Parker. City Gas’ conduct — both in entering into a territorial agreement with Peoples Gas and in refusing to sell gas to Consolidated— was consistent with a clearly articulated and affirmatively expressed state policy that was actively supervised by the state.
Based on the foregoing discussion, I respectfully, but strongly, dissent from this court’s decision today.
. These sections of the Clayton Act authorize private actions under the antitrust laws. Section 4 authorizes private actions for treble damages and provides in pertinent part: “any person who shall be injured in his business or property by reason of anything forbidden in the antitrust laws may sue therefor....” 15 U.S.C. § 15(a). Section 16 authorizes private actions for injunctive relief, and provides in pertinent part: "Any person, firm, corporation, or association shall be entitled to sue for and have injunc-tive relief, in any court of the United States having jurisdiction over the parties, against threatened loss or damage by a violation of the antitrust laws_” Id. § 26.
. The panel, and this court today, have adopted the district court’s findings of facts and conclusions of law. For the sake of convenience, therefore, I refer throughout this dissent to the district court’s decision.
. The district court made much of the fact that FGT assumed the lateral pipe costs. Given the uncertain state of the natural gas market in the 1960s, however, see infra at 1269, that fact is of little significance. Because the natural gas market was limited and its future uncertain, City Gas still took a large degree of initiative and assumed substantial risk in devoting its resources to natural gas distribution.
. The statute has been amended. See 1989 Fla. Laws ch. 89-292 (codified as amended, Fla.Stat. ch. 366 (1989)). See infra note 86, for a discussion of the amendment and its relevance to the present case.
. The district court put an unwarranted spin on its characterization of these facts. Adopting a good-guys/bad-guys motif, the district court cast Consolidated as the good guys and City Gas as the bad guys. Everything that City Gas did was bad — part of large-scale and well-planned effort to drive out all competition from the natural gas industry and thereby grease the wheels of its large, rumbling monopoly. As the good guys, Consolidated naturally suffered, finding itself the weak and unwitting victim of the monstrous monopolist.
In my view, this characterization was unjustified. Viewed objectively, City Gas’ conduct was that of a rational business actor. It adopted an aggressive marketing campaign to increase its customer base — activity that epitomizes competition but that the district court labeled anticom-petitive because the conduct was a competitive success. Consider the district court’s characterization of City Gas’ marketing efforts. The district court noted that, shortly before undertaking this marketing and expansion effort, City Gas’ Board of Directors met and considered acquiring Consolidated — which was then simply *1270a struggling LP gas company — a consideration that hardly constitutes anticompetitive conduct. The district court then stated:
City Gas’ solicitations and pipeline extensions were not simply ordinary business expansions, but rather, were basic components of City Gas' strategy to force Consolidated to sell out. This conclusion is further supported, we think, by the fact that City Gas had to duplicate a portion of Consolidated’s pipeline in order to serve the eighteen Consolidated customers it had recently signed. Yet if City Gas had expected to reach a purchase agreement with Consolidated, it would have acquired Consolidated’s pipelines and would not have had to incur this expense. City Gas’ action of duplicating these pipelines before notifying Consolidated of its interest in acquiring the company seems to have been designed to force Consolidated to sell at a low price or face the resultant certainty of losing its customers and their value as quickly as City Gas could serve them.
This characterization of these events is, in my view, absurd. As the Supreme Court stated in United States v. Grinnell Corp., 384 U.S. 563, 570-71, 86 S.Ct. 1698, 1704, 16 L.Ed.2d 778 (1966), what violates section 2 of the Sherman Act is “the willful acquisition or maintenance of [monopoly] power as distinguished from growth or development as a consequence of a superior product, business acumen, or historic accident." (Emphasis added.) City Gas’ successful expansion effort was clearly the latter: the very type of growth or development that defines competition. If the courts were to characterize as anti-competitive conduct every effort that businesses undertook to expand their business through advertising and building facilities, we would soon find ourselves in a competitive crisis. And surely the district court did not mean to argue that because City Gas legitimately intended to compete for Consolidated’s business — which would in turn reduce the value of Consolidated’s assets — City Gas’ conduct suddenly became anti-competitive. I can think of no legitimately competitive conduct that, when successful, does not adversely affect a less-successful competitor’s assets: when one cola company captures a higher percentage of the market, another cola company’s market share slips.
Nor was the district court justified in characterizing Consolidated as a victim because, in attempting to enter the natural gas market as a retail distributor, it had to face City Gas as a competitor. Consider these passages from the district court's opinion:
In deciding to convert from LP gas to natural gas in 1982, Consolidated suddenly found itself in potential competition with City Gas, which had grown to be a giant in the south Florida area.
... Consequently, Consolidated was forced to compete with City Gas in order to retain its customers or else see them go to City Gas for more economical gas service.
... Thus, it became crucial that Consolidated convert to natural gas for as low a cost as possible in order to be able to offer rates that were competitive with those City Gas could offer.
665 F.Supp. at 1505-06 (emphasis added). City Gas had the foresight and took the initiative to enter the natural gas business at an early stage. By 1982, as the result of "growth [and] development as a consequence of a superior product, business acumen, [and] historic accident,” Grinnell, 384 U.S. at 570-71, 86 S.Ct. at 1704, City Gas found itself in a highly competitive position. For that. City Gas need not feel guilty. And for having failed to jump that train itself, Consolidated can claim no sympathy.
By holding that under these circumstances Consolidated can invoke the antitrust laws to require City Gas to underwrite Consolidated’s entry into the natural gas market, the district court, affirmed by the court today, embarked on an entirely unprecedented course. Before embarking, however, we should be entirely clear that under established law, City Gas has done nothing wrong.
. This expansion effort triggered a lawsuit by Consolidated, seeking an injunction to prevent City Gas from laying pipe in certain utility easements over which Consolidated claimed an exclusive right. The State Circuit Court granted Consolidated a preliminary injunction but then dissolved it, and the State Court of Appeal affirmed the dissolution, holding that Consolidated had no exclusive right to the easement.
. The district court again applied its good-guys/bad-guys characterization to City Gas' intervention into the FERC proceedings. Although the district court held that City Gas’ intervention did not itself constitute a violation of section 2, the court suggested that the action was evidence of City Gas’ bad motives. This was so for two reasons: "first, Consolidated was severely damaged by the delay which occurred as a direct result of City Gas’ intervention; and, second, City Gas’ reason for opposing Consolidated’s application was to protect its own domain.” 665 F.Supp. at 1542. Neither the district court, nor any one else, ever indicated that City Gas had committed an abuse of process by intervening into the FERC proceeding. On the contrary, City Gas, along with some other intervening natural gas suppliers, had legitimate concerns relating to Consolidated’s requested allocation. In my view, a rational business actor has a strong interest in “protect[ing] its own domain.” Id. Thus, City Gas did nothing wrong in intervening in the proceedings. If Consolidated’s request for an allocation took longer than Consolidated expected, that may have harmed Consolidated but the fault is certainly not attributable to City Gas, which merely exercised its right to petition in a legitimate way. Perhaps the problem was that Consolidated’s expectations, which the district court seems to have adopted, were entirely unreasonable.
. As the district court noted, shortly before undertaking this marketing and expansion effort, City Gas’ Board of Directors had met and passed a resolution to acquire "the assets of Consolidated Gas Company and take whatever steps with respect thereto, as they deem appropriate.” 665 F.Supp. at 1507. Based on these facts, the district court concluded
that City Gas' solicitations and pipeline extensions were not simply ordinary business expansions, but rather, were basic components of City Gas’ strategy to force Consolidated to sell out. This conclusion is further supported, we think, by the fact that City Gas had to duplicate a portion of Consolidated’s pipeline in order to serve the eighteen Consolidated customers it had recently signed. Yet if City Gas had expected to reach a purchase agreement with Consolidated, it would have acquired Consolidated’s pipelines and would not have had to incur this expense. City Gas' action of duplicating these pipelines before notifying Consolidated of its interest in acquiring the company seems to have been designed to force Consolidated to sell at a low price or face the resultant certainty of losing its customers and their value as quickly as City Gas could serve them.
Id. at 1507-08.
Faced with this situation, Consolidated began its effort to convert to natural gas. As the district court stated: "In deciding to convert from LP gas to natural gas in 1982, Consolidated suddenly found itself in potential competition with City Gas, which had grown to be a giant in the south Florida area.” Id. at 1505.
. "Therm," as defined in the FPSC’s regulations under chapter 366, "denote[s] a unit of heating value equivalent to one hundred thousand (100,-000) British thermal units.” 1986 Fla.Admin. Code Ann. 25-7.003(15). A British thermal unit is, of course, “the quantity of heat required to raise the temperature of one pound of water one degree Fahrenheit.” Id. at 25-7.003(14).
. Section 16 of the Clayton Act also provides that "[a]ny person, firm, corporation, or association shall be entitled to sue for and have injunc-tive relief ... against threatened loss or damage by a violation of the antitrust laws.” 15 U.S.C. § 26. As I discuss below, see infra at 1272-1275, courts have interpreted section 16 as imposing a causation requirement that is similar to section 4’s. For purposes of convenience, I refer only to the requirement under section 4, but the discussion also applies to Consolidated's request for injunctive relief under section 16 to the extent that its request is based on the territorial agreement.
. The Government brings civil enforcement actions under section 4 of the Sherman Act, 15 U.S.C. § 4, which provides, in pertinent part:
The several district courts of the United States are invested with jurisdiction to prevent and restrain violations of sections 1 to 7 of the title; and it shall be the duty of the several United States attorneys, in their respective districts, under the direction of the Attorney General, to institute proceedings in equity to prevent and restrain such violations. ...
. In the Brunswick case, for example, the plaintiff had already proved that its loss "occurred ‘by reason of the unlawful [conduct],” or as a material result of that conduct, but the Court held that the plaintiff had not proved that the loss occurred ‘"by reason of that which made the [conduct] unlawful.” Brunswick, 429 U.S. at 488, 97 S.Ct. at 697. In short, not every injury materially caused by conduct proscribed by the antitrust laws will justify a recovery under those laws.
. As I discuss below, the district court was also off target in attempting to fit this case into the "willful acquisition” prong of the Grinnell standard. See infra part V.
. Consolidated raised only a section 2 claim, charging in its amended complaint that City Gas’ conduct constituted an attempt to monopolize and monopolization under section 2 of the Sherman Act. Consolidated has not presented a claim under section 1 of the Sherman Act, 15 U.S.C. § 1, attacking City Gas’ conduct as part of a conspiracy in restraint of trade. The territorial agreement might have been relevant to such a claim.
. Consolidated claimed that this agreement was a horizontal market division. As the Court explained in United States v. Topco Assocs., 405 U.S. 596, 608, 92 S.Ct. 1126, 1133, 31 L.Ed.2d 515 (1972), a horizontal market division is an agreement between competitors at the same level of the market structure to allocate territories in order to minimize competition. Such concerted action is usually termed a "horizontal" restraint, in contradistinction to combinations of persons at different levels of the market structure, e.g., manufacturers and distributors, which are termed "vertical” restraints.
.In this respect, section 4’s causation requirement parallels the elements of liability under section 2. As courts have stated, monopoly power alone provides no basis for liability under section 2; rather, the improper exercise of that power justifies the imposition of liability. *1275See, e.g., Byars v. Bluff City News Co., 609 F.2d 843, 853 (6th Cir.1979); Berkey Photo, Inc. v. Eastman Kodak Co., 603 F.2d 263, 273 (2d Cir.1979) (citing Standard Oil Co. v. United States, 221 U.S. 1, 62, 31 S.Ct. 502, 516, 55 L.Ed. 619 (1911)), cert. denied, 444 U.S. 1093, 100 S.Ct. 1061, 62 L.Ed.2d 783 (1980); see infra at 1284.
. As the passage from Grinnell cited by the district court indicates, however, the restrictive agreement in that case was only one of the exclusionary practices supporting the government’s case. See Grinnell, 384 U.S. at 576, 86 S.Ct. at 1706-07. Whether that agreement, standing alone, would have provided an independent basis for liability even in Grinnell remains uncertain.
. Even if Consolidated could satisfy the injury requirement with respect to the territorial agreement, I doubt that the agreement would provide a basis for liability under section 2 of the Sherman Act. As I note above, see supra note 15, Consolidated properly characterized the agreement as a "horizontal market division.” A horizontal market division can, of course, provide the basis for liability under section 1 of the Sherman Act. See, e.g., Topco, 405 U.S. at 596, 92 S.Ct. at 1126. I have found no support for the proposition that a horizontal market division could also provide a basis for liability under section 2. The district court here did not address this point; it merely stated that "a territorial agreement, even between regulated utilities, is a per se violation of the Sherman Act” and concluded that City Gas "has violated § 2 of the Sherman Act.” 665 F.Supp. at 1532. In support of this proposition, however, the district court cited three cases in which territorial agreements were held to violate section 1, and not section 2, of the Sherman Act. These section 1 cases, in my view, do not control in this section 2 case. Thus, the district court’s holding that the territorial agreement provided a basis for liability under section 2 — even if Consolidated could satisfy the Clayton Act’s causation requirement — lacks supporting authority.
. These definitions contain a resonance of what a market historically was: a physical location where merchants sold their goods. The definitions, however, use terms of location that are at once more modern and more abstract, referring to “any context" or "any area.”
. This description would apply as well to the historical market — the physical location where merchants hawk their goods. The mere existence of such a physical market did not guarantee a "market” in the sense of a mechanism for negotiating price. The physical market meant that parties to a potential transaction had a place where they could conduct their negotiations, but it did not necessarily provide the parties with a history of transactions in the context of which they could negotiate a price.
For example, imagine strolling through the agora on a classically beautiful day in ancient Athens. Merchants are selling olives, retsina, and goats. Suddenly, a mysterious looking figure appears in the center of the agora with a creature resembling a goat, but golden in color and with a single, spiral horn rising from the top of its head. The figure announces that this creature is for sale, and a crowd begins to gather. A goatherd, whose interest has been piqued, offers to buy the beast for five drachmas, the going rate for a ram. The mysterious figure laughs and instead offers to sell the unicorn in exchange for the goatherd's entire herd. The goatherd, along with the rest of the crowd, looks at the figure in surprise. Clearly, they have never tried to buy a unicorn before. The figure scans the crowd, but no one attempts another offer. With a sudden wave of his arm, the figure opens a path through the crowd and departs, as quickly as he came.
The historical market, thus, provides a stage for this scene. The history of transactions incorporated in that market, however, are useless as a basis for arriving at a price for a unicorn.
. That criterion must be satisfied before the FERC can order an allocation under section 7(a) of the Natural Gas Act, 15 U.S.C. § 717f(a) (1988), which requires that the allocation be "necessary or desirable in the public interest” standard.
. Ball based his "competitively priced source of gas” figure on the price City Gas paid for wholesale gas from FGT. As I discuss below, see infra notes 60-61 & accompanying text, that figure provided an improper basis for calculating Consolidated’s future profits if Consolidated purchased gas from City Gas. These figures, on the other hand, would provide a proper basis for calculating Consolidated’s future profits if Consolidated purchased gas from FGT, which is the focus of my discussion here.
. The source of these and other present-value figures in this opinion is L. Gitman, Principles of Managerial Finance app. A, tbl. A-4 (3d ed. 1982).
. Monopoly leveraging also refers to a monopolist’s conduct in attempting to integrate vertically, by using its established monopoly power (on a given level of the market) to leverage its way into another level of the market.
. But see Aspen Highlands Skiing Corp. v. Aspen Skiing Co., 738 F.2d 1509 (10th Cir.1984), aff’d, 472 U.S. at 585, 105 S.Ct. at 2847. The Tenth Circuit there applied the essential facilities doctrine to a refusal to deal that did not involve monopoly leveraging. See infra at 1304-1306, for a discussion of the case. In affirming the Tenth Circuit, however, the Supreme Court did not analyze the case under the essential facilities doctrine. In my view, this case is not properly classified as an essential facilities case.
. As Areeda and Hovenkamp note, the essential facilities doctrine can be mechanically applied to a broad range of situations: "the 'essential facility’ is just an epithet describing the monopolist’s situation: he possesses something the plaintiff wants. It is not an independent tool of analysis but only a label — a label that beguiles some commentators and courts into pronouncing a duty to deal without analyzing the implications...’’ P. Areeda & H. Hovenkamp, Antitrust Law, ¶ 736.1a, at 701 (Supp.1989).
. To summarize briefly the terminology, refusal to deal cases under section 2 involve two types of conduct, either monopoly leveraging or non-monopoly leveraging. Courts have developed a variety of formulas for analyzing the circumstantial evidence in these refusal-to-deal cases in order to determine whether the evidence gives rise to an inference of anticompeti-tive intent. These formulas include the essential facilities doctrine, the intent test, and the monopoly-leveraging theory.
. I have discussed in detail above the irrelevance of City Gas abandoned contract with Florida Gas. See supra at 1282-1283.
. In Bonner v. City of Prichard, 661 F.2d 1206, 1209 (11th Cir.1981) (en banc), this court adopted as binding precedent all decisions of the former Fifth Circuit handed down prior to October 1, 1981.
. As I discuss below, see infra part VI, the present case also involves difficulties in fashioning a remedy that do not arise in these monopoly-leveraging cases. In these cases, the defendants were already engaged in providing the service at issue: Kodak sold photographic supplies at wholesale; National Screen sold movie advertising accessories at wholesale. Here, City Gas never sold gas at wholesale to anyone. In fashioning a remedy, therefore, the district court could look to no price established by City Gas for the transaction at issue. The district court instead had to set its own price (a task it delegated to the FPSC). That task presents a set of complications that I explore below. Regarding these problems, however, the monopoly-leveraging, cases provide no guidance.
. As a more general problem, the district court improperly identified the facility at issue here. According to the district court’s description, the facility appears to be a facility (the pipesystem) and the resource contained in that facility (the natural gas for resale). The district court, however, also referred to "City Gas’ lateral pipe” as the essential facility. This lateral pipe connected City Gas' pipesystem to FGT’s pipeline and enabled City Gas to pump a supply of natural gas into its pipesystem. Which of these three elements — the gas, the lateral pipe, or the pipe-system itself — constituted the essential facility? Clearly, all three components, together, formed the essential facility here. Consolidated did not want access just to City Gas’ lateral pipe. That pipe was located a long distance from Consolidated’s pipesystem. Consolidated, however, needed more than just gas: it needed access to City Gas' entire pipesystem in order to obtain the gas. Of course, Consolidated wanted access to that pipesystem only insofar as it contained gas, thus making the lateral pipe (through which City Gas acquired its gas) an indispensable component of the facility. In short, Consolidated wanted to hook into City Gas’ pipesys-tem, at a point near Consolidated’s own pipesys-tem, in order to take gas from City Gas’ pipesys-tem (gas from FGT’s pipeline that entered City Gas’ pipesystem through City Gas’ lateral pipe) in order to supply its own pipesystem. In effect, therefore, Consolidated wanted access to City Gas’ entire distribution operation.
This case is not about a piece of City Gas’ system, but rather about that system as a whole. Consolidated claimed that in order to compete with City Gas, it had to have access to City Gas’ system. Consolidated needed gas from FGT, and, according to Consolidated, City Gas — with its lateral pipe and pipesystem — stood between Consolidated and FGT. Thus, Consolidated either had to go around City Gas’ system or through it. Consolidated argued that going around that system would be too expensive to allow Consolidated to compete (in effect, what the second element of the essential facilities doctrine requires). It therefore claimed that access to the facility — i.e., City Gas’ system as a whole — was "essential." Consolidated, therefore, wanted to be a customer of City Gas’, but with an important difference from the rest of City Gas’ customers: Consolidated wanted City Gas to sell this gas at a lower price than City Gas charged its other customers so that Consolidated itself could turn around and sell the gas.
. The nature of the facilities that courts have held are not practically or feasibly able to be duplicated in other essential facilities cases also supports my conclusion here. See, e.g., Terminal R.R., 224 U.S. at 383, 32 S.Ct. at 507 (major railroad terminal); Fishman, 807 F.2d at 520 (Chicago stadium); Aspen, 738 F.2d at 1509 (ski mountain); Hecht, 570 F.2d at 982 (Robert F. Kennedy Stadium).
. As I explain in detail below, see infra at 1314-1319, by "fair" price, I mean the price required by the just compensation clause of the fifth amendment.
. The government brought suit under section 4 of the Sherman Act, 15 U.S.C. § 4, contending that the arrangement was a combination in restraint of trade in violation of section 1 of the Sherman Act, id. § 1.
. The Court rejected the government’s request for a dissolution of the ownership group. Relying on a proposition that " ‘injury to the public by the prevention of an undue restraint on or the monopolization of trade or commerce is the foundation upon which the prohibitions of the statute rest,' ” id. 224 U.S. at 409, 32 S.Ct. at 515 (quoting Standard Oil, 221 U.S. at 78, 31 S.Ct. at 523), the Court held that mandating equal access to the facility on the part of all railroads would accomplish the requirements of the antitrust laws most consistently with the public's interest in a well-operated railroad terminal.
. The monopoly-leveraging cases that I discuss above, see supra at 1289-1290, would therefore not control here either.
. Indeed, it differs from refusal-to-deal cases in general, under any formula of analysis. See supra at 1283-1288.
. As such, this element would be very different from a legitimate business reason asserted by a company in its defense after a plaintiff has already submitted sufficient evidence to raise an inference of anticompetitive intent.
. That'is, it must satisfy the fifth amendment's •just compensation requirement. See infra at 1314-1319.
. These advertisers, moreover, did not request advertising space on special terms, monetary or otherwise; the subject of their advertisements was unobjectionable; and the advertisements would not have had an adverse impact on the newspaper’s ability to attract other clients. Thus, the newspaper presumably had no legitimate business reason for its refusal.
. Although the FPC could not order a company to wheel involuntarily, the FPC did have authority to order interconnections and sales at wholesale.
. The Court faced the question of whether the federal regulatory scheme limited the district court's power to fashion a decree. The Court concluded that the requirement of FPC approval of interconnections posed a potential limit on the decree. Because the decree mandated wholesaling only subject to Commission approval, however, the facts of the case did not implicate that limit. If, in light of future events, a problem were to arise, the Court indicated that the district court, which retained jurisdiction, could consider the problem then.
The Supreme Court noted, however, that the Federal Power Act gave the FPC no authority to order an electric company to wheel if the company had not voluntarily undertaken an obligation to do so, see id. at 375-76, 93 S.Ct. at 1028-29. Thus, the federal regulatory scheme had no impact on the district court's decree that Otter Tail wheel power. Id. at 376, 93 S.Ct. at 1029; see Florida Power & Light Co. v. FERC, 660 F.2d 668, 674-76 (5th Cir. Unit B Nov. 1981), cert. denied, 459 U.S. 1156, 103 S.Ct. 800, 74 L.Ed.2d 1003 (1983).
. In fact, the court held only that the evidence was sufficient to support the jury’s finding that AT&T’s refusal to deal was unlawful. 708 F.2d at 1132-33.
. MCI merely wanted access to the local exchange facilities on the same terms that AT&T had access to them. The Bell System, which consisted of twenty-three operating companies, provided the local exchange services. These companies were in the business of providing telephone service on the local level and of selling AT&T access to that service for its long-distance operation. Thus, the Bell System was obviously in the business of interconnecting long-distance lines to its local switching services.
. The court reached this conclusion based on its application of antitrust principles. It considered those principles independently of the effect that a regulatory determination regarding such interconnections might have. As the court admitted, its conclusion might have been different if the FCC had authorized or mandated that AT&T provide the interconnections. Without such an FCC order, however, the antitrust laws imposed no duty on AT&T to provide the interconnections. See id. at 1149.
. The government might also bring a civil enforcement action under section 4 of the Sherman Act, 15 U.S.C. § 4, to enjoin the defendant from refusing to deal in violation of section 2. See infra note 11.
. As I elaborate below, see infra at 1321-1325, the private plaintiff may also seek, in addition to an injunction, damages for profits lost between the refusal to deal and the date of final judgment; or, if injunctive relief would not provide a remedy because the refusal to deal forced the plaintiff to go out of business, then the plaintiff may seek damages for loss of its going concern value and damages for profits lost in the interim.
. The district court and panel made much of the fact that City Gas "had, in fact, entered into a contract with a third party [Florida Gas] to sell its gas for resale.” 880 F.2d at 300, see 665 F.Supp. at 1510. This contract, however, was merely executory: City Gas never performed. As the panel claimed, this contract no doubt provides evidence that City Gas had "power” to sell gas at wholesale, 880 F.2d at 301 (emphasis in original), but the contract is not evidence that City Gas ever actually did sell at wholesale. As I discuss in part IV, that City Gas might have possessed the power to sell at wholesale is neither here nor there; rather, what is significant is that City Gas legitimately chose not to exercise that power — not selling to Consolidated, to Florida Gas, or to any other firm. And in making this choice, City Gas did nothing wrong.
As I also discuss in part IV, the contract would not be relevant even if the parties had not abandoned it: the contract would have involved a different geographic market than the market involved in the present case. Perhaps the best evidence of this contract’s irrelevance, though, is the fact that the district court never looked to it as a basis for establishing a selling price in its injunctive order requiring City Gas to supply wholesale gas to Consolidated. The district court delegated the task of establishing a selling price to the FPSC. See 665 F.Supp. at 1512, 1534, 1545. In so doing, the district court implicitly found the Florida Gas contract to be *1313unreliable as evidence of City Gas’ wholesale gas price.
. In articulating this new rule of law, the district court should have distinguished the body of antitrust cases in which courts have held that the antitrust laws impose no positive duties, but only negative duties. See, e.g., Aspen, 472 U.S. at 600-01, 105 S.Ct. at 2856-57; United States v. Colgate & Co., 250 U.S. 300, 307, 39 S.Ct. 465, 468, 63 L.Ed. 992 (1919); Olympia, 797 F.2d at 376 (7th Cir.1986); Florida Fuels, Inc. v. Belcher Oil Co., 717 F.Supp. 1528, 1532 (S.D.Fla.1989). These cases teach that a company has no duty to foster competition or to aid a competitor but that a company may not do anything to prohibit competition or to prevent a company from competing. Under the district court’s new rule, which the court today upholds, section 2 would seem to impose a positive duty on City Gas, requiring it to sell at wholesale to Consolidated so that Consolidated can compete, even though City Gas has never voluntarily sold wholesale gas to any firm. Although I see no grounds for distinguishing these cases from the present case, I assume for purposes of discussion that their inapplicability to the present case could be explained.
. In the typical refusal-to-deal case, the fifth amendment poses no problem: the court requires the defendant to sell at its already-established selling price, applying a presumption that the defendant has set the price that it must receive in order to obtain just compensation.
. The Hope Court approved historical cost as a basis for calculating rates. Under this rule, just compensation is provided by “[r]ates which enable [a] company to operate successfully, to maintain its financial integrity, to attract capital, and to compensate its investors for the risks assumed.” Hope, 320 U.S. at 605, 64 S.Ct. at 289.
The rationale behind this rule, as one judge has stated, is that regulation
was part of the original compact between investors and the state. Rate regulation is, in theory, the substitute for competition. The state stands in the shoes, as it were, of competitors, keeping the utility within bounds that would be drawn by market forces in a non-monopolistic market. Whether economically sound or not, utility rate regulation in itself raises no constitutional concerns.
Jersey Central, 810 F.2d at 1190 (Starr, J., concurring). A crucial principle behind this rule, therefore, is the idea that the utility companies have voluntarily undertaken to operate in a regulated industry, enjoying the benefits of a natural monopoly in return for submitting to reasonable regulation of their rates. As a corollary to this principle, the utility company may, of course, choose not to participate in the industry any longer: although the regulations limit the companies’ rates, those regulations, as this court has itself observed, do not require “the regulated group ... to participate in the regulated industry.” Whitney v. Heckler, 780 F.2d 963, 972 (11th Cir.), cert. denied, 479 U.S. 813, 107 S.Ct. 65, 93 L.Ed.2d 23 (1986).
. Application of established takings cases also supports this conclusion. In this case, the "compact” rationale, see supra note 51, under which utilities voluntarily subject themselves to regulation in return for certain benefits, does not apply. Rather, "the regulated group is ... required [by a federal court implementing the Sherman Act] to participate in the regulated industry,” Whitney, 780 F.2d at 972, and courts have considered such compulsion to be an important indication of a fifth amendment taking. Such a regulation could still be enforced if the company was provided just compensation. Because the company acts under compulsion, however, rather than voluntarily submitting to regulation as in the ratemaking cases, the court should apply a more rigorous standard for just compensation than the relatively broad "zone of reasonableness” standard developed under Hope. Before turning to the question of just compensation, I first analyze this case under established takings principles.
As the Supreme Court has emphasized, "[t]he power to exclude has traditionally been considered one of the most treasured strands in an owner’s bundle of property rights.” Loretto v. Teleprompter Manhattan CATV Corp., 458 U.S. 419, 435, 102 S.Ct. 3164, 3176, 73 L.Ed.2d 868 (1982); see Nollan v. California Coastal *1315Comm'n, 483 U.S. 825, 830-32, 107 S.Ct. 3141, 3145, 97 L.Ed.2d 677 (1987); Kaiser Aetna v. United States, 444 U.S. 164, 179-80, 100 S.Ct. 383, 393, 62 L.Ed.2d 332 (1979); see also FCC v. Florida Power Corp., 480 U.S. 245, 251-53, 107 S.Ct. 1107, 1112, 94 L.Ed.2d 282 (1987). In fact, the Court has held that a permanent physical occupation of property by the government constitutes a per se taking. See, e.g., Loretto, 458 U.S. at 434-36, 102 S.Ct. at 3175-76; see also Florida Power Corp., 480 U.S. at 251-53, 107 S.Ct. at 1112. The severity of such an invasion warrants this conclusion without requiring analysis under the "multifactor inquiry generally applicable to nonpossessory governmental activity.” Loretto, 458 U.S. at 440, 102 S.Ct. at 3178-79; see Penn Central Transp. Co. v. City of New York, 438 U.S. 104, 124, 98 S.Ct. 2646, 2659, 57 L.Ed.2d 631 (1978) (describing multifactor inquiry); see also infra (applying inquiry in present case). The Supreme Court has emphasized, however, that this permanent-physical-occupation rule is "very narrow.” See, e.g., Florida Power, 480 U.S. at 251-53, 107 S.Ct. at 1112; Loretto, 458 U.S. at 441, 102 S.Ct. at 3179.
In my view, this case presents a close question concerning the Loretto rule’s applicability. Lor-etto involved a state regulation that required landlords to allow cable television companies to install cable facilities on their property. The Court found that the installation involved a permanent, physical attachment and thus held that it constituted a per se taking. Loretto, 458 U.S. at 441, 102 S.Ct. at 3179. As the Court has subsequently indicated, a key rationale behind that holding was the fact that the statute required landlords to allow cable installation. On this basis, the Court in Florida Power distinguished a federal statute, the Pole Attachments Act, 47 U.S.C. § 224 (1982), which authorized the FCC to regulate the terms of contracts between utility companies and cable operators for the attachment of cable to utility poles, from the regulation in Loretto. As the Florida Power court discussed,
while the statute we considered in Loretto specifically required landlords to permit permanent occupation of their property by cable companies, nothing in the Pole Attachments Act as interpreted by the FCC in these cases gives cable companies any right to occupy space on utility poles, or prohibits utility companies from refusing to enter into attachment agreements with cable operators.
480 U.S. at 251, 107 S.Ct. at 1112. The Court in Florida Power expressly stated, however, that it did "not decide ... what the application of Loretto ... would be if the FCC in a future case required utilities, over objection, to enter into, renew, or refrain from terminating pole attachment agreements." Id. at 251 n.6, 107 S.Ct. at 1112 n.6. The present case seems to present exactly this question, and to the extent that the element of compulsion explains the rule in Lor-etto, that rule should apply here.
Another key factor in the Loretto decision was the nature of the invasion; the permanence of the facilities and the landlord’s lack of control over them and their installation. In my view, the invasion in the present case is factually similar. As in Loretto, the district court’s mandate requires a permanent, physical intrusion onto City Gas’ property. The intrusion is permanent in two ways: first, a pipe must be constructed permanently to attach City Gas’ pipesystem to Consolidated’s pipesystem; and second, this interconnecting pipe will permanently drain natural gas from City Gas’ pipesys-tem and into Consolidated’s pipesystem. As the Court has previously held, such a continuous arrangement constitutes a "permanent physical occupation" for purposes of the Loretto rule. See Notion, 483 U.S. at 832, 107 S.Ct. at 3145 (easement giving individuals "permanent and continuous right to pass to and fro, so that the real property may continuously be traversed").
I also think that City Gas, like the landlord in Loretto, lacks control over this invasion. City Gas has no control over the invasion in that City Gas cannot refuse to provide gas. But City Gas also lacks control over the invasion in a secondary sense that the Loretto Court also held to be important. The Court there referred to the landlord’s lack of control over the "placement, manner, use, and possibly the disposition of the installation” and suggested that the outcome might have been different if the landlord had such control. 458 U.S. at 440 n.19, 102 S.Ct. at 3179 n.19. Similarly, in Nollan, the Court noted the property owner’s inability to regulate the continuous flow of traffic across his property. See 483 U.S. at 830-32, 107 S.Ct. at 3145. Here, City Gas will have to supply gas according to the terms mandated by the district court: in set amounts, at set times, and at set prices. The fact that City Gas owns its own pipesystem is of no significance. The district court, and not City Gas, decrees the terms of the arrangement. I would think, therefore, that Loretto controls: if section 2 of the Sherman Act requires City Gas to deal under these circumstances, that constitutes a per se taking.
Even if Loretto does not control, this forced supply constitutes a taking under the Penn Central multifactor inquiry. That inquiry looks to (1) the economic impact of the regulation, (2) the extent to which it interferes with investment-backed expectations, and (3) the character of the governmental action. See Penn Central, 438 U.S. at 124, 98 S.Ct. at 2659. Imposing a duty to deal will have a significant economic impact on City Gas. With an investment of time and money, City Gas will have to divert some of its own gas supply to Consolidated and also administer the actual process of supplying gas to Consolidated. In addition, by enabling Consolidated to compete without having to undertake the costs associated with entry into the natural gas supply business, the duty to deal might force City Gas to change its retail distribution business: in order to avoid losing customers to Consolidated, City Gas might (subject to the FPSC's approval) have to lower its retail prices, if it can do so without loosening the safety controls of its operation; it might have to *1316undertake a more aggressive marketing strategy, which would in turn increase its operating costs; it might also lose potential business opportunities — e.g., the ability to extend its services to customers in territory occupied by Consolidated. In the event of a gas shortage, moreover, City Gas might have to curtail delivery to its own customers in order to continue supplying Consolidated.
As these considerations suggest, the duty to deal will also interfere with City Gas' investment-backed expectations. By changing the market’s regulatory structure and altering the nature of competition in that market, the duty to deal will significantly affect these expectations.
Finally, the character of the district court’s invasive action suggests a taking. This is not a case where a government regulation merely restricts the use of property. Rather, the case involves an actual physical invasion of City Gas' property: gas is removed from its pipesystem and transferred to Consolidated's pipesystem. Even under the multifactor inquiry, therefore, this case involves a taking.
. In deciding this case, of course, the court faces the question of just compensation in a different procedural context than it would in the standard ratemaking cases that I cite above. Rather than reviewing a state or federal agency’s rate determination to decide whether it passes constitutional muster, the court, as I explain in the text below, must itself set the just compensation price in order to determine the issue of liability in the case.
In addition, this "taking" must satisfy the fifth amendment’s “public use” requirement. That requirement, however, has been interpreted broadly. As the Supreme Court stated in Hawaii Housing Auth. v. Midkiff, 467 U.S. 229, 240, 104 S.Ct. 2321, 2329, 81 L.Ed.2d 186 (1984), ”[t]he ‘public use' requirement is ... coterminous with the scope of a sovereign’s police powers.” See also Berman v. Parker, 348 U.S. 26, 32-33, 75 S.Ct. 98, 102, 99 L.Ed. 27 (1954). Thus, the courts conduct “‘an extremely narrow’ ” review of "a legislature's judgment as to what constitutes a public use,” Midkiff, 467 U.S. at 240, 104 S.Ct. at 2328 (quoting Berman): "the Court ... will not substitute its judgment for a legislature’s judgment as to what constitutes a public use ‘unless the use be palpably without reasonable foundation,’” id. at 241, 104 S.Ct. at 2329 (quoting United States v. Gettysburg Elec. R. Co., 160 U.S. 668, 680, 16 S.Ct. 427, 429, 40 L.Ed. 576 (1896)). I assume, for the sake of discussion, that the taking in the present case would satisfy this requirement.
. In effect, antitrust doctrine already reflects this idea by holding that a party is not liable under section 2 for a refusal to deal if a valid business justification supports that party’s decision. See, e.g., Mid-Texas Communications Sys. v. AT&T, 615 F.2d 1372, 1388-89 (5th Cir.), cert. denied, 449 U.S. 912, 101 S.Ct. 286, 66 L.Ed.2d 140 (1980); Poster Exchange, 431 F.2d at 339; Gamco, Inc. v. Providence Fruit & Produce Bldg., Inc., 194 F.2d 484, 488 (1st Cir.), cert. denied, 344 U.S. 817, 73 S.Ct. 11, 97 L.Ed. 636 (1952).
. In its “cost" figure, the district court below included only the price that City Gas paid for its wholesale gas and not the overall costs that City Gas incurred in its retail gas operation. See 665 F.Supp. at 1511.
. As I discuss above, see supra notes 5 & 7, the district court below put a spin on these facts, characterizing City Gas as the bad guys and Consolidated as the good guys. The district court suggested that City Gas knew its initial offer of cost plus ten cents per therm was too high and that the FPSC would never approve it. In addition, the district court pointed to testimony by City Gas' Vice President that "he made no attempt to ascertain a reasonable price for direct gas sales to Consolidated, and that he came up with the cost plus ten cents terms 'out of the air.’” 665 F.Supp. at 1510.
This characterization assumes that City Gas had a duty to enable Consolidated to enter the natural gas business at as little cost as possible so that it could most effectively compete with City Gas. Because City Gas did not want, and did not seriously attempt, to do this, the court characterized its behavior as bad. This characterization is counterintuitive. City Gas did exactly what one would expect any rational business firm to do: it continued operating as it normally did, doing nothing illegal to obstruct a potential competitor’s entrance into the market but also doing nothing to facilitate its entrance into the market. To characterize this type of conduct as bad is to adopt a position totally at odds with normal business practices ■ in this country.
. Consolidated must show that City Gas’ price exceeded the fifth amendment price by a substantial amount. If that point is fairly debatable, then City Gas will not be subject to liability.
I suggest that Consolidated might also have to show that its counter offer was reasonable in light of the fifth amendment price. If Consolidated offered a price, indicating that it would pay no more for the gas, and the fifth amendment price substantially exceeded Consolidated’s offered price, then Consolidated has no claim under section 2.
. Consolidated bears the burden of proof on this point. In arranging the order of proof, however, the court might apply a presumption requiring City Gas to come forward with all of the evidence, on the ground that the evidence is exclusively in its possession and that fairness requires City Gas to proceed first. City Gas will then have to present evidence demonstrating that the price it quoted to Consolidated was necessary to provide just compensation. See infra at 1317-1319, for a discussion of the factors relevant to such a showing. The burden, however, will then shift to Consolidated to rebut City Gas’ just compensation price.
. Blackacre’s fair market value can be determined based on the selling price of comparable pieces of property. Even if no comparable sales have occurred in the market, the value can be determined by constructing a hypothetical market and taking expert testimony regarding Blackacre’s best-use price.
. If such a company had existed in the present case, then the district court below, in fashioning its injunction requiring that City Gas sell to Consolidated, would not have had to delegate to the FPSC the task of setting a selling price. See 665 F.Supp. at 1512, 1534, 1545. Instead, it could have looked to that other company’s wholesale gas price as evidence of City Gas' just compensation price.
Even if such a company could be found, its selling price would probably not be controlling. That company would be a regulated company, selling its wholesale gas at rates approved by the FPSC. As such, that company would have voluntarily undertaken to sell at wholesale in a regulated market, subjecting itself to the FPSC's rates in return for the benefits of operating in a regulated industry. See supra note 51. That price would not be presumptively fair to City Gas, a company that has not voluntarily undertaken to deal at those FPSC-approved rates.
. Consolidated obviously could not point to the price charged by wholesalers like FGT, as a basis for the just compensation price. City Gas has opted not to sell in the wholesale market, applying its resources to the retail market instead. As a retail distributor, City Gas’ fixed and operating costs are entirely different from those of a wholesaler like FGT: City Gas’ investments are committed to a variety of expenditures that are unnecessary for a wholesale operation. In order to receive a reasonable return on its investment, therefore, City Gas must charge a different price than FGT charges.
. In order to evaluate this last factor, the court considers, in part, probable changes in the market value of Consolidated’s stock.
. As the Hope Court stated:
[T]he investor interest has a legitimate concern with the financial integrity of the company whose rates are being regulated. From the investor or company point of view it is important that there be enough revenue not only for operating expenses but also for the capital costs of the business. These include service on the debt and dividends on the stock.... By that standard the return to the equity owner should be commensurate with returns on investments in other enterprises having corresponding risks. That return, moreover, should be sufficient to assure confidence in the financial integrity of the enterprise, so as to maintain its credit and to attract capital.
320 U.S. at 603, 64 S.Ct. at 288.
. That is, Consolidated must prove that the price City Gas offered Consolidated was substantially higher than the fifth amendment price. See supra note 57.
. In that context, moreover, the courts’ task is eased by a high initial burden of proof on the utility company to show that the rate is unreasonable before the test even applies. In addition, the courts apply a broad standard of reasonableness in reviewing rates set by regulators. In part, that standard reflects institutional concerns and ensures the courts’ deference to agencies' determinations. That standard also reflects the nature of the compact between utility companies that have voluntarily entered a regulated market and the state, which regulates the market. See supra note 51. In the present case, however, City Gas has not voluntarily entered such a compact. A price that might be satisfactory in the regulatory context, therefore, will not necessarily provide City Gas with a reasonable rate of return on its investment here.
. As I develop below, see infra at 1325-1328, the district court in this case did not find a fifth amendment price, let alone evaluate City Gas’ offered price in light of that fifth amendment price. The district court, therefore, never properly decided whether, in light of the fifth amendment price, City Gas actually refused to deal under section 2. In addition, having failed to find a fifth amendment price, the district court did not properly determine whether Consolidated actually suffered injury as a result of City Gas’ refusal to deal. Finally, because of this failure to find a fifth amendment price, the remedies fashioned by the district court were also erroneous: as I discuss below, see id., proper remedies in this case must be based on the fifth amendment price.
. City Gas’ shareholders will initially absorb these expenses through a depletion of City Gas’ resources but will then pass them off to the customers through rate increases. If the FPSC does not approve rate increases, however, the shareholders will assume the burden themselves. In that case, City Gas will eventually go out of business (when its resources are exhausted).
. Congress surely did not design the antitrust laws merely to prevent companies from going out of business where the defendant has done nothing to undercut that company’s ability to compete. Nor did Congress provide for liquidated damages in the antitrust laws, and I sincerely doubt that Congress intended the antitrust laws to tie up a courtroom for months for the purpose of awarding nominal damages merely to punish a company such as City Gas for not selling gas at the fifth amendment price.
. Curiously, Consolidated did not request a preliminary injunction. If Consolidated's ability to operate as a natural gas supplier — and to do so at a profit. — depended upon a hook-up to City Gas, then presumably Consolidated would have wanted a hook-up as soon as it could obtain one. Nevertheless, for reasons not apparent in the record, Consolidated decided not to request a preliminary injunction. Instead, it waited until trial to specify the remedy it was seeking. There, Consolidated claimed that City Gas, by refusing to supply wholesale gas, had in effect taken Consolidated’s business. Consolidated therefore sought damages in the form of the going concern value of its business. Through its expert witness, Ben Ball, Consolidated asserted that City Gas had deprived it of twenty years of future profits. As it turned out, the district court awarded Consolidated its lost future profits (reduced to their present value) as well as a permanent injunction requiring City Gas to supply Consolidated with gas at a price to be set by the FPSC. The granting of injunctive relief was patently inconsistent with the damages award, an issue that I discuss below, see infra at 1324-1325, 1327.
. If Consolidated cannot show past injury for purposes of the damages award, then Consolidated will probably also be unable to show a likelihood of future injury for purposes of in-junctive relief. In a case such as this one, a plaintiff generally proves a likelihood of future injury based on a showing of past injury.
. Conceivably, Consolidated could have lost a part of its going concern value prior to its total demise. Assume, for example, that Consolidated (as an LP gas distributor) serviced Whiteacre and Blackacre and that, soon after the refusal to deal, City Gas took control of Blackacre. At that point, which preceded its general demise, Consolidated lost the ability to go back into operation in Blackacre even if it could have acquired a supply of gas before its general demise. In that case, Consolidated would still seek damages for the company’s going concern value just prior to the demise, as well as damages for profits lost prior to its demise; it would also seek damages for the going concern value of its operation in Blackacre — the portion of its operation that was lost before the demise. Of course, when the court determined the going concern value of Consolidated’s entire operation just prior to the demise, that operation would no longer include Consolidated’s operation in Blackacre. Cf. infra note 76 (discussing loss of partial going concern value in relation to in-junctive relief).
. I assume that City Gas was Consolidated’s only feasible gas supplier, i.e., that hooking up to FGT’s pipeline was too expensive for Consolidated.
. In the dickering process, the buyer and seller assess the risk of the investment. The buyer emphasizes the possibility that Consolidated will generate profits only for a short term, taking into account the likelihood of changes in the fifth amendment price. Obviously, the more uncertain Consolidated’s profit picture might be, the less the buyer is willing to pay.
. For the sake of discussion, I assume that $250,000 in annual profits represents Consolidated's net operating profits: gross revenues less gas delivery costs with no reserve for depreciation. I also assume that Consolidated’s fixed assets — its plant and equipment — are self-consuming and will be worthless at the end of twenty years.
.In the unlikely event that the demise occurred the moment of City Gas’ refusal, and Consolidated could still show that it could have turned a profit had City Gas not refused to deal, then Consolidated obviously cannot seek damages for lost profits in the past: the damages for loss of future profits will be calculated as of the moment of the refusal and will thus include all profits that Consolidated could have generated had the refusal not occurred.
. These two remedial schemes would be consistent and would not involve a double recovery in a case where a company permanently loses part of its assets as a result of a refusal to deal but retains enough of its assets to continue operating once it gains access to the needed resource. Assume, for example, that Company A operated in Whiteacre and Blackacre and had an opportunity to expand into Beyondacre. As a result of Company B’s refusal to sell resource X to Company A (which would constitute a violation of section 2 under the district court’s new rule, even though Company B has never sold resource X to other companies), Company A can no longer operate in Blackacre and has lost forever the opportunity to expand into Bey-ondacre; Company A’s operation, although still profitable, is now limited to Whiteacre. Thus, the court grants an injunction ordering Company B to deal, which allows Company A to continue its operation in Whiteacre, and the court also awards damages for lost profits between the time of Company B’s refusal to deal and the final judgment in the case. In addition, the court awards Company A the fair market value of its operation in Blackacre and its opportunity to expand into Beyondacre (these business opportunities having been lost forever). If the court, in doing so, made the necessary factual findings — such as what the fifth amendment price was for resource X, whether Company A could turn a profit given that fifth amendment price, and what the degree of uncertainty concerning Company A’s profitability was — these remedies would be proper.
. The district court below found that City Gas refused to deal as of February 20, 1982 — the date when City Gas’ Board of Directors met and first considered acquiring Consolidated. In my view, that finding is clearly erroneous. The refusal could not have occurred until City Gas and Consolidated entered into negotiations and City Gas refused to deal at the fifth amendment price.
. In improperly directing the parties to repair to the FPSC for a determination of the wholesale price, the district court in the present case implied that the FPSC's price would, without further consideration by the district court, operate as a provision (the price term) of the injunction. The district court apparently gave no thought to what should happen if the FPSC’s price did not satisfy the fifth amendment. That is, the district court did not retain jurisdiction to review the FPSC's price in the event that the price failed to provide City .Gas with just compensation.
. Ball’s testimony assumed an entire taking of Consolidated’s business as a company with gas. Thus, the record provided no basis for treating this case as a partial-taking situation, see supra notes 71 & 76, in which damages for loss of going concern value might be consistent with injunctive relief.
. In awarding Consolidated damages in the form of the present-day value of the profits that it would have made over the next twenty years, the district court found — albeit implicitly (by accepting Ben Ball’s testimony) — that City Gas would have supplied gas for that period at a "competitive price.” If Ball’s testimony supported the finding that, with a supply of gas from City Gas, Consolidated could have operated profitably for twenty years, then that testimony certainly supported, with equal force, the district court’s conclusion (implicit in its award of permanent injunctive relief) that an injunction would enable Consolidated to operate profitably for twenty years.
. Municipalities, unlike private parties, must satisfy only the first Midcal requirement in order to have Parker immunity. See Town of Hallie v. City of Eau Claire, 471 U.S. 34, 46-47, 105 S.Ct. 1713, 1720, 85 L.Ed.2d 24 (1985).
. Because the state courts conducted no review of specific committee decisions, the Patrick court found it unnecessary to decide "the broad question whether judicial review of private conduct ever can constitute active supervision." 486 U.S. at 104, 108 S.Ct. at 1664-65; see also Pinhas v. Summit Health, Ltd., 894 F.2d 1024, 1030 (9th Cir.1989) (also avoiding that question), cert. granted, - U.S. -, 110 S.Ct. 3212, 110 L.Ed.2d 660 (1990). I consider this question below, as it applies to this case. See infra at 1336.
. The Court did suggest that a simple "minimum markup” statute, which required retailers to charge 112% of their actual wholesale cost, might satisfy Midcal's second prong. See 479 U.S. at 344 n.6, 107 S.Ct. at 725 n.6.
. This chapter has been amended by 1989 Fla. Laws ch. 89-292, to authorize expressly the FPSC to approve territorial agreements and resolve territorial disputes between natural gas utilities. I consider below the amendment’s effect on the present case, see infra note 86.
. Indeed, once the panel’s decision in this case adopted this inconsistent position, the Florida legislature in effect overruled it by adding language that expressly authorizes the FPSC to approve territorial agreements and resolve territorial disputes between natural gas utilities. The legislature thus made explicit what, as I have observed, was already implicit in the old provision.
. This conclusion finds support in the Florida Supreme Court's holding in the Peoples Gas case that chapter 366 implicitly grants the FPSC power to approve territorial agreements. As the panel noted, 880 F.2d at 302-03, although that decision is not dispositive of this inquiry, it is relevant.
. The district court quoted the relevant passage from the FPSC’s decision:
It is not clear that this Commission has the statutory authority to either establish exclusive service territories for natural gas utilities (as it does for telephone, water and sewer utilities) or resolve territorial disputes between natural gas utilities (as specifically authorized for electric utilities). However, our resolution of this question is not necessary in view of our [other] finding....
665 F.Supp. at 1530. This "expression of doubt” is clearly dictum. Whatever doubt the FPSC might have expressed, moreover, has not effected the FPSC's actions in approving territorial agreements, as I discuss. Also, that expression of doubt has nothing to do with the active supervision issue; rather, the expression merely comments on whether the state has clearly articulated a policy in favor of territorial agreements. As I show above, the state has indeed done so — regardless of whatever doubt the FPSC (an agency without judicial authority to interpret legislative intent) may have expressed.
. As the Supreme Court has made clear, a state court’s determination that conduct is consistent with some state law does not necessarily insulate that conduct from federal antitrust law. See Parker, 317 U.S. at 351, 63 S.Ct. at 314 (citing Northern Sec. Co. v. United States, 193 U.S. 197, 344-47, 24 S.Ct. 436, 459-61, 48 L.Ed. 679 (1904)). Here, however, the Florida Supreme Court has approved that conduct as consistent with the state’s policy to displace competition in the natural gas industry. Such a decision is exactly what invokes Parker immunity.
. Curiously, the panel noted that City Gas challenged the district court’s holding that Parker immunity did not apply to its refusal to deal, see 880 F.2d at 301, but nevertheless failed to address the issue.
. The district court claimed that the legislature, in drafting this provision, had in mind sales of gas for resale to landlords who would re-meter the gas for further distribution to their tenants and that the legislature, therefore, did not specifically contemplate wholesale sales to other natural gas utilities. Although the legislature no doubt had in mind sales to landlords, that fact provides no basis for the district court’s conclusion that this provision does not apply to wholesale sales to other natural gas utilities. Not only is that reading inconsistent with the provision’s plain language, which refers in broad terms to sales of gas for resale generally, but the district court’s construction of the legislature’s intent is extremely crabbed. Wholesale sales to other natural gas companies also place an administrative burden on the FPSC by altering the structure of competition in the industry or of the utility’s assets, thus increasing the complexity of the ratemaking process. See supra at 1332-1333. The district court’s too narrow construction of the legislature’s intent is *1336thus inconsistent with the nature of the regulatory scheme created by chapter 366.