The state of Illinois brought this antitrust suit on its own behalf and on behalf of a class of residential and commercial consumers of natural gas in central Illinois. The state alleges that the Panhandle Eastern Pipe Line Company violated federal and state antitrust laws in the early 1980s by refusing to transport natural gas pur*1472chased by its principal commercial customers (the local distribution companies that distribute gas to residential and most commercial and industrial end-users) through its pipelines. After a bench trial, the district court ruled that Panhandle’s conduct was not anticompetitive. Illinois ex rel. Hartigan v. Panhandle Eastern Pipe Line Co., 730 F.Supp. 826 (C.D.Ill.1990). We agree, and affirm the judgment.
I. Background1
The late 1970s and early 1980s found the natural gas industry in the throes of deregulation. Regulation in the industry dated back to the 1930s when a handful of pipeline companies monopolized the purchase and distribution of natural gas. Congress had responded by regulating the pipelines and controlling natural gas prices at the wellhead. Under regulation, pipelines typically purchased the gas from producers and resold it to their customers; “gas flow[ed] from producer to pipeline to distributor to consumer, with title passing at each change of possession.” Pierce, Reconsidering the Roles of Regulation and Competition in the Natural Gas Industry, 97 Harv.L.Rev. 345, 348 (1983). The pipelines thus traditionally bundled together two commodities — gas and pipeline transportation — for their customers. Their profit, however, derived solely from the return permitted by regulators on the transportation service; the commodity component of pipeline rates reflected only a pass through of the price paid by the pipeline for the gas.
Shortages plagued the natural gas market under regulation, leading Congress to reverse its course. In 1978, Congress embarked on a program of phased deregulation — embodied in the National Gas Policy Act (NGPA), 15 U.S.C. §§ 3301-3432 — that established a graduated series of increases in the maximum permissible price for natural gas and culminated in the complete termination of wellhead (producer) price regulation in 1985. See Mobil Oil Exploration v. United Distr. Co., — U.S.-, 111 S.Ct. 615, 620-21, 112 L.Ed.2d 636 (1991). The NGPA also changed the regulations governing the distribution of natural gas, providing an impetus for the Federal Energy Regulatory Commission (FERC) to loosen the grip of pipelines on the territorial monopolies regulation had preserved. Section 311 of the NGPA authorized FERC to permit interstate pipelines to transport gas purchased by local distribution companies (LDCs) and large industrial end-users directly from producers, although it did not authorize FERC to order the pipelines to do so.
As deregulation progressed, many pipelines entered into long-term contracts to purchase natural gas at high deregulated prices, anticipating continued shortages and continued growth in the demand for natural gas. Those prices were subject to little regulatory control, since FERC reviews pipeline gas acquisition costs only for “fraud or abuse,” 15 U.S.C. § 3431(c)(2), and permits pipelines to include “purchase gas adjustment” clauses (PGAs) in their contracts with distributors. Id. at 350. These clauses permit pipeline companies to adjust their rates regularly to reflect changes in the cost of the gas they purchase. Pierce, supra, at 350 n. 33.
The Panhandle Eastern Pipe Line Company was no exception. Panhandle’s pipeline system stretches northeast from the Gulf of Mexico into Michigan, and during the early 1980s, Panhandle was the exclusive supplier of natural gas to 37 counties in central Illinois. Beginning in 1979, Panhandle launched an aggressive campaign to secure what at that time were still scarce, and expensive, gas supplies. Among its efforts to secure gas were two very expensive projects: Panhandle contracted through Trunkline Gas Company, a subsidiary pipeline, to purchase liquified natural gas from Algeria, and joined a partnership to construct pipelines to import gas from Canada.
*1473Deregulation worked, however, and higher prices spurred increased production of natural gas. But as production was increasing, demand was decreasing because the prices of alternative fuels were dropping and energy conservation measures were intensifying. Despite warnings from its customers that demand was slacking, Panhandle continued to enter into long-term gas purchase contracts at the maximum prices permitted by the NGPA phaseout price levels. To compound the problem, Panhandle agreed to significant take- or-pay provisions2 in virtually all of its purchase contracts without demanding force majeure or other types of market-out clauses to limit their take-or-pay obligations. Increased production and reduced demand produced a glut of natural gas that depressed spot market prices far below the price ceilings established by the NGPA, and well below the levels mandated in Panhandle’s contracts. In 1982, when the expensive gas from its Algerian and Canadian projects became available, Panhandle’s costs increased dramatically, much to the chagrin of its customers.3 By 1984, Panhandle’s rates were among the most expensive in the country.
Consequently, Panhandle’s LDC customers (the parties focus on Central Illinois Light Company (“CILCO”), and so will we) began exploring the possibility of buying gas from sources other than Panhandle. Panhandle’s contracts with these customers presented a huge obstacle, however, for Panhandle had a “sole supplier” provision in the contracts.4 This provision *1474was part of Panhandle’s FERC-approved “G tariff” rate schedule, which obligated Panhandle to use its best efforts to meet its customers’ demands for gas and permitted customers to vary the quantity of gas purchased each month (demand for natural gas typically decreases substantially during the summer). In return, customers agreed to purchase their full requirements of gas from Panhandle. See Panhandle Eastern Pipe Line Co., 10 F.P.C. 185 (Opinion No. 214), modified, 10 F.P.C. 322 (1951), further modified, 13 F.P.C. 53 (1954) (Opinion 269), rev’d in part, 230 F.2d 810 (D.C.Cir.1955). As long as the price of natural gas was regulated, the sole supplier provision was a small price for the G tariff customers to pay for the security of a stable supply of gas. Under price regulation, they could afford to be indifferent to their source of gas, and were: Panhandle’s supply contracts with LDCs were typically long-term affairs (its 1970 contract with CILCO had a term of eighteen years), a fact that also contributed to Panhandle's own willingness to execute long-term purchase contracts from gas producers.
Notwithstanding the G tariff, the move toward deregulation gave Panhandle’s customers some muscle. Beginning in 1979, FERC enacted a series of regulations under § 311 of the NGPA authorizing interstate pipelines to transport natural gas purchased from sources other than the pipeline itself. In 1982 FERC instituted the “Blanket Certificate Program,” under which interstate pipelines were authorized to transport nonsystem gas (gas to which they did not have title) directly to high-priority end-users (hospitals, schools, and essential agricultural users), so long as the pipeline first received a certificate of public convenience and necessity.
Despite these regulatory initiatives, Panhandle declined to transport nonsystem gas. When CILCO made its first formal request that Panhandle transport gas it had purchased directly from producers in March 1983, Panhandle refused on the ground that enabling its customers to obtain gas from other sources would dramatically reduce demand for the expensive gas it was contractually obligated to purchase, exposing it to enormous take-or-pay liability. CILCO had filed a complaint with FERC in 1982, seeking authorization to interconnect with a natural gas pipeline other than Panhandle; after Panhandle refused to transport nonsystem gas, CILCO incorporated an attack on the G tariff into its case. FERC prospectively invalidated the sole supplier provision of the tariff' in 1987 (Opinion No. 265, 38 FERC H 61,164), but that opinion was vacated and the case remanded for a determination of the reasonableness of the sole supplier restriction. Panhandle Eastern Pipeline Co. v. FERC, 881 F.2d 1101 (D.C.Cir.1989) (per curiam).
In 1983, the FERC issued orders 234-B and 319. 48 Fed.Reg. 34872 (1983); 48 Fed.Reg. 34875 (1983). These orders great*1475ly expanded the class of end-users for which interstate pipelines could obtain approval to transport gas the consumers purchased directly from producers (as opposed to gas the pipeline had itself purchased for resale). Distributors remained ineligible. Application under these programs, as under the blanket certificate program, was voluntary; pipelines were not required to transport gas purchased from other sources. After FERC adopted Orders 234-B and 319, it approved “Special Marketing Programs” (“SMPs”) that permitted certain pipelines, including Panhandle, to obtain credits against their take-or-pay contractual obligations by selling gas to “non-captive” customers (those price-sensitive industrial end-users able to switch to different fuel sources) at the lower spot market prices rather than at the contract prices. Panhandle’s version of the SMP was its “PanMark” program. In September 1984, FERC renewed approval for SMPs such as PanMark for another year, but added a requirement that participating pipelines allow LDCs to purchase up to ten percent of their monthly contract demand directly from gas producers. The FERC order specifically provided for a temporary limited waiver of the sole supplier provision of the G tariff to allow LDCs to participate in the SMPs without jeopardizing their G tariff status. Panhandle objected to the extension of the SMPs to captive LDCs, but continued to participate in order to retard the defection of noncaptive industrial customers.
Those customers of Panhandle ineligible for full participation in the PanMark program, including LDCs like CILCO, continued to press Panhandle to transport non-system gas under authority of its blanket certificate. In late 1983, Panhandle developed the Market Area Transport (“MAT”) program, which expanded the class of industrial customers eligible to obtain transportation of gas purchased directly from producers. Panhandle conditioned its agreement to transport for these customers, however, on its right to meet the price of the producer first; if it did, then the customer was obligated to buy the Panhandle gas. The MAT program was not available to LDCs; it too was intended only as a measure to discourage noncaptive industrial consumers from defecting. Panhandle did extend the MAT program to LDCs who did not purchase under the G tariff (those that purchased gas from other pipelines as well, under a nonexclusive tariff), but continued to refuse to include G tariff LDCs in the program.
In 1985, the D.C. Circuit vacated the FERC orders that authorized the SMP and MAT programs. Maryland People’s Counsel v. FERC (MPC I), 761 F.2d 768 (D.C.Cir.1985) (FERC opinions approving an SMP failed to adequately justify the program’s exclusion of the pipeline’s captive customers); Maryland People’s Counsel v. FERC (MPC II), 761 F.2d 780 (D.C.Cir.1985) (FERC orders 234-B and 319 failed to consider antitrust aspects of the SMP programs); Maryland People’s Counsel v. FERC (MPC III), 768 F.2d 450 (D.C.Cir.1985) (per curiam) (no basis for limiting LDC participation in SMPs to 10% of gas purchases).
In response to this series of decisions, FERC issued Order 436. 50 Fed.Reg. 42408 (1985). This order reauthorized pipelines to transport gas purchased from other sources, but on a nondiscriminatory basis. The Order also permitted pipelines to begin transporting on a provisional basis without committing themselves to continue doing so in perpetuity. Having reversed Orders 234-B and 319 in MPC II for failing to address the concerns of captive natural gas consumers, the D.C. Circuit vacated Order 436 because it failed to consider the “take or pay” problem faced by the pipelines. Associated Gas Distribs. v. FERC, 824 F.2d 981, 1021-30 (D.C.Cir.1987) (AGD).5
*1476Panhandle initially shut down its MAT program rather than participate in the Order 436 program. During 1986, however, Panhandle began an interim program that continued to deny transportation services for nonsystem gas to captive G tariff LDCs. It did so on the condition that the captive LDCs would not request that Panhandle transport direct-purchase gas for them. If any G tariff LDC made such a request, Panhandle informed them, it would terminate its interim 436 program for all. The LDCs protested, but ultimately capitulated, reasoning that even if they couldn’t make use of the program themselves, they also had a stake in preventing fuel-switchable industrial consumers receiving gas directly from Panhandle from defecting to other fuels and thereby increasing the percentage of the Panhandle’s fixed costs each would be required to pay.
The state filed suit in February 1984 in its capacity as a natural gas consumer and as parens patriae for a class of plaintiffs consisting of all of Panhandle’s indirect purchasers residing in the central Illinois counties served exclusively by Panhandle. The state’s complaint alleged that Panhandle monopolized the sale of natural gas within central Illinois by refusing to transport nonsystem gas purchased by LDCs directly from independent producers, thereby forcing them to purchase gas from Panhandle. The complaint comprises ten counts, five under federal law, and five parallel counts under Illinois law, alleging unlawful monopolization (counts 1 and 2), attempted monopolization (counts 3 and 4), monopoly leveraging (counts 5 and 6), an unlawful denial of access to an essential facility (counts 7 and 8), and an unlawful tie of gas transportation to gas purchases (counts 9 and 10).6
The district court denied the state’s motion for a preliminary injunction in December 1984. Panhandle then filed a motion to dismiss, predicated on the doctrine of Illinois Brick v. Illinois, 431 U.S. 720, 97 5.Ct. 2061, 52 L.Ed.2d 707 (1977), which held that, in most cases, indirect purchasers cannot bring antitrust claims against upstream suppliers. The district court denied the motion in September 1985, but certified the question for interlocutory appeal. We initially reversed, 839 F.2d 1206 (7th Cir.1988), but subsequently reheard the case en banc and affirmed, 852 F.2d 891 (7th Cir.1988), cert. denied, 488 U.S. 986, 109 S.Ct. 543, 102 L.Ed.2d 573 (1988), holding that Illinois Brick did not bar the claims of the indirect purchasers of Panhandle’s gas who were unable to turn to alternative sources of fuel (CILCO’s captive residential and commercial customers). While the interlocutory appeal was pending, the district court conducted a bench trial on the merits and found in favor of Panhandle on all counts. The State appealed, and briefs were filed in early 1989. On the day after the last brief was filed, the Supreme Court decided Kansas v. Utilicorp United, Inc., — U.S.-, 110 S.Ct. 2807, 111 L.Ed.2d 169 (1990), holding that the “costplus” exception to the Illinois Brick doctrine did not permit indirect purchasers of a regulated utility to sue for antitrust damages. Panhandle moved for supplemental briefing on whether Utili-Corp disposed of state’s suit and the motion was granted.
II. UtiliCorp’s Impact
Our first order of business, then, is to determine whether UtiliCorp controls the outcome of this case. The UtiliCorp case addressed the issue of whether the *1477customers of a public utility had standing to sue when the utility’s suppliers had allegedly violated the antitrust laws by overcharging the utility for natural gas. In Illinois Brick v. Illinois, the Court held that only direct purchasers may sue sellers who violate the antitrust laws; the purchaser’s customers (the “indirect” purchasers) may not. 431 U.S. at 736-47, 97 S.Ct. at 2069-75; see also Hanover Shoe, Inc. v. United Shoe Machinery Corp., 392 U.S. 481, 493-94, 88 S.Ct. 2224, 2231-32, 20 L.Ed.2d 1231 (1968). The rationale for the indirect purchaser bar is that allowing suits by parties all along the distribution chain could impose duplicative liability on the antitrust violator, particularly since it is difficult to apportion damages between successive distribution levels. That rationale suggested an exception to the rule, however, in cases where the cost of the monopolistic overcharge was certain to be passed on to the next distribution level due to the existence of a cost-plus contract between the direct purchaser and its customers. See Illinois Brick, 431 U.S. at 736, 97 S.Ct. at 2069; Hanover Shoe, 392 U.S. at 494, 88 S.Ct. at 2232. The plaintiffs in UtiliCorp sought the shelter of the “cost-plus” exception, arguing that as customers of a regulated public utility, they paid the entire anticompetitive overcharge originally exacted from the utility by the natural gas producer.
The Court declined, however, “to create an exception [to the Illinois Brick doctrine] for regulated public utilities.” 110 S.Ct. at 2813. The Court held that the “cost-plus” exception applies only when “the direct purchaser will bear no portion of the overcharge and otherwise suffer no injury,” id. at 2818, and that is not the case, said the Court, when the direct purchaser is a public utility. Id.
Illinois maintains that its case is factually distinguishable from UtiliCorp. We note initially that the state took a different view of the significance of the factual distinctions between the cases as an amicus curiae in the UtiliCorp case. In its brief in that case, the state observed:
The natural gas system in Illinois operates in essentially the same manner as the systems in place in Kansas and Missouri. It is likely that the outcome of this case will be dispositive of the identical issue in Panhandle notwithstanding that the facts in Panhandle are somewhat stronger than the facts in the instant case.
We agree with this assessment. The cases are factually distinguishable, see 852 F.2d at 893, and Calkins, The October 1989 Supreme Court Term and Antitrust: Power, Access, and Legitimacy, 59 Antitrust L.J. 339, 360-61 (1991), but the Court’s reasoning in UtiliCorp eliminates the significance of these distinctions. The Court made this point explicitly when it stated that it had granted certiorari in UtiliCorp “to resolve a conflict” between the 10th Circuit’s decision in that case (holding that the cost-plus exception was not available to residential consumers) and our en banc decision in this case (holding that it was). 110 S.Ct. at 2811.
In UtiliCorp, it was not clear what portion of the monopoly overcharge was in fact passed on to the consumers; in our case, we know that prices paid by consumers reflect 100% of the overcharge. The Court rejected our view that apportionment of the overcharge between the utility and its customers therefore presents no difficulties, for two reasons. “First, an overcharge may injure a utility, apart from the question of lost business, even if the utility raises its rates to offset its increased costs_‘if a cost rise is merely the occasion for a price increase a businessman could have imposed absent the rise in his costs, the fact that he was earlier not enjoying the benefits of the higher price should not permit the supplier who charges an unlawful price to take those benefits from him without being liable for damages.’ ” Id. at 2813 (quoting Hanover Shoe, 392 U.S. at 493 n. 9, 88 S.Ct. at 2231 n. 9). We recognized this possibility in our opinion, but concluded that “the doubts here are too small to warrant our insisting that this potentially serious antitrust violation ... shall go unremedied_” 852 F.2d at 898. The Supreme Court disagreed. It acknowledged that regulators *1478may prevent a utility from raising rates in the absence of increased costs, but concluded that “state regulation does not simplify the problem but instead imports an additional level of complexity. To decide whether a utility has borne an overcharge, a court would have to consider not only the extent to which market conditions would have allowed the utility to raise its rates prior to the overcharge, as in the case of an unregulated business, but also what the state regulators would have allowed.” Id. 110 S.Ct. at 2813. Second, although acknowledging that many gas utilities can, through the use of PGAs, quickly recoup their costs, the Court observed that even the minimal delays incurred in recouping overpayment from the indirect purchasers works to the detriment of the direct purchaser. “During any period in which a utility’s costs rise before it may adjust its rates, the utility will bear the costs in the form of lower earnings.” Id. at 2814.
Both of these rationales, of course, apply with equal force whether or not there is in force an explicit cost-plus contract between purchaser and indirect purchaser. That fact makes the state’s reliance on that distinction futile, and indeed, raises questions about the viability of the cost-plus exception in any context. See Calkins, supra, at 363 n. 135. The “broader point” made by the Court that, “even assuming that any economic assumptions underlying the Illinois Brick rule might be disproved in a specific case, we think it an unwarranted and counterproductive exercise to litigate a series of exceptions,” id. at 2817, underscores the scope of its ruling in UtiliCorp. The Court’s interpretation of the cost-plus exception appears so narrow (setting up as it does a demand for rigorous proof of a 100% pass through and then suggesting an unwillingness to consider such detailed evidence) as to preclude its application in any case; it seems particularly unlikely, then, that the Court intended to permit application of the exception by the indirect purchasers of a natural gas supplier to turn on subtle factual distinctions among the supplier’s indirect customers.
The state’s assertion that UtiliCorp should not be applied retroactively to bar the indirect purchaser claims in this case has more plausibility than does its claim that UtiliCorp would not control the disposition of those claims if applied. The “threshold test,” United States v. Johnson, 457 U.S. 537, 550 n. 12, 102 S.Ct. 2579, 2587 n. 12, 73 L.Ed.2d 202 (1982), in determining whether a decision should be applied retroactively is whether it establishes “a new principle of law, either by overruling clear past precedent on which litigants may have relied ... or by deciding an issue of first impression whose resolution was not clearly foreshadowed....” Chevron Oil Co. v. Huson, 404 U.S. 97, 106, 92 S.Ct. 349, 355, 30 L.Ed.2d 296 (1971). If Utilcorp eliminated the cost-plus exception sub silentio, there may be something to the state’s assertion that the case should not be applied retroactively.
To adopt the state’s reasoning, however, we would have to ignore the Court’s position that Utilicorp faithfully interprets Hanover Shoe and Illinois Brick. 110 S.Ct. at 2813 & 2817. The Court said its decision was not new law, and we are therefore unwilling to disagree, particularly when getting the state over the “new law” threshold does nothing to prevent the door from slamming shut on the state’s claims anyway. It is not enough that a decision announces a new rule of law; it must also, among other things, work some inequity on the party against whom it is applied. Chevron, 404 U.S. at 107, 92 S.Ct. at 355. Applying Utilicorp, however, works no hardship on the state of Illinois (other than the adverse ruling it requires). Unlike the plaintiff in Chevron, the state has not forgone any legal rights in reliance on its view that its claims were not barred by the Illinois Brick doctrine; the applicability of the doctrine has been contested from the outset of the case. Indeed, “retroactive application” is something of a misnomer here, since this case has not been fully adjudicated. Our decision to permit the state’s case to go forward is the law of the case, but that law is subject to change if it is determined to be clearly erroneous and would work a substantial injustice if *1479uncorrected. Weidner v. Thieret, 932 F.2d 626, 631 (7th Cir.1991). In light of the Court’s decision in UtiliCorp, our previous decision cannot stand. Federal claims are to be decided in accordance with the law existing at the time of decision, Goodman v. Lukens Steel Co., 482 U.S. 656, 662-63, 107 S.Ct. 2617, 2621-22, 96 L.Ed.2d 572 (1987), and UtiliCorp bars the indirect purchaser claims in this case. The state claims that it would be inequitable to apply UtiliCorp at this late stage in the litigation, but cannot provide a convincing reason why. The district court conducted the trial on the merits while the interlocutory appeal on the Illinois Brick issue was pending before this court; the state would thus have incurred virtually the entire costs of litigating its case regardless of the final determination as to the resolution of the Illinois Brick issue. Changing the ruling on that issue at this time works no injustice to the state; it merely results in the dismissal of its Sherman Act claims,7 claims that the district court denied on the merits anyway.
All of which is not to say that UtiliCorp mandates the dismissal of the state’s entire case. The state’s complaint contains pendent state law counts paralleling each federal antitrust violation alleged, and Illinois law explicitly permits indirect purchasers to sue under the state’s antitrust laws to recover monopolistic overcharges passed on to them. See Ill.Rev.Stat. ch. 38, § 60-7(2). In California v. ARC America, 490 U.S. 93, 109 S.Ct. 1661, 104 L.Ed.2d 86 (1989), the Supreme Court ruled that the Illinois Brick rule does not bar indirect purchasers from recovering damages flowing from violations of state antitrust law when, as here, there is an express state statutory provision giving such purchasers a cause of action. Panhandle does not contest the relevance of ARC America. Rather, it resurrects an argument rejected by the district court, namely, that Illinois’ state antitrust law is preempted by the extensive federal regulation of the natural gas industry and that the law violates the commerce clause.
The state maintains that Panhandle waived its preemption claim by failing to raise it in its original brief. Preemption, the state maintains, has nothing to do with the applicability of Utilicorp, the issue on which this Court ordered supplementary briefing. We disagree. Prior to Utilicorp, the independent viability of the Illinois Antitrust Act claims was not at issue in this appeal, and Panhandle cannot be faulted for not addressing an irrelevant issue in its original brief. After the Court decided Utilicorp, however, those claims became critical, a fact the state’s initial supplementary brief made plain. Panhandle’s preemption theory was a direct response to the state’s claim that notwithstanding Utilicorp ’s affect on the federal claims, the state law claims survived, and was both timely and relevant.
That being said, we do not agree with Panhandle’s view that the state law claims are preempted by federal regulation of the natural gas industry. The arguments Panhandle puts forth for preemption would apply with equal force to federal antitrust law, but federal gas regulation does not immunize natural gas companies from application of the federal antitrust laws. California v. Federal Power Comm’n, 369 U.S. 482, 82 S.Ct. 901, 8 L.Ed.2d 54 (1962). When state antitrust law only mirrors federal antitrust law, there is no reason to conclude that Congress intended to preempt the state law. When the two antitrust regimes differ, federal regulation that does not preempt federal antitrust law may preempt state antitrust law, see, e.g., Connell Constr. Co. v. Plumbers & Steamfitters Local Union No. 100, 421 U.S. 616, 632, 635-37, 95 S.Ct. 1830, 1840, 1841-42, 44 L.Ed.2d 418 (1975), but we face no divergent antitrust aims in this case. Sections 3(3) and 3(4) of the Illinois Antitrust Act, on which the state law claims in this case are based, were modeled after sections 2 and 1 of the Sher*1480man Act, respectively, and Illinois law provides that its courts should use the construction of federal antitrust law by federal courts to guide their construction of those state antitrust laws that are substantially similar to federal antitrust law. III. Rev.Stat. ch. 38, § 60-11. Illinois courts therefore look to federal law when construing these provisions. Collins v. Associated Pathologists, Ltd., 844 F.2d 473, 481-82 (7th Cir.1988); People v. College Hills Corp., 91 Ill.2d 138, 150, 61 Ill.Dec. 766, 435 N.E.2d 463 (1982). There is, therefore, no conflict between state and federal antitrust law to create a preemption question. Illinois law does permit indirect purchaser suits while federal law does not, but that difference reflects different judgments about the feasibility of trying such claims and the potential danger of duplicative recoveries rather than different judgments about the implied immunity of the natural gas industry from the application of the substantive provisions of the federal and state antitrust laws which, in this case anyway, are identical. Illinois’ pendent state law claims therefore survive application of UtiliCorp and we must reach the merits of the state’s appeal of the dismissal of those claims.
III. Panhandle’s Conduct
Despite its complex regulatory backdrop, this is a straightforward case. In force between Panhandle and its G tariff customers, like CILCO, was an exclusive dealing contract, approved originally by the Federal Power Commission (FERC’s predecessor) in 1951, that required those customers to purchase all of their natural gas requirements from Panhandle. In exchange for this sole supplier provision, Panhandle incurred an obligation to use its best efforts to meet its customers’ supply requirements. To satisfy that obligation, Panhandle entered into a number of long-term contracts to secure gas for the future. When Congress deregulated wellhead prices, however, a market that historically had been characterized by chronic shortages quickly found itself awash in natural gas; spot market prices soon fell well below the level at which many pipelines, including Panhandle, had contracted to purchase gas. At that point many LDCs, like CILCO, balked at paying above market rates for their gas and sought to escape their contractual obligations under the G tariff by demanding that Panhandle transport gas they wanted to purchase from other sources. At the same time, these customers wanted to hold Panhandle to its obligation to supply their contract demand quantities, should they desire to purchase them. In the words of the district court, “CILCO wanted to have its cake and eat it too.” 8 730 F.Supp. at 886. Panhandle, obligated by its own purchase contracts to buy expensive gas, refused these demands and tried to keep its G tariff in force. The question presented in this case is simply whether Panhandle’s efforts to maintain its G tariff in the face of the regulatory changes sweeping the industry violated the antitrust laws.9 In other words, did Panhandle violate the antitrust laws when it wouldn’t give CILCO another slice?
The district court said no. It found that Panhandle did possess monopoly power in the relevant market, power that was not effectively constrained by regulation.10 Nevertheless, the district court concluded that Panhandle’s efforts to maintain its G tariff did not violate the antitrust laws because it found that Panhandle’s refusal to transport gas for G tariff customers was not an effort to maintain a monopoly in sales of natural gas. It was, the district court found, merely a “lawful refusal to *1481cut its own throat.” 11 730 F.Supp. at 883; see also id. at 915-22.
The district court, like many courts addressing monopolization claims, spoke in terms of Panhandle’s “intent” to monopolize, leading the parties to debate whether “intent” is an element of the offense of monopolization. “Intent” is relevant to the offense of monopolization. Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585, 602, 105 S.Ct. 2847, 2857, 86 L.Ed.2d 467 (1985). But we have to be clear about what is meant by “intent,” for in the context of a monopolization ease “intent” is an elusive concept. The “intent” to achieve or maintain a monopoly is no more unlawful than the possession of a monopoly. Indeed, the goal of any profit-maximizing firm is to obtain a monopoly by capturing an ever increasing share of the market. Virtually all business behavior is designed to enable firms to raise their prices above the level that would exist in a perfectly competitive market. Economic rent — the profit earned in excess of the return a perfectly competitive market would yield — provides the incentive for firms to engage in and assume the risk of business activity. Monopolies achieved through superior skill are no less intentional than those achieved by anticompetitive means (as Learned Hand observed, “no monopolist monopolizes unconscious of what he is doing” 12), so the intent relevant to a § 2 Sherman Act claim is only the intent to maintain or achieve monopoly power by anticompetitive means. Section 2 forbids not the intentional pursuit of monopoly power but the employment of unjustifiable means to gain that power. 3 P. Areeda & D. Turner, Antitrust Law ¶ 626c at 76 (1978); see also Olympia Equip. Leasing Co. v. Western Union Tel. Co., 797 F.2d 370, 373 (7th Cir.), cert. denied, 480 U.S. 934, 107 S.Ct. 1574, 94 L.Ed.2d 765 (1986) (The offense of monopolization requires proof of “conduct designed to maintain or enhance [monopoly power] improperly”).
When courts consider the “intent” of a firm charged with monopolization, they look not to whether the firm intended to achieve or maintain a monopoly, but to whether the underlying purpose of the firm’s conduct was to enable the firm to compete more effectively. Did the firm engage in the challenged conduct for a legitimate business reason? Or was the firm’s conduct designed solely to insulate the firm from competitive pressure? Intent is relevant, then, because intent determines “whether the challenged conduct is fairly characterized as ‘exclusionary’ or ‘anticompetitive.’ ” Aspen Skiing, 472 U.S. at 602, 105 S.Ct. at 2857. When courts speak of a firm’s intent in a monopolization case, they refer to the legitimacy of the firm’s conduct as measured by its intended effect on the competitive process. In United States v. Grinnell Corp., 384 U.S. 563, 570, 86 S.Ct. 1698, 1704, 16 L.Ed.2d 778 (1966), for example, when the Court stated that one of the elements of monopolization is “the willful acquisition or maintenance of [monopoly] power,” it went on to define “willful acquisition or maintenance” by distinguishing it from “growth or development as a consequence of a superior product, business acumen, or historic accident.” Id.; see also Aspen Skiing 472 U.S. at 608 n. 39, 105 S.Ct. at 2860 n. 39 (“ ‘the law can usefully attack [exclusionary conduct] only when there is evidence of specific intent to drive others from the market by means other than superior efficiency.’ ”) (emphasis added) (quoting R. Bork, The Antitrust Paradox 157 (1978)). Conduct that tends to exclude competitors may therefore survive antitrust scrutiny if the exclusion is the product of a “normal business purpose,” Aspen Skiing, 472 U.S. at 608-10, 105 S.Ct. at 2860-61, for the presence of a legitimate business justification reduces the likelihood that the conduct *1482will produce undesirable effects on the competitive process. Id. at 608 n. 39, 105 S.Ct. at 2860 n. 39 (“Proof of specific intent to engage in predation may be in the form of ... evidence that the conduct was not related to any apparent efficiency.”); Olympia Equipment, 797 F.2d at 378. Whether valid business reasons motivated a monopolist’s conduct is a question of fact, Aspen Skiing, 472 U.S. at 604-05,105 S.Ct. at 2858-59; our task, like the Court’s in Aspen Skiing, is simply to determine whether the trial court’s finding that Panhandle’s actions were motivated by a justifiable business purpose is defensible.
The state attempts to read “intent” out of the monopolization equation by resurrecting a notion long discarded in antitrust law, namely, that antitrust laws exist to protect competitors. In the state's view, “[mjaintenance of monopoly power ... is ‘wilful’ ... whenever the conduct of the defendant has caused anticompetitive or injurious effects to the defendant’s competitors and ultimately to consumers_” Appellants’ Reply Brief at 7. The standard aphorism is that antitrust law protects competition and not competitors, Copperweld Corp. v. Independence Tube Corp., 467 U.S. 752, 767 n. 14, 104 S.Ct. 2731, 2740 n. 14, 81 L.Ed.2d 628 (1984); as we reformulated it in Olympia Equipment, “the emphasis of antitrust policy [has] shifted from the protection of competition as a process of rivalry to the protection of competition as a means of promoting economic efficiency.” 797 F.2d at 375. In Olympia Equipment, we reversed a jury verdict that Western Union had monopolized the market for telex terminal equipment on the ground that it had no duty to promote sales of its competitors’ products. Its actions were reasonable, we held, because there was no evidence to suggest that they were part of a scheme designed to exclude competitors from that market; rather, they were motivated only by Western Union’s desire “to liquidate its supply of telex terminals faster.” 797 F.2d at 378. That type of conduct, we said, is not “objectively anticompetitive,” id. at 380, because it was an objectively reasonable business decision. The state’s approach would render business justifications and efficiency considerations irrelevant, an approach we rejected in Olympia Equipment and the Supreme Court rejected in Aspen Skiing. See Easterbrook, On Identifying Exclusionary Conduct, 61 Notre Dame L.Rev. 972, 975 (1986) (“[T]he Court’s position [in Aspen Skiing ] resolves to a conclusion that a dominant firm that imposes large costs on its rival must have a good business justification (one consistent with ‘efficiency’)....”).
The state claims, however, that Panhandle’s reasons for refusing to transport nonsystem gas are irrelevant because Panhandle’s pipelines are “essential facilities.” And indeed, “[s]ome cases hold ... that a firm which controls a facility essential to its competitors may be guilty of monopolization if it refuses to allow them access to the facility.” Olympia Equipment, 797 F.2d at 376. In essential facilities cases, however, liability hinges on the feasibility of competitors developing competing facilities and of the owner providing access to the facility. MCI Communications Corp. v. American Tel. & Tel. Co., 708 F.2d 1081, 1132-33 (7th Cir.1983), cert. denied, 464 U.S. 891, 104 S.Ct. 234, 78 L.Ed.2d 226 (1983). In this case, neither condition is satisfied. The first condition goes to whether access to the facility is, in fact, essential to compete with the monopolist. In this case, access to Panhandle’s pipeline was not essential. The district court found that it would have been economically feasible for competitors to duplicate much of Panhandle’s system within central Illinois by means of interconnections between competing pipelines and the construction of new pipelines. 730 F.Supp. at 928. That finding was not clearly erroneous, as evidenced by the fact that by 1987 Panhandle faced substantial competition within central Illinois from other pipelines for transportation of natural gas to industrial end-users. Id. at 872-73. The primary barrier to entry into the central Illinois natural gas market through the period of this litigation was not capital but the Panhandle’s G tariff, which effectively deterred LDCs from dealing with other *1483pipelines and from buying gas directly from producers.
The state’s essential facility claim also fails because to be liable for monopolizing an essential facility, providing access to the facility must have been feasible for the owner. This second prerequisite to essential facility liability suggests that essential facilities eases are no different conceptually than cases involving other monopolization theories, because it reintroduces “intent” (a.k.a. “business justification”) back into the monopolization equation and excuses refusals to provide access justified by the owner’s legitimate business concerns. See MCI, 708 F.2d at 1133 (evidence supported jury finding that AT & T could have feasibly provided FX and CCSA interconnections as “no legitimate business or technical reason was shown for AT & T’s denial of the requested interconnections”). Panhandle had such concerns — in spades. According to the district court, Panhandle faced over $4 billion in potential take-or-pay liability. 730 F.Supp. at 862 (combined exposure of Panhandle and Trunkline). The state disputed Panhandle’s actual exposure to take-or-pay liability under its long-term contracts, presenting evidence that such claims had been settled, on average, for ten cents on the dollar. But as the court observed, even at that level, Panhandle’s potential liability was extremely large, and there was no guarantee that Panhandle would be able to limit its exposure to that extent in the future.
The district court found that Panhandle’s intransigence regarding the G tariff was genuinely and reasonably motivated by the need to limit its potential take-or-pay liability, not by a desire to maintain its monopoly position by excluding competition in the sale of natural gas. In the district court’s view that concern, when coupled with the regulatory flux the natural gas industry was undergoing at the time, was sufficient to negate the possibility that Panhandle was motivated by anticompeti-tive intent.13 That finding is one of fact to which we must defer since there was ample evidence to support it. Aspen Skiing, 472 U.S. at 611, 105 S.Ct. at 2861-62.
In any event, it is one with which we agree. “A plausible response to the gas shortages of the 1970s, [take-or-pay clauses have] created significant dislocations in light of the oversupply of gas that has occurred since. Today many purchasers face disastrous take-or-pay liability without sufficient outlets to recoup their losses.” Mobile Oil Exploration, 111 S.Ct. at 627. Panhandle assumed its contractual obligations knowing that its LDC customers would buy all the gas they needed from Panhandle; the G tariff thus mitigated some of the risks posed by the take-or-pay provisions. Under “the pattern of contractual arrangements developed by the industry during previous eras of pipeline expansion and supply curtailment, such as take- or-pay ... provisions, customers of pipelines have been allocated a large part of the risk that gas supplies deregulated by the NGPA might prove unmarketable over the life of contracts signed by producers and pipelines.” FERC Order 436, 48 Fed. Reg. at 42421. What the state labels “monopolization” was nothing more than the enforcement of legitimate contracts de*1484signed to allocate risk between Panhandle and its customers; what the state asks us to do is reallocate those risks. We decline the invitation. Panhandle had incurred obligations itself in reliance on the G tariff and to satisfy its regulatory obligations to anticipate and meet future customer demand. City of Mishawaka v. American Elec. Power Co., 616 F.2d 976, 985 (7th Cir.1980), cert. denied, 449 U.S. 1096, 101 S.Ct. 892, 66 L.Ed.2d 824 (1981).14 CILCO and Panhandle’s other LDC customers were, in turn, obligated to buy their full requirements for gas from Panhandle. We do not believe that it was “anticompetitive” for Panhandle to hold them to that deal.
Monopolists needn’t acquiesce to every demand placed upon them by competitors or customers; a monopolist’s duties are negative — to refrain from anti-competitive conduct — rather than affirmative — to promote competition. Olympia Equipment, 797 F.2d at 375-76. Just as the monopolist has no duty to deter the sale of its own equipment by promoting that of a competitor, id., so too it has no duty to incur contractual liability itself by excusing its customers from their contractual obligations. We recognized the sufficiency of exactly that type of “self-serving” business justification in Olympia Equipment, where we observed that “[consumers would be worse off if a firm with monopoly power had a duty to extend positive assistance to new entrants, or having extended it voluntarily a duty to continue it indefinitely. The imposition of such a duty would make firms that possessed or might be thought to possess monopoly power, however laudably obtained, timid about ... competing with new entrants.” 797 F.2d at 379. By the same token, consumers would be worse off if a firm had a duty under the antitrust laws to release customers from their contractual obligations; it is anything but efficient for a firm to abandon its contractual rights at the behest of customers who are no longer happy with their bargain, even when consumers might be better off (at least in the short run) if they did so. Imposing that type of affirmative obligation on a monopolist — whether explicitly or by refusing to acknowledge the legitimacy of such refusals — would penalize the monopolist for refusing to surrender a lawfully obtained monopoly, a result courts have long foresworn.
To cite just one example, we addressed quite similar claims in MCI, when MCI claimed that AT & T violated the antitrust laws by refusing to grant it access to interconnections that would have given it access to AT & T’s entire nationwide long-distance network. In MCI the FCC, like the FERC in this case, appeared to be opening the telecommunications industry to competition, but had not directed AT & T to provide access to its long-distance network to competitors. We concluded that AT & T’s refusal to voluntarily assume “the extraordinary obligation to fill in the gaps in its competitor’s network,” id. at 1149, did not suffice to support a finding that it was trying to maintain its monopoly of long-distance telephone service by anticompetitive means. “[Gjiven the unsettled regulatory status of the telecommunications industry at the time of these events,” we held that MCI had failed to present sufficient evi*1485dence “to permit a finding that AT & T’s denial of interconnection for multipoint service was primarily motivated by an illegal intent to monopolize.” Id.
To support its position the state, not surprisingly, points to cases in which courts have held that a monopolist’s actions evinced an attempt to exclude competition from the market rather than an attempt to minimize costs. The cases it cites are factually distinguishable, however, and do nothing to undermine the validity of the district court’s factual finding in this case. We need not address every case in which courts have found a monopolist’s actions were animated by anticompetitive intent to make the point; the state relies most heavily on Otter Tail Power Co. v. United States, 410 U.S. 366, 93 S.Ct. 1022, 35 L.Ed.2d 359 (1973), and it is enough to point out the distinctions between that case and this one. In Otter Tail, the Court held that an electricity producer’s refusal to sell or “wheel” (transport) electricity to towns that sought to end their reliance on its retail electric distribution system by establishing their own systems constituted illegal monopolization of the retail sale and distribution of electricity. The state says that Panhandle has done the same thing by refusing customer requests to transport gas purchased from other sources. Critical to the Court’s analysis in Otter Tail, however, was the fact that Otter Tail’s franchise contracts with the towns had expired; they were no longer contractually bound to use Otter Tail’s distribution facilities. Panhandle’s G tariff customers, by contrast, were contractually obligated to purchase their gas requirements from Panhandle, and Panhandle had itself entered into contractual obligations with gas producers in reliance on the demand anticipated from its customers.15
These distinctions make all the difference. Otter Tail’s business justification was simply that it could not compete if forced to abandon practices designed to eliminate competition in the market for retail electrical transmission; the district court therefore found, and the Supreme Court agreed, that “Otter Tail’s refusals to sell at wholesale or to wheel were solely to prevent municipal power systems from eroding its monopolistic position.” 410 U.S. at 378, 93 S.Ct. at 1030. Panhandle, by contrast, does not claim that it cannot compete on a level playing field; it merely objects that it is unreasonable to expect it to permit its customers to avoid their contractual obligations when to do so would be to expose itself to enormous take-or-pay obligations. As the district court correctly observed, Otter Tail “does not stand for the proposition that a utility must renegotiate extant long-term service agreements to enable a customer to supplant the utility as its sole supplier.” 730 F.Supp. at 909 (emphasis in original).
Despite the dire predictions of the state, this does not mean that there now exists a “contract immunity” defense to antitrust liability. The existence of a contract in this case does not immunize Panhandle from antitrust liability; it is merely a factor that is relevant to the question of Panhandle’s intent to monopolize. The existence of a contract that was itself an unreasonable restraint of trade, violating § 1 of the Sherman Act, would do little to dispel an inference of anticompetitive intent. In Otter Tail, for example, the utility attempted to invoke contractual provisions in its contracts with other suppliers that forbade the suppliers from providing electricity to any of the utility’s retail customers, past or present. That provision, as the Supreme Court observed, was simply a territorial *1486allocation scheme designed to insulate the utility from competition in the sale of electricity and had no legitimate justification. 410 U.S. at 378-79, 93 S.Ct. at 1030. Panhandle’s exclusive dealing contract with its G tariff customer, by contrast, was a legitimate means of ensuring that it would not be stuck holding expensive natural gas for customers who had decided to purchase unexpectedly plentiful and cheap gas from others, one that had been given regulatory sanction. Contrary to the state’s suggestion, when Congress enacted the NGPA, Panhandle’s tariffs did not become invalid or illegal. Recognizing the obligations Panhandle incurred in reliance on the tariffs does not elevate a private contract above national policy as the state suggests.
The state has its own theory about Panhandle’s motives, but its conjecture does little to make us question the soundness of the district court’s findings. According to the state, Panhandle refused to adopt an open access transportation policy because it wanted to exact monopoly profits from the gas it sold to its G tariff customers. It did so, according to the state, by tying the purchase of its monopolistically priced gas to the purchase of its regulated pipeline capacity and by unlawfully segmenting the central Illinois natural gas market and price discriminating between gas consumers who were able to switch to an alternate fuel and those who did not.
Panhandle, however, didn’t profit on its sales of gas to the LDCs. Panhandle’s gas was priced above the spot market, but that price merely reflected the price it was paying for gas as the result of the long-term contracts it agreed to in order to secure gas that was both high-priced and scarce during the early days of deregulation. Panhandle’s rate of return was based on its transportation service, not its gas prices, a fact that suggests that absent a fear of take-or-pay liability it would have had little reason to object to transporting gas purchased from other sources. The state’s brief acknowledges this point but, inexplicably, goes on to rail against “the profits of Panhandle and its subsidiaries on gas sales.” Brief of Appellant at 35. The inconsistency is explained later, when the state reveals that what it calls “profits” on the sale of gas are “more accurately” characterized not as profits but as losses avoided. Brief of Appellant at 39. Translated, the state’s theory is simply that Panhandle’s desire to avoid take-or-pay liability constituted an antitrust violation because Panhandle enforced the G tariff rather than reducing its rate of return by recouping less than 100% of its gas prices. Panhandle was entitled to pass through 100% of the cost of its gas to its customers, however; it had no duty to voluntarily reduce its rate of return below the “just and reasonable” level authorized by regulators. Cf. Town of Concord v. Boston Edison Co., 915 F.2d 17, 27 (1st Cir.1990). This is not to say, of course, that a utility can engage in anticompetitive conduct in order to increase its earnings to the authorized level. Nor do we say that there can never be a case in which a utility’s refusal to voluntarily take action that would reduce its profit margin is anticompetitive. The plaintiff in that case, however, will have to present a more plausible theory than Illinois has presented here.
The state points out that Panhandle was vertically integrated, which meant that it might have been able to force consumers to pay a supracompetitive price for gas by purchasing gas at above market rates from affiliated producers, but there is no evidence that this was the reason that its costs were high. The evidence suggested that its high costs were due principally to its Algerian and Canadian ventures, neither of which were with affiliated producers. True, Panhandle bought the liquified Algerian gas from an affiliated pipeline, Trunkline, but self-dealing is a danger when a regulated company and an unregulated company are vertically integrated, see Jefferson Parish Hosp. Dist. No. 2 v. Hyde, 466 U.S. 2, 36 n. 4, 104 S.Ct. 1551, 1571 n. 4, 80 L.Ed.2d 2 (1984) (O’Connor, J., concurring), not when two regulated companies are affiliated horizontally. Trunk-line, like Panhandle, merely passed on the above market rate it paid to unaffiliated gas producers.
*1487But what of Panhandle’s willingness to transport for its noncaptive customers? By mollifying them, the state maintains, Panhandle engaged in “price discrimination” and “market segmentation,” facilitating its ability to charge supracompetitive prices for the gas it sold to captive customers and thereby perfecting its monopoly over those customers. This is exactly the argument raised by petitioners when they challenged FERC Orders 234 and 319 in MPCII, see 761 F.2d at 784, and it succeeded there, but there are several reasons why it fails here.
First, we should be clear about the state’s complaint. The discrimination it objects to related not to the price of gas Panhandle sold to consumers who could switch between gas and other fuels (producers, not Panhandle, sold gas at the lower spot market rate), but to the discriminatory access Panhandle gave those consumers to cheaper sources of gas by agreeing to transport it. In this respect, the state’s theory merely restates its claim, discussed above, that the G tariff did not preclude LDCs from purchasing gas directly from producers. See supra note 2. The state maintains that Panhandle selectively applied its interpretation of the G tariff — that the tariff applied to direct sales from producers to consumers — to captive customers, but fails to explain that the end-users who obtained transportation for nonsystem gas were not themselves G tariff customers, and were under no contractual obligation to Panhandle. Of course, neither were the captive residential and industrial consumers to whom the LDCs distributed gas, but those consumers didn’t purchase gas directly from the wellhead. The district court found that the fuel-switchable end-users eligible for the MAT program did, and the state points to no contrary evidence. The captive residential and commercial LDC customers could, in theory, have purchased gas from producers directly, but most LDCs, including CILCO, “had transportation tariffs which either expressly precluded transportation services for residential end-users or effectively precluded transportation for residential end-users by imposing a volumetric limitation.... In addition ... most producers and brokers were unwilling to enter into contracts for small volumes of gas.” 730 F.Supp. at 890. We therefore agree with the district court’s conclusion that Panhandle did not selectively enforce the G tariff; “the ‘discrimination’ apparent in Panhandle’s transportation policy was a legitimate enforcement of that G tariff” against those who were bound by it. 730 F.Supp. at 921.
After FERC issued Order 436, discrimination on the basis of sole supplier clauses was no longer legitimate; the FERC order required pipelines offering transportation to make the option available to all customers, regardless of the existence of full requirements or sole supplier clauses in their gas purchase contracts. See 50 Fed.Reg. 42445. Rather than comply, Panhandle initially shut down its MAT program, and only resumed it after its G tariff customers agreed not to request unbundled transportation services. The G tariff customers agreed to this condition because they, too, had a stake in keeping fuel-switchable industrial consumers on line; keeping the industrials on line helped spread the fixed cost component of Panhandle’s rates among a wider customer base, and helped support their own revenues by maintaining high through-put volumes to these end-users (the LDCs, like Panhandle, were effectively selling transporting services to these customers). This agreement did not, as the state suggests, violate the terms of Order 436, for the Order also required “full requirements” customers to switch to a partial requirements tariff to obtain transportation, recognizing that “[tjhere can be differences in the costs of providing full and partial requirements service.” 50 Fed. Reg. at 42445; see also FERC Order 436-A, 50 Fed.Reg. 52217 (1985) (reiterating requirement that full requirements customers switch to partial requirements tariff to receive pipeline transportation services). Panhandle’s G tariff customers thus had the option to obtain transportation by switching to a partial requirements tariff, but were unwilling to give up the security of the G tariff to do so; Panhandle therefore had no obligation to transport for *1488them. These events effectively demonstrate that if the state (and the residential consumers it represents) have a quarrel with a utility, it should be with CILCO and other LDCs rather than with Panhandle. Faced with a choice of obtaining access to low-priced gas supplies or giving up stable gas supplies, CILCO and other LDCs opted for the latter.
The second reason the state’s price discrimination theory fails is that, as noted above, there is no evidence suggesting that self-dealing was the cause of Panhandle’s high gas prices. The real culprits were long-term supply contracts. When the self-dealing charge is deflated, the state’s price discrimination theory collapses as well because Panhandle had no monopoly profits to hide. Panhandle undoubtedly wanted to pass on the full amount of its gas costs, but that is a far cry from extracting monopoly profits. The state’s theory ignores the fact that, under its PanMark program, Panhandle received take-or-pay credit from producers for volumes its fuel-switchable customers purchased from them directly. Panhandle did not always receive take-or-pay credit for the gas transported under the MAT program (although the MAT program did yield over $50 million in take-or-pay credits), but that program too was designed to help mitigate the problems created by the discrepancy between the spot market price of natural gas and the price Panhandle was contractually obligated to pay. PanMark enabled Panhandle to recover its gas costs by giving it take-or-pay credits for gas sold at low spot market prices, and MAT enabled Panhandle to obtain some take-or-pay relief by keeping large industrial end-users from switching, or converting, to other fuels. FERC may or may not have adequately justified its reasons for approving such programs, see MPC I, 761 F.2d at 774, but that fact is not relevant to the issue of whether Panhandle’s actions under the FERC programs constituted an unlawful exercise of monopoly power. As unbundled transportation became the norm in the industry, the FERC programs were the principal means available to Panhandle for resolving its take-or-pay dilemma. Panhandle’s implementation of these programs reinforces the conclusion that it was the discrepancy between spot market and contract prices for gas, rather than exclusionary animus, that drove Panhandle’s policies. Had Panhandle’s goal been to exclude other sellers from central Illinois, it would not have transported gas under any program, whether or not it provided take-or-pay credit.
IV. Conclusion
This case is essentially a dispute about who should bear the cost of the transformation of the natural gas industry from a regulatory to a competitive regime. Panhandle refused to transport natural gas for its G tariff customers out of concern for its take-or-pay exposure. The state maintains that enforcing the G tariff was anticompet-itive because it was at odds with the changes wrought by enactment of the NGPA and FERC’s moves to give consumers access to a competitive gas market. FERC's reluctance to jump with both feet into an open access transportation policy, however, rebuts the state’s claim that the FERC’s initial sallies in that direction stripped the G tariff of its mantle of regulatory sanction. Panhandle had to respond to those changes mandated by law and by regulation, but was unwilling to go further than required because to do so would have been to expose itself to huge losses. Panhandle abided by the terms of FERC’s transportation initiatives, and relied on them in good faith, a fact that, while not rising to the level of a regulatory justification defense (the FERC did not require pipelines to participate in the programs), leads us to agree with the district court that Panhandle’s programs were the product of legitimate business concerns and not a naked desire to deny natural gas producers access to the central Illinois market. FERC itself was reluctant to move ahead too quickly; it didn’t require pipelines offering unbundled transportation to do so on a nondiscriminatory basis until it adopted Order 436 in late 1985, and that Order was later vacated because it did not adequately address the dilemmas faced by pipelines *1489like Panhandle. None of FERC’s attempts to manage the deregulatory transition have completely satisfied the courts; it is hardly reasonable to expect that Panhandle should have jumped on the open access bandwagon after FERC’s initial, tentative, moves to get that wagon rolling. The district court attributed Panhandle’s reserve in the face of regulatory flux to caution and self-preservation rather than to monopolistic excess, a determination we find eminently reasonable. The decision of the district court is therefore
AFFIRMED.
. We have limited our exposition of the facts to those that are essential to an understanding of the issues raised on appeal. For the complete story, the reader is referred to the district court’s comprehensive opinion at 730 F.Supp. 826.
. "A take-or-pay clause allocates part of the volume risk to the purchaser — the pipeline company — by obligating the purchaser to pay for a specified quantity of gas whether or not the purchaser actually takes delivery of that gas.” Pierce, supra, at 355.
. • Panhandle’s gas purchases during this period have thus far survived regulatory and judicial review. See Office of Consumer's Counsel v. FERC, 914 F.2d 290 (D.C.Cir.1990) (upholding FERC determination that Panhandle’s purchases from gas producers were not abusive; remanding for further fact-finding on issue of whether Panhandle’s purchases from Trunkline were prudent).
. Section 1.9 of the General Terms and Conditions applicable to Panhandle’s G-2 tariff reads:
1.9 General Service Buyer. General Service Buyer is any buyer which does not purchase gas from any other natural-gas company, as defined in the Natural Gas Act [NGA], for distribution in areas served with Seller’s gas; provided, however, a Buyer under the General Service Rate Schedule which seeks from Seller an increase in contract demand and Seller is unable to supply the increase in contract demand, then such Buyer may purchase natural gas from other natural-gas companies but shall remain a General Service Buyer under this Tariff.
See Appellant’s Supplemental Appendix at 347 (emphasis added).
The meaning of the phrase "as defined in the Natural Gas Act” is a bone of contention between the parties, one that could affect the outcome of the case. The state maintains that Panhandle’s characterization of the G tariff as a “sole supplier” contract is erroneous, and asserts therefore that the existence of the G tariff was no justification for Panhandle’s refusal to transport gas during the early 1980s. The state maintains that the G tariff did not restrict G tariff customers from purchasing "new gas,” that is, gas not committed or dedicated to interstate commerce as of November 8, 1978, see 15 U.S.C. § 3431(a)(1)(A) and (B), directly from producers because producers are not "natural gas companies” under the NGPA.
The NGPA didn’t change the definition of "natural-gas company” under the NGA, however; it merely rendered the sale of "new gas” an event over which the FERC is not entitled to exercise jurisdiction. Panhandle Eastern Pipeline Co., 38 FERC ¶ 63,009 at 65,052-53 (1987). Moreover, the state concedes that producers who sell natural gas for resale in interstate commerce were, before enactment of the NGPA in 1978, natural gas companies under the NGA (15 U.S.C. § 717a). Appellant’s Brief at 34. That is the only definition relevant to the interpretation of the G tariff as that is the definition incorporated by the tariff. It is true that in 1951, when the G tariff was first approved, gas producers were not considered to be "natural-gas companies.” But producers sold almost exclusively to pipelines, so the omission of producers in the 1951 tariff was hardly significant.
Moreover, the industry understanding changed with the Supreme Court's decision in Phillips Petroleum v. State of Wisconsin, 347 U.S. 672, 74 S.Ct. 794, 98 L.Ed. 1035 (1954), which held that producers were subject to regulation as natural-gas companies under the NGA. Id. at 677, 74 S.Ct. at 796. Customers who contracted to purchase gas under the G tariff after Phillips were on notice that the tariff precluded purchases from gas producers as well as from other pipelines. CILCO last renewed its G tariff with Panhandle in 1970, well after Phillips, and well before enactment of the NGPA.
It is possible, we suppose, to read the reference to the NGA definition in the G tariff as an *1474attempt to incorporate all future manifestations of the definition, but that is not the interpretation adopted by the parties to the contract. Panhandle’s customers also read the G tariff to require them to buy all their gas from Panhandle; CILCO took that position when it challenged the provision in its FERC complaint, after the NGPA modified the NGA definition of “natural-gas company.” See Panhandle Eastern Pipe Line Co. v. FERC (Panhandle I), 881 F.2d 1101, 1105 (D.C.Cir.1989) (per curiam). The administrative law judge who first ruled in the case, the FERC, and the D.C. Circuit each treated § 1.9 as requiring G tariff customers to purchase all of their requirements from Panhandle. See Panhandle Eastern Pipe Line Co., 32 FERC ¶ 63,085 at 65,321-26 (1985) (the CILCO Complaint case); Panhandle Eastern Pipe Line Co., 38 FERC ¶ 61,164 at 61,465 (1987) (Opinion 265); Panhandle I, 881 F.2d at 1107. Absent language affirmatively contemplating and adopting future changes in that definition, we agree that Panhandle’s interpretation of § 1.9 as a sole supplier provision was reasonable.
The state cites evidence suggesting that Panhandle did not really interpret § 1.9 to bar its customers from purchasing gas directly from producers, but the district court found that Panhandle maintained the position that the G tariff was a sole supplier tariff and that finding is not clearly erroneous. At any rate, whether Panhandle subjectively believed that the tariff was a sole supplier provision is largely irrelevant because the state concedes that concern about take-or-pay liability animated Panhandle’s refusal to transport “new gas” for its customers. The question in this case is simply whether that motivation was anticompetitive within the meaning of the antitrust laws.
. The subsequent history of FERC’s efforts to resolve the take-or-pay problem are not directly relevant to our case because all of Panhandle’s challenged conduct took place before Order 436 was vacated. Nevertheless, the years spent by FERC and the courts grappling with the issue are instructive, for they provide compelling evidence of the intractable nature of the problem. After Order 436 was vacated, FERC next issued Order 500, an interim rule, to temporarily address the take-or-pay problem while FERC conducted rulemaking proceedings to comply with *1476the concerns raised in AGD. That order, too, was remanded. American Gas Ass'n v. FERC, 888 F.2d 136 (D.C.Cir.1989); see also Associated Gas Distribs. v. FERC, 893 F.2d 349 (D.C.Cir.1989), cert. denied sub nom. Berkshire Gas Co. v. Associated Gas Distribs.,-U.S.-, 111 S.Ct. 277, 112 L.Ed.2d 232 (1990) (vacating FERC Orders promulgated under Order 500 provisions). With Order 500-7£ however, FERC finally resolved the take-or-pay problem satisfactorily (with one minor exception). See American Gas Ass’n v. FERC, 912 F.2d 1496 (D.C.Cir.1990).
. The state’s complaint alleges that the tie was an unlawful restraint of trade violating of § 1 of the Sherman Act, rather than an exclusionary practice violating § 2. In its brief on appeal, however, the state describes the tie as an exclusionary practice relevant to its § 2 monopolization claim, and, accordingly, that is how we treat it.
. We do not address the issue of whether Utili-Corp bars the state’s claims for injunctive relief, as we agree with Panhandle that such relief is unavailable because Panhandle now makes its transportation services available to its LDC customers.
.Alternatively, in the words of FERC:
Purchasers of natural gas, seeing the availability of supplies in the field at prices below the rolled-in average cost of all gas, are seeking to purchase gas directly in the field and have it transported to the city-gate or burner-tip in competition with the system supplies which the pipeline retains under certain uneconomic long-term contracts.
FERC Order 436, 48 Fed.Reg. at 42421.
. The state does not claim that the G tariff was originally violative of the antitrust laws.
. Panhandle disputes these findings, but our agreement with the district court’s finding that Panhandle lacked anticompetitive intent makes it unnecessary to address them.
. By contrast, the district court did find that Panhandle’s adoption of its “Transportation Guidelines” for industrial end-users was anti-competitive. Id. at 891-94, 921. The district court concluded, however, that Panhandle did not have monopoly power over those users, saving it from antitrust liability on that score as well. Id. at 888, 922.
. United States v. Aluminum Co. of America (“Alcoa”), 148 F.2d 416, 432 (2d Cir.1945).
. The district court rejected, however, Panhandle’s view that its concern about take-or-pay liability was sufficient to constitute a defense, adopting a more narrow view of the type of business justification that constitutes a defense. In the district court’s view, ”[i]t is not a legitimate business justification for antitrust purposes that the defendant sought to protect itself from added costs or lost profits.” 730 F.Supp. at 932-33.
We do not believe that the distinction among types of business justifications drawn by the district court is a viable one. The district court cited Otter Tail Power Co. v. United States, 410 U.S. 366, 380-81, 93 S.Ct. 1022, 1031, 35 L.Ed.2d 359 (1973), for the proposition that only measures that produce "superior service, lower costs, and improved efficiency" support a business justification defense, without considering that measures designed to avoid higher costs are essentially measures designed to lower costs. Whether the lack of business justification is viewed as an element of the offense or the presence of a business justification constitutes an affirmative defense goes to the allocation of burdens of proof. It says nothing about the type or quality of justification required, and we can see no reason for imposing a more demanding test in one case than the other.
. The state's claim that Panhandle’s status as a utility is irrelevant since it was not required by regulation to enforce its G tariff is misleading. Panhandle was required by the FERC, pursuant to the FERC’s authority under § 7(a) of the NGA, to extend service to many of its customers. Order 436, 50 Fed.Reg. at 42440. To serve all of its customers, Panhandle was required by regulation to anticipate their future requirements for natural gas. See, e.g., 18 C.F.R. § 2.61 (generally requiring pipelines to demonstrate ability to meet projected demand for next twelve years). Moreover, as the district court observed, "[o]nce a pipeline has commenced service ... it may not abandon or terminate that service unless it first obtains a certificate of abandonment from the FERC. A pipeline is required by FERC to continue providing service to an LDC customer even after the expiration of the service agreement with the customer.” 730 F.Supp. at 877. In any event, the state exaggerates the significance of a regulatory imperative. Whether Panhandle’s efforts to enforce the G tariff were required by regulation is not, as the state seems to suggest, dispositive of Panhandle’s liability. The existence of affirmative regulatory obligations is merely a factor to be considered in determining whether a utility's conduct was intentionally anticompetitive.
. Litton Systems, Inc. v. American Tel. & Tel., 700 F.2d 785 (2d Cir.1983), cert. denied, 464 U.S. 1073, 104 S.Ct. 984, 79 L.Ed.2d 220 (1984), is distinguishable for the same reason. In Litton, AT & T had filed a tariff with the FCC (which the Commission ultimately rejected) that required customers that had purchased equipment from its competitors to purchase an unnecessary "interface” device in order to access AT & T’s telephone system. The jury found that AT & T’s tariff evidenced its intent to create barriers to entry in the telephone equipment market and held it (along with AT & T’s efforts to convince the FCC to approve the tariff) to be anticompeti-tive and the Court of Appeals affirmed. Panhandle’s tariff, by contrast, had been approved since 1951, and Panhandle had assumed long-term contractual obligations in reliance on the tariff.