Full opinion text
AMENDED MEMORANDUM DECISION AND ORDER RE: MOTIONS FOR SUMMARY JUDGMENT ON SECTION TWO SHERMAN ACT CLAIMS RAFEEDIE, District Judge. The captioned case came on for hearing before this Court, the Honorable Edward Rafeedie, United States District Judge, presiding, on August 8, 1988. The following motions were heard; Defendant United Airlines (“United”) and American Airlines’ (“American”) Motions for Summary Judgment on plaintiffs’ claims under Section 2 of the Sherman Act, 15 U.S.C. § 2, Plaintiff Continental’s Motions for Partial Summary Judgment: (1) that Defendants’ Computerized Reservations Systems are Essential Facilities, (2) that Defendants did not allow Equal Access to their Essential Facilities, and (3) that Defendants have Monopoly Power. The Court, having read and considered the papers submitted, and the argument of counsel at the hearing, orders that summary judgment should be granted in part, and denied in part, for the reasons stated in this Memorandum Decision and Order. FACTUAL BACKGROUND This case arises out of defendants’ ownership of Computerized Reservation Systems (“CRS”). A CRS is composed of computer terminals and printers in travel agents’ offices which are telephonically linked to the vendor’s computer. This equipment enables the travel agent to send and receive air transportation booking information, book flights and print out a ticket. These CRSs are owned by various airlines and each system contains flight information for airlines other than the vendor airline. The vendor charges the travel agent for the use of its system and they charge other airlines fees for booking air transportation through the CRS. Defendant American owns the world’s largest CRS, SABRE, which is comprised of six IBM mainframe computers that are connected to nearly 100,000 other devices, including computer terminals, ticket printers, and boarding pass printers. More than 11,000 travel agency locations use SABRE to handle airline as well as hotel and car reservations for their clients. SABRE contains schedules for more than 650 airlines and projects more than one year into the future. SABRE processes over 10 million reservations a month. Defendant United was the first company to announce plans to market a CRS. United’s CRS, Apollo, has been the second largest CRS in the world with an estimated market share of 23% of all travel agency locations. In 1981, Apollo claimed a 39% market share. SABRE’s market share is in the 30% range, down from 40% in 1980. The market also includes SystemOne (or SODA), owned by Eastern Airlines, PARS is run by TWA, and DATAS II is owned by Delta. Originally, travel agents paid a fee for CRS equipment rental and other services, while airlines were not charged for participating in the CRS or for bookings. The vendor airline, however, received substantial revenue from additional airline business they received through “biasing” the system. Biasing is the practice of displaying flight information in a way that favors the vendor airline. The travel agent inputs its client’s preferences and the CRS displays, in order of desirability, the various flights. However, each system was biased, to differing degrees, so that the desirability of the vendor’s flights would be artificially inflated. In the late 1970’s, SABRE and Apollo began signing carriers to “cohost contracts” under which the contracting carrier’s product would receive preferential treatment in the CRS in return for a fee paid on each booking which the carrier received through the CRS. Beginning in 1981, vendor airlines began entering into individually negotiated contracts with each airline, and booking fees rose from $0.25 per booking up to $3.00 per booking. In August 1984, the Civil Aeronautics Board (“CAB”) established a number of rules governing the practices of CRS vendors. Those rules required each CRS vendor to make available an unbiased primary display, to charge all carriers participating in its CRS the same booking fees for the same level of service, and to make CRS marketing data available for sale. The CAB declined to regulate booking fees. 49 C.F.R. 255. In response to the CAB rules, American announced it would charge $1.75 for booking, and United followed with a $1.85 booking fee. Competing CRS vendors are currently charging the following fees: the SystemOne fees are $1.75 and $2.00 for direct access, the PARS fees are $1.75 and $2.00 for direct access, and the DATAS II fees are $1.50 and $1.75 for direct access. I. Standard for Summary Judgment Rule 56 of the Federal Rules of Civil Procedure states that the court shall enter judgment if “the pleadings, depositions [and] affidavits show that there is no genuine issue as to any material fact and that the moving party is entitled to judgment as a matter of law.” Fed.R.Civ.P. 56(c). The nonmoving party must show there is a genuine issue of fact if the specific facts set forth by the moving party, coupled with undisputed contextual or background facts, are such that a rational or reasonable jury might return a verdict in its favor based on the evidence. T.W. Electrical Service v. Pacific Electrical Contractors, 809 F.2d 626 (9th Cir.1987). The nonmoving party can meet this burden with any kind of evidence listed in Federal Rule 56(c), but the pleadings alone are not enough. Celotex Cory. v. Catrett, 477 U.S. 317, 106 S.Ct. 2548, 91 L.Ed.2d 265 (1986). All that is required at this stage is that sufficient evidence supporting the claimed factual dispute be shown to require a factfinder to resolve the parties’ disputing versions of the truth at trial. T.W. Electrical, 809 F.2d at 630. The judge is not to weigh conflicting evidence with respect to disputed material facts, nor should the judge make credibility determinations with respect to statements in the affidavits. At summary judgment, the judge must view the evidence in the light most favorable to the nonmoving party. Id. The Supreme Court has plainly stated that the standard for summary judgment applies equally to antitrust cases as it does to any other case. Matsushita Electric Industrial Co. v. Zenith Radio Corp., 475 U.S. 574, 106 S.Ct. 1348, 89 L.Ed.2d 538 (1986). II. Essential Facilities Doctrine The essential facilities doctrine imposes on a business that controls an essential facility the obligation to provide its competitors reasonable access to that facility. Byars v. Bluff City News Co., 609 F.2d 843, 856 (6th Cir.1979). An essential facility is one which cannot be reasonably duplicated and to which access is necessary if one wishes to compete. Fishman v. Estate of Wirtz, 807 F.2d 520, 539 (7th Cir.1986). A refusal to deal in this context violates section 2 because control of an essential facility can “extend monopoly power from one stage of production to another, and from one market into another.” MCI Communications Corp. v. American Tel. & Tel. Co., 708 F.2d 1081, 1132 (7th Cir.1983). A facility or resource is “essential” when competitors must have access to the facility to meaningfully be able to compete with the firm controlling the facility. Holmes, 1987 Antitrust Law Handbook, section 2.06, at 175. The doctrine has been applied to electric transmission lines, football and basketball stadiums, downhill ski hills, natural gas pipelines, and local telephone facilities. To be essential, “a facility need not be indispensable; it is sufficient if duplication of the facility would be economically infeasible and if denial of its use inflicts a severe handicap on potential market entrants.” Fishman, 807 F.2d at 539 (quoting, Hecht v. Pro-Football, Inc., 570 F.2d 982, 992 (D.C.Cir.1977), cert. denied, 436 U.S. 956, 98 S.Ct. 3069, 57 L.Ed.2d 1121 (1978)). “The point of the essential facilities doctrine is that a potential market entrant should not be forced simultaneously to enter a second market, with its own large capital requirements.” Fishman, 807 F.2d at 540. A participant in the downstream market, who also controls a facility which is deemed essential to the downstream market, has the power to increase the costs of market entry through its control of the essential facility. The facility owner has the power to monopolize the market to which his facility is the “bottleneck.” Id.; Gamco, Inc. v. Providence Fruit & Produce Bldg., Inc., 194 F.2d 484, 487 (1st Cir.), cert. denied, 344 U.S. 817, 73 S.Ct. 11, 97 L.Ed. 636 (1952). Plaintiff s contend that SABRE is an essential facility for domestic airlines; that is, SABRE confers exclusive access to a large number of travel agents, and all domestic airlines are dependent on SABRE because no airlines can afford to forgo access to the SABRE travel agents, therefore, rival CRSs are not substitutes for SABRE. It is argued that American, as the major CRS vendor, holds all other airlines captive by virtue of the essential need for access to the SABRE travel agents. American contends that this is simply not an essential facilities case. In a normal case brought under this doctrine, the facility at issue constitutes a bottleneck to competition in the downstream market. Control of the bottleneck by a competitor can foreclose competition in the underlying market. For example, in United States v. Terminal R.R. Ass’n, 224 U.S. 383, 32 S.Ct. 507, 56 L.Ed. 810 (1912), the railroad combination’s denial of the use of its terminal facilities to non-member railroads would have foreclosed “any other reasonable means of entering the city” to the non-members, making it impossible to gain access to St. Louis. In the MCI case, foreclosure of access to local telephone distribution channels precluded MCI from competing in the long distance telephone service market because access to the local distribution channels is necessary to providing long distance service. The court stated that, given present technology, local telephone service is generally regarded as a natural monopoly and is regulated as such. Therefore, it would not be economically feasible for MCI to duplicate AT & T’s local distribution facilities. Such duplication would involve laying millions of miles of cable and line to individual homes and businesses, and regulatory authorization could not be obtained for such an uneconomical duplication. MCI, 708 F.2d at 1133. Similarly, in Otter Tail Power Co. v. United States, 410 U.S. 366, 93 S.Ct. 1022, 35 L.Ed.2d 359 (1973), a regulated electric utility refused to sell power wholesale, or to transmit power purchased from other sources to municipalities which had chosen to own their own retail distribution systems. The electric utility’s facility for transmitting electric power was essential to competition in the downstream market for the retail distribution of electric power. Again, the facility was along the lines of a natural monopoly. That is, duplication of the facility would be an inefficient waste of resources where the one facility is all that is needed, absent anticompetitive conduct by the facility owner with respect to a downstream market. The antitrust laws will relieve a competitor of the need to build its own production facilities only where the market will not support more than one. Fishman, 807 F.2d at 574 (Easterbrook, dissenting). See, Consolidated Gas Co. of Fla. v. City Gas Co. of Fla., 665 F.Supp. 1493 (S.D.Fla.1987) (a gas pipeline was deemed essential based upon the increased costs of gas which would be caused by the uneconomical duplication of an existing pipeline). In Fishman, the Seventh Circuit upheld the district court’s finding that the Chicago Stadium was the only stadium in the Chicago Metropolitan area which was suitable for the exhibition of professional basketball, and that it could not feasibly be duplicated by plaintiffs. Since building a new stadium would cost $19 million, and the basketball franchise was substantially less expensive than a stadium, the court held that duplication of the facility would be unreasonable. On this basis, the court held that the Chicago Stadium was essential to competition in the market for the exhibition of professional basketball in Chicago. Fishman, 807 F.2d 539-40. See, Hecht v. Pro-Football, Inc., 570 F.2d 982 (D.C.Cir. 1977), cert. denied, 436 U.S. 956, 98 S.Ct. 3069, 57 L.Ed.2d 1121 (1978) (football stadium in Washington D.C. deemed essential to market for professional football exhibition). In this case, there are channels of distribution other than SABRE because there are computerized reservation systems in competition with SABRE. Thus, plaintiffs can look to other systems for alternative channels of distribution and are not forced to enter the market for CRSs in order to compete in the downstream market of airline transportation. Nevertheless, plaintiffs claim that SABRE is essential because the other CRSs are not effective substitutes for SABRE. Under ideal market conditions, the absence of an airline from SABRE would decrease SABRE’s attractiveness to travel agents and cause them to seek out better systems. Continental, however, points to a number of market imperfections which allegedly impair better CRSs from attracting SABRE agents. American’s contracts with SABRE travel agents contain provisions which plaintiffs argue make it very difficult for an agent to switch CRS systems. The contracts contain a SABRE usage clause (“parity of usage rule”), a liquidated damages provision and roll-over provisions. American’s “minimum of parity” rule required new subscribers to have at least as many SABRE Computerized Reservation Terminals (“CRTs”) as there are non-SABRE CRTs at each location. American also regularly employ five year term contracts which contained a provision requiring a SABRE agent to sign a new five year contract whenever any new piece of equipment is installed or even replaced during the life of the existing contract. Finally, the liquidated damages provision in the contracts (installed after the anti-bias rules became effective) is (typically) as follows; 80% of the agent’s contract fee plus 200 (the average monthly segment booking per SABRE CRT) multiplied by the $1.75 segment fee multiplied by the number of SABRE CRTs in place, multiplied by the remaining months of the contract. Plaintiffs continue the attack by arguing that travel agents have strong economic incentives to carry only one CRS. First, the installation of a CRS, or additional CRSs, imposes significant costs on the agency in terms of both equipment and training. For example, the average SABRE and Apollo subscriber fees per year per agency location (not agency) in 1986 were $10,596 and $12,763, respectively. Thus, agents are unlikely to switch to another CRS unless the costs of conversion are paid for by the new vendor. Second, the minimum of parity rule constitutes a contractual restriction on the subscribing agents use of other CRSs. Finally, plaintiffs point out that the airline industry has always been characterized by narrow operating margins, therefore, even slight reductions in sales have a disproportionately adverse impact on profits and the carrier’s ability to compete. These contractual provisions and the structure of the air transportation market are claimed to prevent effective competition in the CRS market, and consequently, justify plaintiffs’ argument that SABRE is an essential facility to the air transportation market despite the existence of competing CRSs. Plaintiffs’ point to the conclusions of government regulatory bodies and Judge Posner’s opinion for the court in United Airlines, Inc. v. C.A.B., 766 F.2d 1107 (7th Cir.1985) where the court upheld the CAB’s antibias rule. In United, the court wrote: Unless an airline limits its operations to one small region, it must, whether or not it has its own computerized reservation system, persuade several of the largest airlines to list its flights in their [CRS] systems if it is to have a fair chance of success. It is thus dependent for an essential facility on what may be its principal competitors. Id. at 1114. Judge Posner’s opinion has a number of helpful observations. First, it is necessary to put the case in its proper perspective. The Seventh Circuit held, as a matter of administrative law, that the CAB’s exercise of rulemaking authority was within the scope delegated by Congress through the Federal Trade Commission Act, section 5, as amended 15 U.S.C. § 45, and the Federal Aviation Act of 1958, section 411, 49 U.S.C. App. § 1381. Under these statutory provisions the CAB can forbid anticompetitive practices “before they become serious enough to violate the Sherman Act.” United, 766 F.2d at 1114. The court ruled that the Board’s ruling was “plausible, if not compelling, [and its] rules can not be set aside as arbitrary and capricious.” Id. at 1116. Therefore, the Seventh Circuit did not rule that Apollo was an essential facility, it merely held that the Board’s analysis was not arbitrary and capricious in light of its statutory authority. Thus, the Board has the power to outlaw conduct which may restrain competition. More interesting, however, are Judge Posner’s comments' on the CRS market: If the owner of a computerized reservation system used the system to weaken competition from other airlines, it is a little hard to see why those airlines would not simply switch their patronage to a competing system that was not biased against them. Competition would, (one might have thought) force at least some of the owners of competing systems to offer unbiased listings in order to expand the market for their systems. Even if every airline owner refused, because of the impact on its air transportation revenues, to give equal prominence to a competitor’s flights, there is nothing to stop independent companies from offering a computerized reservation system with no such inhibitions — and one does. The court, however, goes on to state that an airline needs to be listed “at least in the largest” CRSs. “Of course, if the owner of a system charges such a high price that no competing airline will pay it, the owner is hurt. It not only loses revenues from that airline; its system will be worth less to travel agents if it contains less information. But the owner may be able to extract a high enough price from competitors to slow their growth; indeed, that may be the purpose of the high prices.” Id. at 1115. In sum, the CRS vendor will charge as high a price as it can without losing participating airlines (and thereby decreasing the attractiveness of its’ CRS). Similarly, with respect to display bias, the CRS vendor “hope[s] to gain more business by complicating the marketing of competing services than it would lose to travel agents by providing them with less flight information.” Id. The question is whether the entry barriers to a competing CRS (with respect to access to SABRE agents) are sufficient to qualify SABRE as an essential facility. An entry barrier is either (a) “a condition that makes the long run costs of a new entrant into the market higher than the long run costs of the existing firms in the market”, or more broadly, (b) “heavy start-up costs required for entry into the market.” Posner, Economic Analysis of Law, section 10.9, p. 290 (3rd ed. 1986). In a hypothetical market without any entry barriers, a seller charging more than his marginal costs (which includes “competitive profits” —profits sufficient to maintain the seller’s investment in the industry) reaps “monopoly profits” which will attract increased output in the form of new firms entering the market and increased output from firms already in the market. Entry barriers, however, serve to stave off increased industry output beyond the threshold point established in the hypothetical market: that point being when a seller’s price is greater than his marginal costs. The seller’s ability to reap monopoly profits depends upon the significance of the entry barriers. In this case, the contractual provisions of the SABRE leases raise the cost of entering the market for access to SABRE agents. However, the entry barriers do not foreclose competition in the market. Rather, they raise the cost of entry to the extent that it is expensive for a travel agent to attain dual-CRS capacity. The significance of the entry barriers, however, are exaggerated by plaintiffs. Assuming that American uses SABRE to weaken competition in the air transportation market by providing biased information, or by charging exorbitant booking fees, there will be an increased incentive for travel agents to attain dual or multiple CRS capability. Moreover, American’s anticompetitive action will create an incentive for other CRSs to reduce bias and/or booking fees in an attempt to capture the SABRE agent market. The costs of making SABRE agents dual-CRS-capable can be shared by the agents, airlines and CRS vendors, all of whom have an incentive to widen the market if SABRE is an unattractive product. While American may be able to extract supracompetitive booking fees in the CRS market, and such price increases do create entry barriers in the air transportation market by raising competitors costs, the entry barrier in the air transportation market is not significant enough to create a danger of monopolization under the circumstances. Just because defendant may be extracting monopoly profits in one market does not inevitably lead to the conclusion that its power in an affected market is also significant enough to confer monopoly power in the related market. Under Section 2 of the Sherman Act, a firm must provide reasonable access to an essential facility in order to prevent that firm from extending its monopoly power from one market into another. MCI, 708 F.2d at 1132. The evil at which the doctrine is aimed is monopolization of the underlying market. In Monsanto Co. v. Spray-Rite Service Corp., 465 U.S. 752, 104 S.Ct. 1464, 79 L.Ed.2d 775 (1984), the Court recognized that the Sherman Act contains a “basic distinction between concerted and independent action,” and that distinction is embodied in sections 1 and 2 of the Act. In Copperweld Corp v. Independence Tube Corp., 467 U.S. 752, 104 S.Ct. 2739, 81 L.Ed.2d 628 (1984), the Court considered the policy differences between claims brought under section 1 and section 2 of the Sherman Act. Section 2 provides: Every person who shall monopolize, or attempt to monopolize, or combine or conspire with any other person or persons, 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. In contrast, section 1 provides: “[e]very contract, combination in the form of trust or otherwise, or conspiracy, in restraint of trade or commerce among the several States ... is declared to be illegal ...” 15 U.S.C. § 1. The “conduct of a single firm is governed by section 2 alone and is unlawful only when it threatens actual monopolization.” Copperweld, 467 U.S. at 767, 104 S.Ct. at 2739. “It is not enough that a single firm appears to ‘restrain trade’ unreasonably, for even a vigorous competitor may leave that impression.” Id. Partially because “it is sometimes difficult to distinguish robust competition from conduct with long-run anticompetitive effects, Congress authorized Sherman Act scrutiny of single firms only when they pose a danger of monopolization.” Id. at 767-68, 104 S.Ct. at 2740; The Jeanery, Inc. v. James Jeans, Inc., 849 F.2d 1148 (9th Cir.1988) (“a single firm’s conduct, absent the danger of monopolization, is not the object of antitrust scrutiny because to treat it with such scrutiny would heighten ‘the risk that the antitrust laws will dampen the competitive zeal of a single aggressive entrepreneur.’ ”) The essential facilities doctrine must be tied to the particular statutory provision under which relief is sought. The doctrine itself is an abstraction designed to assist courts in determining whether the statute has been violated. The starting point of statutory construction is the language of the statute itself. Landreth Timber Co. v. Landreth, 471 U.S. 681, 105 S.Ct. 2297, 2301, 85 L.Ed.2d 692 (1985); United States v. Hoffman, 794 F.2d 1429, 1431-32 (9th Cir.1986). The plain meaning of the statute is controlling “absent a clearly expressed Congressional intention to the contrary.” Hoffman, 794 F.2d at 1432 (citing, North Dakota v. United States, 460 U.S. 300, 312, 103 S.Ct. 1095, 1102, 75 L.Ed.2d 77 (1983)). While section 1 of the Act precludes conspiracies in “restraint of trade,” section 2 is targeted at “monopolization.” As the Supreme Court has stated, the section 2 standard is less stringent than section 1 because it is often difficult for a court to distinguish between vigorous competition and anticompetitive conduct when analyzing the actions of a single firm. Copperweld, 467 U.S. at 768, 104 S.Ct. at 2740. Thus, the Sherman Act authorizes scrutiny of single firms only when they pose a danger of monopolization. Id. Therefore, when applying the essential facilities doctrine in the context of section 2 of the Sherman Act, a facility should be deemed essential to the downstream market only where control of the facility by a competitor poses a danger of monopolization of the downstream market. In this case, there is no danger that American will monopolize the market for air transportation. Looking at the facts in a light most favorable to the plaintiffs, American’s conduct may restrain trade, but it does not threaten monopolization. Even if “entry barriers” to accessing SABRE agents are substantial enough to preclude a single airline from withdrawing from SABRE because SABRE agents will not look to other CRSs (or other sources of information) when making reservations, there is no danger that American will monopolize the air transportation market. As Judge Posner recognized, American is restrained by the existence of competing CRSs. Although there are costs (entry barriers) associated with leasing more than one CRS, these costs only partially insulate American from competition in the SABRE agent market. “[I]f the owner of a system charges such a high price that no competing airline will pay it, the owner is hurt. It not only loses revenues from that airline; its system will be worth less to travel agents if it contains less information.” Although the CRS owner may be able to “extract a high enough price from competitors to slow their growth,” such a restraint of trade does not violate section 2 of the Act unless it poses a danger of monopolization. The existence of competing CRSs constitute a substitute for SABRE and constrain American’s ability to weaken competition in the air transportation market. American has never had more than a 14 percent share of the air transportation market. Although market share is, in itself, an insufficient indicator of market power, such a minimal share precludes a reasonable jury from finding monopolization. “There is substantial merit in a presumption that market shares below 50 or 60 percent do not constitute [market] power.” P. Areeda & H. Hovenkamp, Antitrust Law, section 518.3c (1986 Supp.) (hereinafter P. Areeda & H. Hovenkamp). A claim that a twelve to fourteen percent market share confers monopoly power is absurd, absent a showing that the air transportation market is characterized by a low elasticity of demand or supply. The large number of competing airlines belies any contention that the industry is faced with low elasticity of supply. The court has also been presented with uncontroverted evidence that the airline industry is highly competitive from the demand side. The USAir plaintiffs have argued that requiring a dangerous probability of successful monopolization of the underlying market renders the doctrine superfluous. This argument is not well taken. First, whenever courts have applied the doctrine, the facility at issue constituted a bottleneck to competition in the underlying market which created a danger that the facility owner would monopolize the underlying market. Moreover, a competitor’s failure to provide its competitors with access to a facility does not constitute “anticompetitive conduct” absent an essential facilities doctrine. The general rule is that “even a firm with monopoly power has no general duty to engage in a joint marketing program with a competitor.” Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585, 600, 105 S.Ct. 2847, 2856, 86 L.Ed.2d 467 (1985). Thus, a firm with monopoly power has no general duty to help competitors by “pulling its competitive punches.” Olympia Equip. Leasing v. Western Union Tel., 797 F.2d 370, 375 (7th Cir.1986), cert. denied, 480 U.S. 934, 107 S.Ct. 1574, 94 L.Ed.2d 765 (1987). However, in some circumstances, failure to cooperate with a competitor may indicate probable anticompetitive effect where the monopolist lacks a business justification for its action. Aspen, 105 S.Ct. at 2859-60; Olympia, 797 F.2d at 378. The particular circumstances in which firms are required to provide a helping hand to its competitors are not well defined in antitrust law. The essential facilities doctrine can be interpreted as a set of requirements triggering a competitor’s obligation to cooperate with a competitor. Absent the essential facilities doctrine, a firm’s refusal to provide its competitors with access to its facility would not constitute anticompetitive conduct. Therefore, requiring plaintiffs to show that denial of access to an essential facility creates a danger of monopolization does not render the essential facilities doctrine superfluous. Rather, the doctrine established the conditions for requiring a firm to cooperate with its competitors and thereby rendering the failure to provide reasonable access an anticompetitive act under section 2. Moreover, satisfaction of the doctrine’s four internal requirements eliminates the need to prove that the firm had a specific intent to monopolize. III. Monopolization A. Relevant Service Market A determination of the relevant market usually requires an inquiry into the nature of the product, and the geographical area of effective competition. Oahu Gas Serv., Inc. v. Pacific Resources, Inc., 838 F.2d 360, 364 (9th Cir.1988). Plaintiffs claim that the relevant market in this case is the nationwide market for access to SABRE agents. Defendants argue that the relevant service market is the nationwide market for access to all CRS automated travel agents. A relevant service market is the smallest service market for which (1) the elasticity of demand and (2) the elasticity of supply are sufficiently low that a firm with 100% of that market could profitably reduce output and increase price. “What is called for is an appraisal of the ‘cross-elasticity’ of demand in the trade.” Fount-Wip, Inc. v. Reddi-Wip, Inc., 568 F.2d 1296, 1301 (9th Cir.1978). Thus, products and services which are “reasonably interchangeable” for the same or similar uses normally should be included in the same product market for antitrust purposes. Kaplan v. Burroughs Corp., 611 F.2d 286, 291 (9th Cir.1979). A particular brand may define a relevant market; “a relevant submarket in one’s own products will exist if such a submarket is (a) sufficiently distinct in commercial reality, and (b) is relatively immune from competition of substitutes, or (c) was acquired by means that show an attempt to monopolize.” United States v. CBS, Inc., 459 F.Supp. 832 (C.D.Cal.1978). As a legal matter, a firm’s own product can constitute a distinct market only where the product is “so unique or so dominant” in the market that control over the product would virtually assure that competition in the market would be “destroyed.” Bushie v. Stenocord Corp., 460 F.2d 116, 121 (9th Cir.1972). A “market” is “any grouping of sales whose sellers ... could raise prices significantly above the competitive level. If the sales of other producers substantially constrain the price-increasing ability of the [seller], they are part of the market.” P. Areeda & H. Hovenkamp, at 311-12. It is true that products differing in brand name may have certain physical or service based differences which may be substantial. Furthermore, brand recognition may be substantial in many cases. Consequently, differentiated products are not perfectly interchangeable for consumers. “Unlike a perfectly competitive firm, a producer of a successfully differentiated product can raise his price above that of his rivals without losing all of his sales.” P. Areeda & H. Hovenkamp, at 316. The question is whether “the degree of power inherent in each differentiated product is sufficient to make each brand a separate market for ... monopolization purposes? The answer is almost uniformly negative.” Id. at 316-17. The reason is, from a policy perspective, legal rules limiting a monopolist cease to make sense when applied to every producer of a differentiated product. From an economic perspective, producers engage in product differentiation because they are involved in a highly competitive market and are merely attempting to compete with other brands. Product differentiation is a public good created by competitive markets, and not an indication that competition is lacking in the product market. Id. The ultimate question is whether other existing or potential CRSs significantly constrain the price-raising power of American. If the answer is yes, then these other CRSs must be grouped in the same “market” as SABRE. However, if the other CRSs do not impair SABRE’s ability to control prices or exclude competition, then the other CRSs can not be viewed as substitutes for SABRE. The Supreme Court has suggested the following factors when determining the relevant market: The boundaries of such a submarket may be determined by examining such practical indicia as industry or public recognition of the submarket as a separate economic entity, the product’s peculiar characteristics and uses, unique production facilities, distinct customers, distinct prices, sensitivity to price changes, and specialized vendors. Brown Shoe v. United States, 370 U.S. 294, 325, 82 S.Ct. 1502, 1524, 8 L.Ed.2d 510 (1961); Kaplan, 611 F.2d at 292-93. Continental supports its claim that access to SABRE agents is a distinct service market by applying these factors. There is evidence that the industry recognizes access to SABRE as a separate service from access to other CRSs. The fact that almost all airlines participate in virtually all CRSs is evidence that each airline perceives each CRS as a different service. Moreover, when United, TWA and American raised their booking fee to $1.75, no airline elected to participate in one of the lower priced CRSs (Eastern’s System-One and Delta’s DATAS II had rates of $ .75) and bypass Apollo or SABRE. Rather, all airlines continued to participate in both systems. Evidence that American engaged in price discrimination prior to enactment of the 1984 CAB rules is also evidence that carriers perceive the different CRSs as constituting different services. Finally, Continental points to evidence that American set its booking fees without regard to prices set by other CRS vendors. Despite disparity in pricing, demand for access to SABRE has not gone down. The apparent reason, according to Continental, for the low elasticity of demand is that access to SABRE provides a carrier with virtually the only practical access to SABRE-automated travel agents, since more than 90% of all SABRE agents only use SABRE. The evidence amassed by Continental, however, does not support its contention that a reasonable jury could not find a broader market (such as the market for access to all CRS-automated travel agents) to be relevant in this case. Market definition is essentially a question of fact dependent upon the special characteristics of the industry involved. Oahu, 838 F.2d at 363 (citations omitted). The evidence provided does not compel a factfinder to conclude that other CRSs do not constrain American when it establishes booking fees for SABRE. Continental’s argument is based upon a static reading of the CRS market. The historical fact that almost all SABRE agents have only one CRS and that almost all carriers participate in virtually all of the major CRSs is not indicative of market power. The apparent low elasticity of demand is meaningless unless it is complemented by low elasticity of supply. [Djefining a market ... on the basis of demand considerations alone is erroneous. The function of defining a market is to determine that grouping of sales which, if controlled by a single firm ..., could charge noncompetitive prices. But that is not possible if either demand or supply elasticity are high. Exploitative prices cannot be maintained on, say, children’s shoes, if high prices induce consumers to shift to other products or induce producers of men’s or women’s shoes to make children’s shoes. P. Areeda & H. Hovenkamp, at 319. ■Thus, although current market conditions suggest low demand elasticity, the existence of alternative channels of distribution such as competing CRSs indicate high supply elasticity. CRSs such as Apollo, PARS, DATAS II, and SystemOne provide essentially the same services as SABRE, except that SABRE is used in more travel agencies than the other CRSs. If SABRE began charging supracompetitive prices, then airlines would abandon SABRE, causing SABRE-automated travel agents to look to other CRSs. The existence of the other CRSs, therefore, constrain American’s ability to set supracompetitive prices for SABRE. The existence of alternative CRSs suggests high supply elasticity, so there exists at least a factual dispute as to whether the relevant market is “access to SABRE agents.” The more interesting question is whether the court should characterize the market for CRS access as the relevant market as a matter of law. Although market definition is a factual matter, it is often properly decided on summary judgment because issues of market definition do not involve questions of motive or intent. General Business Systems v. North American Phillips Corp., 699 F.2d 965 (9th Cir.1983) (affirming granting of summary judgment for failure to present evidence of a relevant market). The question is whether a reasonable jury could conclude that access to SABRE agents is a distinct service market, when viewed in the light most favorable to plaintiffs. Entry Barriers to Accessing SABRE-Agents Plaintiffs have presented evidence that certain entry barriers preclude other CRS vendors from competing with American for access to SABRE agents. The existence of entry barriers suggest that elasticity of supply is low, despite the presence of alternative distribution channels. This conclusion is based upon the alleged “captive customers” relationship between SABRE and the SABRE-automated agents, maintained through long-term contracts and high liquidated damage clauses. As stated earlier, plaintiffs argue that access to SABRE-automated agents is impaired through contractual provisions and the costs involved in making travel agents dual-CRS-capable. In defining a “market” the court must look to all products which substantially constrain the price-setting ability of the seller. Even if certain SABRE agents are stuck with long term contracts, American must make its product attractive to new travel agencies, SABRE agents whose contracts are expiring, and SABRE agents who can afford the extra expense of obtaining a second CRS. SABRE-automated agents can become ex-SABRE-automated agents as their contracts expire, or as they chose to become dual-CRS-capable. Competing CRSs can quickly expand output merely by setting up a terminal at the travel agent’s office. Indeed, the virtual non-existence of start-up costs would permit a competing CRS to bear some of the financial burden of making a SABRE-agent dual CRS capable. Defendants argue that plaintiffs’ claim that SABRE is a separate market for purposes of section 2 must be based upon a further, unstated, assumption: that SABRE-automated travel agents have a “captive” group of customers who buy all their air transportation through the SABRE-agents, even if the SABRE-agents do not provide a competitive level of service. If consumers are not somehow held captive by their travel agents, then basic economic theory holds that they will seek out the best service available. There is no evidence that consumers are held captive by travel agents. The consumer’s freedom effects the entire system so that travel agents will demand better service from its CRS vendor, and if such service is not given, the agent will either seek alternative methods for booking flights (other CRSs, direct booking by telephone, etc.), or go out of business and make way for more competitive travel agents. This competitive pressure constrains American from charging supracompetitive prices to the carriers because the CRS vendor needs to provide access to its CRS in order to make the service attractive to travel agents. Therefore, as long as the consumer is not held captive, the CRS vendor must compete with other CRS vendors. In this case, there is no evidence that consumers are held captive by travel agents. Defendants’ theory accurately shows that American is constrained by the existence of competing CRSs. However, the question is not whether there is any form of constraint, but whether American is precluded from charging monopoly prices. While entry barriers exist and impose additional costs for entering the market, the difficult, factual question is whether the barriers are high enough to allow monopoly pricing. American may be able to extract booking fees high enough to slow the growth of their airline competitors, but low enough to make participation in the system cost efficient for the carrier. See, United Airlines v. C.A.B., 766 F.2d at 1114. The test for monopoly pricing is not whether participation in the system is cost efficient for the carrier (it must be, otherwise the carrier would not participate), but whether the price is significantly higher than the seller’s marginal cost. The existence of entry barriers for access to SABRE agents creates a material issue of fact as to whether other CRSs are effective substitutes for access to SABRE. The evidence, interpreted in a light most favorable to plaintiff, could permit a reasonable jury to conclude that SABRE is a separate service market because there is evidence that airlines view each CRS as offering a unique service: access to a particular set of travel agents. A reasonable jury may conclude from the evidence that SABRE automated agents are locked into SABRE (through contractual provisions) and that American is free to extract supra-competitive booking fees from participating airlines, although the fees are not high enough to justify abandoning the system. Thus, defendants are able to set booking fees at a supracompetitive rate, and consumers will not abandon SABRE agents because the travel agency’s services are not effected since supracompetitive prices are high enough to slow an airlines growth, but low enough to remain economically cost-efficient for the airlines. B. Elements of Monopolization The elements of monopolization under section 2 of the Sherman Act are (1) possession of monopoly power in a relevant market; (2) wilful acquisition or maintenance (“use”) of that power; and (3) causal antitrust injury. Catlin v. Washington Energy Co., 791 F.2d 1343, 1347 (9th Cir.1986) (citations omitted). 1. Monopoly Power Monopoly power focuses generally on a firm’s ability to “control prices or exclude competition.” Oahu, 838 F.2d at 366 (citations omitted). Courts have employed the following evidentiary methods in order to determine whether a defendant has market power in the relevant market; proof that defendant accounts for a high percentage of total firm sales within the market (“market share”), defendant’s actual exercise of price leadership control over the industry, affirmative actions taken by defendant that has excluded actual or potential competitors, profit levels, barriers to competition in the industry that would thwart new entry at the expansion of existing competitors, and historical trends within the industry. Holmes, supra, at 148-49. Inquiries in this area often depend heavily upon market share and barriers to entry. Oahu, 838 F.2d at 366. “A firm with a high market share may be able to exert market power in the short run, but ‘[substantial market power can persist only if there are significant and continuing barriers to entry.’ ” Id. (quoting, 2 P. Areeda & D. Turner, Antitrust Law, at section 505). A high market share will not raise an inference of market power in a market with low entry barriers or other evidence of defendant’s inability to control prices or exclude competitors. Oahu, 838 F.2d at 366. “By the same token, ‘[a]declining market share may reflect an absence of market power, but it does not foreclose a finding of such power.’ ” Id. (quoting, Greyhound Computer Corp. v. International Business Machines Corp., 559 F.2d 488, 496 n. 18 (9th Cir.1977), cert. denied, 434 U.S. 1040, 98 S.Ct. 782, 54 L.Ed.2d 790 (1978)). Market share is used as a surrogate for information which, if available, would provide a more accurate assessment of a firm’s market power. Thus, if a firm’s market elasticities (supply and demand) could be shown directly, there would be no need for extrapolating market power from market share. Landes & Posner, Market Power in Antitrust Cases, 94 Harv.L.Rev. 938, 953 (1981). If a firm’s marginal cost could be ascertained, its exercise of market power would be revealed merely by subtracting the price actually charged from the marginal cost of production. Id. These figures, however, are notoriously difficult (if not impossible) to determine. Therefore, courts look to market share in a relevant market as an effective surrogate for direct measurement of market power. P. Areeda & D. Turner, section 507, at 330-31. If the relevant market may be access to SABRE agents, then defendant controls 100% of the market, and included in the determination of “relevant market” is a finding that other CRSs may not be effective substitutes for SABRE. Thus, there is a material issue of fact as to whether SABRE has market power because there is a material issue as to whether access to SABRE automated agents is a relevant market. If the relevant market is the national CRS market, then plaintiffs must show that American has market power despite its relatively small share of the market. SABRE’s share of Computerized Reservation Terminals has never exceeded 44.2%, and it appears to be declining. Plaintiffs argue, however, that the market share figures do not accurately represent American’s market power. Thus, plaintiffs point to the actual exercise of market power as evidence of American’s monopoly power. Evidence of Market Power First of all, there is direct evidence of the defendants’ exercise of monopoly power in the CRS market. The recent Department of Transportation Report shows that defendants price booking fees above marginal cost. The report indicates that SABRE’s booking fees equal 233 per cent of their average costs for producing reservations during 1986, and Apollo’s booking fee constitutes 192 per cent of its costs for producing reservations. These results, however, do not mandate a finding of market power, as a matter of law, because the Report itself details the complications in allocating costs between booking transactions and supporting agency subscribers. Nevertheless, these factual findings preclude the court from finding, as a matter of law, that defendants do not have market power in the CRS market. Plaintiffs also point to evidence that American engaged in price discrimination prior to the enactment of the 1984 CAB rules. Price discrimination is “a term that economists use to describe the practice of selling the same product to different customers at different prices even though the cost of sale is the same to each of them.” R. Posner, Antitrust Law, 63 (1976). Price discrimination in sales of the same product can not last long in a competitive market because other sellers of the same product have an incentive to sell the product at a competitive rate to the victims of price discrimination. R. Posner, Antitrust Law, at 63 (“persistent price discrimination is very good evidence of monopoly because it is inconsistent with a competitive market”). Therefore, price discrimination usually can not occur in a competitive market for very long. P. Areeda & D. Turner, section 513, at 341-42.; Coal Exporters Ass’n of the United States, Inc. v. United States, 745 F.2d 76, 91 (D.C.Cir.1984), cert. denied, 397 U.S. 907, 105 S.Ct. 2151, 85 L.Ed.2d 507 (1985) (“it is well established that the ability of a firm to price discriminate is an indicator of significant monopoly power”). There is evidence that American engaged in price discrimination. For example, Eastern Airlines became a co-host on SABRE in March 1981, at a fee of 24 cents per booking. Delta Airlines, which competed with American at the Dallas/Ft. Worth hub, became a participating carrier on SABRE in October 1982 only after agreeing to a fee of $1.32. New York Air agreed to pay a $2.00 booking fee in December 1981. The issue in both the market definition and monopoly power area is whether American is constrained from charging monopoly prices by the existence of alternative distribution channels. The ability to engage in price discrimination is an indication that American has market power. Whether the degree of market power rises to the level of monopoly power, however, is a detailed factual question which involves a weighing of the evidence. In response to charges of price discrimination, American argues that different prices were the result of individually negotiated contracts where some airlines had more bargaining power based upon the benefits to American created by their participation in SABRE. In this case, evidence of price discrimination is not a compelling indication of market power. In a typical case of price discrimination, the value of the product in issue is not effected by the purchase of the product. That is, the' fact that Buyer purchases a widget does not increase the value of the widget such that the market value of the widget increases because of the purchase by the particular Buyer. However, the value of SABRE is enhanced by each subscriber because the number of subscribers directly increases the market value of the product to travel agents, and consequently, to other airlines. In effect, a major airline gives American something of value merely by subscribing to SABRE. A lesser airline’s participation does not increase the value of the CRS as much because that flight information is not as valuable to the CRS. Therefore, the subscription price does not reflect American’s entire compensation for providing SABRE services to its subscriber. The evidence shows that Eastern became a co-host at a fee of 24 cents per booking in March 1981. SABRE’s booking fee increased, so that New York Air was charged a $2.00 booking fee in December 1981, and Delta was charged $1.32 per booking in October 1982. However, neither New York Air nor Delta became “co-hosts”, and their participation began after SABRE had established itself in the market. In 1981 and 1982 the CRS market was just establishing itself, and PARS was really the only other CRS, aside from Apollo and SABRE, in the market. Above cost pricing and price discrimination is expected at this stage of market development. If the pricing was in fact supra-competitive it would serve to attract new entrants into the market. In fact, such new entry did take place when DATAS II and SystemOne entered the market and began to chip away at SABRE’s and Apollo’s market shares. Defendant’s exercise of price discrimination supports plaintiffs’ allegation that defendant had market power, although under the circumstances it is not compelling. A reasonable jury could conclude that defendant’s exercise of price discrimination was indicative of market power, however, a reasonable jury could also come to the opposite conclusion. Finally, whether the relevant market is deemed to be the market for access to SABRE automated travel agents, or the market for access to CRS automated travel agents, plaintiffs have presented evidence indicating that competing CRSs may not be effective substitutes for SABRE. Thus, the discussion relating to whether SABRE is a distinct service market also supports plaintiffs’ argument that defendants have monopoly power in the CRS market. The evidence indicates that the other CRSs may not constrain SABRE since the other CRSs are not substitutes for SABRE. Although not compelling, the evidence precludes a finding that defendants do not have monopoly power as a matter of law. 2. Wilful Acquisition or Maintenance of that Power The second element of monopolization under section 2 of the Sherman Act is “the wilful acquisition or maintenance of [monopoly] power as distinguished from growth or development as a consequence of superior product, business acumen, or historic accident.” United States v. Grinnell, 384 U.S. 563, 570-71, 86 S.Ct. 1698, 1704, 16 L.Ed.2d 778 (1966); Oahu, 838 F.2d at 367-68. The question of whether conduct is anticompetitive is a question of law. Oahu, 838 F.2d at 368 (reversing judgment against defendant monopolist on the ground that conduct was not anticompetitive as a matter of law). Section 2 of the Sherman Act does not prohibit “monopoly” itself (“monopoly in the concrete”), Standard Oil Co. of New Jersey v. United States, 221 U.S. 1, 62, 31 S.Ct. 502, 516, 55 L.Ed. 619 (1911); rather, it prohibits conduct directed at “smothering competition.” Berkey Photo, Inc. v. Eastman Kodak, Inc., 603 F.2d 263, 275 (2d Cir.1979). Conduct which may not be unlawful in itself may be illegal when done by a monopolist because it tends to destroy competition. Lorain Journal & Co. v. United States, 342 U.S. 143, 72 S.Ct. 181, 96 L.Ed. 162 (1951). Plaintiffs contend that American engaged in predatory pricing through which it attained its monopoly power. “Predatory pricing” refers to a firm’s attempt to drive a competitor out of business, or to discourage a potential competitor from entering the market, by selling its output at an artificially low price. The theory is that once the rival has been dispatched from the market, the predator will be able to reap monopoly profits which will more than pay for the losses incurred during the predatory period. Price reductions that constitute a legitimate, “competitive response” to market conditions are competitive, not predatory. William Inglis and Sons Baking Co. v. ITT Continental Baking Co., 668 F.2d 1014, 1031 (9th Cir.1981), cert. denied, 459 U.S. 825, 103 S.Ct. 58, 74 L.Ed.2d 61 (1982). “Pricing is predatory only where the firm forgoes short-term profits in order to develop a market position such that the firm can later raise prices and recoup lost profits ...” Janich Bros., Inc. v. American Distilling Co., 570 F.2d 848, 856 (9th Cir.1977), cert. denied, 439 U.S. 829, 99 S.Ct. 103, 58 L.Ed.2d 122 (1978). While the Ninth Circuit has adopted the Areeda-Turner theory, it has also modified the test. Inglis, 668 F.2d at 1032 (citing cases). The general rule is that a price should be considered predatory “if its anticipated benefits depend on its tendency to eliminate competition.” Inglis, 668 F.2d at 1034. If the justification for a price reduction did not depend upon this anticipated effect, then it does not support a claim of illegal monopolization or attempted monopolization. Predatory pricing may be proven, without reference to evidence of subjective intent, by examining the relationship between the defendant’s prices and costs. “But such proof must tend to show that the anticipated benefits of the prices, at the time they were set, depended on their anticipated destructive effect upon competition and the consequent enhanced market position of the defendant.” Inglis, 668 F.2d at 1034. Ultimately, “cost-based” inquiries serve merely as “an aid in determining the ultimate question: Did the justification for the defendant’s price depend upon its anticipated destructive effect on competition, or was the price justified as a reasonably calculated means of maximizing profits minimizing losses, or achieving some other legitimate end?” Inglis, 668 F.2d at 1038. However, the cost-based analysis establishes the burden of proof on the issue of predation: “If the defendant’s prices were below average total cost but above average variable cost, the plaintiff bears the burden of showing defendant’s pricing was predatory. If, however, the plaintiff proves that the defendant’s prices were below average variable cost, the plaintiff has established a prima facie case of predatory pricing and the burden shifts to the defendant to prove that the prices were justified without regard to any anticipated destructive effect they might have on competitors.” Inglis, 668 F.2d at 1036. In this Circuit,- therefore, all of the Areeda-Turner presumptions are rebuttable by a showing of predatory or nonpredatory purpose. Continental claims that American engaged in predatory pricing in the following manner; (1) American provided SABRE services to other airlines for free, and (2) American provided SABRE to travel agents below cost. In determining whether American was forgoing short-term profit in order to later reap monopoly profits, it is first necessary to examine the nature of SABRE. SABRE is a “joint” product which provides services to both travel agents and airlines. Thus, a determination that American was pricing below cost must account for American’s cost allocation decisions. The cost initially charged to travel agents was designed merely to recover the costs of automating the agent. However, Apollo and SABRE began cutting prices as they competed for access to travel agents. American also began by providing free SABRE services to other airlines in order to increase the value of SABRE to travel agents. This evidence indicates that American was pricing SABRE below its marginal costs in order to gain a foothold (or chokehold) in the market. The next question is whether American engaged in below cost pricing so that it could later increase its prices and recoup lost profits. The evidence permits an inference that American provided free services to airlines and sold SABRE to travel agents below costs in order to allow penetration of the CRS market through which it could recoup its lost profits by engaging in display bias (and later, supracompetitive booking fees). It appears that American decided to provide SABRE to travel agents at less than marginal cost, and planned on allocating to the participating airlines the bulk of its costs through display bias. However, it is clear that American provided SABRE services for free at its inception. A reasonable jury could infer that below cost pricing was taking place, and that it was undertaken for the purpose of later recouping lost profits through monopoly overcharges primarily through display bias. Therefore, there is a material issue of fact as to whether American engaged in anti-competitive conduct to attain or maintain its monopoly power. Continental argues, as another ground for finding a factual issue as to anticompetitive conduct, that American’s contracts with travel agents were meant to lock in SABRE subscribing agents through long term obligations and high liquidated damages provisions. American allegedly engaged in this type of behavior in order to lock in travel agents so that it would be free to increase display bias without seriously risking loss of subscribing travel agents. The contractual practices, in conjunction with the evidence of below cost pricing, permits an inference that American wilfully attempted to attain or maintain its monopoly power by tying up its participating travel agents. Possibility of Wilful Acquisition or Maintenance of Monopoly Power Plaintiffs’ theory is that SAB