In this extraordinary antitrust case,1 defendant American Telephone and Telegraph Company (“AT & T”) appeals from a judgment in the amount of $1.8 billion, entered on a jury verdict, in a treble damage suit brought by plaintiffs MCI Communications Corporation and MCI Telecommunications Corporation (collectively “MCI”) under section 4 of the Clayton Act, 15 U.S.C. § 15 (1976).2
I. FACTS
MCI’s original complaint, filed March 6, 1974, contained four separate counts: monopolization, attempt to monopolize, and conspiracy to monopolize — all under section 2 of the Sherman Act3 — and conspiracy in restraint of trade — under section 1 of the Sherman Act. MCI alleged that AT & T had committed twenty-two types of misconduct, classifiable into several categories including predatory pricing, denial of interconnections, negotiation in bad faith and unlawful tying. MCI claimed at trial, on the basis of a lost profits study originally prepared in part for financing purposes, that it had suffered damages of approximately $900 million as a result of AT & T’s allegedly unlawful actions.4
The case was tried to a jury between February 6 and June 13, 1980. After completion of MCI’s case in chief, the district court directed a verdict in favor of AT & T on seven of the twenty-two alleged acts of misconduct.5 The remaining fifteen *1093charges — all based on section 2 of the Sherman Act — were submitted to the jury. A special verdict form required the jury to make a separate finding of liability as to each of the fifteen charges, but permitted the jury to award damages in a single lump sum, without apportioning MCI’s claimed financial losses among AT & T’s various lawful and unlawful acts. The jury found in favor of MCI on ten of the fifteen charges submitted, and awarded damages of $600 million — a sum equal to two thirds the total damage figure claimed in MCI’s aggregated lost profits study.6 The district court trebled this damage award, as required by section 4 of the Clayton Act, resulting in a judgment of $1.8 billion, exclusive of costs and attorneys’ fees.
AT & T filed motions for judgment notwithstanding the verdict or, in the alternative, for a new trial on June 23, 1980. These motions were denied without opinion on July 29, 1980. On August 25, 1980, AT & T filed its notice of appeal. On September 8, 1980, MCI filed a notice of cross-appeal.7 In this opinion, we reject challenges to certain jury findings upon which AT & T’s liability was based, sustain other challenges, and remand for a new trial on the issue of damages.
A. Background and Initial Entry of MCI
Prior to 1969, the telecommunications industry was regulated as a lawful monopoly. Local exchange service was and still is provided exclusively by one of the twenty-three Bell System operating companies or by one of some 1600 independent telephone companies, depending upon the geographical area involved.8 Long distance service was provided by the Long Lines Department of AT & T in partnership with these operating companies.9 The network of long distance transmission facilities was owned in substantial part by Long Lines; however, the interexchange facilities of the local telephone companies, including both transmission and switching facilities, were used in conjunction with Long Lines facilities whenever efficiency required. The local exchange facilities and switching machines belonging to the local companies were also used at each end of a regular long distance call.
This same nationwide network was used as well by AT & T to provide other intercity telephone services, including point-to-point private lines, foreign exchange lines (“FX”) and common control switching arrange-*1094merits (“CCSA”). Point-to-point private lines (also called tie lines) are connections between two locations that do not require the use of local switching machines because the lines are available to the customer on a continuing and exclusive basis. FX and CCSA, although classified for tariff purposes as private line services, do require interconnection with local switching machines.10
In 1963, Microwave Communications, Inc., the predecessor corporation to MCI,11 requested permission from the Federal Communications Commission (“FCC”) to construct and operate a long distance telephone system between Chicago and St. Louis. The proposed system consisted of a terminal in each city and microwave radio relay towers connecting the terminals. Through this system, MCI intended to provide long distance, private line telephone service to business and industrial subscribers whose needs justified the exclusive or semi-exclusive use of a long distance telephone line. MCI also sought interconnections from its terminals to ordinary local telephone facilities, principally telephone wires running in conduits beneath the street. These interconnections were essential to MCI’s ability to do business, since they provided the telephone or computer linkage between MCI’s terminals and its individual customers in each city.
In 1969, after lengthy administrative proceedings in which AT & T and the other general service carriers opposed MCI’s application, the FCC approved MCI’s proposal. Microwave Communications, Inc., 18 F.C. C.2d 953, 966 (1969); 21 F.C.C.2d 190 (1970).12 The FCC’s decision specifically authorized MCI to provide only point-to-point private line service not requiring connection to the nationwide switched network — that is, tie lines that would connect two or more locations without the use of switching machines. 18 F.C.C.2d at 953-54. The FCC also retained jurisdiction to order appropriate local interconnections.
The MCI decision resulted in a deluge of new applications to the FCC for authority to construct and operate facilities for specialized common carrier services. MCI filed applications for authority to provide specialized services among more than 100 cities. Other companies filed similar applications, creating a situation in which, in many instances, more than one carrier was seeking to provide specialized services over the same route. To deal with this situation, the FCC instituted a broad rulemaking inquiry designed to permit consideration in one proceeding of the policy questions raised by these numerous applications. *1095Specialized Common Carriers, 24 F.C.C.2d 318 (1970) (Notice of Inquiry).
In June 1971, the FCC handed down its Specialized Common Carriers decision, approving in principle the entry of specialized carriers into the long distance telecommunications field, and declaring as a matter of policy that there should be open competition in the specialized services to which the decision applied: 29 F.C.C.2d 870 (1970). Because AT & T, reversing its earlier position, agreed to negotiate with MCI and other new entrants for local interconnections, the FCC elected to defer consideration of MCI’s claim that AT & T was misusing its power over local telephone service to gain a competitive advantage over potential specialized competitors.
The FCC’s Specialized Common Carriers decision was hardly a model of clarity.13 The decision did not define the specialized services to which it referred, nor did it define the corresponding obligations that the FCC expected the general carriers (primarily AT & T) to assume in order to assist the new carriers. AT & T contended, both at the time of the FCC decision and throughout the pendency of this lawsuit, that the Specialized Common Carriers decision authorized only point-to-point private line services not requiring switched network connections, and that the obligations of the Bell System extended only to providing local distribution facilities for these point-to-point private line services. MCI, by contrast, has consistently taken the position that the Specialized Common Carriers decision authorized it to provide FX and CCSA type services, as well as point-to-point private lines, and that AT & T had a corresponding obligation to provide it with the switched network connections required for these services. MCI also contended, both before and after the Specialized Common Carriers decision, that AT & T was obligated to provide it with local distribution facilities at the same rate at which AT & T provided such facilities to Western Union, under a longstanding contract between those two carriers. AT & T disagreed, claiming that the contract then in effect with Western Union did not reflect AT & T’s current costs, and that the price charged to MCI for local distribution facilities should be set so as to recover AT & T’s costs on a current basis.
In September 1971, AT & T entered into interim contracts with MCI defining the kinds of interconnections that AT & T would provide for MCI’s initial Chicago-St. Louis route and establishing the price for those interconnections. These contracts did not permit switched network connections for FX or CCSA type services,, nor was the price set by the contracts for local distribution facilities comparable to that charged to Western Union.
During this same time period, the original MCI investors joined forces with William McGowan, an experienced business executive and engineer, to form a venture that envisioned the eventual construction and operation of a nationwide long distance telephone system. After scrutiny of the market it believed had been opened by the Specialized Common Carriers decision, MCI created a plan contemplating sales of 74,000 circuits (leased telephone lines) having an average length of 500 miles per circuit, or approximately 37 million circuit miles14 by the end of 1975. According to this plan, MCI expected its revenues to average $1.00 per circuit mile excluding AT & T’s local connection charges, which MCI intended to pass on to its customers. Projected annual revenues for 1975 were approximately $350 million. Armed with these projections, MCI proceeded to raise $110 million by June 1972, making it one of the largest start-up ventures in the history of Wall Street. The funds were raised after review and analysis by leading lenders and large equipment suppliers who were either lending the funds or underwriting or guaranteeing the financing.
*1096MCI commenced operations over its Chicago-St. Louis route on January 1,1972. In the fall of 1972, MCI began construction of the first segment of its nationwide system, extending east and south from the original Chicago-St. Louis route. MCI initially expected to complete the first portion of its national network and commence customer service over major parts of the system by late summer 1973. Expansion to a second and a third group of smaller cities was to follow over the next three years. MCI planned to fund these capital expenditures from its initial $110 million capitalization, from substantial additional anticipated financing and from operating revenues.
B. The Interconnection Disputes
During late 1972, while construction was progressing, MCI entered into negotiations with AT & T over the provision by AT & T of interconnections and local distribution facilities on the expanded MCI system. Because MCI had previously experienced difficulty obtaining satisfactory interconnections for its Chicago-St. Louis segment, MCI hired an experienced lawyer-negotiator to secure a national interconnection agreement with AT & T that would permit MCI to serve the entire market it believed the FCC had opened. These negotiations began in September 1972, and continued with little progress for the next nine months.
During this same period, MCI appealed to the FCC for help in breaking down what it viewed as AT & T’s unreasonable negotiating stance. Through a series of informal complaints and conferences with FCC staff, MCI charged that AT & T was treating it unfairly, on the question of interconnections, in at least three respects:
(1) MCI claimed that AT & T was unlawfully denying it interconnections to the switched network for FX and CCSA services and for point-to-point service to customers located outside a local distribution area,15 including multipoint service;16
(2) MCI claimed that it was being charged excessive and discriminatory prices for the local distribution facilities provided by the Bell System; and
(3) MCI claimed that it was being harassed by Bell System employees in the provision of local distribution facilities through delays, improper installation, improper maintenance and other similar practices.
AT & T denied each of these charges. Both in its direct dealings with MCI and in its responses to FCC staff members, AT & T adhered to the position that the Specialized Common Carriers decision authorized only private line service not requiring switched network connections. AT & T also contended that it was providing MCI with all the interconnections to which MCI was entitled and that the prices it was charging for those interconnections were not excessive or unfair.
In August 1973, with negotiations still pending, and without informing MCI, AT & T decided to file with forty-nine of the state utility commissions interconnection tariffs that would be equally applicable to all carriers — including MCI and Western Union. By filing interconnection tariffs with the state commissions rather than with the FCC, AT & T made it more difficult for MCI to oppose the tariffs, since, in the words of one AT & T official, the interconnection “controversy would spread to 49 jurisdictions.” PX 2148 at 2031. Even after making this unilateral tariff decision, AT & T continued to “negotiate” with MCI. After MCI accidentally learned of the state tariff plan, however, AT & T formally broke off all contract negotiations.
*1097In early October 1973, several top MCI officials met with Bernard Strassburg, Chief of the FCC Common Carrier Bureau, to discuss a plan designed to resolve the interconnection controversies between MCI and AT & T. Pursuant to this plan, FCC Chairman Burch, on October 4, 1973, issued a letter on behalf of the Commission, rejecting AT & T’s resort to state regulatory agencies as unlawful and asserting exclusive FCC jurisdiction over the interconnection dispute. Shortly thereafter, MCI wrote to Mr. Strassburg, inquiring as to the nature and scope of the services that MCI was authorized to provide and for which AT & T was obliged to supply interconnections under the Specialized Common Carriers decision. Mr. Strassburg replied by letter dated October 19, 1973, that these services included FX and CCSA, as well as services outside local distribution areas and multi-point services. On November 2, 1973, MCI filed a complaint in federal district court under section 406 of the Communications Act asking that AT & T be ordered to provide interconnections for these services.
On December 31,1973, the United States District Court for the Eastern District of Pennsylvania issued a preliminary injunction ordering AT & T to provide all of the interconnections sought by MCI, on the theory that such interconnections were contemplated and required by the FCC’s Specialized Common Carriers decision. MCI Communications Corp. v. AT & T, 369 F.Supp. 1004 (E.D.Pa.1973). AT & T provided the required interconnections, but immediately appealed the district court’s injunction. Meanwhile, the FCC, on December 13, 1973, issued its own order requiring AT & T to show cause why it should not be held to have violated the Specialized Common Carriers decision by refusing to provide the interconnections requested by MCI.
On April 15, 1974, the Third Circuit reversed the preliminary injunction issued against AT & T. MCI Communications Corp. v. AT & T. 496 F.2d 214 (3d Cir.1974). On April 16, 1974, despite assurances that the FCC’s “show cause” decision was expected “any day now,” and despite FCC warnings that disconnection of MCI’s customers would violate the Communications Act, AT & T ordered its local operating companies to disconnect MCI’s customers on twenty-four hours notice. MCI alleged that the resulting disconnections caused turmoil among its customers and seriously damaged its reputation for reliable service. On April 23,1974 — eight days after the Third Circuit had vacated the injunction obtained by MCI — the FCC issued a decision ordering AT & T to provide the disputed interconnections.17 Bell System Tariff Offerings of Local Distribution Facilities for Use by Other Common Carriers, 46 F.C.C.2d 413, aff’d sub nom. Bell Telephone Co. v. FCC, 503 F.2d 1250 (3d Cir.1974), cert. denied, 422 U.S. 1026, 95 S.Ct. 2620, 45 L.Ed.2d 684 (1975). The FCC held that it had intended to include both FX and CCSA services within the terms “specialized” or “private line” services as those terms were used in the Specialized Common Carriers decision. 46 F.C.C.2d at 425-27. AT & T provided the requested interconnections within ten days of the FCC’s order.
C. The Execunet Decision
In October 1974, MCI filed a tariff with the FCC for what the tariff referred to as metered use private line services, principally a service called “Execunet.” Although the FCC did not immediately perceive it as such, this tariff was apparently designed to permit MCI to provide ordinary switched long distance service to users in any city to which its microwave system extended. See MCI Telecommunications Corp., 60 F.C.C.2d 25, 40-43 (1976) (the “Execunet decision”). When the FCC discovered the nature and purpose of the new tariff, it declared the tariff unlawful and ordered MCI to discontinue providing ordinary long distance message service on the ground that the Special*1098ized Common Carriers decision limited MCI’s authorization to the provision of private line services. 60 F.C.C.2d at 35-44, 58.
MCI appealed the FCC’s Execunet decision to the Court of Appeals for the District of Columbia Circuit and, in July 1977, the Court of Appeals set the decision aside. MCI Telecommunications Corp. v. FCC, 561 F.2d 365 (D.C.Cir.1977), cert. denied, 434 U.S. 1040, 98 S.Ct. 781, 54 L.Ed.2d 790 (1978). In its opinion, the Court of Appeals assumed, without deciding, that “a service like Execunet was not within the contemplation of the [FCC] when it made the Specialized Common Carriers decision,” 561 F.2d at 378, but held that the FCC had not conducted a sufficient hearing — either during the Specialized Common Carriers proceeding or at any subsequent time — to justify any limitation on the operating authority of MCI and the other new specialized carriers. Id. at 378-80.
This decision by the District of Columbia Circuit — handed down long after the events involved in the instant case occurred — rendered virtually meaningless the debate between MCI and AT & T over the proper interpretation and definition of the specialized private line services to which the Specialized Common Carriers decision applied. AT & T also claims that it was only by virtue of this Court of Appeals decision that MCI was able to achieve profitability since, according to AT & T, MCI’s costs for private line services (including FX and CCSA) substantially exceeded the rates AT & T was then charging its large users under the Telpak tariff. See infra, pp. 1099-1100.
D. The Pricing Controversies Between MCI and AT&T
From the time of MCI’s entry into the telecommunications field, AT & T’s prices for specialized long distance services had been a significant source of controversy. Initially the principal controversy centered on AT & T’s Telpak tariff. The Telpak tariff, which accounted for most of AT & T’s private line circuits at the time MCI commenced operations, offered private line service to large users under two schedules: (1) the user could obtain the right to up to 60 circuits between any two points for $30 per mile per month, or an average of $.50 per circuit mile per month if all circuits were being used; or (2) the user could obtain the right to up to 240 circuits between any two points for $85 per mile per month, or an average of $.35 per circuit mile per month if all 240 circuits were being used. PX 821.
AT & T originally instituted its Telpak tariff in 1961 as a competitive response to the FCC’s decision to permit large telephone users to construct and operate their own private microwave systems.18 At the time MCI entered the industry, in 1969, a number of microwave manufacturers were contending in proceedings before the FCC that Telpak rates were too low and unfairly hindered efforts to interest large users in building their own microwave systems. At the same time, however, a number of large users, including the federal government, were resisting any efforts to increase the Telpak tariff and, indeed, were contending that Telpak rates were already too high.
In 1968, shortly before MCI obtained its first authorization to enter the telecommunications industry, AT & T was permitted to increase its Telpak rates on an interim basis. During the period 1969-1972, AT & T was able — over the strenuous objections of some Telpak users — to obtain FCC approval for two additional rate increases. Although MCI contended strongly before the FCC that AT & T’s Telpak tariff did not cover its fully distributed costs and was therefore predatory, the FCC, in 1977, ultimately rejected all of the attacks upon the Telpak tariff.19
*1099Concurrent with MCI’s entry into the telecommunications field, AT & T also initiated studies to consider nationwide deav-eraging of its rates for individual private line service. Pursuant to these studies, AT & T formulated a plan known as the Hi-Lo tariff, which provided for the deaveraging of AT & T’s individual private line service into two principal rate categories.20 Under Hi-Lo, AT & T would lower its rates on certain “high density” long distance routes, many of which MCI planned to serve. At the same time, AT & T would increase its rates between so-called “low-density” cities, most of which MCI was not planning to serve. In February 1973, the month after MCI had announced its plans and prices for nationwide service, AT & T announced Hi-Lo to the public and sought permission from the FCC to file the new tariff. AT & T did not actually receive permission to file its Hi-Lo tariff until November 15, 1973, and the new tariff finally became effective on June 13, 1974.
E. MCI’s Damage Evidence
Faced with unproductive negotiations, a “chilled” market caused by AT & T’s early announcement of Hi-Lo, and curtailed sales commitments stemming in part from customer awareness of MCI’s interconnection difficulties, MCI in mid-1973 began to pare down its construction program. Because the company’s revenues were substantially lower than originally anticipated, MCI decided to defer construction on fifteen of the thirty-four routes contained in its original plan. ■ In addition, MCI terminated almost one-third of its employees and renegotiated its bank loans to secure permission to use loan proceeds for working capital rather than for additional construction. Although MCI survived and eventually prospered, it alleges in the instant lawsuit that by the time the interconnection dispute was finally resolved, in May 1975, it had a far smaller system, slower growth rate and related lower net cash flows and profits than it would have had absent AT & T’s unlawful interference.
At trial, MCI’s proof of damages was based almost entirely on a lost profits study authored by MCI’s former controller, Mr. Uhl. This study compared the profits that a hypothetical MCI — undamaged by AT & T’s allegedly unlawful actions — would have enjoyed with MCI’s actual and projected profit figures for the years 1973-1984.21 The revenues posited for the “undamaged” MCI were based upon projections made by MCI in 1971-1972 and previously used for financing purposes. Among other presumptions, these revenue projections assumed that AT & T’s Telpak service — which the jury in this case found to be lawfully priced and marketed — would not be in existence during the relevant time period. Costs for the “undamaged” MCI were derived from MCI’s actual operating experience. These revenue and cost projections were then used to compute MCI’s “lost profits,” measured in net cash flow, for each of the years 1973-1994.22 These computations resulted in an aggregated before-tax damage claim of $900,468,000.
AT & T, at trial, sharply disputed the accuracy of MCI’s revenue projections. AT & T argued that MCI’s own lost profits study demonstrated that MCI could never *1100have achieved profitability in the private line business, even including FX and CCSA services, since MCI’s costs for such services substantially exceeded the rates AT & T was then charging its large business users under the Telpak tariff. According to AT & T, MCI’s lost profits study showed MCI’s costs to be $.63 per circuit mile per month assuming that it could obtain local distribution facilities at the Western Union contract rates and $.74 per circuit mile per month, if it had to pay for those facilities on the basis of the current prices charged by AT & T. On either basis, AT & T argued that MCI's costs were substantially in excess of the Telpak rates and, hence, that MCI could not have undercut these rates and still have covered its costs.23 AT & T also argued that because its ordinary long distance rates are averaged on a nationwide basis, and because state and federal regulatory policy has traditionally required AT & T to set its long distance rates high enough to subsidize its less profitable local telephone service, MCI and other specialized carriers, by competing exclusively in the most lucrative long distance markets, could easily undercut AT & T’s artificially elevated long distance rates.
II. REGULATION AND THE ANTITRUST LAWS
A. The Federal Regulatory Scheme for Telecomm unications
The first venture of the federal government into the regulation of telecommunications was section 7 of the Mann-Elkins Act of 1910,24 which added telephone and telegraph companies to the list of common carriers regulated by the Interstate Commerce Commission (“ICC”). The Mann-Elkins Act imposed upon the newly-designated common carriers the obligation to provide service upon request at just and reasonable rates, without unjust discrimination or undue preference.25 The Act did not, however, subject the telecommunications industry to the broad tariff and regulatory jurisdiction enjoyed by the ICC over railroads. See Essential Communications Systems v. AT & T, 610 F.2d 1114, 1117-19 (3d Cir. 1979) (detailing early regulation of telecommunication and railroad industries).
Competition among telephone services in the same geographic area was, in the early part of the century, a fact of life. Thus, the enactment, in 1914, of the Clayton Act’s antimerger provisions26 presented a serious obstacle to the development of an integrated national telephone network. The Willis-Graham Act addressed this problem by authorizing the ICC to approve the consolidation of telephone company properties into single companies if such consolidation was “of advantage to the persons to whom service is to be rendered and in the public interest.” Willis-Graham Act of 1921, ch. 20, 42 Stat. 27 (1921) (current version at 47 U.S.C. § 221(a) (1976)). The statute granted express immunity from the antitrust laws for such consolidations. Id.
Thus, as of 1921, federal law recognized the telecommunications industry as a common carrier, subject to the consumer protection and non-discrimination provisions of the Mann-Elkins Act and exempt from antitrust liability for consolidations of competing local service systems. In other respects, however, the industry was subject to the antitrust laws. Indeed, in 1914, a government antitrust suit produced a consent decree against AT & T. See Essential Communications, 610 F.2d at 1119 & n. 19. Aside from the ICC’s jurisdiction to enforce AT & T’s common carrier obligations, AT & T was free to determine its own rates, return on investment and service obliga*1101tions. Federal law did not even impose upon AT & T an obligation to interconnect with other communications common carriers, although AT & T’s local subsidiaries were subject to regulation at the state level. Id. at 1119.
In 1934, Congress enacted the Federal Communications Act, 47 U.S.C. § 151 et seq. (1976), which constitutes the primary-federal regulatory mechanism for the telecommunications industry today. The 1934 Act severed regulation of the telephone, telegraph and radio industries from the ICC, and vested regulatory jurisdiction over those industries in the newly created Federal Communications Commission. The Act carried forward, almost verbatim, many provisions of the Mann-Elkins Act of 1910— for example, the just and reasonable tariff requirement and the prohibition against unjust or unreasonable discrimination.27 The 1934 Act also imposed certain new obligations on the telecommunications industry— for example, the requirement that regulated carriers interconnect or establish through routes with other common carriers. See 47 U.S.C. § 201(a) (1976).
With respect to tariffs, the 1934 Act continued the prior practice that tariffs be generated, at least in the first instance, by the carriers themselves. Under section 203(a) of the Act, these tariffs must be filed with the FCC, and carriers must give the FCC and the public ninety days notice of any proposed changes. 47 U.S.C. § 203(a) (1976); 47 U.S.C.A. § 203(b) (West Supp. 1982). No charge may be demanded or collected, or any service rendered, except in accordance with a filed tariff. Id. § 203(c). Section 204 of the Act further authorizes the FCC, either sua sponte or upon request, to conduct a hearing concerning the lawfulness of the rates embodied in a proposed tariff and to suspend operation of the tariff for up to five months. Id. § 204. If the Commission determines that the new tariff does not meet the requirements of the Act, it may prescribe a “just and reasonable” substitute, or set maximum and/or minimum charges to be observed. Id. § 205; see American Broadcasting Companies v. FCC, 643 F.2d 818, 822 (D.C.Cir.1980). Any carrier which knowingly fails to obey an FCC order issued under this section is liable for a fine of $1000 per violation per day. In addition, any common carrier which does or causes to be done any act prohibited or declared unlawful by the Communications Act shall be liable “to the person or persons so injured thereby for the full amount of damages,” plus attorneys’ fees. 47 U.S.C. § 206 (1976).
B. Implied Immunity
AT & T contends that the district court should have dismissed this suit on its motion because the FCC’s regulatory control over AT & T’s conduct renders AT & T immune from antitrust liability.28 The trial court denied the motion in a well-reasoned memorandum opinion. MCI Communications Corp. v. AT & T, 462 F.Supp. 1072 (N.D.Ill. 1978). Judge Grady traced the legislative history of the Federal Communications Act, and concluded that while AT & T is subject to considerable regulatory control and supervision, there is no indication that the Act was meant to immunize a carrier such as AT & T from the antitrust laws. 462 F.Supp. at 1086-87. Moreover, he concluded, the regulatory scheme to which AT & T is subject is not so wholly inconsistent with the antitrust laws as to require immunity. AT & T is not subject to conflicting requirements, nor would it be held liable for decisions which were not its own business judgment. The district court noted that the FCC did not sanction AT & T’s conduct with regard to interconnections nor dictate its tariffs. Thus, while certain actions might ultimately have been subject to agency review, the initial decisions were the product of AT & T’s private business judg*1102ment, and were not so heavily regulated as to remove them from AT & T’s control.
On appeal, AT & T contends that the district court’s decision incorrectly focused on blanket immunity rather than immunity for the particular actions of which MCI complained. Thus, AT & T argues that the critical question left unconsidered by the district court is “whether the charges in this case do in fact relate to matters basic to the pervasive regulatory scheme to which AT & T is subject.” Appellant’s Br. at 188. Our reading of the district court’s opinion, however, convinces us that it did not, as AT & T insists, miss the point now raised on appeal. While the district court did address the question of “blanket immunity” (i.e., whether regulation by the FCC under the public interest standard contained in the Communications Act is wholly inconsistent with the antitrust laws), 462 F.Supp. at 1078, 1080-82, it also fully considered AT & T’s “fall back position ... that even though all of AT & T’s conduct may not be immunized, the FCC, in its pervasive regulation, has approved each of the allegedly anticompeti-tive activities of which MCI complains and that therefore AT & T should obtain at least ad hoc immunity from antitrust laws.” Id at 1078, 1082-1102. For the reasons largely set forth in the district court’s memorandum opinion denying AT & T’s motion to dismiss, we reject AT & T’s assertion of implied immunity.
As the district court recognized, the Communications Act of 1934 does not expressly grant AT & T immunity from the antitrust laws for the conduct challenged in the instant case. Nor does the legislative history of the Communications Act indicate how Congress intended that the Act and the antitrust laws were to be reconciled. See United States v. AT & T, 461 F.Supp. 1314, 1321 (D.D.C.1978); Comment, AT & T and the Antitrust Laws: A Strict Test for Implied Immunity, 85 Yale L.J. 254, 269 (1975). It is well established, however, that regulated industries “are not per se exempt from the Sherman Act.” Georgia v. Pennsylvania R.R., 324 U.S. 439, 456, 65 S.Ct. 716, 725, 89 L.Ed. 1051 (1945). “Repeal of the antitrust laws by implication is not favored and not casually to be allowed. Only where there is a ‘plain repugnancy between the antitrust and regulatory provisions’ will repeal be implied.” Gordon v. New York Stock Exchange, 422 U.S. 659, 682, 95 S.Ct. 2598, 2611, 45 L.Ed.2d 463 (1975) (quoting United States v. Philadelphia National Bank, 374 U.S. 321, 350-51, 83 S.Ct. 1715, 1734, 10 L.Ed.2d 915 (1963)). As a further limitation, repeal is to be regarded as implied only where necessary to make the regulatory scheme work, and even then, only to the minimum extent necessary. Silver v. New York Stock Exchange, 373 U.S. 341, 357, 83 S.Ct. 1246, 1257, 10 L.Ed.2d 389 (1963); see National Gerimedical Hospital & Gerontology Center v. Blue Cross, 452 U.S. 378, 101 S.Ct. 2415, 69 L.Ed.2d 89 (1981).
Application of these general principles to a particular claim of implied immunity requires an evaluation of the specific regulatory scheme involved and the administrative authority exercised pursuant to that scheme. Northeastern Telephone Co. v. AT & T, 651 F.2d 76, 83 (2d Cir.1981), cert. denied, 455 U.S. 943, 102 S.Ct. 1438, 71 L.Ed.2d 654 (1982); see National Gerimedical Hospital & Gerontology Center v. Blue Cross. Thus, in our case, the inquiry must focus upon (1) whether the activities that are the subject of MCI’s complaint were required or approved by the Federal Communications Commission, pursuant to its statutory authority, in a way that is incompatible with antitrust enforcement, see, e.g., Gordon v. New York Stock Exchange, 422 U.S. 659, 95 S.Ct. 2598, 45 L.Ed.2d 463 (1975); Pan American World Airways, Inc. v. United States, 371 U.S. 296, 83 S.Ct. 476, 9 L.Ed.2d 325 (1963), or (2) whether these activities are so pervasively regulated “that Congress must be assumed to have forsworn the paradigm of competition.” Northeastern Telephone, 651 F.2d at 82; see United States v. AT & T, 461 F.Supp. 1314, 1324 (D.D.C.1978).
With respect to interconnections, we conclude, as did the district court, that the FCC’s regulatory authority under the *1103Communications Act does not preclude application of the Sherman Act. See 462 F.Supp. at 1089-96. The mere pervasiveness of a regulatory scheme does not immunize an industry from antitrust liability for conduct that is voluntarily initiated. Otter Tail Power Co. v. United States, 410 U.S. 366, 374, 93 S.Ct. 1022, 1028, 35 L.Ed.2d 359 (1973); see Comment, The Application of Antitrust Law to Telecommunications, 69 Calif.L.Rev. 497, 509 (1981). Although the FCC has authority to compel interconnection under section 201(a) of the Act, the initial decision whether to interconnect rests with the utility, and the record shows that the FCC did not control or approve of AT & T’s actions here. Nor has the FCC supervised AT & T’s interconnection practices so closely that the FCC’s approval could be inferred. Cf. Gordon v. New York Stock Exchange, 422 U.S. 659, 95 S.Ct. 2598, 45 L.Ed.2d 463 (1975).
Other circuits that have considered AT & T’s implied immunity in interconnection-type disputes have uniformly rejected arguments the same as or similar to those made by AT & T in the instant case. See, e.g., Northeastern Telephone Co. v. AT & T, 651 F.2d 76 (2d Cir.1981), cert. denied, 455 U.S. 943, 102 S.Ct. 1438, 71 L.Ed.2d 654 (1982); Phonetele, Inc. v. AT & T, 664 F.2d 716 (9th Cir.1981), cert. denied,-U.S. -, 103 S.Ct. 785, 74 L.Ed.2d 992 (1983); Mid-Texas Communications Systems v. AT & T, 615 F.2d 1372, 1377-82 (5th Cir.), cert. denied, 449 U.S. 912, 101 S.Ct. 286, 66 L.Ed.2d 140 (1980); Sound, Inc. v. AT & T, 631 F.2d 1324, 1327-31 (8th Cir.1980) (citing with approval Judge Grady’s memorandum opinion); Essential Communications Systems v. AT & T, 610 F.2d 1114 (3d Cir.1979); see also United States v. AT & T, 461 F.Supp. at 1320-30. But see Southern Pacific Communications Co. v. AT & T, 556 F.Supp. 825 (D.D.C. 1982). We agree with the reasoning of these decisions and are not persuaded that a contrary result is warranted here.
AT & T relies heavily on Hughes Tool Co. v. Trans World Airlines, Inc., 409 U.S. 363, 93 S.Ct. 647, 34 L.Ed.2d 577 (1973), and Pan American World Airways, Inc. v. United States, 371 U.S. 296, 83 S.Ct. 476, 9 L.Ed.2d 325 (1963), to support its claim that “matters at the heart of a pervasive scheme of common carrier, or public utility, regulation [here, presumably, AT & T’s interconnection and pricing policies] are immune from antitrust liability.” Appellant’s Br. at 183. In both of these cases, however, the Supreme Court found that the transactions challenged as violative of the antitrust laws fell precisely within the detailed scheme of administrative oversight established by Congress. Thus, in Hughes Tool, the Court held that where the Civil Aeronautics Board (CAB) had specifically authorized certain transactions between a parent and its subsidiary, those transactions were immunized from antitrust liability by section 414 of the Federal Aviation Act, 49 U.S.C. § 1378 (1976). Similarly, in Pan American Airways, the Court held that section 411 of the Federal Aviation Act granted to the CAB the very jurisdiction over the division of territories and allocation of air carrier routes that was the subject of the government’s antitrust complaint. In the instant case, by contrast, neither AT & T’s interconnection decisions nor its price structure policies are dictated, in the first instance, by the FCC (although, of course, AT & T’s overall rate of return is subject to continuing surveillance). Moreover, to the extent that any FCC decisions are relevant to AT & T’s claim of implied immunity, those decisions disapprove of, rather than condone, AT & T’s actions. Thus, this is not a case like Hughes Tool or Pan American Airways, where the refusal to grant antitrust immunity could subject AT & T to conflicting and potentially irreconcilable liability standards. See also Phonetele, Inc., 664 F.2d at 732-34.
AT & T also cites the case of FCC v. RCA Communications, Inc., 346 U.S. 86, 73 S.Ct. 998, 97 L.Ed. 1470 (1953), for the proposition that the public interest standard embodied in the Communications Act is inconsistent and thus presumably irreconcilable with the policy of the antitrust laws favoring competition. However, the Eighth Circuit, in Sound, Inc. v. AT & T, 631 F.2d 1324 (8th Cir.1980), recently rejected precisely *1104this irreconcilability argument. In Sound, Inc., AT & T argued that it was exempt, by virtue, inter alia, of the public interest standard contained in the Communications Act, from antitrust liability arising out of its rate structure and marketing practices for terminal telephone equipment. In rejecting AT & T’s assertion that the public interest standard of the Communications Act was necessarily inconsistent with the pro-competition standard of the antitrust laws, the Eighth Circuit noted that the FCC had exercised its supervisory authority so as to encourage rather than discourage competition in the terminal equipment market. In light of this policy, the court concluded that “the maintenance of an antitrust suit will not conflict with the operation of the regulatory scheme authorized by Congress but will supplement that scheme.” 631 F.2d at 1330. Similarly, in the instant case, the interconnection policies adopted by the FCC during the time period relevant to this litigation appear designed to promote rather than inhibit competition in the specialized telecommunications field. Thus, the allowance of antitrust liability is likely to complement rather than undermine the applicable statutory scheme.
AT & T’s assertion of implied immunity with respect to MCI’s predatory pricing allegations presents a closer question. Because section 201(b) of the Communications Act requires that AT & T’s rates be “just and reasonable,” and because both AT & T’s rates and rate making methodology are subject to continuing supervision by the FCC, it is probable that AT & T enjoys less flexibility in setting rates than it does, for example, in making initial interconnection decisions. Moreover, it can be argued that the hearing and enforcement provisions of the Communications Act itself afford competitors such as MCI an adequate opportunity to contest and seek relief from tariffs they consider unreasonable or unfair.29 Although these arguments are not entirely without merit, we believe that, under the particular circumstances of this case, AT & T is not entitled to antitrust immunity for the competitive rate filings which form the basis of MCI’s predatory pricing claims.
Although the Communications Act grants the FCC potentially broad authority over interstate and foreign telephone rates, in practice this authority is considerably more circumscribed. First, as the district court in this case noted, the Act gives the carrier sole responsibility for filing a tariff, and a carrier may file a new or revised tariff at any time. See 47 U.S.C. § 204 (1976). Thus, it is AT & T, not the FCC, that has the primary responsibility for initiating and setting both regular and private line telephone rates. See Sound, Inc., 631 F.2d at 1330. “When [such decisions] are governed in the first instance by business judgment and not regulatory coercion, courts must be hesitant to conclude that Congress intended to override the fundamental national policies embodied in the antitrust laws.” Otter Tail, 410 U.S. at 374, 93 S.Ct. at 1028.30
Moreover, although the Communications Act gives the FCC the right to conduct hearings on proposed tariffs, a new tariff automatically goes into effect after 90 days unless acted upon by the FCC in its discretion. See 47 U.S.C. § 203(b)(1) (Supp.1981). Thus, the FCC does not expressly approve or adopt as agency policy every tariff it permits to become effective. “By permitting a tariff to go into effect, the FCC does not assert that it has examined the content of the tariff and found it necessary or appropriate to effectuate the regulatory program, nor does it have an obligation under the Act to make such a finding.” Phonetele, 644 F.2d at 733; see Essential Communications, 610 F.2d at 1124; MCI Telecommunications Corp. v. FCC, 561 F.2d 365, 374 (D.C.Cir.1977), cert. denied, 434 U.S. 1040, 98 S.Ct. 781, 54 L.Ed.2d 790 (1978).
*1105The less than comprehensive nature of the FCC’s authority over tariffs is further reinforced by the huge volume of tariff filings received by the Commission. During the twelve month period between September 1974 and August 1975, for example, the FCC received 1,371 tariff filings, totaling 11,491 pages. Because of this volume, it was able to investigate only a small percentage of the tariffs filed. See United States v. AT & T, 461 F.Supp. at 1326. Recognizing these practical limitations on its regulatory jurisdiction, the FCC has acknowledged, in an antitrust case involving implied immunity questions similar to those at issue here, that “rate filings generally proceed from the carrier’s independent judgment ...” Id. at 1326 (quoting Memorandum of FCC, filed December 30, 1975, pp. 19-20). Moreover, the FCC has consistently maintained — in contrast to the SEC in the stock exchange cases relied upon by AT & T — that antitrust enforcement is not precluded in this area.31 United States v. AT & T, 461 F.Supp. at 1326. Finally, as is the case in the interconnection context, the actual FCC decisions relevant to the pricing policies challenged as predatory in the instant ease have tended to disapprove of, rather than support, those policies.32 We thus conclude that where, as here, the pricing decisions complained of are more the result of business judgment than regulatory coercion, and the FCC has neither dictated nor approved of those decisions, the challenged rate filings are not immune from antitrust scrutiny.33 See City of Kirkwood v. Union Electric Co., 671 F.2d 1173, 1176-79 (8th Cir.1982), cert. denied, - U.S. -, 103 S.Ct. 814, 74 L.Ed.2d 1013 (1983) (no immunity for rate filing under similar provisions of Federal Power Act); City of Mishawaka v. Indiana & Michigan Electric Co., 560 F.2d 1314, 1318-21 (7th Cir.1977), cert. denied, 436 U.S. 922, 98 S.Ct. 2274, 56 L.Ed.2d 765 (1978) (denying immunity for price squeeze claim arising out of relationship between electric utility’s filed wholesale and retail rates); cf. Cantor v. Detroit Edison Co., 428 U.S. 579, 96 S.Ct. 3110, 49 L.Ed.2d 1141 (1976) (denying state action immunity for light-bulb-exchange program contained in tariff approved by state public utility commission).
C. The Impact of Regulation
Our conclusion that AT & T is not entitled to antitrust immunity in the instant ease does not mean that AT & T’s status as a regulated common carrier is irrelevant to our evaluation of AT & T’s conduct. On the contrary, an industry’s regulated status is an important “fact of market life,” the impact of which on pricing and other competitive decisions “is too obvious to be ignored.” ITT v. General Telephone and Electronics Corp., 518 F.2d 913, 935-36 (9th Cir.1975) (footnote omitted). For this reason, the Supreme Court has repeatedly recognized that consideration of federal and state regulation may be proper even after the issue of antitrust immunity has been resolved. United States v. Marine Bancorporation, 418 U.S. 602, 627, 94 S.Ct. 2856, 2872, 41 L.Ed.2d 978 (1975) (application of antitrust doctrine to bank mergers *1106“must take into account the unique federal and state restraints on [defendant’s conduct]. Failure to do so would produce misconceptions that go to the heart of the doctrine itself”); see Silver v. New York Stock Exchange, 373 U.S. 341, 360-61, 83 S.Ct. 1246, 1258-1259, 10 L.Ed.2d 389 (1963) (although applicable statutory scheme not sufficiently pervasive to create antitrust immunity, particular acts of self regulation — even if in restraint of trade — may be justified with reference to that scheme); Otter Tail, 410 U.S. at 381, 93 S.Ct. at 1031 (court, in fashioning antitrust remedy, “should [not] be impervious to [regulated utility’s] assertion that compulsory interconnection or wheeling will erode its integrated system and threaten its capacity to serve adequately the public”).
Similarly, several recent decisions of the courts of appeals involving regulated industries have emphasized the “continuing significance of regulation” in evaluating alleged antitrust violations. Mid-Texas Communications Systems v. AT & T, 615 F.2d 1372, 1385 (5th Cir.1980), cert. denied, 449 U.S. 912, 101 S.Ct. 286, 66 L.Ed.2d 140 (1980) (antitrust laws “are not so inflexible as to deny consideration of government regulation.”); Almeda Mall, Inc. v. Houston Lighting & Power Co., 615 F.2d 343, 354 (5th Cir.), cert. denied, 449 U.S. 870, 101 S.Ct. 208, 66 L.Ed.2d 90 (1980) (“Monopolization cases involving ... regulated industries are special in nature and require close scrutiny.”); Jacobi v. Bache & Co., 520 F.2d 1231, 1237-39 (2d Cir.1975), cert. denied, 423 U.S. 1053, 96 S.Ct. 784, 46 L.Ed.2d 642 (1976) (rejecting application of per se liability rule in light of regulation of stock exchange); ITT, 518 F.2d at 935-36 (impact of regulations must be assessed as fact of market life). As Professors Areeda & Turner have stated:
[Antitrust courts can and do consider the particular circumstances of an industry and therefore adjust their usual rules to the existence, extent, and nature of regulation. Just as the administrative agency must consider the competitive premises of the antitrust laws, the antitrust court must consider the peculiarities of an industry as recognized in a regulatory statute.
1 P. Areeda & D. Turner, Antitrust Law ¶ 223d (1978).
Whether in a regulated context or not, the broad outline of the offense of monopolization is well understood. Most recently, the Supreme Court has stated:
The offense of monopoly under § 2 of the Sherman Act has two elements: (1) the possession of monopoly power in the relevant market and (2) the willful acquisition or maintenance of that power as distinguished from growth or development as a consequence of a superior product, business acumen, or historic accident.
United States v. Grinnell, 384 U.S. 563, 570-71, 86 S.Ct. 1698, 1703-1704, 16 L.Ed.2d 778 (1966); see Berkey Photo, Inc. v. Eastman Kodak Co., 603 F.2d 263, 274-76 (2d Cir.1979), cert. denied, 444 U.S. 1093, 100 S.Ct. 1061, 62 L.Ed.2d 783 (1980). Cases dealing with non-regulated industries have developed a number of analytic tools designed to aid courts in identifying each of these elements. In many instances, however, these tools are of only limited value in resolving monopolization charges against regulated monopolies. See Watson & Brun-ner, Monopolization by Regulated “Monopolies”: The Search for Substantive Standards, 22 Antitrust Bull. 559, 563 (1977). In particular, the presence of a substantial degree of regulation, although not sufficient to confer antitrust immunity, may affect both the shape of “monopoly power” and the precise dimensions of the “willful acquisition or maintenance” of that power. Id.
According to the Supreme Court, monopoly power may be defined as “the power to control prices or exclude competition” in a relevant market. United States v. E.I. duPont de Nemours & Co., 351 U.S. 377, 391, 76 S.Ct. 994, 1004, 100 L.Ed. 1264 (1956). In many cases involving unregulated industries, however, courts have eschewed examination of the ostensible monopolist’s actual degree of control over prices or competition, and have relied solely on statistical data concerning the accused firm’s share of the *1107market. Where that data reveals a market share of more than seventy to eighty percent, the courts have inferred the existence of monopoly power. See, e.g., United States v. Grinnell, 384 U.S. at 571, 86 S.Ct. at 1704; American Tobacco Co. v. United States, 328 U.S. 781, 797, 66 S.Ct. 1125, 1133, 90 L.Ed. 1575 (1946); Standard Oil Co. v. United States, 221 U.S. 1, 33, 31 S.Ct. 502, 505, 55 L.Ed. 619 (1911).
Such a heavy reliance on market share statistics is likely to be an inaccurate or misleading indicator of “monopoly power” in a regulated setting. In many regulated industries, each purveyor of service, regardless of absolute size, is in a monopoly position with regard to its customers. Indeed, while a regulated firm’s dominant share of the market typically explains why it is subject to regulation, the firm’s statistical dominance may also be the result of regulation. See United States v. Marine Bancorporation, 418 U.S. at 633, 94 S.Ct. at 2875. For these reasons, the size of a regulated company’s market share should constitute, at most, a point of departure in assessing the existence of monopoly power. Ultimately, that analysis must focus directly on the ability of the regulated company to control prices or exclude competition — an assessment which, in turn, requires close scrutiny of the regulatory scheme in question.34
In the instant case, the district court properly instructed the jury that, in determining whether AT & T possessed monopoly power in the relevant market,
you may consider the effect of the FCC’s exercise of regulatory authority over prices and entry, including interconnection. Similarly, you may consider the effect of the exercise by state regulatory agencies of regulatory authority over prices and entry in connection with the provision of local services and facilities. That AT & T may have had the largest share or the entire share of the telephone business in certain areas would not be sufficient to establish that AT & T possessed monopoly power if in fact regulation by regulatory agencies prevented AT & T from having the power to restrict entry or control prices.
App. 1200.
Although the district court’s instructions in this area might have been more helpful if they had described, in more detail, the specific regulatory scheme to which AT & T was subject, see Mid-Texas, 615 F.2d at 1386-87, we believe the instructions, taken as a whole, adequately apprised the jury of its duty “to take into account the unique federal and state regulatory restraints” to which AT & T was subject. Id. at 1387. We, therefore, reject AT & T’s contention that the trial court’s instructions on this issue left the jury without any meaningful way to assess the impact of regulation on the existence or non-existence of AT & T’s monopoly power and constituted reversible error.
AT & T’s status as a regulated public utility also bears on the second element of a monopolization offense: the willful acquisition or maintenance of monopoly power. The precise dimensions of the “willfulness” standard have been the subject of considerable litigation and varying formulations even in cases involving unregulated industries. Some courts, building upon Judge Learned Hand’s noted opinion in United States v. Aluminum Co. of America, 148 F.2d 416 (2d Cir.1945), have concluded that monopolistic conduct can be presumed from the possession of monopoly power unless the accused firm affirmatively demon*1108strates that its monopoly position has been “thrust upon it.” Id. at 432; see American Tobacco Co. v. United States, 328 U.S. at 813-14, 66 S.Ct. at 1140-1141. Under this analysis, if the ordinary business conduct of a dominant firm leads to the acquisition or maintenance of monopoly power, that conduct is presumed to reflect the requisite willful monopolistic intent. Whatever merit this presumption may have in other contexts,35 we believe it is a particularly inappropriate means of identifying monopolistic conduct by a regulated utility or common carrier. For these industries, anticipating and meeting all reasonable demands for service is often an explicit statutory obligation. See, e.g., 47 U.S.C. § 201(a) (1976) (“It shall be the duty of every common carrier ... to furnish such communication service upon reasonable request therefor.”). To apply the Alcoa presumption to such conduct would be tantamount to holding that adherence to a firm’s regulatory obligation could, by itself, constitute improper willfulness in a section 2 monopolization case.
This circuit has already declined to endorse such an anomalous result. In City of Mishawaka v. American Electric 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), we specifically held that “[i]n the particular circumstances of a regulated utility ... entitled to recover its cost of services and provide its investors with a reasonable rate of return, we believe that something more than general intent should be required to establish a Sherman Act violation.” See also Watson & Brunner, supra, at 574-79 (willfulness by a regulated monopoly should be demonstrable only by evidence of predatory conduct or other exclusionary acts contrary to public policy). We reaffirm our holding in Mishawaka and, therefore, reject MCI’s contention on cross-appeal that the trial court erred in requiring MCI to prove that each allegedly anti-competitive act or practice attributed to AT & T was done with the intent to maintain a monopoly in the relevant market.36
The impact of regulation was also an important element of AT & T’s defense in the instant case. Particularly with regard to the interconnection controversy, AT & T argued that its dealings with MCI were reasonable and that they represented a good faith attempt to comply with AT & T’s regulatory obligations under section 201 of the Communications Act. AT & T claims that the trial court’s instructions improperly prevented the jury from considering this defense, in that the instructions were fatally “silent concerning the overall structure of the Communications Act, the public interest standards under which the provisions of that Act are administered by the FCC and to which common carriers are *1109required to conform their conduct, and the requirements set forth in the Act relating to the particular interconnection and pricing controversies presented to the jury for resolution.” Appellant’s Br. at 138.
:[13] MCI, by contrast, argues in- its cross-appeal that the district court gave too much credence to AT & T’s regulatory defense. In particular, MCI claims that the district court improperly held it to an “over-rigorous burden of proof” by instructing the jury that, if AT & T believed in good faith that interconnection with MCI would have violated established regulatory policies, then AT & T’s refusal to interconnect could not be considered anticompetitive conduct. We reject both parties’ contentions. The district court in this case properly allowed AT & T to assert a defense based on good faith adherence to its regulatory obligations. See Mid-Texas, 615 F.2d at 1388-90. The district court also properly articulated this defense in its instructions to the jury. Thus the district court instructed the jury that
MCI must prove' more, however, than the fact that AT & T refused to provide the interconnections. As you know, AT & T contends that it refused to provide the connections because it believed that it had not been ordered to do so, that MCI was not authorized to provide the service, and that it would have violated established regulatory policies for MCI to receive the connections. If AT & T refused the interconnections because of such reasons, believing in good faith that they justified the refusal, then the refusal to provide the interconnections was not anti-competitive conduct and cannot be considered conduct engaged in for the purpose of maintaining a monopoly.
MCI has the burden of proving that in refusing the FX and CCSA interconnections AT & T acted with anti-competitive intent, for the purpose of maintaining a monopoly, rather than for what it in good faith regarded as legitimate reasons.
App. 1201.
Similarly, with respect to the charge that AT & T unlawfully pre-announced its Hi-Lo tariff, the jury was told to consider AT & T’s contention that the time interval involved was reasonable and required by applicable regulations. In addition, the district court instructed the jury that:
With respect to those facilities and interconnections which AT & T did not provide, its position is that its failure to do so was based upon a good faith belief that it would have violated established regulatory policies and therefore that it acted reasonably in all the circumstances.37
App. 1200.
We believe these instructions adequately conveyed to the jury the substance of AT & T’s regulatory defense and, thus, allowed the jury to “consider the effect of regulation in ascertaining whether Bell misused its monopoly power.” Mid-Texas, 615 F.2d at 1389. We reject AT & T’s contention that the trial court’s failure to provide a more detailed exposition of the standards contained in the Communications Act constitutes reversible error. We also reject MCI’s counter-argument that the district court’s instructions in this area improperly placed upon MCI the burden of disproving AT & T’s subjective good faith. See California Computer Products, Inc. v. IBM Corp., 613 F.2d 727, 736 (9th Cir.1979) (holding that a verdict must be directed in favor of defendant when plaintiff’s evidence is insufficient to establish that defendant acted unreasonably). In the particular context of an industry subject to extensive and rapidly changing regulatory demands, we believe that an antitrust defendant is entitled both to raise and to have the jury *1110consider its good faith adherence to regulatory obligations as a legitimate antitrust defense. See Mid-Texas, 615 F.2d at 1389-90; City of Mishawaka, 616 F.2d at 985.
Finally, we believe the fact of FCC regulation is relevant to our analysis of antitrust principles in another, more subtle way. AT & T, as the dominant firm in a regulated industry recently opened in part to competition, is subject to dual, and sometimes conflicting, principles of regulatory and antitrust law. As already indicated, we believe the trial court properly reconciled these bodies of law by allowing AT & T to present evidence as to its good faith belief in its compliance with regulatory requirements. In addition, the fact of FCC regulation to some extent affects our view of the appropriate purposes and proper scope of antitrust law in the present context — specifically, whether we should focus our examination on economic efficiency and consumer benefit or whether we should more expansively consider the political and social consequences of bigness or concentration of economic power. Compare R. Bork, The Antitrust Paradox (1978) and R. Posner, Antitrust Law (1976) with L. Sullivan, Handbook of the Law of Antitrust § 2 (1977) and Pitofsky, The Political Content of Antitrust, 127 U.Pa.L.Rev. 1051 (1979) and Schwartz, “Justice” and Other Non-Economic Goals of Antitrust, 127 U.Pa.L. Rev. 1076 (1979).
Certain factors may tend to distinguish this from ordinary monopolization cases. AT & T is a public utility subject to public regulation, occupying a unique place in the American industrial scene. To the extent that it may have enjoyed economies of scale and significant technological resources, the political and regulatory judgment, until recently, has been to tolerate the political and social consequences of its size in the ostensible interest of reliable, effective and economic telecommunications service. Now this regulatory judgment has been drastically modified, and competition — with all its economic, political and social consequences — is transforming the telecommunications industry.
Certainly this transformation, carried out at the behest of regulatory authorities, is meeting the broadest objectives of the antitrust laws at least as effectively as they might be pursued by this court in this case. The FCC has exercised its powers under the Communications Act and has instituted sweeping pro-competitive changes in the telecommunications industry to accommodate the broad demands of national communications policy. We also note the role of the Justice Department, AT & T itself and the federal district court in the consent decree entered recently between AT & T and the government in the District Court for the District of Columbia. See United States v. AT & T, 552 F.Supp. 131 (1982). The massive restructuring of AT & T accomplished in that decree is an additional avenue through which the issues of the concentration of economic power in the Bell System, as well as its political power, are' being addressed.
We acknowledge with approval the populist origins of the antitrust laws as well as the preeminent role of the Sherman Act as a charter of economic freedom.38 But we also believe that, as we have pointed out, larger concerns about broad pro-competitive policy, economic concentration and political power have been, and are being at this very moment, effectively addressed by the regulators, and possibly by the Congress. Hence, we have tended to believe it appropriate to focus at this time and in this case upon the specific issues of economic efficiency and consumer benefit which are directly presented. Thus, our resolution of the allegations of predatory pricing and unlawful failure to interconnect MCI to Bell’s local distribution facilities has centered on the questions whether prices cover costs and whether the denied facilities are essential. We are, of course, not insensitive to broader social and political issues, but as indicated, *1111we think that our principal task is to deal in depth with the specific questions presented.
III. PREDATORY PRICING
At trial MCI alleged that AT & T had engaged in predatory pricing of both its Telpak and Hi-Lo services for long distance business communications. The jury found that Telpak was lawfully priced, but that Hi-Lo was priced below its fully distributed costs and was predatory. We disapprove this finding with respect to Hi-Lo because of erroneous instructions, the use of an improper cost standard and insufficiency of the evidence. We also disapprove the jury’s finding that AT & T unlawfully pre-an-nounced its Hi-Lo tariff. Further, we reject MCI’s cross-appeal on Telpak’s marketing plan and sustain the jury’s finding that Telpak was lawfully priced and marketed.
A. Jury Instructions
One of the crucial issues presented at trial concerned the proper standard for determining predatory pricing. Both parties presented expert testimony on this issue. AT & T argued that unless its prices for a particular service failed to cover that service’s long-run incremental costs the price could not be found predatory. MCI contended that proof of price below fully distributed cost was sufficient to establish predation.
At trial Judge Grady refused to instruct the jury as to which cost measure was the correct legal standard to determine predatory pricing. Instead, he left the choice of a cost-based standard for predation — in this case fully distributed costs (“FDC”) or long-run incremental costs (“LRIC”) — for the jury to decide as a question of fact. Judge Grady instructed the jury:
[Yjou’re going to have to decide whether it should be fully distributed costs on the one hand, or incremental costs on the other hand; and in doing that you’ll have to look at all the evidence and decide which is the cost that truly reflects the actual cost of producing the service.
The test for determining whether Hi-Lo was predatory is the same as for Tel-pak. Again, it is a question of whether the price covered what you consider the applicable cost. If it did, you may not infer predatory intent. If it did not, you may infer predatory intent.
Tr. 11486-87.39 As a result, the special verdict required the jury to check which cost standard it felt was appropriate and then decide whether AT & T’s prices were below that measure of cost: either LRIC or FDC.
This we hold to be error. The choice of a cost-based standard for evaluating claims of predatory pricing is a question of law to be decided by the trial judge. Thus, while several courts have stated that the appropriate cost-based standard for predation may differ depending on the facts of the case, see, e.g., Chillicothe Sand & Gravel Co. v. Martin Marietta Corp., 615 F.2d 427 (7th Cir.1980), both courts and commentators are united in regarding the selection of that standard as a question of law. Indeed, the entire judicial and academic struggle to enunciate an appropriate definition of predatory pricing reflects the legal rather than factual nature of the question. Since a finding of below-cost pricing permits the jury to infer, or even presume, anticompetitive intent, it is imperative that the judge instruct the jury on the relevant cost standard to compare with defendant’s prices. See 2 P. Areeda & D. Turner, supra, at ¶ 315.
MCI relies on Greenville Publishing Co. v. Daily Reflector, Inc., 496 F.2d 891 (4th Cir. 1974), to support the proposition that the jury may select the appropriate cost standard to evaluate a predatory pricing claim. MCI’s reliance here reflects an overly broad *1112interpretation of that case. In Greenville Publishing the Fourth Circuit reversed a grant of summary judgment for the defendant in a case charging monopolization and attempted monopolization. On the issue of predatory pricing the court took note of affidavits by the defendant purporting to show that the operation of the advertising guide in question was profitable and that prices covered average variable costs. Plaintiffs challenged both the actual calculation of these costs and the failure of the defendant to include in its cost calculations “any portion of the company’s fixed expenses or personnel costs.” Id. at 397 & n. 10.
The court in Greenville Publishing reversed the grant of summary judgment in favor of the defendant stating “[t]he sum of this evidence presents an issue of disputed fact.” Id. at 398. It is misleading, however, in the context of the instant case, to place much reliance on this sentence. The court in Greenville Publishing was not concerned with which entity — judge or jury — is empowered to select the proper cost-based standard for determining predation. Rather, the Greenville Publishing court was addressing the much more general issue of the propriety of summary judgment in a complex antitrust case. At the summary judgment stage, the plaintiffs in Greenville Publishing had presented no evidence on the issue of anticompetitive intent other than the pricing policies of the defendant. Hence, the propriety of summary judgment on plaintiffs’ monopolization and attempted monopolization claims turned entirely on whether any inferences of intent could be drawn from the relationship between the defendant’s prices and costs. The Fourth Circuit’s refusal to uphold the grant of summary judgment in Greenville Publishing merely represents the traditional view that summary judgment is generally inappropriate in complex antitrust cases where intent may be difficult to discern. Id. at 398 (citing Poller v. Columbia Broadcasting System, 368 U.S. 464, 82 S.Ct. 486, 7 L.Ed.2d 458 (1962)).
There is no support in the cases for the proposition that a jury may simply choose the cost-based standard it feels is most appropriate. Indeed, the only other purportedly apposite case cited by MCI in its brief, the district court’s opinion in Northeastern Telephone Co. v. AT & T, 497 F.Supp. 230 (D.Conn.1980), has been reversed on this very point, with the Second Circuit stating that the cost standard used to determine whether a monopolist’s prices were predatory was a legal question. 651 F.2d 76, 87 (2d Cir.1981), cert. denied, 455 U.S. 943, 102 S.Ct. 1438, 71 L.Ed.2d 654 (1982), rev’g in part 497 F.Supp. 230, 240-41 (D.Conn.1980). Judge Grady himself acknowledged that it is inappropriate for the jury to consider all possible economic theories of predation in ruling that MCI’s originally proffered profit-maximizing theory was inadequate as a matter of law.
B. Below Cost Pricing
Liability for predatory pricing represents an exception to the general antitrust regime which contemplates that no limits on price competition shall be imposed. Predatory pricing is prohibited because of the fear that a monopoly or dominant firm will deliberately sacrifice present revenues for the purpose of driving rivals from the market and then recoup its losses through higher profits earned in the absence of competition. See Northeastern Telephone Co. v. AT & T, 651 F.2d 76, 86 (2d Cir.1981), cert. denied, 455 U.S. 943, 102 S.Ct. 1438, 71 L.Ed.2d 654 (1982); Areeda & Turner, Predatory Pricing and Related Practices Under Section 2 of the Sherman Act, 88 Harv.L. Rev. 697, 698 (1975). [Hereinafter cited as Areeda & Turner, Predatory Pricing].
There is at present, in cases such as the one before us, no reliable way to determine whether predatory pricing has occurred without some comparison between the prices charged and a properly defined measure of the cost of production. A subjective test based wholly upon intent is almost incapable of distinguishing between pro- and anticompetitive price cuts by a monopolist. Areeda, Predatory Pricing (1980), 49 Antitrust L.J. 897, 899 (1980); R. Posner, supra, at 188. Nor is a subjective test *1113capable of identifying which pricing strategies represent rational business decisions and which have no legitimate business purpose and are designed only to injure competition.
In addition, a test based wholly on intent is unworkable.40 Even if it were possible to identify those persons within a firm whose intentions are relevant, the meaning of the evidence will usually be obscure. After all, competition consists of winning business from rivals. The intent to preserve or expand one’s market share is presumptively lawful. To encourage judges and juries to rely overly on nonprobative data allegedly bearing on a firm’s “state of mind” invites the twin mischiefs of (1) burdening litigation with thousands of documents about the firm’s motives and calculations; and (2) encouraging inconsistent and quixotic results. Areeda, Predatory Pricing (1980), 49 Antitrust L.J. 897, 899 (1980); see R. Posner, supra, at 189-90.41
In the absence of an objective standard, firms making pricing decisions in the presence of competition would be unable to ascertain what price reductions may be legally undertaken. Because the antitrust laws are designed to encourage vigorous competition, as well as to promote economic efficiency and maximize consumer welfare, such uncertainty seriously undermines the goals of antitrust enforcement. As one commentator has recently emphasized:
It is imperative that courts timely establish objective and understandable pricing standards which bring into sharp focus the line which separates commendable price reductions from predatory pricing practices. Such standards are necessary for the guidance of businessmen ... Businessmen should not be put into the position where they must either forego competitive price decreases or risk treble damages in Sherman Act suits.
Sherer, Predatory Pricing: An Evaluation of its Potential for Abuse Under Government Procurement Contracts, 6 J.Corp.L. 531, 539 (1981).
Within the past decade, both economists and lawyers have recognized the need for an objective standard to evaluate predatory pricing claims.42 Advocates of an objective test agree that price cuts by a dominant firm or monopolist are nothing more than lawful competition on the merits if prices remain above costs. Since such price cuts benefit consumers by providing greater output of desired goods at lower prices, they are pro-competitive and cannot result in the elimination of equally efficient competitors.
Similarly, the courts have nearly unanimously adopted some form of a cost-based standard in deciding questions of predation. E.g., Northeastern Telephone Co. v. AT & T, 651 F.2d 76 (2d Cir.1981), cert. denied, 455 U.S. 943, 102 S.Ct. 1438, 71 L.Ed.2d 654 (1982); Chillicothe Sand & Gravel Co. v. Martin Marietta Corp., 615 F.2d 427, 430-32 (7th Cir.1980); California Computer Products, Inc. v. IBM Corp., 613 F.2d 727, 742-43 (9th Cir.1979); Janich Bros. v. American Distilling Co., 570 F.2d 848, 857 (9th Cir. 1977), cert. denied, 439 U.S. 829, 99 S.Ct. 103, 58 L.Ed.2d 122 (1978); Pacific Engi*1114neering & Production Co. v. Kerr-McGee Corp., 551 F.2d 790, 797 (10th Cir.), cert. denied, 434 U.S. 879, 98 S.Ct. 234, 54 L.Ed.2d 160 (1977); National Association of Regulatory Utility Commissioners v. FCC, 525 F.2d 630, 637-38 & n. 34 (D.C.Cir.), cert. denied, 425 U.S. 992, 96 S.Ct. 2203, 48 L.Ed.2d 816 (1976).
MCI nonetheless argues in its cross-appeal that the district court erred in requiring it to prove that AT & T priced its Hi-Lo service below any measure of cost. MCI contends that, if AT & T knowingly sacrificed revenue (i.e., failed to maximize its profits) with the intent to injure competition, this court should hold that behavior to constitute unlawful predatory pricing. In support of this “profit maximization” theory, MCI cites a trio of cases. Hanson v. Shell Oil Co., 541 F.2d 1352, 1358 n. 5 (9th Cir.1976), cert. denied, 429 U.S. 1074, 97 S.Ct. 813, 50 L.Ed.2d 792 (1977); International Air Industries, Inc. v. American Excelsior Co., 517 F.2d 714, 724 (5th Cir.1975), cert. denied, 424 U.S. 943, 96 S.Ct. 1411, 47 L.Ed.2d 349 (1976); ILC Peripherals Leasing Corp. v. IBM Corp., 458 F.Supp. 423, 432 (N.D.Cal.1978), aff’d per curiam sub nom. Memorex Corp. v. IBM Corp., 636 F.2d 1188 (9th Cir.1980), cert. denied, 452 U.S. 972, 101 S.Ct. 3126, 69 L.Ed.2d 983 (1981).
Each of these cases contains language to the effect that a price may be predatory if it is below the short-run profit-maximizing price and barriers to new entry are great. Assuming, arguendo, that these statements are more than mere dicta, we must reject such a “profit maximization” theory as incompatible with the basic principles of antitrust. The ultimate danger of monopoly power is that prices will be too high, not too low. A rule of predation based on the failure to maximize profits would rob consumers of the benefits of any price reductions by dominant firms facing new competition.43 Such a rule would tend to freeze the prices of dominant firms at their monopoly levels and would prevent many pro-competitive price cuts beneficial to consumers and other purchasers. In addition a “profit maximization” rule would require extensive knowledge of demand characteristics — thus adding to its complexity and uncertainty. Another, and related, effect of adopting the “profit maximization” theory advocated by MCI would be to thrust the courts into the unseemly role of monitoring industrial prices to detect, on a long term basis, an elusive absence of “profit maximization.” Such supervision is incompatible with the functioning of private markets. It is in the interests of competition to permit dominant firms to engage in vigorous competition, including price competition. See Berkey Photo, Inc. v. Eastman Kodak Co., 603 F.2d 263, 273 (2d Cir. 1979), cert. denied, 444 U.S. 1093, 100 S.Ct. 1061, 62 L.Ed.2d 783 (1980). We therefore reject MCI’s “profit maximization” theory, and reaffirm this Circuit’s holding that liability for predatory pricing must be based upon proof of pricing below cost. Chilli-cothe Sand & Gravel Co. v. Martin Marietta Corp., 615 F.2d 427 (7th Cir.1980).
C. Defining Measures of Cost
The first commentators to propose a specific cost-based standard for predatory pricing were Professors Areeda and Turner, who argued that pricing below a firm’s short-run marginal cost should be deemed unlawful, and that prices above that level should be deemed lawful. Areeda & Turner, Predatory Pricing at 709-13. See also 3 P. Areeda & D. Turner, supra, at ¶ 711-15. In economic terms, short-run marginal cost represents the increment to total cost that results from producing an additional unit of output, where some inputs of production are variable and others are fixed. 3 P. Areeda & D. Turner, supra, at ¶ 712 at 155. Because short-run marginal cost is an economic concept that cannot be derived by conventional accounting methods, Areeda and Turner advocate the use of “average *1115variable cost” (“AVC”) as a proxy in predatory pricing cases. Variable costs, as the name implies, are costs that vary with changes in output. They typically include such items as materials, fuel, maintenance, and labor directly used to produce the product. Id. A product’s average variable cost is the sum of all its variable costs divided by the number of units of output.
The Areeda-Turner rule has engendered much discussion about whether short-run marginal (or, its proxy, average variable) cost represents the proper cost standard for evaluating predatory pricing claims.44 Much of the economic literature, as well as the case law, has examined whether average variable cost or some measure of average total cost (which includes “fixed” as well as variable costs) represents the better cost standard for measuring predatory pricing.
It is unfortunate that in the course of trial the case before us was characterized as a contest between the supporters and opponents of the Areeda-Turner rule. At trial long-run incremental cost was incorrectly equated with average variable cost while fully distributed cost was incorrectly equated with average total cost. In fact, the validity of the Areeda-Turner rule, based on short-run marginal costs, is not at issue in this case because neither party ever argued for a short-run cost standard. Rather, AT & T introduced evidence, unrefuted by MCI, showing that its prices for both Telpak and Hi-Lo were above those services’ long-run incremental costs.
There are important economic differences between long-run incremental cost and short-run marginal cost. First, incremental costs (LRIC) represent the average cost of adding an entire new service or product rather than merely the last unit of production. Professor Alfred Kahn has highlighted this distinction by stating:
[Mjarginal cost, strictly speaking, refers to the additional cost of supplying a single, infinitesimally small additional unit, while “incremental” ... refer[s] to the average additional cost of a finite and possibly a large change in production or sales.
1 A. Kahn, The Economics of Regulation 66 (1970) (emphasis in original).
Second, and more important, long-run incremental cost differs from average variable cost in that it is a long-run rather than a short-run cost measure. Because variable costs, by definition, are associated with the limited time period in which a firm cannot replace or increase its plant or equipment, the cost of plant and equipment is regarded as fixed and is not included in the calculation of a product’s short-run marginal, or average variable, cost. Long-run incremental cost, by contrast, measures all the costs of adding a new product or service — “fixed” as well as variable costs (and “capital” as well as “operating” items).45 Essentially, the LRIC approach assumes that all costs become variable in the long run. Hence, a number of the criticisms that have been leveled against the choice of a short-run marginal cost standard are not applicable to the use of long-run incremental cost. The use of long-run cost analysis may be particularly appropriate to capital-intensive processes where growth of plant and equipment is marked.
*1116In addition to incorrectly equating long-run incremental cost with short-run marginal (or average variable) cost, the district court (without adequate guidance from the parties) incorrectly equated fully distributed cost (“FDC”) with average total cost. Both these notions are incorrect because LRIC and FDC can be viewed as simply different ways of defining the average total cost (“ATC”) of a particular product or service for a firm that produces multiple products or services.46
For a single product firm, average total cost can be easily defined as the sum of all costs, both fixed and variable, divided by the total units of output produced by the firm. Such simple concepts of average total cost, however, lose their meaning when one considers a multi-service firm such as AT & T. Joint and non-joint common costs shared among products of the same firm render it impossible to calculate ATC simply by adding up costs and dividing by the number of units of output. This is possible only in a firm which produces a single product. One cannot proceed in this fashion for multiproduct firms because the total number of units produced include many different products each with different costs and different price and sales data. It is therefore necessary, in the multiproduct context, to determine what costs are caused by which products and services, and this requires some sort of differential (e.g., incremental) methodology.
In an antitrust context, fully distributed cost is not an economically relevant definition of average total cost and must be rejected as determinative.47 First, FDC is a quite arbitrary allocation of costs among different classes of service. There are countless FDC methods, each allocating costs by a different mathematical formula.48 Despite trenchant criticism on economic grounds,49 FDC continues to be widely used for regulatory purposes, inter alia, because of its ease of application in dividing an authorized total revenue requirement among individual products or services— much as a pie is divided into slices. But FDC cannot purport to identify those costs which are caused by a product or service, and this is fundamental to economic cost determination.
FDC also fails as an economically relevant measure of cost for antitrust purposes because it relies on historical or embedded costs. For it is current and anticipated cost, rather than historical cost that is rele*1117vant to business decisions to enter markets and price products. The business manager makes a decision to enter a new market by comparing anticipated additional revenues (at a particular price) with anticipated additional costs. If the expected revenues cover all the costs caused by the new product, then a rational business manager has sound business reasons to enter the new market. The historical costs associated with the plant already in place are essentially irrelevant to this decision since those costs are “sunk” and unavoidable and are unaffected by the new production decision. This factor may be particularly significant in industries such as telecommunications which depend heavily on technological innovation, and in which a firm’s accounting, or sunk, costs may have little relation to current pricing decisions.50
In particular, FDC fails as a relevant measure of cost in a competitive market. FDC is, at best, a rough indicator of an appropriate rate ceiling for regulatory purposes and should not be used as a measure of the minimum price permissible in a competitive market. The justifiable fear of monopoly, and the basis of section 2 of the Sherman Act, is that a firm enjoying monopoly power will not be constrained by market forces; it will raise prices and decrease output in such a manner that its own profit will be maximized but that consumers will be subject to higher prices and a less efficient allocation of resources than would be the case in a competitive market. A standard making predatory pricing illegal and subject to treble damages must be carefully structured to fit the needs of the Sherman Act and its encouragement of competition on the merits. See Janich Bros. v. American Distilling Co., 570 F.2d 848, 855 (9th Cir.1977), cert. denied, 439 U.S. 829, 99 S.Ct. 103, 58 L.Ed.2d 122 (1978). When a price floor is set substantially above marginal or incremental cost a price “umbrella” is created which allows less efficient rivals to remain in the market sheltered from full price competition. A fully distributed price floor may thus misallocate resources and force consumers to pay more for less production than competition would dictate.
The economic literature that has considered the problem of predatory pricing has rejected almost entirely the notion that fully distributed costs are a relevant measure of ATC.51 To the contrary, long-run incremental cost has been approved as an economically relevant measure of average total cost for one product produced by a multiproduct firm. Professor Baumol has stated in reply to the sloppy use of the term “average total cost”:
By average total cost, [one] surely does not mean fully allocated cost, which is a mare’s nest of arbitrary calculations parading as substantive information ... Consequently, I assume that when [one] requires the price of a good in the long-run to exceed its “average total cost,” [one] defines the latter to mean the average incremental cost of the product including any fixed cost outlays required by the item.
Baumol, Quasi-Permanence of Price Reductions: A Policy for Prevention of Predatory Pricing, 89 Yale L.J. 1, 9 n. 26 (1979). Professors Joskow- and Klevorick agree with this critique of fully distributed cost as a measure of average total cost:
*1118For a single-product firm, average total cost is easily defined. In the more likely multiproduct context, we are using “average total cost” to signify the average incremental cost of the commodity of concern and not any arbitrary “fully allocated cost measure.”
Joskow & Klevorick, A Framework for Analyzing Predatory Pricing Policy, 89 Yale L.J. 213, 252 n. 79 (1979). Since all costs are variable in the long run it is long-run incremental costs (including return on investment) which most closely measure anticipated average total cost. 3 P. Areeda & D. Turner, supra, at ¶ 712 at 156. Cf. R. Posner, supra, at 190.52
A simplified example of some of. these cost relationships can be found in Judge Wilkey’s dissenting opinion in Aeronautical Radio, Inc. v. FCC, 642 F.2d 1221, 1236 (D.C.Cir.1980), cert. denied, 451 U.S. 920, 101 S.Ct. 1998, 68 L.Ed.2d 311 (1981) (Wilkey, J., dissenting).53 In Judge Wilkey’s example, a judge accepts an invitation to participate in a law school moot court, with the school paying for his hotel room costing $125 per night. He later decides to bring his wife along even though the school’s moot court representative cannot assure him that his wife’s expenses will also be paid. Judge and Mrs. X attend the moot court, and their hotel bill for two is $150 per night. Upon his return, Judge X sends the moot court board his itemized expenses, noting that if the board has decided to pay for his wife’s trip, he should be reimbursed at $150 a day; if not, he should receive $125 a day, the amount it would have cost him had he attended the moot court alone. The moot court board sends back a check for $75, noting that it is unable to absorb the expenses of Mrs. X, and explaining that, using a fully distributed cost methodology, it has allocated one half of the couple’s daily $150 hotel bill to Judge X and the other half to his wife. Judge X is understandably both annoyed and confused; he knows that if he had attended the moot court alone, he would have been reimbursed at $125 a day, because this is what his actual hotel charge would have been, and because the moot court board’s original invitation had been extended on this basis. Judge- X is also angry because, had he known that the moot court board was going to penalize him in this manner, he would not have asked Mrs. X to accompany him, but would have come by himself at the agreed all-expenses paid rate of $125 a day. Thus, both practical considerations and economic theory dictate that the relevant cost of Mrs. X’s stay is $25 and that a marginal cost methodology should be used to analyze the judge’s travel expenses and other real world problems.54
*1119D. The Proper Cost Standard
This case, insofar as predatory pricing is concerned, is truly one of first impression for this circuit. The case law in this and most other circuits has thus far largely addressed the merits of short-run marginal cost (or its proxy, average variable cost) as compared with average total cost; the cases have not discussed the choice between long-run incremental cost and fully distributed cost as a way to measure average total cost. See, e.g., Chillicothe Sand & Gravel Co. v. Martin Marietta Corp., 615 F.2d 427 (7th Cir.1980); Borden, Inc. v. FTC, 674 F.2d 498, 515 (6th Cir.1982), petition for cert. filed, 51 U.S.L.W. 3271 (U.S. Aug. 25, 1982) (No. 82-328); O. Hommel Co. v. Ferro Corp., 659 F.2d 340 (3d Cir.1981), cert. denied, 455 U.S. 1017, 102 S.Ct. 1711, 72 L.Ed.2d 134 (1982); Americana Industries v. Wometco de Puerto Rico, Inc., 556 F.2d 625 (1st Cir.1977); International Air Industries v. American Excelsior Co., 517 F.2d 714 (5th Cir.1975), cert. denied, 424 U.S. 943, 96 S.Ct. 1411, 47 L.Ed.2d 349 (1976). Cf. Northeastern Telephone Co. v. AT & T, 651 F.2d 76, 89-90 (2d Cir.1981), cert. denied, 455 U.S. 943, 102 S.Ct. 1438, 71 L.Ed.2d 654 (1982) (rejecting use of fully distributed cost standard). The Supreme Court has not spoken on the entire issue of predatory pricing except to note a firm’s below cost pricing in a price discrimination case. Utah Pie Co. v. Continental Baking Co., 386 U.S. 685, 698-99, 87 S.Ct. 1326, 1333-1334, 18 L.Ed.2d 406 (1967).
Recently, several courts have questioned whether short-run marginal cost should be the exclusive standard for predatory pricing and have expressed a willingness to consider other factors. William Inglis & Sons Baking Co. v. ITT Continental Baking Co., 668 F.2d 1014 (9th Cir.1981), cert. denied, -U.S.-, 103 S.Ct. 58, 74 L.Ed.2d 61 (1982); International Air Industries v. *1120American Excelsior Co., 517 F.2d 714 (5th Cir.1975), cert. denied, 424 U.S. 943, 96 S.Ct. 1411, 47 L.Ed.2d 349 (1976); see generally Note, Predatory Pricing: The Retreat from the AVC Rale and the Search for a Practical Alternative, 22 B.C.L.Rev. 467 (1981) (hereinafter cited as Note, Retreat from AVC). Exclusive reliance on AVC (a proxy for short-run marginal cost) has been criticized primarily on the grounds that it focuses on short-run rather than long-run price cost comparisons, a criticism which, as noted earlier, is not fairly applicable to LRIC. See Note, Retreat from AVC at 484-85, 489-94.
This court in Chillicothe Sand & Gravel Co. v. Martin Marietta Corp., 615 F.2d 427 (7th Cir.1980), affirmed the use of an incremental cost methodology as the starting point for predatory pricing analysis. In that case, the Areeda-Turner standard based upon short-run marginal cost was cited as “both a relevant and an extremely useful factor” in identifying predatory conduct. 615 F.2d at 432. We are now required to move away from the AreedaTurner rule because we are faced with a choice between two different cost standards — LRIC or FDC — each of which may be argued to measure average total cost. If average total cost is the objective (and the principle of cost causation is to be honored), we think that LRIC is and FDC is not an appropriate method of getting at it.55 We, of course, do not close the door on such other methods — as yet undeveloped and undisclosed — as may be firmly based on the relation of cause and effect between the product or service involved and the costs it produces.
It is not surprising that no court has ever adopted fully distributed cost as the appropriate cost standard in a predatory pricing case. Most recently, the Second Circuit rejected fully distributed cost and adopted marginal cost as the test for predation in a case involving AT & T. In Northeastern Telephone Co. v. AT & T, 651 F.2d 76 (2d Cir.1981), cert. denied, 455 U.S. 943, 102 S.Ct. 1438, 71 L.Ed.2d 654 (1982), the court considered allegations that a Bell System affiliate had engaged in predatory pricing in the equipment market. The Second Circuit, in reversing the portion of the judgment relating to predatory pricing, stated:
Adopting marginal costs as the proper test of predatory pricing is consistent with the pro-competitive thrust of the Sherman Act. When the price of a dominant firm’s product equals the product’s marginal costs, “only less efficient firms will suffer larger losses per unit of output; more efficient firms will be losing less or even operating profitably.” ... Marginal cost pricing thus fosters competition on the basis of relative efficiency. Establishing a pricing floor above marginal cost would encourage underutilization of productive resources and would provide a price “umbrella” under which less efficient firms could hide from the stresses and storms of competition.
Id. at 87 (citation omitted).
The Second Circuit explicitly rejected the trial court’s reasoning that because AT & T was a multiservice regulated utility the use of fully distributed cost was appropriate. Id. at 89-90. The Second Circuit reiterated its conclusion that maintaining a price floor above marginal cost provided a haven for inefficient competitors. It then detailed the perverse effects of FDC pricing on consumer welfare and the competitive process itself. Finally, the court examined and rejected the argument that FDC was required to prevent cross-subsidization, explaining that if prices were above marginal cost no subsidies could exist and in fact contributions would be made to the overhead of the other Bell services. Id. at 90. See also Southern Pacific Communications Co. v. AT & T, 556 F.Supp. 825 (D.D.C. 1982).
*1121The Eighth Circuit has also rejected the use of fully allocated costs, although in a less definitive manner than the Second Circuit. In Superturf, Inc. v. Monsanto Co., 660 F.2d 1275 (8th Cir.1981), the court held that pricing below fully allocated cost but above average variable cost was not predatory, particularly in the absence of predatory intent or other conduct sufficient to render the pricing unreasonable.
Nor has FDC gained any adherents among district courts in the Sixth Circuit, which has not decided the validity of the Areeda-Turner short-run marginal cost approach. See Borden, Inc. v. FTC, 674 F.2d at 515 (affirming violation of section 5 of the FTC Act where monopolist had engaged in selective price cutting and promotional allowances in competitive markets only); cf. Brodley and Hay, Predatory Pricing at 780-86. In Hillside Dairy Co. v. Fairmont Foods Co., 1980-2 Trade Cas. ¶ 63,313 (N.D. Ohio 1980), the Northern District of Ohio considered a meeting competition defense to a price discrimination charge where the facts indicated that the defendant had inadvertently beaten rather than met its competitor’s dairy prices. The court held that such a defense would still prevail if the defendant had made substantial efforts to verify the actual price offered by its competitor, and did not operate at a loss in supplying the product. Id. at p. 75,625. The court, in holding for the defendant, explicitly chose average variable cost over fully allocated cost as the proper standard. Id. at p. 75,626.
The other circuits have been virtually unanimous in their endorsement of a marginal cost standard for predatory pricing. The Third Circuit stated recently in O. Hommel Co. v. Ferro Corp., 659 F.2d 340 (3d Cir.1981), cert. denied, 455 U.S. 1017, 102 S.Ct. 1711, 72 L.Ed.2d 134 (1982), that although the record before it obviated the need to choose explicitly among competing economic theories of predation, it was “inclined to accept the basic premise of the Areeda and Turner thesis that predatory intent may not be inferred from sales at or above average variable cost.” 659 F.2d at 352.56 Similarly, in International Air Industries v. American Excelsior Co., 517 F.2d 714 (5th Cir.1975), cert. denied, 424 U.S. 943, 96 S.Ct. 1411, 47 L.Ed.2d 349 (1976), the Fifth Circuit held that, except where barriers to entry are “extremely high,” a plaintiff claiming predatory pricing must show that the defendant “is charging a price below his average variable cost in the competitive market.” 517 F.2d at 724 & n. 31.57 *1122The First, Tenth and District of Columbia Circuits have also expressed their approval of the Areeda-Turner marginal cost test. See Americana Industries v. Wometco de Puerto Rico, Inc., 556 F.2d at 628; Pacific Engineering & Production Co. v. Kerr-McGee Corp., 551 F.2d 790, 797 (10th Cir.), cert. denied, 434 U.S. 879, 98 S.Ct. 234, 54 L.Ed.2d 160 (1977); AT & T v. FCC, 602 F.2d 401, 410 n. 49 (D.C.Cir.1979); National Association of Regulatory Utility Commissioners v. FCC, 525 F.2d 630, 638 n. 34 (D.C.Cir.), cert. denied, 425 U.S. 992, 96 S.Ct. 2203, 48 L.Ed.2d 816 (1976); Southern Pacific Communications Co. v. AT & T, 556 F.Supp. 825 (D.D.C.1982).
Only one circuit has ever permitted a jury to hear evidence of predation based on pricing above average variable but below average total cost. In William Inglis & Sons Baking Co. v. ITT Continental Baking Co., 668 F.2d 1014 (9th Cir.1981), cert. denied, -U.S.-, 103 S.Ct. 58, 74 L.Ed.2d 61 (1982), the Ninth Circuit held that it was permissible for a jury to find predation based on evidence that demonstrated pricing above average variable cost, if accompanied by intent. In Inglis the trial court had entered judgment n.o.v. for the defendant as a result of plaintiff’s failure to introduce evidence that prices were below marginal costs. Id. at 1026.
The Ninth Circuit reversed, noting its reluctance to adopt the Areeda-Turner rule as the exclusive test for predatory pricing. Id. at 1032. In place of Areeda-Turner, the court stated a new rule:
[W]e hold that to establish predatory pricing a plaintiff must prove that the anticipated benefits of defendant’s price depended on its tendency to discipline or eliminate competition and thereby enhance the firm’s long-term ability to reap the benefits of monopoly power. 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.
Id. at 1035-36. See also D & S Redi-Mix v. Sierra Redi-Mix and Contracting Co., 692 F.2d 1245 (9th Cir.1982).
Nothing in this statement supports the use of fully distributed cost in a predatory pricing case. The Ninth Circuit established a rule which allows a jury to hear evidence of pricing between ATC and AVC without any reference to FDC at all. The court in Inglis defined average total cost as the “portion of the firm’s total cost — both fixed and variable — attributable on an average basis to each unit of output.” Id. at 1035 n. 30. This definition is consistent with the use of long-run incremental cost as a measure of ATC, as advocated by Professors Baumol, Joskow and Klevoriek; it does not support the use of non-economic cost measures such as FDC. Moreover, the Inglis rule must be read narrowly to avoid conflict with prior Ninth Circuit decisions endorsing a marginal cost standard and with the specific reason given by the Ninth Circuit in Inglis for reversing the district court. See id. at 1032-33, 1036.
It is important to understand that the “average total cost,” to which some courts and commentators refer, should not be equated with FDC; it is, when properly understood, best measured in the multipro-duct context by average long-run incremental cost. Essentially, this is the case because LRIC, unlike FDC, only measures costs which are causally related to the service or product in question.58
*1123This is not an economist’s quibble or a theoretical musing; it is a matter of principled analysis and practical reality in the market place. Pricing at or above long-run incremental cost in a competitive market is a rational and profitable business practice. Because there are legitimate, and in fact compelling, business reasons for pricing products at or above their long-run incremental cost, no predatory intent should be presumed or inferred from such conduct.59
E. Cross-subsidization
MCI makes one final argument to support the use of fully distributed cost. MCI argues at considerable length that an FDC methodology is required to prevent AT & T from subsidizing its competitive services with revenues derived from services in which it retains a monopoly. MCI claims that such “cross-subsidization” injures AT & T’s competitors as well as AT & T’s local monopoly customers, who must pay higher rates in order to “subsidize” the company’s less profitable private line services. Nowhere does MCI define precisely what it means by a “cross-subsidy,” although it presented evidence at trial that different AT & T services earned differing rates of return. In particular, MCI noted that AT & T’s Telpak and other private line long distance services, although showing a posi-five rate of return, earned on an allocated rate base a lower rate of return than did certain other AT & T long distance services.
Such differing rates of return, however, even if correctly and meaningfully derived, do not support the imposition of antitrust liability. The fact that different services may earn different rates of return largely reflects the realities of a competitive market.60 Where a firm faces competition, demand is more elastic — that is, more sensitive to changes in prices — because of the presence of other firms producing substitute products to which buyers may turn. Lower returns on investment are to be expected in competitive markets because each firm, in accordance with classical competitive theory and practice, will be forced to lower prices toward marginal costs in order to maintain its market share.
MCI’s argument presumes that customers of monopoly services will have to pay higher prices if AT & T prices below FDC in markets where competition is present. See In Re American Telephone & Telegraph Co., 61 F.C.C.2d 587, 624, 652 (1976). Such arguments ignore the nature of costs and revenues in a multi-service enterprise. AT & T’s unattributable overhead costs do not increase when AT & T offers a new service, nor do they decrease when such a service is *1124discontinued. When a multiproduct firm prices a competitive service above its long-run incremental cost, no cross-subsidy can occur because the additional revenues produced exceed all additional costs associated with the competitive service and provide a contribution to the unallocable common costs otherwise borne by the firm’s existing customers. For this very reason the Second Circuit in Northeastern Telephone Co. v. AT & T rejected a cross-subsidization argument identical to that advanced by MCI here:
[The plaintiff’s] argument in favor of the fully distributed cost test is based on a misunderstanding of the economic notion of subsidization. [The plaintiff] seems to believe that whenever a product’s price fails to cover fully distributed costs, the enterprise must subsidize that product’s revenues with revenues earned elsewhere. But when the price of an item exceeds the costs directly attributable to its production, that is, when price exceeds marginal or average variable cost, no subsidy is necessary. On the-contrary, any surplus can be used to defray the firm’s non-allocable expenses.
651 F.2d 76, 90 (2d Cir.1981), cert. denied, 455 U.S. 943, 102 S.Ct. 1438, 71 L.Ed.2d 654 (1982). See also Southern Pacific Communications Co. v. AT & T, 556 F.Supp. 825 (D.D.C. 1982).
Judge Wilkey of the District of Columbia Circuit amplified this point in his dissent in Aeronautical Radio, Inc. v. FCC, 642 F.2d 1221, 1222 (D.C.Cir.1980), cert. denied, 451 U.S. 920, 101 S.Ct. 1998, 68 L.Ed.2d 311 (1981):
AT & T’s common or joint unattributable costs will exist whether or not it offers services in the competitive market. These costs existed and were borne by AT & T’s monopoly service customers before AT & T entered the competitive market, and would again be borne fully by them if AT & T were forced out of the competitive market.
When AT & T considers whether to enter or to expand sales in a competitive market the old monopoly service customers stand to benefit so long as the new customers bear any part of the common or joint costs. To determine whether monopoly customers will benefit from the firm’s operations in the competitive market, one need only calculate whether the revenues from the new competitive market operations pay fully for the incremental or additional costs the firm incurs for these operations. If revenues cover these costs (including cost of capital as measured by LRIC or any similar variant of marginal cost measurement) then ANY additional revenue earned above the LRIC level is a bonus for the monopoly customers.
642 F.2d at 1240 (Wilkey, J., dissenting). See also 2 P. Areeda & D. Turner, supra, at ¶ 719; 1 A. Kahn, supra, at 150-58.
If AT & T were forced to price at FDC levels in competitive markets, its monopoly customers would probably be worse rather than better off. Because of the elasticity of demand in competitive markets, any rate substantially above LRIC would cause AT & T to lose business against an equally efficient competitor and, hence, decrease AT & T’s total revenue from competitive markets. There would thus be less revenue available from competitive services to contribute to the firm’s joint or common costs, and monopoly customers would be required to provide a greater share of these costs.61
For a regulated utility such as AT & T, fully distributed cost methodology may be used to establish a regulatory rate ceiling, in order to provide no more than a “fair rate of return” for the enterprise as a whole. If FDC is adopted as a floor for predatory pricing purposes, as well as a ceiling for ratemaking purposes, the regulated utility will be effectively prohibited from materially raising or lowering prices to engage in competition. This result flies in the face of a major objective of the *1125antitrust laws — the promotion of price competition. It is also inconsistent with the FCC’s explicit endorsement of price competition in its Specialized Common Carriers decision. An antitrust rule requiring a dominant firm to price at or above FDC in competitive markets may effectively require the firm to forego price competition and gradually abandon its market share, i.e., lose its business. Constraining AT & T to FDC pricing of its competitive services thus runs the risk of permitting actually or potentially less efficient competitors to serve a growing segment of the telecommunications market and thus deprive consumers of the benefits of price competition.62
F. Insufficiency of the Evidence
In addition to reliance on an incorrect cost standard, the jury’s finding that Hi-Lo was predatory is disapproved and must be set aside because MCI failed to produce sufficient evidence to create a jury question that Hi-Lo was priced below cost under any standard. The testimony of Dr. William Melody, a regulatory economist, accompanied by certain documents, constitutes the only evidence MCI presented on the issue of predatory pricing. Dr. Melody testified twice, first in the latter part of February 1980 and again on June 3 and 4, 1980. On neither occasion did his testimony produce evidence sufficient to sustain a jury verdict that Hi-Lo was priced below cost under any standard.
Testifying the first time, Dr. Melody presented no evidence whatsoever that Hi-Lo was priced below any measure of cost. Dr. Melody introduced a chart, Plaintiff’s Exhibit 933, which purported to prove that Telpak was predatory by comparing its price with the costs associated with Hi-Lo service. • Dr. Melody argued that the costs attributable to both services were identical because each service was simply a different marketing plan for the same private lines. This chart shows the cost of the Hi Density (Hi-D) circuits to be $.65 per circuit mile. Thus, MCI’s own proof on this issue establishes that AT & T’s Hi-D circuits63, which were sold for $.85 per circuit mile, were priced $.20 above even their fully distributed costs.
This admission was reinforced on cross-examination in an exchange between counsel for AT & T and Dr. Melody:
Q: Now turning back to Hi-Lo, you are not contending, are you, that the high density portion of the Hi-Lo tariff is below cost by any measure?
A: I have not contended that the high density rate is below cost. I have not assessed the high density rate in terms of costs.
Tr. 2593. Despite expressing misgiving about the costs reflected in PX 933, Dr. *1126Melody repeatedly adopted these costs, including the $.65 figure, as the best evidence available. Tr. 2576,10481. Dr. Melody also stated that, in examining Bell’s cost data, he was unable to make the adjustments necessary to demonstrate that Hi-D costs were any greater than $.65. Tr. 2594.
On rebuttal, Dr. Melody purported, for the first time, to suggest that Hi-D was below cost. MCI introduced PX 3915, reproduced below, which is a table computed by Dr. Melody showing various •alleged revenue deficiencies for AT & T’s entire private line telephone service.
AT&T PRIVATE LINE TELEPHONE SERVICE
(IN $ MILLIONS)
1971 1972 1973 1974 1975
REVENUE NECESSARY TO COVER AT&T’S COST OF CAPITAL $109 $120 $153 $164 $172
REVENUE AVAILABLE $ 37 $ 47 $ 65 $ 68 $ 73
DEFICIENCY OF REVENUE BELOW COST ($ 72) ($ 73) ($ 88) ($ 96) ($ 99)
Dr. Melody explained the preparation of his chart as follows:
On the basis of Mr. Johnson’s [sic] [an AT & T witness’s] study, what I did was I examined every revenue that would be available after deducting all of the normal operating expenses of business....
What I did was I calculated the revenue that would be available for paying the cost of capital on the basis of Mr. Johnson’s [sic] studies....
I then calculated the revenue that would be necessary to cover AT & T’s cost of capital as earned by the business as a whole. That is indicated by the first row. The revenues that would be necessary if private line telephone service were to provide sufficient revenue to pay the cost of capital.
I then subtracted the revenue necessary from the revenue available and was able to calculate the deficiency of revenue below costs for private line telephone service.
Tr. 10474-76.
Neither Dr. Melody’s testimony nor his private line telephone chart reflect the sort of analysis and presentation necessary to support a claim of predatory pricing. The summary nature of Dr. Melody’s chart would make it very difficult for the jury to determine the basis of his calculations. Dr. Melody purports to be making adjustments to a series of fully distributed cost exhibits introduced by AT & T. Each of these exhibits consisted of voluminous cost studies (using several different methods of cost distribution), or summaries of such studies, which on their face stated that private line telephone service and Telpak earned positive rates of return64 under each FDC method used in the studies. Dr. Melody provides us with no calculation (or even specification) of AT & T’s overall cost of capital (including, presumably, embedded cost of debt), which establishes a deficiency in contribution by private line telephone to that cost.65
*1127Thus, Dr. Melody’s testimony is deficient as to the reasons why he selected one of the FCC’s at least seven cost methods or how he may have adjusted AT & T’s cost studies to produce the revenue deficiencies derived on his chart. Dr. Melody does not state what percentage he used to calculate AT & T’s cost of capital rate, nor what plant items he attributed to private line services for purposes of calculating these capital costs. Similarly, there is no definition or description of AT & T’s “normal operating expenses of business.” Thus, we do not know if, or how, Dr. Melody allocated capital costs and normal operating expenses.66 More importantly, Dr. Melody’s chart fails to isolate Hi-D circuits or even Hi-Lo service as a whole; instead it calculates alleged revenue deficiencies for AT & T’s entire private line sector. As Dr. Melody acknowledged, this sector includes many services besides Hi-D, as well as several types of switching equipment.67
Dr. Melody’s chart, together with his testimony, is therefore an insufficient basis for a jury verdict that the Hi-D portion of the Hi-Lo rate is “below cost.” We assume that Dr. Melody’s testimony was designed to demonstrate that, during the years in question, under some FDC method (presumably Method 1, see supra, note 65), “private line telephone” was returning less than AT & T’s overall cost of capital. Whatever the merits of such a demonstration as a measure of predation, see supra, text and note at note 47, we think the attempted demonstration is defective for lack of specificity and explanation of key elements and because “private line telephone” is inadequately related to the high density portion of the Hi-Lo rate.68 This evidence falls below the legal standard of proof necessary to support a finding of predatory pricing. Indeed, the Second Circuit recently rejected summary evidence of this sort in a predatory pricing case. In Broadway Delivery Corp. v. United Parcel Service of America, 651 F.2d 122 (2d Cir.), cert. denied, 454 U.S. 968, 102 S.Ct. 512, 70 L.Ed.2d 384 (1981), plaintiffs had attempted to demonstrate predation using only lump summaries of defendants’ costs, revenues, and profits, which purported to show below-cost operations. The district court granted defendants’ motion to dismiss, after a jury trial. The Second Circuit affirmed the dismissal, noting that plaintiffs had offered no explanation of how they had adjusted the defendants’ cost figures to show below-cost pricing:
Whether or not one agrees that proof of pricing below marginal or average variable cost is essential to a predatory pricing claim, the plaintiffs could not demonstrate price predation by the defendants without proof permitting a careful assessment of the relationship between the defendants’ prices and costs.... The plaintiffs’ proof did not permit a reasonable fact-finder to make this assessment. The summaries of [one defendant’s] operations lump all its traffic figures in one *1128category, all its revenues in another, and all its profits in a third. It may be that an expert in cost accounting could have discerned in these gross figures a basis for the required analysis, but the plaintiffs presented no such testimony.
651 F.2d at 131 (citations omitted). See Chillicothe Sand & Gravel Co. v. Martin Marietta Corp., 615 F.2d 427 (7th Cir.1980) (affirming district court’s grant of directed verdict on ground that plaintiff’s cost data and other evidence failed to present a prima facie case of predation).
Because MCI’s evidence is similarly inadequate to establish that Hi-Lo was priced below cost under any standard, the jury’s finding that AT & T engaged in predatory pricing of its Hi-Lo private line services cannot be sustained.
G. Pre-announcement
Our conclusion that Hi-Lo was not shown to be predatory largely dictates our disapproval of the jury’s finding that AT & T unlawfully pre-announced its Hi-Lo service. MCI argued that, if Hi-Lo was in fact predatory, AT & T would incur major losses in cutting its rates to stifle competition. At least part of these losses would result from AT & T customers shifting from more remunerative services such as WATS to private line. According to MCI, AT & T could maintain its monopoly without actually incurring any losses by simply announcing the proposed Hi-Lo tariff and then delaying its implementation for a significantly long period of time — thus discouraging AT & T customers from switching to MCI’s more economical services during the fifteen month period between Hi-Lo’s pre-an-nouncement and its effective date. Insofar as this claim is predicated upon the alleged predatory nature of the announced Hi-Lo price, the jury verdict on this count is disapproved and must be set aside.
MCI also claims that by pre-announcing Hi-Lo, AT & T knowingly misled the public into believing that the new tariff would be implemented without “undue delay,” when, in fact, AT & T’s actions significantly contributed to the fifteen month lag period between Hi-Lo’s announcement and effective date. How to determine whether and under what circumstances the pre-an-nouncement of a lawful price might constitute an act of monopolization is an extremely delicate task. Unnecessarily restrictive rules are likely to inhibit the flow of valuable information to the market; they also risk infringing upon protected commercial speech and first amendment rights. See Central Hudson Gas & Electric Corp. v. Public Service Commission, 447 U.S. 557, 100 S.Ct. 2343, 65 L.Ed.2d 341 (1980); Cox Broadcasting Corp. v. Cohn, 420 U.S. 469, 491-97, 95 S.Ct. 1029, 1044-1047, 43 L.Ed.2d 328 (1975). In commenting on liability for pre-announcement, Professors Areeda and Turner have stated that while a “knowingly false statement designed to deceive buyers” could qualify as an exclusionary practice,
no liability should attach to statements that truly reflect the monopolist’s expectations about future quality or availability where that expectation is both actually held in good faith and objectively reasonable. Such reasonable good faith statements about research, development, and forthcoming production serve the social interest in maximizing the relevant information available to buyers.
3 P. Areeda & D. Turner, supra, at 1738 p. 284.
Such a standard comports with the policies of the antitrust laws and with existing case law. In Americana Industries v. Wom-etco de Puerto Rico, Inc., 556 F.2d 625 (1st Cir.1977), the plaintiffs alleged that the defendant movie theaters had advertised in advance prices for the film “Godfather II” which maliciously undercut Americana. Plaintiffs also alleged that the pre-an-nouncement was done with the intent of putting Americana out of business. The court held that, absent allegations of pricing below marginal cost, the announcement represented the sort of healthy competition which the antitrust laws were designed to foster. 556 F.2d at 628. The court stated:
[The defendant] was not obliged to follow Americana’s prices, nor to hide the fact *1129that it would, in three months time, show the same film in another city for less. Americana’s complaint not only alleges facts that are completely consistent with perfectly lawful conduct, but falls short of alleging facts that, by themselves, constitute an antitrust violation.
Id.
It is certainly not unusual for competitors to announce a new product, service or price before its actual introduction in the market. The courts have upheld such “pre-announce-ments” in a variety of factual contexts. In Ronson Patents Corp. v. Sparklets Devices, 112 F.Supp. 676 (E.D.Mo.1953), the court considered an antitrust counterclaim to a patent infringement suit where the plaintiff was charged with advertising its new butane lighter in advance of even obtaining a patent. 112 F.Supp. at 688-89. The court declined to find liability even though the actual lighter advertised never appeared on the market. The court refused to infer predatory intent given the use of such an announcement to test demand in the market and production problems which subsequently led the company to decide not to put the product on the market. Id.
Similarly, in Berkey Photo, Inc. v. Eastman Kodak Co., 603 F.2d 263, 287-88 (2d Cir.1979), cert. denied, 444 U.S. 1093, 100 S.Ct. 1061, 62 L.Ed.2d 783 (1980), the Second Circuit held that, absent actual deception, a monopolist’s vigorous and even one-sided advertising of a new product or service does not constitute anticompetitive conduct violative of the Sherman Act. See also ILC Peripherals Leasing Corp. v. IBM Corp., 458 F.Supp. 423, 442 (N.D.Cal.1978), aff’d per curiam sub nom. Memorex Corp. v. IBM Corp., 636 F.2d 1188 (9th Cir.1980), cert. denied, 452 U.S. 972, 101 S.Ct. 3126, 69 L.Ed.2d 983 (1981) (defendant’s premature announcement of new product not actionable absent evidence of knowing falsehood); Solargen Electric Motor Car Corp. v. American Motors Corp., 530 F.Supp. 22, 26 n. 3 (S.D.N.Y.1981), aff’d, 697 F.2d 297 (2d Cir. 1982) (competitor’s allegations that dominant car company “knowingly exaggerated” success of its experimental battery do not support antitrust liability).
These cases suggest that AT & T’s early announcement of Hi-Lo must be found to be knowingly false or misleading before it can amount to an exclusionary practice.69 Applying this standard here, the issue of pre-announcement should never have been sent to the jury. Neither AT & T’s application to the FCC for permission to file the Hi-Lo rate, nor the accompanying press release contains any false or misleading information about Hi-Lo or its availability. MCI claims that AT & T’s statement in its FCC application that the Hi-Lo tariff “should be made effective without undue delay” is inconsistent with contemporaneous internal AT & T memoranda suggesting that AT & T had originally planned a voluntary extension of the effective date of the tariff until May 1974.70 We, however, see no deliberate deception or misleading conduct here.71
At the time AT & T first sought permission to file its Hi-Lo tariff, in February 1973, it could not simply file the tariff and *1130automatically set the regulatory review in motion. Rather, because of an FCC “special permission” rule in effect at the time, AT & T was required to obtain FCC approval before it could file any new tariff. Thus, as of February 1973, AT & T was not permitted to file or effectuate its new tariff. It was not until October 1973, when, at the behest of AT & T, the Second Circuit invalidated the “special permission” rule, that AT & T was free to file its Hi-Lo tariff. See AT & T v. FCC, 487 F.2d 865 (2d Cir.1973).
AT & T actually filed its Hi-Lo tariff on November 15, 1973, and the tariff became effective on June 15, 1974. Once the tariff was filed, the provisions of Sections 203 and 204 of the Communications Act accounted for much of the remaining delay. This delay-included the two-month notice period prescribed by FCC regulations and a three-month suspension of the tariff ordered by the FCC. Hence, the only delay that MCI can fairly attribute to AT & T is AT & T’s voluntary extension of the tariff for a period of about two months at the request of the FCC chairman. We believe that this rather minimal delay, considered in conjunction with the absence of deception or knowing falsehood in AT & T’s filings and public statements regarding Hi-Lo, is insufficient to create a jury question on the issue of unlawful pre-announcement.72 We therefore hold that the trial court erred in failing to direct a verdict on this count in favor of AT & T.
H. Telpak Marketing Plan
MCI in its cross-appeal contends that even if Telpak was lawfully priced the jury should have been instructed to consider whether AT & T maintained its monopoly by marketing its Telpak service in a way that excluded competition. MCI alleges that two aspects of the Telpak marketing scheme were anticompetitive: the so-called “free” circuits, and “fictional” routing.
MCI first argues that Telpak customers had free spare circuits available from AT & T because AT & T only offered Telpak in bundles of 60 or 240 circuits. Because AT & T’s rates were quite low, this encouraged customers to order a bundle of Telpak circuits, even if the customer did not need all the circuits immediately. MCI argues that the remaining unused circuits of the Telpak bundle were “free spares” that could be used later at no additional charge, thus discouraging customers from purchasing MCI services.
This argument has no factual or legal merit. The unused capacity was in no sense free. Telpak customers had already paid for all their circuits in advance. The fact that the customer chose to use only a portion of the circuits at first does not render the remaining circuits free.73 Indeed, in this regard Telpak is no different than any volume discount or package pricing plan. The mere existence of volume or package pricing does not support antitrust liability. See Chillicothe Sand & Gravel Co. v. Martin Marietta Corp., 615 F.2d 427, 433 (7th Cir. 1980). Moreover, absent predatory pricing, the sale of bundles of Telpak circuits could have little exclusionary or anticompetitive significance since Bell offered many different marketing plans for long distance business telecommunications users. Since the jury found that Telpak was not predatorily priced, the fact that customers preferred this marketing plan because it was more responsive to their needs than MCI’s alternate marketing plan is a matter of little concern to the enforcement of the antitrust laws.74
*1131MCI also contends that Telpak’s “fictional routing” feature tended to restrict MCI from serving large private line users. MCI argues that the routing was fictional in the sense that a Telpak customer who purchased a bundle of circuits, for example, from New York City to Columbus, Ohio and another bundle of circuits from Columbus to Toledo, Ohio could combine these circuits and make a call from New York to Toledo, which would be billed under Telpak. MCI contends that it has been injured because, although its Toledo to New York service is cheaper than AT & T’s, it does not serve Columbus and a customer requiring communications to all three cities is penalized for choosing MCI.
MCI’s competitive disadvantage in this regard stems from the fact that it entered the market only on a limited geographic scale, and does not reflect unlawful predation by AT & T. What MCI calls “fictional routing” merely represents the customer’s ability to tack one Telpak circuit onto another in such a way that a call can be placed between two distant cities connected by a string of circuits — or between intermediate cities. A customer building a private microwave system linking the same cities would be able to achieve precisely the same result. As AT & T points out, Telpak was designed — and upheld by the FCC — as an alternative to private microwave systems, and the “fictional routing” of which MCI complains involves no more than billing the Telpak customer as if service were provided over the same kind of facilities that the customer would have available from its own private microwave system.75 We hold, as in the case of the volume pricing of Telpak, that in the absence of predatory pricing such a billing system represents no violation of the antitrust laws.76
IY. INTERCONNECTIONS
In the years following the 1971 Specialized Common Carriers decision, a major source of contention between MCI and AT & T was the extent to which AT & T was obliged to interconnect with MCI’s facilities to accommodate MCI’s needs. The interconnection issue arose in part because MCI had facilities in place to serve only a limited number of cities and in part because MCI was unable to provide the local circuits necessary to connect its long distance service to the telephone customer. MCI’s telecommunications system consists of transmission towers that relay microwave impulses between terminals in the cities MCI serves. In each of those cities MCI must connect its terminals to telephones at its customers’ locations. In order to provide full end-to-end transmission, MCI’s equipment at some point must make contact with AT & T’s equipment because AT & T, through its operating companies, controls the local service to MCI’s customers. AT & T also provides long distance service to locations not covered by MCI. The dispute thus focuses on the local interconnections between MCI towers and its customers’ premises and on “multipoint” interconnections (discussed infra, at pp. 1147-1150) between MCI towers and certain AT & T long distance circuits.
*1132When MCI sought interconnections that would give it access to AT & T’s switched network,77 AT & T balked. AT & T contended that the FCC’s 1971 decision limited the new carriers to providing “point-to-point private lines,” which require no switching because each line is dedicated to the exclusive use of a specific customer and runs between only two designated premises. MCI complained that AT & T unlawfully refused interconnections for FX and CCSA services, both of which use switching machines,78 and for essentially local lines that led beyond a limited, defined geographical area. MCI also complained that AT & T unlawfully refused interconnections for multipoint service. Although AT & T supplied some interconnections when required by a 1973 district court injunction, it promptly terminated those connections when the injunction was vacated on appeal because the same issues were pending before the FCC. MCI alleged that these terminations were aimed at maintaining AT & T’s monopoly by injuring MCI’s reputation as a reliable firm and were improper since an FCC decision on the very matter of interconnections was imminent. MCI also maintained that AT & T illegally tied the provision of long distance service to local service.
The interconnections that were actually implemented between AT & T and MCI also gave rise to dispute. MCI asserted that the entire interconnection procedure required by AT & T was unreasonable because the physical interconnections utilized materials inadequate for the volume of business MCI was doing and because it involved unduly complex and ineffective installation and maintenance procedures.
All of these acts, MCI claimed, were committed deliberately by AT & T to damage MCI’s conduct of its business and constituted an abuse of AT & T’s monopoly power over facilities essential to MCI’s success.
A. FX-CCSA Interconnections
1. The Essential Facilities Doctrine
The jury found that AT & T unlawfully refused to interconnect MCI with the local distribution facilities of Bell operating companies — an act which prevented MCI from offering FX and CCSA services to its customers. A monopolist’s refusal to deal under these circumstances is governed by the so-called essential facilities doctrine. Such a refusal may be unlawful because a monopolist’s control of an essential facility (sometimes called a “bottleneck”) can extend monopoly power from one stage of production to another, and from one market into another. Thus, the antitrust laws have imposed on firms controlling an essential facility the obligation to make the facility available on non-discriminatory terms. United States v. Terminal Railroad Association, 224 U.S. 383, 410-11, 32 S.Ct. 507, 515-516, 56 L.Ed. 810 (1912); Byars v. Bluff City News Co., 609 F.2d 843, 856 (6th Cir. 1979).
The case law sets forth four elements necessary to establish liability under the essential facilities doctrine: (1) control of the essential facility by a monopolist; (2) a competitor’s inability practically or reasonably to duplicate the essential facility; *1133(3) the denial of the use of the facility to a competitor; and (4) the feasibility of providing the facility. Hecht v. Pro-Football, Inc., 570 F.2d 982, 992-93 (D.C.Cir.1977), cert. denied, 436 U.S. 956, 98 S.Ct. 3069, 57 L.Ed.2d 1121 (1978). See Otter Tail Power Co. v. United States, 410 U.S. 366, 93 S.Ct. 1022, 35 L.Ed.2d 359 (1973); United States v. Terminal Railroad Association, 224 U.S. at 405, 409, 32 S.Ct. at 513, 515; City of Mishawaka v. American Electric Power Co., 465 F.Supp. 1320, 1336 (N.D.Ind.1979), aff’d in relevant part, 616 F.2d 976 (7th Cir.1980), cert. denied, 449 U.S. 1096, 101 S.Ct. 892, 66 L.Ed.2d 824 (1981).
The Supreme Court in Otter Tail, considered the refusal of a regulated electric utility to sell power wholesale or to transmit power purchased from other sources to municipalities which had chosen to own their own retail distribution systems. This refusal to sell or transmit was held to violate section 2 of the Sherman Act. The Court noted the district court’s determination that “Otter Tail had ‘a strategic dominance in the transmission of power in most of its service area,’ ” id. 410 U.S. at 377, 93 S.Ct. at 1029, and, in effect, was concerned that market power in one market (transmission) was being used to further a monopoly in another market (retail distribution). Id. at 377-79, 93 S.Ct. at 1029-1030.
Otter Tail provides an analogy to the instant problem. AT & T had complete control over the local distribution facilities that MCI required. The interconnections were essential for MCI to offer FX and CCSA service. The facilities in question met the criteria of “essential facilities” in that MCI could not duplicate Bell’s local facilities. Given present technology, local telephone service is generally regarded as a natural monopoly and is regulated as such. It would not be economically feasible for MCI to duplicate Bell’s local distribution facilities (involving millions of miles of cable and line to individual homes and businesses), and regulatory authorization could not be obtained for such an uneconomical duplication.
Finally, the evidence supports the jury’s determination that AT & T denied the essential facilities, the interconnections for FX and CCSÁ service, when they could have been feasibly provided. No legitimate business or technical reason was shown for AT & T’s denial of the requested interconnections. Cf., Gamco, Inc. v. Providence Fruit & Produce Building, Inc., 194 F.2d 484, 487-88 & n. 3 (1st Cir.), cert. denied, 344 U.S. 817, 73 S.Ct. 11, 97 L.Ed. 636 (1952) (defendants may deny access to their building because of limited space or the applicant’s financial unsoundness). See generally, Comment, Refusals to Deal by Vertically Integrated Monopolists, 87 Harv.L.Rev. 1720, 1740 (1974). MCI was not requesting preferential access to the facilities that would justify a denial. See Town of Massena v. Niagara Mohawk Power Corp., 1980-2 Trade Cas. (CCH) ¶ 63,526 (N.D.N.Y.1980). Nor was MCI asking that AT & T in any way abandon its facilities. See American Football League v. National Football League, 323 F.2d 124 (4th Cir.1963). MCI produced sufficient evidence at trial for the jury to conclude that it was technically and economically feasible for AT & T to have provided the requested interconnections, and that AT & T’s refusal to do so constituted an act of monopolization.
2. The Meaning of the Specialized Common Carriers Decision
AT & T never challenged the assertion' that it had denied MCI access to the local distribution network for FX and CCSA service. Instead, AT & T maintained that MCI had never been authorized to receive these interconnections. AT & T also argued that its own regulatory obligations prohibited it from interconnecting MCI with its local switched network for FX and CCSA.
The meaning of the FCC’s 1971 Specialized Common Carriers decision lies at the heart of the dispute over interconnections. As indicated, the FCC’s 1971 decision was extremely opaque. Following the Specialized Common Carriers decision, AT & T contended that it was required to provide local *1134interconnections only for point-to-point private line service, in accordance with service descriptions contained in documents submitted by the specialized carriers in the 1971 case. But MCI argued that the language of the Specialized Common Carriers decision also required AT & T to provide interconnection with its local switched network for FX and CCSA.
In November 1973, following more than a year of futile negotiations, MCI sought an injunction requiring AT & T to accede to MCI’s demand for interconnections. A preliminary injunction was issued but was later vacated on appeal because of a pending FCC show-cause proceeding on the interconnection issue. MCI Communications Corp. v. AT & T, 369 F.Supp. 1004 (E.D.Pa. 1973), injunction vacated on primary jurisdiction grounds, 496 F.2d 214 (3d Cir.1974). AT & T immediately dismantled all the interconnections it had provided pursuant to the preliminary injunction. This action obviously adversely affected MCI’s ability to serve its customers. AT & T claimed its filed tariffs required the disconnection.
Eight days after the Third Circuit decision vacating the injunction, the FCC issued its ruling on the show cause order. The FCC interpreted the Specialized Common Carriers decision, although conceding its lack of clarity, as requiring the interconnections MCI sought. Bell System Tariff Offering of Local Distribution Facilities for Use by Other Common Carriers, 46 F.C.C.2d 413, aff’d sub nom. Bell Telephone Co. v. FCC, 503 F.2d 1250 (3rd Cir.1974), cert. denied, 422 U.S. 1026, 95 S.Ct. 2620, 45 L.Ed.2d 684 (1975).
The Specialized Common Carriers decision was interpreted three years later by the District of Columbia Circuit in the Exe-cunet case. MCI Telecommunications Corp. v. FCC, 561 F.2d 365 (D.C.Cir.1977), cert. denied, 434 U.S. 1040, 98 S.Ct. 781, 54 L.Ed.2d 790 (1978). In Execunet the court held that the FCC had not conducted a sufficient hearing in Specialized Common Carriers to justify any limits on the services MCI could provide. MCI was then permitted to offer all forms of long distance service. Thus, with the benefit of hindsight, the FCC’s 1974 order and Execunet estab--lished MCI’s right to interconnections for FX and CCSA.
At trial, MCI contended that in denying interconnections, AT & T intended to prevent competition. AT & T argued that the Specialized Common Carriers decision was so vague that it gave no guidance on AT & T’s obligation to interconnect or even on MCI’s authority to provide FX and CCSA service. AT & T argued that it reasonably believed that MCI was not authorized to provide the services for which it sought interconnections, and that this good faith belief in the regulatory requirements was a complete defense to antitrust liability for the denial of interconnections.
The propriety of Judge Grady’s instructions on the meaning of the Specialized Common Carriers decision became an extraordinary issue in this case. Although noting that “[ojrdinarily what a decision means is a question of law for the court,” Judge Grady stated that under the circumstances of this case the jury must decide as a fact question whether the Specialized Common Carriers decision required AT & T to provide the requested FX and CCSA interconnections. Tr. 11463. He further told the jury that, even if it found that the Specialized Common Carriers decision, ordered AT & T to make the.connections, AT & T would not be liable unless it “knew or had good reason to believe that the decision constituted such an order.” App. 1200. Judge Grady then explained the holding in Execunet and concluded by stating “I instruct you as a matter of law, therefore, that at the time MCI requested FX and CCSA interconnections it was authorized to render those services and AT & T was obligated under the Communications Act to provide the interconnections.” App. 1201. The judge cautioned the jury that a violation of the Communications Act does not establish a violation of the antitrust laws. Finally, Judge Grady instructed that AT & T would not be liable for failure to provide the interconnections if “it believed that it had not been ordered to do so, that MCI was not autho*1135rized to provide [FX and CCSA service], and that it would have violated established regulatory policies for MCI to receive the connections.” Id.
AT & T asserts that the nature of obligations imposed by a regulatory statute or agency order is a question of law, not of fact, and thus that the jury should not have been permitted to decide whether the Specialized Common Carriers decision required the interconnections MCI sought. AT & T’s argument at trial centered on the contention that it denied the interconnections in good faith.79 That good faith, AT & T said, was based on its belief that neither the Communications Act nor the Specialized Common Carriers decision ordered the interconnections because that decision and the general regulatory policy relevant to the issue were so unclear that they offered little guidance.
By instructing on the Specialized Common Carriers decision, Judge Grady attempted to take AT & T’s argument into account. The unusual dilemma facing the district court, however, was that in this case it concluded that antitrust liability turned not so much on what the Specialized Common Carriers decision ordered, but on what AT & T believed it ordered. Declining to take a legal position on the meaning of the decision, Judge Grady first permitted the jury to decide entirely as a fact question what the decision actually ordered. Although that approach improperly allowed the jury to decide what amounted to a question of law, the only party that could have been harmed by that error was MCI, whose case would have been cut short only by a legally incorrect finding that no interconnections were required. Judge Grady’s approach could not have prejudiced AT & T, and only that possibility concerns us on appeal.
The remainder of the instruction on the Specialized Common Carriers decision took into account the thrust of AT & T’s theory of defense: namely, that regardless of what the 1971 decision ordered as a matter of law, AT & T believed in good faith that MCI was not authorized to provide FX and CCSA and that AT & T was not required to provide the interconnections and could make an independent assessment of the public interest with respect to these interconnections. The jury could decide that AT & T was free to deny interconnections, *1136based upon a good faith determination that interconnection was contrary to the public interest, by finding that AT & T believed no interconnections were required because (a) the 1971 decision by its terms did not explicitly order them or (b) the context in which the interconnections were ordered was so vague that AT & T did not know that they were ordered. In either case, AT & T’s defense could be fully considered by the jury. As indicated, only MCI and not AT & T could have been injured by transforming into a fact question an issue decided against AT & T as a matter of law by the FCC in 1974 and by the District of Columbia Circuit in 1977.
3. “Retroactive” Application of Execu-net
AT & T also contends that the court erred in instructing on the meaning of the Execunet decision. AT & T argues that the district court improperly gave retroactive application to the 1977 Execunet decision by instructing the jury that the case meant MCI had been authorized to provide FX and CCSA, and AT & T had been obligated under the Communications Act to provide the FX and CCSA interconnections since the date MCI’s permits issued.
The Supreme Court has in part analyzed the retroactivity question as follows:
[T]he decision to be applied nonretroac-tively must establish 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) (citations omitted).
Although Execunet may have taken AT & T and the FCC by surprise, its holding did not establish a new principle of law, nor was the case one of first impression. While the FCC had concluded that Specialized Common Carriers did not authorize Execu-net service, 60 F.C.C.2d at 38-40, the appeals court noted that this view represented “a substantial departure from prior administrative practice.” 561 F.2d at 373. The court carefully set out FCC regulations and practices, and the case law from both the District of Columbia and Second Circuits to support its conclusion. Id. at 374-76.80 In a later case dealing with FCC proceedings on the interconnection issue pursuant to the Execunet decision, the District of Columbia Circuit reaffirmed its earlier view. MCI Telecommunications Corp. v. FCC, 580 F.2d 590, 593 (D.C.Cir.), cert. denied, 439 U.S. 980, 99 S.Ct. 566, 58 L.Ed.2d 651 (1978). The reasoning and holding of the District of Columbia Circuit in Execunet, while perhaps startling, did not create a “new principle of law.”81
*1137Unlike the usual situation in which a question of retroactive application is raised, the Execunet decision was not determinative of the outcome in this case.82 Even though Judge Grady correctly instructed the jury that by the reasoning of Execunet MCI was entitled to the interconnections, the instructions clearly stressed that entitlement under the Communications Act did not establish liability under the antitrust laws. More importantly, the careful instruction that the jury was to consider “the historical context and all of the facts and circumstances known to the parties at the time” of AT & T’s allegedly improper acts83 preserved AT & T’s good faith defense (which was emphasized in the instruction immediately following the explanation of Execunet), and rendered Judge Grady’s instruction on Execunet an accurate representation of the controversy between MCI and AT & T rather than a “retroactive” application of a new legal principle resulting in prejudice to AT & T.
Nor do we believe that the explanation of Execunet misled or confused the jury. AT & T argues that it “placed AT & T in the anomalous position of being required to prove that it was ignorant of the law.” AT & T does not argue that Judge Grady misstated the law. His instruction placed in proper context AT & T’s position that it believed at the time that MCI was not authorized to provide FX and CCSA service. It removed any misleading implication that MCI was not, as a matter of law, so authorized, leaving only the crucial question of AT & T’s good faith belief, which was the focus of the remainder of the instruction. Moreover, as previously stated, the instructions, when read as a whole, did not mislead the jury as to the date Execunet was decided in relation to the date of the conduct under scrutiny here.
4. Instructions on Regulatory Policy
As we have already indicated the regulatory constraints governing the behavior of a public utility are an important factor to be weighed in assessing the potential antitrust liability of a regulated firm. See supra, at pp. 1105-1111. Ordinarily, antitrust liability should not be imposed when a firm acts in compliance with its regulatory obligations. See Watson & Brunner, Monopolization by Regulated “Monopolies”: The Search for Substantive Standards, 22 Antitrust Bull. 559 (1977).
AT & T contends that, since it asserted at trial that public interest considerations under Section 201(a) of the Communications Act prompted its refusal to provide interconnections, the district court should have instructed the jury on the details of the regulatory statute as they relate to the controversy involved here. Judge Grady instructed the jury that MCI had to prove that AT & T denied the FX and CCSA interconnections for anticompetitive reasons, and not because of its good faith belief “that it would have violated established regulatory policies for MCI to receive the connections.”
AT & T argues that this instruction was nonetheless inadequate because it failed to explain in detail the regulatory provisions governing interconnection. In making this argument AT & T relies heavily on Mid-Texas Communications Systems, Inc. v. AT & T, 615 F.2d 1372 (5th Cir.), cert. denied, 449 U.S. 912, 101 S.Ct. 286, 66 L.Ed.2d 140 (1980). In Mid-Texas, the defendant Bell System had refused to make available to a local telephone company certain essential interconnections. The Fifth Circuit reversed a verdict for the plaintiff on the basis of erroneous instructions to the jury. The jury was instructed to assume the ex*1138istence of monopoly power, thus precluding it from consideration of state and federal regulation. On the issue of willful misuse of monopoly power, the instruction allowed for consideration of “legitimate telephone business reasons,” but did not specifically direct the jury to take into account the effect regulation might have had on the conduct of the Bell System. The Fifth Circuit concluded that the jury must be instructed as to the relevant regulatory framework to determine whether Bell’s conduct was reasonable. Id. at 1390-91.
AT & T argues that more is required than the charge that regulation is to be taken into account. In Mid-Texas, the court stated that “the district court should have instructed the jury on the applicable regulatory provision.” Id. at 1387. From this, AT & T finds support for a rule requiring the court to describe and explain in detail all relevant regulations. But this fails to take into account the rule, cited with approval in Mid-Texas, that “the question on appeal is not whether an instruction was faultless in every respect, but whether the jury, considering the instruction as a whole, was misled. Thus, only in those cases where the reviewing court has substantial doubt whether the jury was fairly guided in its deliberations should the judgment be disturbed.” Id. at 1390-91 n. 16 (citations omitted). Accord, Alloy International Co. v. Hoover-NSK Bearing Co., 635 F.2d 1222, 1226 (7th Cir.1980); Commercial Iron & Metal Co. v. Bache Halsey Stuart, Inc., 581 F.2d 246 (10th Cir.1978), cert. denied, 440 U.S. 914, 99 S.Ct. 1229, 59 L.Ed.2d 463 (1979); Allers v. Bohmker, 199 F.2d 790, 792 (7th Cir.1952).
Mid-Texas does not state how detailed a regulatory framework instruction must be. All we know is that it was error to omit all reference to how evidence on regulation was to be used in the jury’s deliberations. 615 F.2d at 1390-91 n. 16. An ideal instruction would very briefly explain, for example, that a carrier has an obligation under the Communications Act to interconnect, but may deny interconnections if it determines that the public interest is to the contrary; and that if the carrier at the time had a reasonable basis in regulatory policy to conclude, and in good faith concluded, that denial of interconnections is required by concrete, articulable concerns for the public interest, then there is no liability under the antitrust laws. AT & T’s proposed instruction went well beyond this concise model. The court’s instruction fell short to the extent that it did not explain the particular provision of the Act, but we do not consider this fatal. Unlike the jury in Mid-Texas, the jury here was clearly instructed to consider the impact of regulation and, where applicable, AT & T’s defense of good faith. Both parties presented evidence on regulation and thoroughly argued the application of that evidence to the jury. The thrust of AT & T’s position under the Communications Act was made plain, and nothing in the pertinent instructions misled or confused the jury in this regard.84
AT & T makes a second argument in favor of specific instructions on the meaning of its regulatory obligations. AT & T contends that the jury instructions omitted specific reference to and explanation of the Rule of Reason required to be applied in antitrust cases. AT & T argues that pursuant to the Rule of Reason the jury must consider all the facts and circumstances bearing on the reasonableness of the challenged conduct. Such consideration would require detailed instructions on the meaning of the Communications Act.
*1139AT & T’s argument misapprehends the proper role of the Rule of Reason in antitrust eases. The Rule of Reason is a rule of construction which applies to section 1 of the Sherman Act. The need for such a rulé arose because a literal reading of section 1 would prohibit virtually every private contract. See National Society of Professional Engineers v. United States, 435 U.S. 679, 687-88, 98 S.Ct. 1355, 1363, 55 L.Ed.2d 637 (1978); Chicago Board of Trade v. United States, 246 U.S. 231, 238, 38 S.Ct. 242, 243, 62 L.Ed. 683 (1918). Thus, under the Rule of Reason only agreements which are unreasonably restrictive of competition violate section 1 of the Sherman Act, Standard Oil Co. v. United States, 221 U.S. 1, 31 S.Ct. 502, 55 L.Ed. 619 (1911); the Rule of Reason does not directly apply as such to the offense of monopolization under section 2 of the Sherman Act.
Even if the Rule of Reason were construed to apply to analysis of section 2 of the Sherman Act, the Rule does not stand for the proposition that “all the facts and circumstances which bear on the reasonableness of the challenged conduct be considered.” For example, a defendant cannot claim as a defense that a decision to fix prices was reasonable, United States v. Socony-Vacuum Oil Co., 310 U.S. 150, 60 S.Ct. 811, 84 L.Ed. 1129 (1940); United States v. Addyston Pipe & Steel Co., 85 F. 271 (6th Cir.1898), modified, 175 U.S. 211, 20 S.Ct. 96, 44 L.Ed. 136 (1899), or claim that it was reasonable to replace competition with monopolistic arrangements. United States v. Joint Traffic Association, 171 U.S. 505, 573-77, 19 S.Ct. 25, 33-34, 43 L.Ed. 259 (1898); United States v. Trans-Missouri Freight Association, 166 U.S. 290, 17 S.Ct. 540, 41 L.Ed. 1007 (1897).
The Rule of Reason was analyzed comprehensively in National Society of Professional Engineers v. United States, 435 U.S. 679, 98 S.Ct. 1355, 55 L.Ed.2d 637 (1978). Justice Stevens writing for the Court stated:
Contrary to its name, the Rule does not open the field of antitrust inquiry to any argument in favor of a challenged restraint that may fall within the realm of reason. Instead, it focuses directly on the challenged restraint’s impact on competitive conditions.
Id. at 688, 98 S.Ct. at 1363. The inquiry under the Rule of Reason is thus confined to a consideration of the impact of the challenged conduct on competitive conditions and does not inquire whether a policy favoring competition is in the public interest. Id. at 690, 692, 98 S.Ct. at 1364-1365.
The district court properly instructed the jury on the applicable provisions of the Sherman Act and how the facts fit within that legal framework. As part of the instruction, the jury was cautioned that violation of the Communications Act did not prove that antitrust laws also were breached. If the jury has been adequately instructed how to apply the pertinent law to the evidence, there is no reversible error. Alloy International, 635 F.2d at 1226-27. The instructions given comport with AT & T’s view that the impact of regulation must be considered. We further find no reversible error in the refusal to instruct on the specific details of the Communications Act, or on the Rule of Reason.
5. Insufficient Evidence
AT & T asserted that the evidence was insufficient to sustain the jury’s finding that the FX and CCSA interconnection denials were made in bad faith. The first contention is that MCI never proved AT & T knew in 1973 and 1974 that the court would rule as it did in the 1977 Execunet decision. Therefore, says AT & T, its denial of interconnections was made in good faith. MCI argued, however, that AT & T made its interconnection decisions without reference to its understanding of the state of the law. For support, MCI introduced internal AT & T documents showing AT & T’s expectation that a less limited interconnection policy would be propounded in the near future (one AT & T official anticipated “demands for unrestricted interconnections”), and suggesting that AT & T’s approach be one of “buying time” through *1140“delaying tactics.”85 One AT & T document noted Bell’s adoption of policies designed to “limit flexibility of [the specialized carriers] and [their] ability to sell services utilizing our central office switching capacity.” From this evidence, the jury could reasonably infer that AT & T’s delays in permitting FX and CCSA interconnections were evidence of AT & T’s improper intent “to limit competitive encroachments through devices such as restricting the use other carriers may make of our facilities.” AT & T’s denial of interconnections must also be judged in context with its other actions. See City of Mishawaka v. American Electric Power Co., 616 F.2d 976, 986 (7th Cir.1980). The evidence supports the inference that, regardless of whether it anticipated Execunet, AT & T intended to obstruct MCI’s entry into the market and used the public interest standard in bad faith.86
AT & T also urges that the evidence was insufficient to support the jury’s rejection of AT & T’s defense that it had in good faith interpreted the Specialized Common Carriers decision as limiting its duty to provide interconnections. The jury was entitled to credit the evidence showing that generally AT & T officials intended to impede competition regardless of the meaning of the 1971 decision. AT & T presented testimony of its officials and others familiar with the telecommunications industry. That evidence indicated that some persons familiar with the telecommunications industry viewed the FCC’s use of the term “private line” service in the Specialized Common Carriers opinion as vague enough to be read as limiting MCI to “point-to-point” private line service (an AT & T interpretation that would exclude FX and CCSA interconnections and limit required interconnections only to “tie lines” through local wire “links” that would not provide access to the public switched network).87 In response, however, MCI introduced evidence that AT & T always classified FX and CCSA as “private line service,” which was the principal descriptive term used in the 1971 order to describe the scope of service authorized; that the limiting term “point-to-point private line service” did not appear in the 1971 decision; and that “tie lines” could afford access to a switched network. The “tie line” point, which raised a fact issue, would sustain a jury conclusion that, even if MCI was limited to receiving tie line interconnections, there still was no basis for AT & T to conclude that access to the switched network was forbidden to MCI by the 1971 order. Moreover, the Specialized Common Carriers decision discusses options for the provision of “local loop” interconnections, an approach which assumes that such interconnections would be freely available.88 MCI demonstrated on eross-exami-*1141nation of an AT & T technical witness that the term “local loop” encompassed any local interconnection (including connections to switching equipment used in an FX termination) between an AT & T central office and an MCI terminal. This demonstration of AT & T’s awareness that its own interpretation of Specialized Common Carriers was on highly questionable technical ground is sufficient to sustain a finding of improper intent,89 especially when viewed in light of the arguably anticompetitive comments of AT & T management.
In any event, regardless of whether AT & T reasonably believed that Specialized Common Carriers did not require interconnections, the jury was entitled to conclude, based on the evidence, that AT & T did not act in good faith when it purportedly determined that the public interest justified its denial of interconnections.
6. Evidentiary Rulings
AT & T asserts that the district court erred in admitting into evidence the FCC’s 1974 cease and desist order and accompanying decision, 46 F.C.C.2d 413, which required AT & T to provide FX and CCSA interconnections. That decision noted that the scope of the FCC’s 1971 order may have been unclear (prompting the agency to proceed under a section of the Act different from that usually employed for cease and desist orders), but concluded that AT & T unlawfully denied the interconnections in contravention of the Specialized Common Carriers decision. Although Judge Grady refused to admit the decision and order for the truth of the matters asserted, he admitted the document for non-hearsay use as revealing the FCC’s “state of mind” about the scope of its Specialized Common Carriers decision which AT & T brought into question. AT & T’s good faith defense, based as it was on an asserted ambiguity in the Specialized Common Carriers decision, necessitated that MCI be allowed to show what the FCC believed its own order meant. AT & T recognized, when questioning a witness, the risk of providing a basis of admissibility for the 1971 order. Judge Grady warned AT & T’s counsel that his questions were opening the door for MCI’s use of the 1971 decision. AT & T’s counsel responded, “I don’t think there is any doubt about that,” and continued to pursue his line of questioning.
How the FCC viewed its 1971 decision was relevant to the reasonableness of AT & T’s professed good faith interpretation of that decision. This limited, non-hearsay use was not erroneous. Further, the use of the decision comports with our recognition of the relevance of showing the full regulatory environment within which AT & T operated.
Judge Grady did not, as AT & T claims, reverse his evidentiary rulings on the hearsay use of the opinion; it was not admitted for the truth of the statements it contained.90 Cf. United States v. AT & T, 498 F.Supp. 353 (D.D.C.1980) (finding that same FCC decision inadmissible if offered for the truth of the matters asserted in it). Nor was the admission of the 1974 decision unduly prejudicial under Federal Rule of Evidence 403. To the contrary, it provided highly probative evidence of how the FCC viewed one of its own decisions, a subject discussed by AT & T witnesses who explained their views of the “clear meaning” *1142of the 1971 opinion.91 The entire opinion was admitted, including the portion that characterized the Specialized Common Carriers case as “unclear,” and AT & T was permitted to draw the jury’s attention to that fact.92 Judge Grady carefully instructed the jury on the limited purposes for which it could use the decision, clearly stating that it was not conclusive of the illegality of AT & T’s actions but was merely one piece of evidence relevant to whether AT & T should have known that interconnections were required. In his final instructions, Judge Grady also reminded the jurors that the FCC’s expression of intent is relevant only to the extent that it was based on the language of the 1971 decision or other facts known to the parties at the time. Potential prejudice is not a reason to keep out evidence where probative value outweighs prejudice. Fed.R.Evid. 403. See generally 10 J. Moore & H. Bendix, Moore’s Federal Practice ¶ 403.10[1] (2d ed. 1982). Judge Grady did not abuse his discretion in admitting the opinion for limited purposes. See Forro Precision, Inc. v. IBM, 673 F.2d 1045, 1057 (9th Cir.1982).93
AT & T contends that the district court improperly refused admission of two documents, one of which emanated from Bechtel Corp., MCI’s construction firm, and one of which came from Collins Radio, Inc., one of MCI’s radio equipment suppliers. The documents contained the writers’ impressions of the meaning of the Specialized Common Carriers opinion. Judge Grady properly refused the admission of the Bechtel letter under the business records rule, Fed.R.Evid. 803(6), because it did not reveal how the author drew his conclusions. Judge Grady correctly ruled that the author’s vague statement that “in the beginning” FX and CCSA interconnections were not required “as I understood it” was insufficient to meet the foundation requirement of Rule 803(6). Because no adequate foundation could be established for admission of the Bechtel letter, the district court properly exercised its discretion in refusing to admit it.
The district court also refused to admit a memorandum prepared by the chief legal officer of Collins Radio, which was circulated to certain managers of that firm. That memorandum concerned the attorney’s belief about the scope of interconnection required under the Specialized Common Carriers decision and discussed in part the injunction MCI recently had obtained from the federal district court in Pennsylvania. In a deposition, the attorney stated that he regularly prepared such memoranda on similar topics relevant to Collins Radio’s business. AT & T sought to introduce the memorandum under the business records exception to the hearsay rule to support its claim that it denied the interconnections in *1143good faith because the scope of the Specialized Common Carriers opinion was unclear. Judge Grady refused admission of the memorandum because he believed that it was not a business record and that it would be unduly prejudicial. Although Judge Grady may have erred in refusing to admit the document on those grounds, the error was harmless. The evidence was merely cumulative of the opinions of six other people who expressed the same view of the Specialized Common Carriers decision as that contained in the Collins Radio memorandum. AT & T even introduced a similar statement by the president of one of MCI’s companies. AT & T does not draw our attention to anything particularly novel or significant about the comments in the Collins Radio document.
AT & T objects to the admission of portions of documents comprising an AT & T internal study, called the Interbusiness Relations Report. The report contains statements attributed to AT & T management. The statements express management’s attitude that AT & T’s tactics in regard to interconnection were or should have been designed to obstruct commercial rivals’ legitimate competitive progress by denying interconnections. AT & T asserts that the documents are not admissions of the corporation because they were written by “low level” employees. While AT & T is correct that opinions of such employees without management responsibility are not properly considered to be admissions of the corporation, see, e.g., United States v. Siemens Corp., 621 F.2d 499 (2d Cir.1980), the employees whose opinions were admitted were not “low level.” Rather, the opinions admitted were those of high level management compiled in the course of the study. Precise identification of the persons who made the statements was not possible because AT & T had misplaced the interview sheets. The proposal for the format of the study, however, makes clear that the source of the report’s information was to be “the highest possible management level.” The study indicates that key managers (such as those who were “interfacing” on AT & T’s behalf with firms such as MCI) were involved in the report. Moreover, materials made available to MCI list numerous high-ranking managers as having participated at various stages in the report’s preparation. Cf. Siemens Corp., 621 F.2d at 508 (viewing documents as corporate admissions is proper when there is some indication that senior management has seriously considered and endorsed views stated in them). At any rate, AT & T agreed that the identity of the interviewees was appropriate for argument to the jury.
If the report is viewed as the work of agents of the corporation, it is admissible as an admission, since, at least as shown by circumstantial evidence, it was made while the managers were agents of the corporation; and it concerns matters within the scope of their agency, namely relations with competitors. Fed.R.Evid. 801(d)(2)(D); Mahlandt v. Wild Canid Survival & Research Center, Inc., 588 F.2d 626, 630 (8th Cir.1978); McCormick On Evidence § 267 at 642 (E. Cleary, ed., 2d ed. 1972). The report was the basis for corporate action since it evaluated alternative methods of dealing with competitors and involved a vast expenditure of corporate time. Pekelis v. Transcontinental & Western Air, Inc., 187 F.2d 122, 128-29 (2d Cir.) (A. Hand, J.), cert, denied, 341 U.S. 951, 71 S.Ct. 1020, 95 L.Ed. 1374 (1951). The fact that the report is based on hearsay or reflects opinion goes to its weight and credibility, not its admissibility.94 Pekelis, 187 F.2d at 129. The district court did not abuse its discretion in admitting the report.
7. Substantial Impact
AT & T contends that the evidence fails to sustain the jury’s finding that its FX and CCSA interconnection denials caused MCI substantial harm. AT & T argues that MCI had a significant backlog of uninstalled or*1144ders and had only been in operation for seven months. AT & T’s arguments lack merit.
From its inception, MCI was uncertain whether it could obtain the necessary interconnections. Consequently, MCI decided to defer its construction activities since it was unsure whether it could generate needed working capital, the anticipated source of which was cash flow from service revenues. The deferral decision in turn limited MCI’s scope of operations once interconnections became available. The jury was entitled to find that the time consumed by the denial of interconnection and MCI’s consequent entry into the market on a reduced scale were sufficient to substantially harm MCI, coming as they did during MCI’s start-up period.95
MCI’s backlog of circuit installations, the jury reasonably could believe, was caused in significant part by low morale resulting from layoffs that were made necessary by AT & T’s anticompetitive denial of interconnections. This was reflected by the comments in three internal MCI memoran-da. Other possible causes for the installation backlog (some of which concerned MCI mismanagement) were noted in those mem-oranda, and the jury was asked to consider those reasons as well as the ones that reflected adversely on AT & T. Contrary to AT & T’s contentions, there is nothing inconsistent between these jury determinations and the jury’s finding that AT & T did not harass MCI through its process of handling installations, and did not provide late or faulty installations. If there were other reasons for the backlog, such as low MCI employee morale brought on by AT & T’s anticompetitive actions, it hardly mattered how expeditiously AT & T responded with the installations that were effected. It was reasonable for the jury to find that, based on all the evidence, the refusals to interconnect had a substantial adverse impact on MCI.
The evidence was sufficient to sustain the jury’s conclusions that MCI suffered substantial harm from AT & T’s denial of FX and CCSA interconnections.
B. Tying
The jury found AT & T guilty of tying local to intercity telecommunications. AT & T contends that the jury’s findings are unsupported by substantial evidence and vitiated by erroneous instructions. MCI’s tying theory was based upon AT & T’s monopoly control over local interconnections. MCI argued that AT & T had sold these interconnections as a separate product to local customers, independent telephone companies and others for years. MCI claims that where FX and CCSA services were involved AT & T took the position that the local services could not be purchased separately, but could be used only in conjunction with AT & T’s long distance services.
MCI’s tying claim is, in reality, simply an alternate legal characterization of its claim relating to the FX-CCSA controversy. The only specific act allegedly involving tying that MCI can point to is AT & T’s unconditional refusal to interconnect specialized common carriers with its local distribution system. The facts underlying MCI’s theory are identical to those underlying the entire interconnection dispute. These claims have already been dealt with under the rubric of the essential facilities doctrine. Whether we label AT & T’s violation of the antitrust laws as tying or the denial of an essential facility, our prime concern is that AT & T used its monopoly power in local telephone service as a lever to impede or destroy competition in other markets. Nothing in this case hinges on which theory one uses to condemn AT & T’s conduct. We therefore need not reach the issue of tying given our disposition of the claims on other grounds.96
*1145C. Disconnections
AT & T contends that erroneous instructions account for the jury’s finding that AT & T improperly disconnected MCI customers after the Third Circuit’s decision overturning the district court injunction which ordered interconnection. AT & T specifically asserts that Judge Grady erred in failing to instruct the jury that Bell system tariffs precluded the interconnections, because Bell was obliged to follow the terms of those tariffs once the injunction was vacated.
The instruction given precluded liability unless the jury found that AT & T “did not believe it was acting lawfully in disconnecting these lines.” The instruction also conditioned liability on a finding that the purpose of the disconnections was to keep “MCI out of the market or unfairly [to limit] its ability to compete with AT & T.” AT & T contends that the instruction effectively directed a verdict against it, because even innocent disconnections obviously could keep MCI out of the FX and CCSA market. This argument ignores the portion of the instruction that required the jury to find that, before AT & T could be liable, AT & T must have believed it was disconnecting MCI’s customers unlawfully.
In any event, the tariffs filed by AT & T do not insulate it from the disconnection claim. The tariffs were issued by AT & T and took effect automatically unless the FCC rejected or suspended them— no special approval being required. 47 U.S.C. § 203(a), (b)(1) (1976). If the terms of the tariffs were themselves simply another device, designed and used by AT & T to inflict antitrust injury on MCI, then it does not matter that AT & T merely adhered to those terms. The jury was entitled to infer that AT & T failed to propose changes in the tariff terms. The existing terms were not instituted or required by the FCC, and AT & T could presumably modify them upon due notice, see Ambassador, Inc. v. United States, 325 U.S. 317, 323, 65 S.Ct. 1151, 1154, 89 L.Ed. 1637 (1945). The tariffs could not thus provide an excuse for AT & T’s knowingly anticompetitive conduct. See Cantor v. Detroit Edison Co., 428 U.S. 579, 592-96, 96 S.Ct. 3110, 3118-3120, 49 L.Ed.2d 1141 (1976) (even where utility could neither maintain nor alter allegedly anticompetitive tariff without state regulatory agency’s permission, antitrust laws still apply if “the option to have, or not to have, such a program [in the first instance] is primarily” the utility’s). Since AT & T presented to the jury its theory that its actions, purportedly based on its legal obligations arising under the tariffs, were lawful, and the substance of that defense was communicated to the jury in the instructions, no further elaboration of the specifics of AT & T’s position was necessary. See Beard v. Mitchell, 604 F.2d 485 (7th Cir.1979).
AT & T also argues that there was no evidence to support the jury’s conclusion that MCI sustained significant injury from the disconnections. Consumer and investor confidence in MCI was surely shaken by the abrupt interruption in its provision of service. This reaction was evidenced by testimony about the many telephone calls received from confused and distressed MCI customers. Even if MCI gave its customers advance notice that the Third Circuit might rule as it did, such a warning would not necessarily render the impact of AT & T’s unlawful conduct de minimis. AT & T’s anticompetitive actions created the impression that AT & T would continue to disrupt MCI’s efforts to serve its customers and consequently may have influenced MCI customers who feared that prospect. We approve the jury’s finding on the disconnection charge.
D. Denial of Interconnections for Service Outside of Local Distribution Areas
In the Specialized Common Carriers decision the FCC ordered AT & T to *1146provide local distribution facilities to the new carriers, and authorized the new carriers to construct inter-city communications facilities. AT & T developed geographic boundary maps for the cities MCI served. The territory within the boundaries was known as a local distribution area (“LDA”), and AT & T refused to provide MCI interconnections to reach its customers outside those boundaries.
' MCI adduced evidence that AT & T insisted on boundaries for the LDA’s that bore no relation to those AT & T normally used to define service limits. For example, AT & T’s explanatory letters to its local operating companies concerning LDA maps note the existence of “base rate boundaries” and “exchanges,” terms which refer to common or normal demarcations for provision of service. Yet the LDAs in virtually all the maps, as the letters point out, bear no apparent relationship to those demarcations.97 This is important because MCI asked for a reasonable definition of the LDAs. As one of AT & T’s witnesses acknowledged on cross-examination, MCI was dissatisfied with the various boundaries drawn up even after it participated in negotiations over the establishment of those boundaries. The boundaries were not what MCI wanted, but it was forced to accept them given the time constraints under which it operated. As an AT & T memorandum detailing the history of the new carriers describes, the carriers’ efforts to expand the number of customers they could serve were stymied by AT & T’s insistence on narrowly drawn boundaries. It was for the jury to determine in light of those and other circumstances whether AT & T’s methods of establishing LDAs were anti-competitively arbitrary and unduly limiting, or on the other hand, whether the boundaries were drawn with a proper motive to conform to the reasonable limits mandated by the FCC.
AT & T also argues that there was no evidence to show that the facilities denied were essential, as required under Judge Grady’s instructions. There was, however, testimony that MCI needed the facilities beyond the local distribution areas proposed by AT & T to reach the populations in reasonably delimited metropolitan areas surrounding the major cities MCI served and that MCI could not duplicate these facilities.98 The fact that MCI was *1147building its own facilities in at least one of the contested areas99 may properly have been taken by the jury as evidence that MCI was exercising its option under the 1971 Specialized Common Carriers decision to build its own facilities for end-to-end service, see 29 F.C.C.2d at 940, not that it could duplicate AT & T’s local service facilities at every point. Also, the FCC itself characterized service within the LDAs as “essential” even though the new carriers had an option to duplicate them. Id. We find no fault with the instruction as given.100
E. Multipoint Service
The jury found that AT & T denied interconnections for multipoint service to MCI with the intent to retain its monopoly. AT & T asserts that the instructions on this claim were incorrect because they were inconsistent with the instruction given under the LDA interconnection charge that MCI was not entitled to geographically unlimited interconnections. AT & T also asserts that it was incorrect to allow the jury to find the denial of multipoint interconnections unlawful under the essential facilities doctrine.
Multipoint service described the situation where AT & T provided a private line to a customer between city A and city B, and MCI provided a private line between city B and city C. MCI sought an interconnection in city B between its own line and AT & T’s line so that MCI’s customer in city C could have uninterrupted service between city C and city A. MCI claimed that its ability to compete in the market for city B to city C communications was substantially impaired if it was not able to offer its customers through service over AT & T’s lines to other cities which MCI did not serve itself. AT & T contended that multipoint interconnections effectively allowed MCI to provide its customers service to cities MCI could not reach with its existing equipment although *1148MCI was itself authorized to build the facilities that would provide the service.101
We find that, as a matter of law, the evidence was not sufficient to support a jury finding that AT & T denied multipoint interconnections with the intent to monopolize. We also agree with AT & T that the jury instruction was insufficient, although on grounds different from the one advanced by AT & T.
There are two independent theories upon which the denial of multipoint interconnections could have violated the antitrust laws. First, as we have discussed in terms of interconnections for FX-CCSA service and service beyond a local distribution area, the denial of multipoint interconnections could have been a violation of the antitrust laws if sufficient evidence had been presented that these were “essential services.” In general, a business has no legal obligation to deal with its competitors. There are situations, however, in which the federal courts have found a duty under section 2 of the Sherman Act for a monopolist to trade with all on nondiscriminatory terms. One of these instances is where the monopolist controls an “essential service” or “bottleneck.” Supra at pp. 1132-1133. We hold, however, that, as a matter of law, there was not sufficient evidence presented at trial to permit a finding that interconnection for multipoint service involved “essential services.”
MCI’s principal testimony on interconnections came from its president, William McGowan. While Mr. McGowan testified that interconnections were essential to connect MCI’s metropolitan terminal with the local Bell distribution system (the basis for the FX, CCSA, and LDA counts), none of this testimony addressed the need for interconnection to Bell’s intercity circuits. Tr. 372-80. Similarly, Mr. McGowan testified that it would be “physically impossible” to duplicate Bell’s local telephone system, but did not address the practicability of duplicating the private long distance circuits with which it had requested to be interconnected for multipoint service. Tr. 379. The evidence presented did not demonstrate either that the duplication of Bell’s intercity lines was economically infeasible or that the denial of access inflicted a severe handicap on market entrants. MCI’s primary business was to build precisely the type of facilities to which it sought access from the Bell System. There was no sufficient explanation as to why MCI, on the one hand, was building its own network, and, on the other, was entitled to access in the interim to AT & T’s facilities. Thus, the jury lacked sufficient evidence to conclude that these interconnections were essential.102
A second possible basis for imposing antitrust liability upon AT & T for the denial of multipoint interconnection would be a determination that AT & T’s actions in this respect were sufficient evidence of an intent to monopolize. In addition to the cases finding liability for a refusal to deal when an essential service is involved, there are cases which find liability when a monopolist’s refusal to deal with a competitor is shown to be evidence of an illegal intent to destroy competition. See Lorain Journal Co. v. United States, 342 U.S. 143, 72 S.Ct. 181, 96 L.Ed. 162 (1951); Eastman Kodak Co. v. Southern Photo Materials Co., 273 U.S. 359, 47 S.Ct. 400, 71 L.Ed. 684 (1927); United States v. Colgate & Co., 250 U.S. 300, 39 S.Ct. 465, 63 L.Ed. 992 (1919). These cases focus on the intent and competitive effect of the refusal to deal; not on whether the facility itself is “essential.” It is settled law that actions which might be lawful in another context can constitute a *1149violation of section 2 of the Sherman Act if they are done with the purpose of benefiting a monopolist as against its competitors, and have the effect of smothering competitors, either in the market where the monopoly power exists or in adjacent markets. See United States v. Griffith, 334 U.S. 100, 68 S.Ct. 941, 92 L.Ed. 1236 (1948); Official Airline Guides, Inc. v. F.T.C., 630 F.2d 920 (2d Cir.1980), cert. denied, 450 U.S. 917, 101 S.Ct. 1362, 67 L.Ed.2d 343 (1981); Berkey Photo, Inc. v. Eastman Kodak Co., 603 F.2d 263 (2d Cir.1979), cert. denied, 444 U.S. 1093, 100 S.Ct. 1061, 62 L.Ed.2d 783 (1980); Sargent-Welch Scientific Co. v. Ventron Corp., 567 F.2d 701 (7th Cir.1977), cert. denied, 439 U.S. 822, 99 S.Ct. 87, 58 L.Ed.2d 113 (1978); Town of Massena v. Niagara Power Corp., 1980-2 Trade Cas. ¶ 63,526 (N.D.N.Y.1980).
We do not think that, given the unsettled regulatory status of the telecommunications industry at the time of these events, sufficient evidence was presented to the jury to permit a finding that AT & T’s denial of interconnection for multipoint service was primarily motivated by an illegal intent to monopolize. Contrary to MCI’s assertion, multipoint interconnection was substantially different in character from the other types of interconnections sought by MCI. Multipoint interconnection was the device through which MCI sought access to the full scope of AT & T’s nationwide long distance network. Granting MCI multipoint interconnections would have enabled MCI to compete with AT & T for long distance traffic into areas where MCI may have made no significant capital investment. At the time in question, the FCC may or may not have intended (or indeed may or may not now or in the future intend) to effect such a significant change in the structure of the national telecommunications industry, and to impose upon AT & T the extraordinary obligation to fill in the gaps in its competitor’s network.
As a matter of antitrust liability only, however, can an entrant which actually builds its own facilities between Chicago and Milwaukee, for example, thereby gain entitlement to use all the far-flung facilities of the Bell System? Is its entitlement based on its expressed intention to duplicate major portions of the Bell System on a national basis? Could it claim entitlement before (or without) building any facilities of its own? We think the ramifications of the demand for multipoint service are troubling and complex, and that under the circumstances of this case and without reliance on a regulatory determination, the denial of interconnections for multipoint service cannot form a basis of liability. Therefore, as a pure matter of antitrust law (without any regulatory component), we decline to hold AT & T liable for a refusal to make available its full nationwide network to a competitor. Instead we find that AT & T could only have been liable for denying multi-point interconnections under a theory that this denial was sufficient evidence of monopolistic intent, if the FCC had authorized or mandated multipoint interconnections. Because we find no such order or authorization to have been clearly made, we decline to inject the complications of a generally amorphous antitrust doctrine into what is, at heart, a regulatory matter.
We also find that the instructions which the jury received were insufficient to allow liability for the denial of multipoint interconnection to be premised on the proposition that the FCC’s Specialized Common Carriers decision actually ordered AT & T to provide such interconnections.103 The district court’s instruction on multipoint interconnection essentially stated that, in order to prevail, MCI had to “prove that AT & T unreasonably denied those interconnections with the intent of *1150maintaining a monopoly.”104 Instruction No. 32, App. 1203. This instruction, unlike the instructions given under the FX, CCSA, and LDA interconnection claims, entirely failed to instruct the jury as to the relevance or applicability of the FCC’s determinations, if any, on the subject. No mention was made either of any guidance which may have been afforded to AT & T by the FCC’s decision or of AT & T’s beliefs as to its obligations (or lack thereof) under the Specialized Common Carriers decision. We thus hold that the jury could not have reasonably found, based upon the instructions it received, that AT & T’s refusal to provide multipoint interconnections was in violation of the FCC’s decision and thus evidence of the intent necessary for a violation of the antitrust laws.105
F. Inappropriate or Inefficient Interconnections
AT & T contends that there was no support for the jury’s finding that AT & T provided MCI with “inappropriate or inefficient equipment or procedures for interconnection.” Specifically, AT & T charges that the jury based its findings on generalized complaints, impermissible references to a 1975 agreement arising from an FCC proceeding,106 and vague instructions that allowed the jury to consider an impermissible theory. Because of these errors, AT & T claims that the jury imposed liability despite adequate technical arrangements. We find, however, that the jury was properly instructed and that there was sufficient documentary and testimonial support for the finding of technically inadequate interconnections.107
The evidence indicates that AT & T’s insistence on a “clean interface”108 between MCI and AT & T equipment was at the heart of many of MCI’s technical complaints. To facilitate that requirement, AT & T insisted that all interconnection with MCI be made at the customer’s premises. The customer premises restriction precluded the central office terminations needed to provide FX and CCSA service. E & M signaling109 was provided because it facilitated the “clean interface” concept. Unlike the other available forms of signaling, E & M could be separated and tested independently at the point of conversion, and was *1151best suited for short distance transmission on the customer’s premises. The evidence indicated that E & M signaling caused increased cost to MCI and decreased reliability of service. As a result, MCI provided slow response time in repairing customer equipment.
An internal AT & T report demonstrated that AT & T itself recognized that interconnection at the customer’s premises would cause problems. The report stated:
Our study shows that an interface at an intermediate customer’s premises results in redundant local facilities; degraded service, higher costs to the customer, and a more difficult maintenance arrangement for both Bell and the [other common carriers].
PX 323 at p. 275.
Despite AT & T’s knowledge and despite testimony that MCI requested DX, a different signaling method, AT & T would only provide E & M signaling and insisted on connections at the customer’s premises. The jury was entitled to conclude that AT & T’s actions were deliberately taken to impede MCI’s progress.
AT & T also argues that MCI approved of the equipment and procedures employed by AT & T. In support, AT & T points out that an MCI vice-president initially agreed in an internal memorandum that E & M signaling was “the best long term solution.” The MCI vice-president, however, recognized even in the early memorandum that E & M signaling was not appropriate for MCI’s needs as it started business:
Bell ... want[s] to interface on an E & M basis with MCI. We agree, this is the best long term solution. However, this usually involves Western Electric and for example, in Ohio MCI has been told that it will take 15 weeks to complete and accept the work, and cost $300 to $400 which will be passed on directly to the customer by Bell. This, of course, is not acceptable to MCI in a start up situation.
DX 768. (Emphasis in original).
The trial testimony also indicates that the MCI vice-president qualified his opinion as to the long-run viability of E & M after discussing the alternative forms of signaling.
As to physical equipment, MCI asserted that the 66-type connecting blocks, which were used to connect MCI’s wires to those of AT & T, were inadequate. MCI’s technical witness testified that the 66-type connecting block was designed to be used at a customer’s premises where the customer had more than one line, or in conjunction with the customer’s switchboard unit. The witness testified that the 66-type block was not designed to be used in an area of high activity. MCI sought the 300-type blocks because they were more appropriate for high volume areas and could be centralized at MCI’s own terminals. MCI contended that the 300-type block was needed because MCI was a communications carrier and not merely an AT & T customer seeking connections at separate points of service, as AT & T preferred to think. There was conflicting testimony about whether the 300-type connecting block differed from the 66-type connecting block.110 There was sufficient evidence, however, from which the jury could find that the 66-type block was inappropriate for MCI’s needs and was provided as a means to harm MCI.
MCI also complained that AT & T did not maintain sufficient trouble reporting procedures. There was evidence that AT & T only provided MCI with essentially the same procedures AT & T provided for individual customers. MCI claimed that these procedures were inadequate to handle the special, problems occurring in the network MCI provided its customers. AT & T contends that MCI was satisfied with the procedures because the chairman of MCI’s board, in a 1973 letter to the FCC, expressed his approval of “the fair price of *1152service [and] the technical details of accomplishing interconnection to date.” The letter, however, did not specifically address trouble reporting procedures. Rather, it simply, enumerated five requests for interconnection between MCI and either customer premises or AT & T’s facilities. In addition, there was testimony that MCI’s dissatisfaction with the trouble reporting procedures continued after 1973.111 The jury was entitled to take this into account in evaluating AT & T’s argument.
AT & T asserts that the trial court erred by allowing the jury to hear references to a 1975 agreement in an FCC proceeding initiated to consider MCI’s complaints and resolved through negotiation. AT & T relies on Federal Rules of Evidence 407 and 408, which limit the evidentiary use of subsequent remedial repairs and settlement negotiations, respectively. On direct examination, an MCI vice-president mentioned the settlement agreement. . Instead of objecting on the basis of Rule 407 or 408, however, AT & T’s counsel moved for a mistrial on the ground that the reference violated a provision in the agreement that stated, “[N]othing contained in this settlement agreement ... shall constitute an admission by any party.... ” The district court denied the mistrial motion, finding that AT & T had not been prejudiced. AT & T cannot argue other grounds for reversal on appeal, because “if a specific objection is overruled, only the ground stated in the objection [can] be raised on review.” 21 C. Wright & K. Graham, Federal Practice and Procedure § 5036 at 183 (1977).
Later in the trial, an MCI technical witness characterized the post-1975 technical arrangements MCI had with AT & T as superior to those before 1975, but did not mention the settlement. The court then informed the jury of the agreement in order to explain why services had changed so suddenly in 1975. Judge Grady explained how the jury should view the agreement:
The reason that I allowed in the evidence about [the agreement] is ... that I just don’t see any way you can try the case without knowing what happened. It would be an artificial situation for you not to understand the course of the events. But it is very important for you to understand that the defendant doesn’t admit anything and its having made those changes and having agreed to those changes does not constitute an admission.
Prior to this testimony, AT & T objected, arguing that the settlement provision and the doctrine of subsequent remedial repairs precluded the testimony. On appeal, AT & T argues that the testimony was inadmissible because a court may not admit evidence of a settlement agreement for the purpose of proving fault or liability. Settlement negotiations, however, are admissible to explain another dispute and to assist the trier of fact in understanding the case. Subsequent repairs are also admissible to demonstrate technical feasibility, which was at issue here (a point AT & T does not contest on appeal), and which was a basis for Judge Grady’s decision to admit the testimony. At the time the evidence was introduced, the district court properly instructed the jury on the limited use of the testimony.
Finally, AT & T contends that the district court’s instruction failed to establish any meaningful standard to guide the jury in its deliberations.112 The submis*1153sion and form of instructions, however, are matters within the discretion of the trial court. Instructions must be viewed in their entirety and verdicts will not be overturned by picking and choosing words from an instruction without regard to the whole trial. E.I. du Pont de Nemours & Co. v. Berkley & Co., 620 F.2d 1247, 1270 (8th Cir.1980). The instruction gave sufficient guidance to the jury by setting forth the technical areas in contention. It also stated that in order to find liability, the jury must determine that AT & T knowingly provided inefficient or inappropriate equipment or services with improper intent. This covered AT & T’s claim that it believed even MCI was satisfied that the equipment was adequate. The district court did not abuse its discretion in so charging the jury.113
V. BAD FAITH NEGOTIATIONS AND NOERR-PENNINGTON
A. The State Tariff Filings
The jury found that AT & T filed tariffs in bad faith with state utility commissions as an act in willful maintenance of its monopoly position. AT & T contends that this activity merely constitutes the petitioning of the government protected under the First Amendment and is therefore immune from antitrust scrutiny.
Under the so-called Noerr-Pennington doctrine, activities such as state tariff filings are immune from antitrust liability where their purpose is to influence government action. The doctrine arose from the need to construe the antitrust laws in such a way as to avoid a conflict with the right to petition the government protected under the First Amendment. This immunity doctrine was first enunciated by the Supreme Court in Eastern Railroad Presidents Conference v. Noerr Motor Freight, Inc., 365 U.S. 127, 81 S.Ct. 523, 5 L.Ed.2d 464 (1961). While in Noerr the Court held that “no violation of the [Sherman] Act can be predicated upon mere attempts to influence the passage or enforcement of laws,” Id. at 135, 81 S.Ct. at 528, it explicitly excluded from immunity activities which it labeled “mere sham” and defined as “nothing more than an attempt to interfere directly with the business relationships of a competitor.” Id. at 144, 81 S.Ct. at 533. As the Court restated in United Mine Workers of America v. Pennington, 381 U.S. 657, 85 S.Ct. 1585, 14 L.Ed.2d 626 (1965), Noerr immunity obtains so long as the attempt to influence government action is made in good faith.
The Court in California Motor Transport Co. v. Trucking Unlimited, 404 U.S. 508, 92 S.Ct. 609, 30 L.Ed.2d 642 (1972), extended Noerr-Pennington to administrative and adjudicatory proceedings. The Court also applied the “sham litigation” exception for the first time, holding that a common plan to oppose every application for a trucking permit, regardless of the merits, stated a cause of action under the antitrust laws. The Court stated that, “One claim, which a court or agency may think baseless, may go unnoticed, but a pattern of baseless, repeti*1154tive claims may emerge which lead the fact-finder to conclude that the administrative and judicial processes have been abused.” Id. at 513, 92 S.Ct. at 613. . AT & T seizes upon this language to define the scope of the sham exception, arguing that since no pattern of baseless, repetitive claims was shown, Noerr-Pennington applied as a matter of law to immunize the state tariff filings.
■Although the Court in California Motor Transport specifically mentioned repetitive, baseless claims by way of example, we believe its rationale is not so limited. In Vendo Co. v. Lektro-Vend Corp., 433 U.S. 623, 97 S.Ct. 2881, 53 L.Ed.2d 1009 (1977), the defendant had filed a single state suit, allegedly for the purpose of harassment and the elimination of competition. Four Justices in a strong dissent held fast to the view that California Motor Transport did not limit the sham exception to “a pattern of baseless, repetitive claims,” but indicated that it may include a single use of the adjudicatory process to violate the antitrust laws. Id. at 661-62, 97 S.Ct. at 2902-2903. A plurality of three other Justices reversed the district court’s injunction against enforcement of the state judgment on the basis of the anti-injunction statute, 28 U.S.C. § 2283, but noted that:
Any “disadvantage” to which the federal plaintiff is put in the [state court] proceeding is diminished by his ability to set up the federal antitrust claim as an affirmative defense ... and his ability to sue for treble damages resulting from the vexatious prosecution of that state-court litigation.
433 U.S. at 636 n. 6, 97 S.Ct. at 2890 n. 6.
Thus the Court has left open the question whether a pattern of several claims is required to constitute a “sham.” A close reading of the plurality and dissent in Lek-tro-Vend suggests that a majority of the Court believes that a single claim, lawsuit or petition can be “sham litigation” actionable under the antitrust laws.
The recent Ninth Circuit case of Clipper Exxpress v. Rocky Mountain Motor Tariff Bureau, Inc., 690 F.2d 1240, 1254-57 (9th Cir.1982), provides the most thorough examination of the rationale for permitting antitrust liability to rest on the prosecution of a single claim before an administrative agency. In Clipper Exxpress the defendant opposed the filing of a single tariff before the ICC. The district court granted summary judgment for the defendants holding that the defendants’ conduct was immunized by the Noerr-Pennington doctrine. The Ninth Circuit reversed the summary judgment and held that a single baseless claim could constitute sham litigation. The court analyzed the issue as follows:
The sham exception ... reflects a judicial recognition that not all activity that appears as an effort to influence government is actually an exercise of the first amendment right to petition. At times this activity, disguised as petitioning, is simply an effort to interfere directly with a competitor. In that case, the “sham” petitioning activity is not entitled to first amendment protection, because it is not an exercise of first amendment rights.
If the activity is not genuine petitioning activity, the antitrust laws are not suspended and continue to prohibit the violating activities. Because application of the antitrust laws is not suspended, it will prohibit sham activity, whether that activity consists of single or multiple sham suits. This analytical framework does not permit a conclusion that single sham suits are protected under Noerr.
690 F.2d at 1255 (emphasis in original).
Clipper Exxpress joins a growing list of federal cases which have held that a single lawsuit or claim may constitute sham litigation. Feminist Women’s Health Center v. Mohammad, 586 F.2d 530, 543 n. 6 (5th Cir.1978), cert. denied, 444 U.S. 924, 100 S.Ct. 262, 62 L.Ed.2d 180 (1979); First National Bank of Omaha v. Marquette National Bank, 482 F.Supp. 514, 519-21 (D.Minn. 1979); Technicon Medical Information Systems Corp. v. Green Bay Packaging, Inc., 480 F.Supp. 124 (E.D.Wis.1979); Colorado Petroleum Marketers Association v. Southland Corp., 476 F.Supp. 373, 377-78 (D.Colo. 1979); Cyborg Systems v. Management Sci*1155ence America, Inc., 1978-1 Trade Cas. ¶ 61,927 (N.D.Ill.1978); Associated Radio Service Co. v. Page Airways, Inc., 414 F.Supp. 1088 (N.D.Tex.1976). Cf. Walker Process Equipment, Inc. v. Food Machinery & Chemical Corp., 382 U.S. 172, 86 S.Ct. 347, 15 L.Ed.2d 247 (1965) (a single instance of enforcing a fraudulently procured patent can violate Section 2 of the Sherman Act). See generally Balmer, Sham Litigation and the Antitrust Laws, 29 Buff.L.Rev. 39, 55-56 (1980); Fischel, Antitrust Liability for Attempts to Influence Government Action: The Basis and Limits of the Noerr-Pennington Doctrine, 45 U.Chi.L.Rev. 80 (1977). But see Hydro-Tech Corp. v. Sundstrand Corp., 673 F.2d 1171 (10th Cir.1982). We therefore find that the bringing of baseless claims — even the undertaking of a single sham state court lawsuit — is devoid of the constitutional significance that warrants immunity from the antitrust laws.114
AT & T also challenges the jury instructions on this issue. In its instructions the court told the jury that the claim based on the state tariff filings “involves a resort by AT & T to a regulatory process, and involves certain First Amendment questions, that is, freedom of expression and freedom to seek redress from the appropriate regulatory body.” Tr. 11483. The court went on to explain that consequently the filings would have to be found to be “an actual use of the regulatory process in a way that is really a sham and not a legitimate effort to present an arguable question, or legitimate question to a regulatory body.” Tr. 11484. The court stressed that MCI’s burden on this claim is higher (clear and convincing evidence) because of these First Amendment concerns. While the instruction’did not mention the Noerr-Pennington doctrine by name, the essentials of the doctrine and the exception from it for sham litigation were clearly and correctly explained. We find the instruction adequate.
MCI still bears the burden of proof that AT & T’s tariff filings were, in fact, “sham.” The Noerr and Pennington cases themselves provide little definition of what a “sham” may be other than to indicate immunity for “genuine efforts” and “good faith” attempts to influence governmental bodies. In California Motor Transport the Court held that the allegation that the defendants “instituted the proceedings and actions ... with or without probable cause, and regardless of the merits of the case” falls within the meaning of sham litigation. 404 U.S. at 512, 92 S.Ct. at 612. The Court also defined sham litigation as the filing of baseless claims. Id. at 513, 92 S.Ct. at 613. In concurring in the judgment, Justice Stewart indicated a willingness to allow an antitrust cause of action if the defendant “had made misrepresentations of fact or law to [the] tribunals, or had engaged in perjury, or fraud, or bribery.” Id. at 517, 92 S.Ct. at 615 (Stewart, J., concurring in the judgment).
The lower courts have tended to read California Motor Transport narrowly so as not to tread on the First Amendment freedoms underlying the Noerr-Pennington doctrine. Mid-Texas Communications Systems *1156v. AT & T, 615 F.2d 1372, 1384 (5th Cir.), cert. denied, 449 U.S. 912, 101 S.Ct. 286, 66 L.Ed.2d 140 (1980). The Ninth Circuit has stated that, “The sham exception permits the imposition of antitrust liability on those seeking action from government agencies only when the activity in question can serve no useful purpose, and is undertaken for purely anticompetitive reasons.” Forro Precision, Inc. v. IBM Corp., 673 F.2d 1045, 1060 (9th Cir.1982). The Tenth Circuit apparently has taken the most restrictive view among the Circuits, holding that sham denotes “misuse or corruption of the judicial process.” Hydro-Tech Corp. v. Sundstrand Corp., 673 F.2d 1171, 1176-77 (10th Cir. 1982) (citing Semke v. Enid Automobile Dealers Association, 456 F.2d 1361, 1366-67 (10th Cir.1972)). The Tenth Circuit in Hydro-Tech applied its restrictive rule to hold that the mere absence of probable cause in the initiation of a lawsuit is not enough to invoke the sham exception to Noerr-Pen-nington, 673 F.2d at 1176.
One of the more cogent definitions of sham litigation may be found in Gainesville v. Florida Power & Light Co., 488 F.Supp. 1258 (S.D.Fla.1980):
Without a doubt, the intention to harm a competitor is not sufficient to make litigation or administrative proceedings a sham. That anticompetitive motive is the very matter protected under Noerr-Pennington. Rather, the requisite motive for the sham exception is the intent to harm one’s competitors not by the result of the litigation but by the simple fact of the institution of litigation.
488 F.Supp. at 1265-66 (emphasis in original).
This Circuit has had several opportunities to construe California Motor Transport but has not set forth any general definition of “sham” litigation. Most recently this court has analogized sham litigation to the common law torts of malicious prosecution and abuse of process. Grip-Pak, Inc. v. Illinois Tool Works, Inc., 694 F.2d 466 (7th Cir. 1982). In City of Mishawaka v. American Electric Power Co., this court held that the utility’s rate filings with the Federal Power Commission were not protected by Noerr-Pennington because they were “an abuse of the administrative process” and denied the plaintiff municipalities “fair and effective access to the regulatory process.” 616 F.2d 976, 982-83 (7th Cir.1980), cert. denied, 449 U.S. 1096, 101 S.Ct. 892, 66 L.Ed.2d 824 (1981). See also Kurek v. Pleasure Driveway and Park District, 557 F.2d 580, 594 (7th Cir.1977) (proof that economically unrealistic proposal was submitted to the Park District might support an inference that the proposal was not a genuine effort to persuade public officials).
In the present case MCI alleges that AT & T filed tariffs with state regulatory commissions in bad faith as part of a continuing effort to deny MCI interconnections for local facilities. Complaint Paragraph 23(c)(5) and (1). AT & T argues that there was no basis upon which the jury could have found the state tariff filings a sham or an abuse of agency process. MCI contends that these tariffs were sham proceedings in the sense that AT & T filed the tariffs knowing that the state commissions lacked jurisdiction over long distance interconnection matters. Deliberately bringing administrative actions with the knowledge that the agencies involved lacked jurisdiction would seem to be a strong indication that AT & T did not engage in a genuine effort to influence public officials. There can be no genuine attempt to petition the government when the petitioners know in advance that the governmental body lacks the authority to take the action desired. The sham activity, as alleged by MCI and as found by the jury, is analogous to the institution of numerous appeals when the defendants know that they lack standing. The Second Circuit recently held that this sort of baseless and frivolous activity falls within the sham litigation doctrine. Landmarks Holding Corp. v. Bermant, 664 F.2d 891 (2d Cir.1981).
Although the question of FCC jurisdiction over interconnection tariffs had not been judicially determined at the time *1157AT & T filed the tariffs,115 MCI did present substantial evidence to support the jury’s finding that, at the time the tariffs were filed, AT & T believed that the state commissions lacked jurisdiction to approve them. This evidence includes the testimony of Mr. Strassburg, the FCC’s Common Carrier Bureau Chief, AT & T’s own expectations expressed in internal memoranda, and AT & T’s previous position in FCC proceedings that state commissions did not have jurisdiction over tariffs applicable to MCI. These factors, together with testimony that MCI remained uninformed of the filings despite ongoing interconnection negotiations with AT & T, allowed the jury reasonably to infer that AT & T filed these tariffs solely as a means of undermining the negotiations.
While the filings of tariffs with state commissions are not “immune” under Noerr-Pennington, on the other hand, their filing by AT & T was not necessarily unlawful. Any number of actions are not immune from Sherman Act scrutiny but are entirely lawful because they have no anti-competitive effect or purpose. AT & T has many thousands of tariffs on file with various state utility commissions around the country. Even if these filings are for some reason not immune under Noerr-Penning-ton, the vast majority of them cannot possibly constitute the basis of an antitrust violation merely because the agency with which they are filed may be found to “lack jurisdiction.” Some tariffs are filed with commissions for purely informational purposes. See supra, note 114. Other tariffs relate to rates and services which have no competitive effect whatsoever and are therefore not illegal even if they are not immune.
The tariffs filed with the state commissions which are at issue in this case relate to the charges for interconnection to Bell’s local distribution facilities. AT & T claims that the tariffs were filed to supplant a long-standing contractual arrangement with Western Union in order to set uniform terms and conditions for nationwide interconnections with other common carriers like MCI. MCI, on the other hand, claims that the filings were part and parcel of AT & T’s continuing efforts to unlawfully deny all interconnections related to FX and CCSA.
MCI alleged in its complaint that:
AT & T, acting in bad faith, caused to be filed with various state regulatory commissions sham tariffs which purportedly regulated the provision by the Bell System Companies of interconnection to plaintiffs for use in interstate commerce. Defendant AT & T thereby imposed upon plaintiffs a substantial financial burden for the purpose of exhausting plaintiffs’ resources and destroying plaintiffs as potential competitors.
Complaint 23(1).
These allegations were supported by the testimony of William McGowan and Laurence Harris. Mr. McGowan testified that after twelve months of negotiations, during which time MCI was not receiving any interconnections besides Chicago-St. Louis, AT & T broke off talks completely and insisted that MCI would now have to take up the subject of interconnections with each of the state commissions. Tr. 414. MCI was then forced to undertake lengthy proceedings before the FCC and in the courts to vindicate its rights to interconnections for FX and CCSA service.
*1158This testimony was corroborated by that of Laurence Harris, who was MCI’s chief negotiator on interconnections with AT & T. Mr. Harris described the filings of the state tariffs as the culmination of over twelve months of bad faith negotiations. Early in the discussions, in September 1972, he had informed AT & T of the urgency of working out interconnections to permit MCI to begin its expanded service in the summer of 1973. Tr. 893-94. Harris was subjected to months of fruitless negotiations involving the purchase rather than the lease of facilities, the creation of local distribution areas, pricing and technical disputes. During the entire period MCI received no new interconnections whatsoever.
Then on September 1, 1973, Mr. Harris learned that AT & T was not going to resume negotiations but instead had filed tariffs which did not include FX service, CCSA service or service outside narrowly circumscribed local distribution areas, Tr. 1015, 1049, with forty-nine state utility commissions. Prior to the filing of the state tariffs MCI was negotiating with AT & T for a single contract with a single set of terms and conditions that would be used by all of the AT & T operating companies. Tr. 1013. Because of the filings in state commissions, MCI faced the prospect of negotiations and litigation before forty-nine separate state regulatory agencies.
As one commentator has noted, lawsuits and administrative actions by powerful firms “can tie up smaller businesses in uncertain and expensive proceedings, thereby increasing the cost of doing business and preventing or delaying new entries into a particular market.” Balmer, Sham Litigation and the Antitrust Laws, 29 Buff.L.Rev. 39 (1980). See generally R. Bork, The Antitrust Paradox, 347-64 (1978). Based on the record in this case, the anticompetitive aspects of AT & T’s course of conduct are apparent. AT & T’s bad faith negotiations culminating in the tariff filings harmed MCI in a number of significant ways. First, MCI had to face the prospect of litigation in forty-nine different forums to establish its right to offer its private line services. Any changes in its own charges or services might also then have to be brought before these forty-nine state commissions. This requirement alone would greatly increase litigation costs by requiring MCI to spread its resources across the country and would necessitate the retention of numerous local legal counsel. Perhaps the most harmful aspect of this strategy is the possibility that at least one commission would refuse to allow the proposed service, which might undermine the creation of a nationwide communication system. AT & T’s strategy also forced MCI to bear additional costs and expenses in connection with litigation before the courts and the FCC to establish its right to interconnections.
Finally, and most importantly, the filings with the state commissions (as a culmination of the negotiations) added additional months of delay before MCI could enter the market. The combined effect of the bad faith negotiations and the filing of the state tariffs may have cost MCI one and a half years of time and revenue while it fought for interconnections, first with AT & T at the bargaining table and later in the FCC and the courts. In October 1973, MCI had its terminals in place and had expended substantial amounts of capital only to have its revenue flow obstructed because of its inability to operate while AT & T denied interconnection. Tr. 1067. On this basis the jury properly found that the bad faith negotiations culminating in the filing of state tariffs were unlawful acts committed in order to maintain AT & T’s monopoly position.
B. Bad Faith Negotiations
AT & T also challenges other aspects of the jury’s finding that bad faith contractual negotiations culminated in sham filings before state regulatory commissions. On the issue of bad faith negotiations AT & T argues that the court erroneously instructed the jury that MCI need only “establish that defendant negotiated in bad faith for purposes of delaying plaintiff’s entry . .. and with the intent of maintaining a monopoly.” Appellant’s *1159Brief at 128. AT & T argues that this instruction allowed the jury to hold the entire course of negotiations unlawful without considering the First Amendment implications of filing tariffs with the states. The finding of bad faith negotiations is, however, an entirely separate count from the state tariff filings.
The negotiations themselves were a business transaction carried on by two private parties, AT & T and MCI, and involved no resort to any governmental processes. The jury had ample evidence to hold that these negotiations were conducted in bad faith based on the record of delay, the bargaining position of AT & T, and the ultimate futility of the negotiations. These findings did not depend on the state tariff filings. The lack of candor with respect to the decision to file the tariffs was at most additional circumstantial evidence of bad faith. And, for whatever it may be worth, our conclusion that the state tariff filings were not protected by the Noerr-Pennington doctrine would apply with equal force here, were we somehow to conclude that Noerr-Penning-ton applied to the bad faith negotiations.
AT & T also contends that the verdict on this charge cannot stand because the court failed to define “bad faith” in its instruction. It is only error to fail to define “enigmatic terms” that leave the jury to speculate on their meaning. See Kocher v. Creston Transfer Co., 166 F.2d 680 (3d Cir. 1948). Here, however, the use of the term “bad faith” comports with its ordinary meaning116 and needs no further explanation.
AT & T also claims that there was no evidence that injury resulted from the alleged bad faith negotiations since the facilities eventually were provided. As we have previously indicated MCI introduced sufficient evidence to show that AT & T’s course of conduct delayed its entry into the market and affected MCI’s ability to attain financial viability. Therefore, there was sufficient evidence to sustain the jury’s finding of injury as a result of bad faith negotiations.
Finally, AT & T attacks the jury’s finding of bad faith negotiations as inconsistent with other related findings. But the fact that AT & T was exonerated of charges of discrimination against MCI in favor of Western Union and of charging excessive prices for local interconnections does not foreclose the possibility that bad faith negotiations were found on the basis of other acts. Evidence such as the filing of “sham tariffs,” the failure to disclose this action and other delaying tactics could have contributed to the jury’s conclusion. Given the fair inferences that the jury could have drawn from the other evidence, we cannot engage in the type of speculation AT & T urges to construe the jury’s special findings as irreconcilable. See Stockton v. Altman, 432 F.2d 946, 951 (5th Cir.1970), cert. denied, 401 U.S. 994, 91 S.Ct. 1232, 28 L.Ed.2d 532 (1971).
C. Other Conduct
AT & T additionally argues that á general Noerr-Pennington instruction was necessary because “virtually every liability issue found adversely to AT & T by the jury involved conduct subject to regulation and the Bell System’s participation in the regulatory process.” Appellant’s JSdgf at 142-43. AT & T notes that MCI pointed to many instances of AT & T’s participation in~~ the administrative process as evidence of'" bad faithr-and anticompetitive intent. The court’s failure to give such an instruction, AT & T insists, is reversible error.
The Noerr-Pennington doctrine is concerned solely witl^.Ue~~right to attempt to influence government action. It thus immunizes only those actions'lIirecfeiT toward governmental agencies or officials. *1160The fact that a common carrier’s decision may eventually provoke agency action or review does not alone call the Noerr-Pen-nington doctrine into play. Except for the claim based on state tariff filings (and perhaps the pre-announcement of Hi-Lo), none of the claims sought to impose liability for resort to agencies or courts. Tn Mid-Texas Communications Systems v. AT & T, 615 F.2d 1372 (5th Cir.1980), the Fifth Circuid considering a somewhat similar refusal ta interconnect, held that Bell’s action “was' not an attempt to influence governmental action so as to warrant protection under Noerr-Pennington.” Id. at 1383; see Cantor v. Detroit Edison Co., 428 U.S. 579, 601-02, 96 S.Ct. 3110, 3122-3123, 49 L.Ed.2d 1141 (1976).
Rather, MCI referred to AT & T’s actions before the FCC only as evidence of the purpose and character of business decisions which had already been made and which were relevant to chames other than the filings of state tariffs. “Evidence of activity that is protected by the Noerr doctrine may be admitted to show the purpose and character of other activities if doing so is not overly prejudicial to the defendants.” Feminist Women’s Health Center v. Mohammad, 586 F.2d 530, 543 n. 7 (5th Cir. 1978), cert. denied, 444 U.S. 924, 100 S.Ct. 262, 62 L.Ed.2d 180 (1979); see also United Mine Workers of America v. Pennington, 381 U.S. 657, 670 n. 3, 85 S.Ct. 1585, 1593 n. 3, 14 L.Ed.2d 626 (965). We find no error in the district court’s refusal to give a general Noerr-Pennington instruction.
VI. DAMAGES
We come at last to the substantial award of damages in this case. The jury, after deliberation, found for MCI on ten of the fifteen counts it considered. After completing the special verdict the jury awarded MCI $600 million in general damages which were trebled in accordance with the antitrust laws to total $1.8 billion in damages. Because this determination awards damages for both lawful and unlawful conduct, the damage award must be set aside and the case remanded for a proper determination of damages.
A. MCI’s Proof of Damages
At trial, MCI’s proof of the amount of damages centered on the introduction of a lost profits study. The study sought to calculate the difference between the revenues received by MCI as damaged by the alleged exclusionary acts of AT & T, and those revenues which could have been expected for an undamaged MCI. The data for the “damaged” MCI came from the actual operating figures of the company, plus projections into the future. The data for the “undamaged” MCI purportedly came from the original business plans for the corporation,117 modified for certain events unattributable to AT & T.
After the differences between anticipated and realized revenues were derived, the differentials in income over a twenty year period were reduced to present value using MCI’s estimated cost of capital. The study showed losses for MCI totalling $452,215,-000. This figure was then adjusted to $900,000,000 to produce an after-tax result to MCI equal to the alleged financial losses. The jury, upon finding liability on ten of the fifteen alleged acts of monopolization brought by MCI, returned a general verdict for MCI in the amount of $600,000,000, which was trebled to produce a damage award of $1,800;000,000. AT & T challenges the jury award on the grounds that the lost profits study failed to separate injury caused by lawful competition from that caused by unlawful conduct and that the study was based upon unsupportable assumptions not found in the record. AT & T specifically challenges the study’s assumptions that MCI could have achieved revenues of $.85 a circuit mile per month; that MCI could have attained a market share of thirty-seven million circuit miles by 1975; and that MCI could have financed the communications systems envisioned in the study. MCI relies on case law to the effect that a successful antitrust plaintiff need *1161not “disaggregate” or “tightly compartmentalize” its damages once the fact of injury has been established. MCI also defends the specific assumptions relied on by the authors of the lost profits study.
B. Causation of Damages
If a plaintiff has suffered financial loss from the lawful activities of a competitor, then no damages may be recovered under the antitrust laws. It is a requirement that an antitrust plaintiff must prove that his damages were caused by the unlawful acts of the defendant. See 15 U.S.C. § 15 (1980). This is the essence of “antitrust injury” as set forth by the Supreme Court:
Plaintiffs must prove antitrust injury, which is to say injury of the type the antitrust laws were intended to prevent and that flows from that which makes defendants’ acts unlawful.
Brunswick Corp. v. Pueblo Bowl-O-Mat, Inc., 429 U.S. 477, 489, 97 S.Ct. 690, 697, 50 L.Ed.2d 701 (1977) (emphasis in original).
Once causation of damages has been established, the amount of damages may be determined by a just and reasonable estimate as long as the jury verdict is not the product of speculation or guess work. J. Truett Payne Co. v. Chrysler Motor Corp., 451 U.S. 557, 566-67, 101 S.Ct. 1923, 1929, 68 L.Ed.2d 442 (1981); Zenith Radio Corp. v. Hazeltine Research Inc., 395 U.S. 100, 123-24, 89 S.Ct. 1562, 1576-1577, 23 L.Ed.2d 129 (1969). Since the Supreme Court has been willing to accept a degree of uncertainty in the calculation of damages, strict proof of what damages have been caused by which acts has not been required. Continental Ore Co. v. Union Carbide & Carbon Corp., 370 U.S. 690, 699, 82 S.Ct. 1404, 1410, 8 L.Ed.2d 777 (1962); Story Parchment Co. v. Paterson Parchment Paper Co., 282 U.S. 555, 51 S.Ct. 248, 75 L.Ed. 544 (1931); Locklin v. Day-Glo Color Corp., 429 F.2d 873 (7th Cir.1970), cert. denied, 400 U.S. 1020, 91 S.Ct. 582, 584, 27 L.Ed.2d 632 (1971). Not requiring strict disaggregation of damages among the various unlawful acts of the defendant serves to prevent a defendant from profiting from his own wrongdoing and makes sense when damages arise from a series of unlawful acts intertwined with one another. See Bigelow v. RKO Radio Pictures, Inc., 327 U.S. 251, 264-65, 66 S.Ct. 574, 579-580, 90 L.Ed. 652 (1946); City of Mishawaka v. American Electric Power Co., 616 F.2d 976 (7th Cir. 1980), cert. denied, 449 U.S. 1096, 101 S.Ct. 892, 66 L.Ed.2d 824 (1981).
It is essential, however, that damages reflect only the losses directly attributable to unlawful competition. The Supreme Court in Bigelow emphasized this point by allowing a more lenient standard for calculating the amount of damages only upon “proof of defendant’s wrongful acts and their tending to injure plaintiffs’ business, and from evidence in the decline of prices, profits and values, not shown to be attributable to other causes.” 327 U.S. at 264, 66 S.Ct. at 579.
The courts have always distinguished between proof' of causation of damages and proof of the amount of damages. Thus, the courts have been consistent in requiring plaintiffs to prove in a reasonable manner the link between the injury suffered and the illegal practices of the defendant. The First Circuit affirmed a directed verdict for the defendant on this ground in Momand v. Universal Film Exchanges, Inc., 172 F.2d 37 (1st Cir.1948), cert. denied, 336 U.S. 967, 69 S.Ct. 939, 93 L.Ed. 1118 (1949). The plaintiff in Momand had lost on summary judgment in a previous suit on eighteen of the twenty counts in his complaint. Once the court determined that this prior action was res judicata it held that the plaintiff was required to prove that the two remaining acts were the cause of injury. Id. at 43.
The Third Circuit has also overturned a jury verdict and ordered a new trial when the evidence showed that the damage calculations were based in part on lawful competition by the defendant. Coleman Motor Co. v. Chrysler Corp., 525 F.2d 1338 (3d Cir.1975). In Coleman Motor, a former independent automobile dealer brought suit against Chrysler and its wholly-owned dealers alleging that Chrysler subsidized its *1162wholly-owned factory dealers at the expense of its independent dealers and otherwise discriminated against the independents. To establish damages the plaintiff offered expert testimony which purported to project annual net earnings based upon the volume of sales which could have been expected if no factory dealerships had been established. To arrive at a measure of lost profits the experts, as in the case before us, subtracted from this projected figure the plaintiffs actual net earnings before taxes. Id. at 1351-52.
The Court held that the failure of the sales projection to account for lawful competition from the factory dealers required reversal and a new trial. The Court stated:
The damage figures advanced by plaintiff’s experts may be substantially attributable to lawful competition. In the absence of any guidance in the record, we cannot permit a jury to speculate concerning the amount of losses resulting from unlawful, as opposed to lawful, competition.
Id. at 1353. See also Murphy Tugboat Co. v. Crowley, 658 F.2d 1256 (9th Cir.1981), cert. denied, 455 U.S. 1018, 102 S.Ct. 1713, 72 L.Ed.2d 135 (1982); Van Dyk Research Corp. v. Xerox Corp., 631 F.2d 251, 255 (3d Cir.1980), cert. denied, 452 U.S. 905, 101 S.Ct. 3029, 69 L.Ed.2d 405 (1981).
Coleman Motor was followed in a well-reasoned district court opinion in R.S.E., Inc. v. Pennsy Supply, Inc., 523 F.Supp. 954 (M.D.Pa.1981). The district court granted judgment n.o.v. for the defendants based on plaintiff’s failure to present sufficient damage evidence. The defendants argued that the plaintiff’s proof failed to distinguish between their lawful and unlawful acts. The plaintiff responded, in a manner almost identical to MCI’s position, that it need not disaggregate their damage nor prove exact losses. The court rejected the lesser standard of proof urged by the plaintiff holding:
[I]t is apparent that the arguments made by the defendants . .. attack the causation of damages, not amount of damages, and therefore require a higher measure of proof.
Id. at 964.
The court in deciding the issue of causation of damages stated:
Plaintiff did not address the effect lawful competition may have had upon its damage model, choosing instead to throw the ball back in defendant’s court by citing Story Parchment for the proposition that damage estimates need not be perfect.
Id. at 965. A lost profit study was then rejected because of its “failure to account for any lawful competition.” Id. at 966.
In a case very similar to the case before us, a California district court ordered summary judgment for the defendant in ILC Peripherals Leasing Corp. v. IBM Corp., 458 F.Supp. 423 (N.D.Cal.1978), aff’d per curiam sub nom. Memorex Corp. v. IBM Corp., 636 F.2d 1188 (9th Cir.1980), cert. denied, 452 U.S. 972, 101 S.Ct. 3126, 69 L.Ed.2d 983 (1981), because of the nature of the plaintiff’s proof of damages. After holding that IBM was not guilty of predatory pricing, the district court considered the plaintiff’s damage study which was structured in such a way that one could not separate injury caused by the pricing practices of IBM from injury from other practices. In ordering summary judgment for IBM the court stated:
The way Memorex structured its damage claim there was no basis in the record for the jury to determine what the effect on damages would be if it found one or none of the challenged acts lawful. Thus, if one of IBM’s acts was not a violation of the antitrust laws, much of the damage claim would become invalid.
458 F.Supp. at 434.
When a plaintiff improperly attributes all losses to a defendant’s illegal acts, despite the presence of significant other factors, the evidence does not permit a jury to make a reasonable and principled estimate of the amount of damage. This is precisely the type of “speculation or guesswork” not permitted for antitrust jury verdicts. Bigelow, 327 U.S. at 264, 66 S.Ct. at 579. To allow otherwise would force a de*1163fendant to pay treble damages for conduct that was determined to be entirely lawful. Momand, 172 F.2d at 43; ILC Peripherals Leasing Corp. v. IBM Corp.
There is nothing inconsistent between requiring proof that damages were caused by illegal acts and the rule that a plaintiff need not disaggregate damages among those acts found to be unlawful. In this case, the trial court granted summary judgment for AT & T on seven of the twenty-two counts in the complaint. In addition, the jury found for AT & T on five of the fifteen counts it considered. The jury found for MCI on two counts relating to Hi-Lo tariffs, two counts relating to tariffs filed with state agencies, and six counts relating to interconnection. In addition, this court has now determined that the jury’s findings for MCI on the pricing and pre-announcement of Hi-Lo as well as the finding related to denial of multipoint interconnections must be set aside.
MCI assumed in the preparation of its damage study that all twenty-two of AT & T’s acts charged were illegal. In fact, liability has now been established with respect to only seven of the twenty-two counts of alleged monopolization. MCI’s lost profits study does not establish any variation in the outcome depending on which acts of AT & T were held to be legal and which illegal. On the contrary, the study was prepared well in advance of trial on the assumption that all of AT & T’s actions constituted the willful maintenance of a monopoly. MCI’s brief states:
Section I of the lost profits study allowed a computation for each year between 1973 and 1984, of the difference in financial results between the MCI company subjected to AT & T’s exclusionary acts, and the company as it would have been if allowed to pursue its original business plan subject only to the normal business risks of a competitive marketplace.
Appellee’s Brief at 141 (emphasis supplied).
Even assuming that the instructions, as a whole, were sufficient, the jury was left with no way to adjust the amount of damages to reflect lawful competition from AT & T. This is contrary to MCI’s assertion on appeal that the jury had ample evidence from the lost profits study itself to make such an adjustment. The study provides detailed cost data on specific components of MCI’s operations but does not set forth any information that would permit the jury to adjust the damages in the event that AT & T were successful on any of the counts in the complaint. In particular, the study does not contain any information indicating how to adjust MCI’s projected revenues and profits to reflect a possible finding that Telpak and Hi-Lo were lawfully priced.
Pursuant to Rule 11 of this court and Rule 28(j) of the Federal Rules of Appellate Procedure, MCI has further cited to us this court’s recent decision in Spray-Rite Service Corp. v. Monsanto Co., 684 F.2d 1226 (7th Cir.1982), cert. granted, - U.S. -■, 103 S.Ct. 1249, 75 L.Ed.2d 479 (1983), as support for the proposition that it may recover damages for the lawful and unlawful acts of a defendant if disaggregation is impracticable. In Spray-Rite this court faced a factually peculiar situation in which the special interrogatory sent to the jury was phrased in the alternative so that it was unclear whether the jury had found liability on one act or all three of the acts which were alleged to have been part of the conspiracy which resulted in the termination of the plaintiff as a dealer. Id. at 1233. This court affirmed the jury verdict and damage award because it was able to conclude that there was substantial evidence to support the verdict on all three acts. Id. at 1242 n. 11. This situation is clearly distinguishable from the case at bar for several reasons. First, there is insufficient evidence to support the jury verdict on Hi-Lo predatory pricing. Second, Spray-Rite is also distinguishable since the jury in the present case expressly found Telpak to be lawful. Finally, unlike Spray-Rite, where the extent of the lawful conduct was insubstantial in relation to the damages and the burden of requiring strict proof of causation was relatively great, in the instant case *1164Telpak alone was apparently a major element of the injury giving rise to the damage award.
Thus, in Spray-Rite, the shipping policies, compensation policies and territorial restrictions, which were complained of, were of secondary importance to the major injury— the termination of dealers. In contrast, the lawful conduct in this case, the establishment of AT & T’s pricing policies, represents a central competitive reality of the telecommunications industry and became an important focus of this case. Thus, it would be unjust and contrary to the policies of the treble damage remedy to award MCI damages which may compensate it for the effects of such quantitatively significant lawful competition.
The predatory pricing allegations play such a significant role in MCI’s case that failure to establish either the Telpak or the Hi-Lo allegations mandates re-examination of any damages which may have been predicated on what has been held to be lawful pricing. Throughout MCI’s case-in-chief the alleged predatory pricing of both Tel-pak and Hi-Lo were emphasized as key anticompetitive practices used by AT & T to injure MCI’s business. Telpak’s pricing structure and supposed “free” circuits were alleged to be the major stumbling blocks to MCI in attracting and holding large volume customers. Hi-Lo was alleged to be the specific response adopted by AT & T to unlawfully compete against MCI along the long distance telephone routes MCI had initially chosen to enter.
Mr. Uhl, the principal author of the lost profit study, testified that the study’s purpose was to calculate the net profits MCI could have earned if “it had not been interfered with by AT & T.” Tr. 3199. When asked specifically what types of interference he meant, Mr. Uhl stated:
Well, there were a number of things that occurred ...
MCI was forced to compete against certain services offered by AT & T that it didn’t anticipate it would have to compete against. Two that came to my mind are Telpak service offered by AT & T which was priced at a rate which was too low. In fact, we had proven it to be not cost justified.
Secondly, further on the second point, it was caused to compete against other services such as Hi-Lo service, which I believe was priced in a manner to preclude MCI from serving small customers, where Telpak made it difficult for MCI to serve large customers. And even though Hi-Lo didn’t come into effect, I believe it was 1975, it was early 1973 when AT & T announced the Hi-Lo rates and at that moment in time, of course, MCI had some difficulty trying to market against an AT & T announced rate.
Tr. 3199-3200.
The assumptions concerning Telpak alone are so critical to MCI’s proof of damages that the damage award must be overturned. Telpak and the interconnection controversy were always the twin prongs of MCI’s case. See MCI’s opening statement, Tr. 178-80. Telpak was the single largest component of the private line telephone industry, comprising over half the market. Tr. 426. In discussing the effect of Telpak on MCI, Mr. McGowan, the chief executive of MCI, stated:
[I] certainly always was aware that it had a major impact, significant impact, on MCI because without it being in existence, there would be a significant benefit and with it, there was harm.
Tr. 705. Telpak had all the more significant impact on MCI because it was aimed at the very largest customers, whom MCI hoped to attract.
Since a major premise of the study, illegality of Telpak and Hi-Lo, was incorrect, the study must be rejected. This defect might have been cured if MCI had offered evidence on the adjustment of the damage award to reflect findings of non-predation, but no such evidence was offered.
C. The Flawed Assumptions of the Lost Profits Study
In addition to failing to prove that MCI’s damages were caused by the unlawful con*1165duct of AT & T, MCI’s lost profits study also fails to substantiate adequately the assumptions which provide the foundation for the study. The lost profits study is based on the assumptions that MCI would receive average revenues of $.85 a circuit mile per month; that MCI wouid only have to pay local distribution charges at the same rate as Western Union; that MCI would have in place a capacity of thirty-seven million circuit miles; and that MCI could have raised the $500 million necessary to finance this system. While the validity of all of these assumptions may be questioned, the basis for the average revenue assumption is so lacking in substance that the lost profits study must be rejected as inherently untrustworthy and lacking in foundation.118
Our conclusion that AT & T maintained lawful pricing policies casts grave doubt on MCI’s ability to earn $.85 per circuit mile in the face of legitimate price competition by AT & T and additional specialized common carriers. Mr. Uhl confirmed this point on cross-examination when he stated that “when viewing the 85 cents, there was an assumption that Telpak would not be in existence.” Tr. 3336. Once the assumption that Telpak was unlawful was eliminated the jury was left with no way to adjust the amount of damages to reflect this lawful price competition by AT & T. Similarly, the study provides no guidance to this court in adjusting the award to reflect a finding that Hi-Lo was also lawfully priced.
The main assumption that falls as a result of Telpak and Hi-Lo’s being determined to be lawful is the $.85 revenue assumption. The ability of MCI to earn $.85 per circuit mile over an extended time period is undermined by price competition by AT & T which, contrary to MCI’s assumption, may maintain or conceivably even lower its long distance rates for large volume business users. In addition to failing to address lawful competition from AT & T, the lost profits study does not analyze price competition from other specialized common carriers but merely assumes that this does not pose a problem to MCI in achieving the $.85 revenue target.
The most damaging piece of evidence concerning MCI’s $.85 revenue assumption is the fact that Telpak is priced considerably lower than $.85, at a Telpak price which the jury found to be lawful. If MCI had wanted to compete with Telpak it would presumably have had to lower its prices well below the assumed $.85 rate. MCI argues that this price comparison is inaccurate because it prices its services on a different basis than AT & T. MCI contends that the price stated for Telpak does not include service terminal charges which are already included in MCI’s $.85 assumed revenue. MCI further contends that Telpak is billed by AT & T in such a manner that the mileage for billing purposes is nineteen percent greater than would be the case for the identical circuit owned by MCI. Finally, MCI points to the fact that Telpak customers only use seventy-five percent of their circuits thus producing a seventy-five percent “fill factor” and thereby raising the pro rata cost of the circuits used. MCI adjusts the cost of Telpak for the different billing systems, the “fill factor” and the addition of a pro rata service terminal charge to produce a calculation showing that the “true cost” of Telpak as a whole is $.93. MCI relies on this calculation to show that its projected $.85 service would, using comparable numbers, be priced lower than the competing Telpak, costing $.93. Thus MCI says that its $.85 revenue assumption is competitive and, hence, reasonable.
MCI’s argument fails for a number of reasons. MCI’s claim that the average Tel-pak user utilizes only seventy-five percent of its circuits is seriously misleading because seventy-five percent is not a weighted average reflecting the high percentage of the Telpak business represented by the *1166United States government and the very largest private corporations. For example, the fill factor for the United States government, the largest Telpak customer which alone accounts for over half the Telpak circuit miles, has always been at least ninety-five percent. The evidence also shows that the fill factor for Telpak D as a whole increased to ninety percent by 1977. Hence, we think MCI has not demonstrated that the total price of Telpak exceeds $.85 per circuit mile, or that MCI’s original revenue assumption of $.85 was reasonable.119
Further, MCI’s damage witness, Mr. Uhl, conceded on cross-examination that the $.85 revenue assumption included revenue from Execunet as well as FX, CCSA and other private line services. Execunet is a fully switched MCI service which was not approved until 1977, when the District of Columbia Circuit reversed the FCC and legitimated MCI’s offering of Execunet to its customers. See MCI Telecommunications Corp. v. FCC, 561 F.2d 365 (D.C.Cir.1977), cert. denied, 434 U.S. 1040, 98 S.Ct. 781, 54 L.Ed.2d 790 (1978). Execunet has ostensibly been a profitable business for MCI while FX, CCSA and other private line services might not have been. At least, the evidence before the jury was (in the face of Telpak) insufficient to establish the profitability of these latter services. The jury thus had no rational basis for concluding that, without Execunet, MCI could have earned $.85 per circuit mile during the period up to 1977.120
In addition MCI failed to explain how it derived its $.85 revenue assumption. Purportedly, the $.85 figure came from adjustments made to the original business plan for MCI by Mr. Uhl. The other source for the $.85 figure was an MCI consultant, Dr. Norman Lerner, but he did not testify at the trial. MCI’s business plan was never introduced at trial. Mr. Uhl, who did testify at trial, had expertise solely in the financial and accounting techniques necessary to prepare a lost profits study once the basic assumptions on the revenue side had been made. Mr. Uhl did not attempt to offer any fundamental justification for the original assumptions contained in the business plan. Thus, the $.85 figure represents nothing more than an adjustment made to essentially unsupported data. MCI did not produce at trial any other witnesses or documentary evidence which sufficiently demonstrated how the revenue assumption in the original business plan was derived. Thus while not necessarily hearsay, the lost profits study and the assumptions it contains lack a foundation from which a jury could reasonably have determined the damages which were found in the instant case. Cf. Locklin v. Day-Glo Color Corp., 429 F.2d 873, 879 (7th Cir.1970), cert. denied, 400 U.S. 1020, 91 S.Ct. 582, 584, 27 L.Ed.2d 632 (1971) (assumptions in damage evidence must rest upon an adequate base).
D. Remand for a Partial New Trial
These defects in the proof of damages require that the jury verdict be set aside. This does not, however, necessarily mandate a new trial on all issues. Rule 42(b) of the Federal Rules of Civil Procedure permits the separate trial of any issue when separation would be “in furtherance of convenience or to avoid prejudice, or when separate trials will be conducive to expedition and economy.” Fed.R.Civ.P. 42(b). Only one of these conditions need be met for the court to order a Rule 42(b) separate trial. United States v. IBM Corp., 60 F.R.D. 654 (S.D.N.Y.1973). A new trial *1167on liability is unwarranted since this court has affirmed liability for monopolization on the basis of most of AT & T’s actions involving interconnection with MCI.121 No new trial is required on those issues where we have set aside the findings at trial because our conclusion in these areas was based on a lack of evidence or other legal deficiency and not mere trial error. See Woods Exploration & Producing Co. v. Aluminum Co. of America, 509 F.2d 784 (5th Cir.), cert. denied, 423 U.S. 833, 96 S.Ct. 59, 46 L.Ed.2d 52 (1975). A new trial on damages is required to reflect the determinations by the jury, and by this court, that AT & T’s pricing policies were not predatory. The fact that evidence of damage frequently changes depending upon the findings of liability is one reason that split trials are permitted by the Federal Rules and endorsed by the Manual for Complex Litigation and the rules of the Northern District of Illinois. Manual for Complex Litigation § 4.12 (1981); Rule 21 of the United States District Court for the Northern District of Illinois. See generally 5 J. Moore, J. Lucas, & J. Wicker, Moore’s Federal Practice ¶ 42.03 (2d ed. 1982). The bifurcation of trials has been approved as an effective method of simplifying factual presentation, reducing costs, and saving time. Schwartz, Severance — A Means of Minimizing the Role of Burden and Expense in Determining the Outcome of Litigation, 20 Vand.L. Rev. 1197 (1967); Zeisel and Callahan, Split Trials and Time Saving: A Statistical Analysis, 76 Harv.L.Rev. 1606 (1963); Miner, Court Congestion: A New Approach, 45 A.B.A.J. 1265 (1959); Note, Original Separate Trials on Issues of Damages and Liability, 48 Va.L.Rev. 99 (1962); Note, Separate Trial of a Claim or Issue in Modem Pleading: Rule 42(b) of the Federal Rules of Civil Procedure, 39 Minn.L.Rev. 743 (1955).122
MCI’s proof of damages was quite distinct from its proof on the question of liability. In attempting to prove damages during trial MCI used different witnesses and proof of a somewhat different nature from its evidence on the question of liability. MCI’s principal damage witnesses, Mr. Laros and Mr. Uhl, testified at the close of the case-in-chief. The testimony of Mr. La-ros concerned the business forecast of MCI, which was incorporated into the lost profits study. Mr. Uhl’s testimony concerned only the data and the preparation of the lost profits study. MCI’s two other damage witnesses, Dr.* Hamada and Dr. Lorie, presented evidence only on MCI’s cost of capital and the adjustment of MCI’s damages to present value, while accounting for taxes. This proof of damages is sufficiently distinct from MCI’s proof of monopolization *1168to permit a separate and fair trial on remand to determine damages without prejudice to either party.123 Franklin Music Co. v. American Broadcasting Co., 616 F.2d 528 (3d Cir.1979); In Re Master Key Antitrust Litigation, 528 F.2d 5 (2d Cir.1975); In re Ampicillin Antitrust Litigation, 88 F.R.D. 174 (D.D.C.1980); LoCicero v. Humble Oil & Refining Co., 52 F.R.D. 28 (E.D.La.1971); Fischer & Porter Co. v. Sheffield Corp., 31 F.R.D. 534 (D.Del.1962).
In order to be effective, separate trials of liability and damages in antitrust cases must be grounded upon a clear understanding between the court and the parties of the issues and proof involved in each phase of the trial. The most difficult part of the decision to remand for a partial new trial on damages is the formulation of rules to guide such a proceeding. It is critical to realize what issues have not been remanded. The issues relating to jurisdiction, liability and immunity have been conclusively decided by this opinion and are not subject to further proceedings on remand. To the extent it is necessary to educate the fact finder on these issues, evidence which might normally be associated with a determination of liability may have to be introduced or reintroduced. We suggest, however, that stipulations be heavily relied upon by the parties in accordance with the sound discretion of the trial judge.
The issue to be re-tried on remand is solely the amount of damages that MCI is entitled to receive for those acts by AT & T which have been found to be unlawful. Any new lost profits study introduced on remand must be supported by an adequate foundation for all assumptions critical to the calculation of damages. MCI damage evidence must also make provision for impairments or losses stemming from competition that has been found to be lawful — Tel-pak pricing, Hi-Lo pricing, the pre-an-nouncement of Hi-Lo, the denial of multi-point interconnections and other acts or practices held to be lawful in the district court. The level at which the competition may price its product or service would appear to be an important determinant of profitability. It seems difficult to determine how much MCI lost from the alleged slow-down of its growth without knowing something about the lawful price environment in which it and its competition could reasonably expect to operate. As we have already indicated MCI need not disaggre-gate its proof of damages among individual unlawful acts which have caused financial loss, but MCI must be able to rationally separate these damages from those losses (or reductions in profit) which are caused by the purely lawful competitive actions of AT & T.124
*1169Given the massive discovery which has already been taken in this case little further discovery should be necessary on remand, but this is a matter for the trial court’s discretion. A presumption should also exist that previous rulings on documents and evidence offered at the last trial are not rear-guable, but the trial judge, of course, has broad discretion to determine, or redetermine, these issues.125
VII. THE CONDUCT OF THE TRIAL
AT & T argues that the manner in which the district court presided over the case amounted to a denial of due process. AT & T complains of the court’s failure to set forth the issues for trial in its pretrial order, its imposition of a time limit upon the parties’ presentations and its reversal of several evidentiary rulings during trial. AT & T in addition contends that the jury instructions erroneously excluded certain theories of its defense.
A. Fair Trial
We find that the district court could properly have refused to enter a pretrial order in this case. Federal Rule of Civil Procedure 16 gives the trial court discretion to hold pretrial conferences. Fed.R. Civ.P. 16.126 Although the rule is phrased in mandatory language (“the court shall *1170make an order”) the unique facts of each case should determine the necessity for a pretrial order. United States v. IBM Corp., 68 F.R.D. 358 (S.D.N.Y.1975). As one commentator has stated: “The failure of the court to enter an order after a pretrial conference does not render the ensuing trial a nullity if, at the pretrial conference, the opposing attorneys were in such disagreement that the conference was completely unproductive, and it would be an undue burden on the court if it were forced to forge a pretrial order.” 3 J. Moore, Moore’s Federal Practice ¶ 16.18 (1982).
Under such circumstances, a requirement that a court catalogue all points of dispute and agreement would create an unnecessary burden and cause undue delay. See, e.g., Video Components, Inc. v. Laird Telemedia, Inc., 574 F.2d 1061, 1062 (10th Cir.1978). In the instant case the district court would have been forced to conduct a trial within a trial as the parties argued over how the court should frame each issue. Thus, a pretrial order is not mandatory under Rule 16 when it is clear that the pretrial conference, as in this case, failed to produce an agreed statement and definition of issues. See IBM, 68 F.R.D. at 359.
The district court here encouraged the parties to narrow the issues whenever possible.127 Judge Grady acknowledged the failure of the parties to do so when AT & T pressed for a pretrial order.128 There was no absolute requirement for, and in fact, apparently little benefit in, such an order in this situation. Nonetheless, we stress the important potential benefits of such a document and the need to fully explore its feasibility before rejecting it. In the instant case, the court urged the parties to stipulate and agree to the authenticity and admissibility of evidence. Compliance with the district court’s directions would have obviated the need for any formal statement of the issues.129
We also reject AT & T’s argument that the district court did not allow AT & T sufficient time to present its case in an intelligible manner. Originally, AT & T predicted that it would take approximately eighteen months to try the case. Understandably chagrined, the district court directed the parties to submit lists of their witnesses and a summary of the testimony of each, together with a more precise estimate of the time required for trial. MCI’s list named seventeen witnesses and predicted that it would require twenty-six days to present its case-in-chief. AT & T’s list, by contrast, named 162 witnesses and described a minimum of twenty-one more by category. At that time, AT & T predicted that trial of the entire case would take eight to nine months. The district court reviewed those materials and only then imposed a twenty-six day time limit on the presenta*1171tion of each side’s case-m-chief.130 The district court did not place a limit on the time allotted for rebuttal or surrebuttal. MCI Communications Corp. v. AT & T, 85 F.R.D. 28, 32 (N.D.Ill.1979). On appeal, AT & T argues that the limits which were imposed were wholly arbitrary and amounted to a denial of due process. We cannot agree.
Litigants are not entitled to burden the court with an unending stream of cumulative evidence. United States v. Article of Drug Consisting of 572 Boxes, More or Less, 415 F.2d 390, 392 (5th Cir.1969); Manbeck v. Ostrowski, 384 F.2d 970, 973 (D.C.Cir.1967), cert. denied, 390 U.S. 966, 88 S.Ct. 1077, 19 L.Ed.2d 1170 (1968). As Wig-more remarked, “it has never been supposed that a party has an absolute right to force upon an unwilling tribunal an unending and superfluous mass of testimony limited only by his own judgment and whim.... The rule should merely declare the trial court empowered to enforce a limit when in its discretion the situation justifies this.” 6 Wigmore, Evidence § 1907 (Chadbourne Rev.1976). Accordingly, Federal Rule of Evidence 403 provides that evidence, although relevant, may be excluded when its probative value is outweighed by such factors as its cumulative nature, or the “undue delay” and “waste of time” it may cause. Whether the evidence will be excluded is a matter within the district court’s sound discretion and will not be reversed absent a clear showing of abuse. Hamling v. United States, 418 U.S. 87, 127, 94 S.Ct. 2887, 2912, 41 L.Ed.2d 590 (1974); Chapman v. Kleindienst, 507 F.2d 1246, 1251 n. 7 (7th Cir. 1974).
The time limits ordered by Judge Grady had the effect of excluding cumulative testimony, although in setting those limits the district court apparently fixed a period of time for the trial as a whole. This approach is not, per se, an abuse of discretion. This exercise of discretion may be appropriate in protracted litigation provided that witnesses are not excluded on the basis of mere numbers. See Padovani v. Bruchhausen, 293 F.2d 546, 549-50 (2d Cir. 1961).131 Moreover, where the proffered testimony is presented to the court in the form of a general summary, the time limits should be sufficiently flexible to accommodate adjustment if it appears during trial that the court’s initial assessment was too restrictive.132
*1172The limits set by the district court were not absolute. As Judge Grady stated in his order, “[t]hese limits are subject to change if events at the trial satisfy the court that any limit is unduly restrictive. It is my intention to allow each party sufficient time to present its case; I have no interest in speed for the sake of speed.” 85 F.R.D. at 31-32. Similarly, at a pretrial hearing the court told the parties that there was “nothing absolutely hard and fast” about the limits. After MCI completed presentation of its case in fifteen and one-half days, the court expressed an unwillingness to permit AT & T to exceed its twenty-six day limit, yet it later tempered this by reminding the parties, “I want to make it very clear that nobody is being pushed to do anything that is inconsistent with what he perceives to be the best interest of his client.”133 We cannot say that the district court was prepared to adhere strictly to its preliminary time limits without regard to possible prejudice to either party.
Insisting that the twenty-six day limit was too restrictive, AT & T cites SCM Corp. v. Xerox Corp., 77 F.R.D. 10 (D.Conn.1977), where the court imposed a six-month limit on the plaintiff’s presentation when the plaintiff had failed to make a prima facie showing of liability after fourteen weeks of trial. AT & T in effect suggests that whenever time limits are imposed in a complex case, the limits should involve months, not days. Although there may be some validity to this suggestion in most complex cases, we cannot say that it should necessarily control in the case before us. Obviously, there must be specific attention to the substance of the testimony and the complexity of the issues, but it does not follow that several weeks for each side will never suffice. The circumstances of each individual case must be weighed by the trial judge, who is in the best position to determine how long it may reasonably take to try the case. MCI was confident that it could establish liability in twenty-six days, and in fact finished eleven days ahead of schedule. We recognize, as did the district court, that presentation of a competent defense may require more time than presentation of a plaintiff’s case-in-chief. In light of the substance of AT & T’s proffered testimony, however, and the district court’s considered view that an efficient, yet effective, presentation of AT & T’s defense would take no longer than the time MCI used to present its case, we conclude that the district court did not manifestly abuse its discretion in limiting the time for AT & T’s case-in-chief.
AT & T also asserts that it was prejudiced when the district court reversed an earlier ruling and allowed AT & T to present evidence concerning the public interest, because that decision came too late for AT & T to include that evidence in its case-in-chief. Instead, AT & T had to present this evidence on surrebuttal. It is unclear precisely what harm AT & T believes it suffered. To the extent that the *1173harm it perceives derives from the jury’s hearing the testimony out of context, the district court offered to “make any concession to the defendant” in order to assure that AT & T’s evidence would be accorded its full impact. Moreover, as Judge Grady pointed out, “most of this evidence ha[d] been presented already” in one form or another during the testimony of AT & T witnesses. To the extent that AT & T believes it was harmed by the jury’s becoming disconcerted by being held for an even longer trial, we note that AT & T refused Judge Grady’s offer to allow AT & T’s evidence in before MCI finished its rebuttal case, and to explain the situation to the jury, including the fact that AT & T had wanted to put its evidence in earlier. Particularly since Judge Grady reversed his ruling on “public interest” evidence as soon as AT & T brought the only fully apposite authority for its position — the Mid-Texas case — to his attention, the circumstances surrounding the late reversal of the district court’s position, considered as a whole, do not demonstrate prejudice to AT & T’s substantial rights. The district court did not abuse its discretion. See Nanda v. Ford Motor Co., 509 F.2d 213, 223 (7th Cir.1974). See also Zenith Radio Corp. v. Hazeltine Research, Inc., 401 U.S. 321, 331, 91 S.Ct. 795, 802, 28 L.Ed.2d 77 (1971); Ditter v. Yellow Cab Co., 221 F.2d 894, 899 (7th Cir. 1955); Patton v. Roane-Anderson Co., 192 F.2d 965, 966 (6th Cir.1951).
AT & T suggests that the district court’s actions in managing the trial combined to deny AT & T a fair trial. The proper inquiry is whether it affirmatively appears from the record that some prejudice to the substantial rights of a party has resulted. McCandless v. United States, 298 U.S. 342, 347-A8, 56 S.Ct. 764, 766, 80 L.Ed. 1205 (1936); Citron v. Aro Corp., 377 F.2d 750, 753 (3d Cir.), cert. denied, 389 U.S. 973, 88 S.Ct. 473, 19 L.Ed.2d 466 (1967); United States v. Dressier, 112 F.2d 972, 978 (7th Cir.1940). We conclude that AT & T suffered no substantial prejudice from any of the instances of alleged error. “It is axiomatic that litigation parties are entitled only to a fair trial, not to a perfect one.” Ohio-Sealy Mattress Mfg. Co. v. Sealy, Inc., 585 F.2d 821, 843 (7th Cir.1978), cert. denied, 440 U.S. 930, 99 S.Ct. 1267, 59 L.Ed.2d 486 (1979). Considering the complexities of this ease and the near impossibility of coping with them adequately with traditional procedures, AT & T received a fair trial.
B. Theories of Defense
In its instructions to the jury, the district court refused to articulate AT & T’s position on each charge. Reasoning that since the plaintiff had the burden of proof and that AT & T’s basic theory was that “it did not do these things,” the district court concluded that the jury would be adequately instructed without “explaining in detail what the defendant’s position is.”
AT & T nevertheless argues that the refusal was prejudicial because it, in effect, deprived AT & T of an opportunity to present its defense theories to the jury. Citing Florists’ Nationwide Telephone Delivery Network v. Florists’ Telegraph Delivery Association, 371 F.2d 263 (7th Cir.), cert. denied, 387 U.S. 909, 87 S.Ct. 1691, 18 L.Ed.2d 627 (1967) (“FTD ”), AT&T argues that it was entitled to a specific instruction on each theory for which there was record evidence. The court in FTD found error in the trial court’s refusal to instruct on each defense theory, and AT & T points to that case as controlling. In FTD, however, it is clear that the reversal was prompted by an overall inadequacy in the instructions. If the instructions are too general to encompass the theory which a party espouses, reversal is mandated. This court in Beard v. Mitchell, 604 F.2d 485 (7th Cir.1979), recognized, on the other hand, that if the instructions were given “in substance,” there is no error. Id. at 497. Further, “even if the jury is only given a general instruction without being given the specific theory of the ease, the refusal will not be reversible if the party’s theory of the case was apparent throughout the course of the trial.” Id.
In addition, our inquiry must focus on whether as a whole the jury was adequately instructed, and not on whether every specif*1174ic aspect of AT & T’s defense was covered in detail. Id. at 498. So long as the instructions, “when viewed in the light of the evidence, show no tendency to confuse or mislead the jury,” there is no error. Allers v. Bohmker, 199 F.2d 790, 792 (7th Cir. 1952). FTD does not mandate reversal where the instructions as a whole encompass the theories of defense that are clearly incorporated in testimony and argument at trial.
Thus, a review of the record leads us to conclude that the instructions meet both these tests of adequacy. AT & T cannot be said to have so failed in making its position known during trial that the refusal to instruct left the jury with no understanding of AT & T’s defense. In these circumstances a general instruction would be acceptable. The district court also communicated the substance of AT & T’s theories. We agree that much of AT & T’s defense amounts merely to negating MCI’s contentions. AT & T, for example, asserts that it was entitled to an instruction that it believed the relevant geographic market was not nationwide but limited to the areas where MCI was authorized to do business. The court in fact charged, “The geographic scope of the relevant market is the area in which the firms involved do business and the area from which they draw customers.... MCI claims that ... the geographic scope of the market was nationwide.” App. 1198. The court further instructed the jury on what factors to consider in determining whether the market was, in fact, nationwide. These instructions in context lead us to believe that the jury understood AT & T’s view. Considering the trial evidence and argument as well as the instructions tendered, the jury’s obvious choice was between a nationwide market (espoused by MCI) or a more limited market (advocated by AT & T). Similar examples may be found throughout the instructions. Essentially we find no error in the instructions since by eliminating redundancies and excess verbiage, the district court avoided potential confusion while preserving the substance of the theories of defense offered by AT & T.134 See Southern Railway v. Shealey, 382 F.2d 752, 755 (5th Cir.1967).
VIII. CONCLUSION
As indicated in this opinion, we conclude that the jury’s award of damages and certain jury findings on the merits lack eviden-tiary support or are otherwise improper as a matter of law, so that they must be set aside. In addition, we approve the other jury findings on the merits. Accordingly, the judgment is reversed insofar as it is based on the award of damages and disapproved findings. In all other respects it is affirmed, and the cause is remanded for a new trial on the issue of damages only, in accordance with this opinion. Circuit Rule 18 shall not apply on remand. Each party shall bear its own costs on appeal. All arguments raised in this appeal which are not expressly addressed herein have been considered and determined to be without merit.
. The author of this opinion recognizes his debt and expresses his appreciation to Judge Wood, whose draft of this opinion plowed important ground and formed a basis for what has become the majority opinion. Although Judge Wood and the author disagree on some of the points at issue here, we agree fully about the Herculean joint efforts required to produce such a “weighty” finished product. Further, we both recognize the important contributions of Judge Fairchild to this product.
. Section 4 of the Clayton Act, 15 U.S.C. § 15 (1976), provides as follows:
Any person who shall be injured in his business or property by reason of anything forbidden in the antitrust laws may sue therefor in any district court of the United States in the district in which the defendant resides or is found or has an agent, without respect to the amount in controversy, and shall recover threefold the damages by him sustained, and the cost of suit, including a reasonable attorney’s fee.
. Section 2 of the Sherman Act, 15 U.S.C. § 2 (1976), provides in relevant part: “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 felo-
. AT & T also filed a counterclaim against MCI alleging that MCI attempted and conspired to monopolize the relevant market and actually monopolized the St. Louis-Chicago segment, conspired to restrain trade in the relevant market, and wrongfully acquired stock or share capital of other corporations, which substantially lessened competition. The district court did not permit any of these allegations to go to the jury, and AT & T does not challenge the propriety of that action on appeal.
. The district court directed a verdict in favor of AT & T on the following seven allegations: (1) inducing Western Union to file a tariff which mirrored the St. Louis-Chicago charges of‘MCI; (2) increasing AT & T’s capacity to conduct business in data communications for the purpose of destroying competition; (3) introducing experimental service to discourage potential customers from dealing with MCI; (4) *1093disparaging MCI; (5) bringing sham proceedings before certain administrative and judicial bodies; (6) participating in a massive public propaganda campaign conducted against MCI; and (7) refusing to provide Joint Telpak (a special tariff) to MCI. MCI does not challenge the propriety of the district court’s directed verdict on any of these issues.
. The Special Verdict is reprinted in the Appendix. See infra, p. 1207.
. Because we hold that Judge Grady did not err in refusing to admit evidence alleging destruction of documents by AT & T, we do not reach a motion filed by AT & T to strike a portion of MCI’s reply brief on this issue.
. Local exchange telephone service is the ordinary service provided in nearly all homes and businesses. From a technical standpoint, it involves a wire connection between the telephone set and a switching machine in a nearby telephone company central office which is connected by transmission trunks to the switching machines in other central offices within the exchange area. When the telephone is taken off the hook — or in the case of multiple telephones behind a private branch exchange, when a designated access code such as “9” is dialed — a signal is sent to the central office. The switching machine in the central office responds to this signal by sending a dial tone that enables the calling party to dial any telephone connected to the switched network within that exchange area. See DX 1828.
. Long distance service operates in a manner similar to local exchange service but typically involves a two-step process in which the user first gains access to the local switching machine through a dial tone and then requests access to the long distance toll switching machine by dialing an area code plus the number of the telephone the calling party wishes to reach. If a circuit is available to handle the call, it is routed through the calling party’s central office to a toll office nearby, over an intercity circuit to a toll office in the city that is being called, and finally through the centred office that serves the called telephone to that telephone. Wade, Tr. 3903-04; Marshall, Tr. 4893; DX 1828, 1833.
. From a technical standpoint, FX and CCSA services are similar to local exchange service in that they provide a connection into a switching machine in a telephone company central office which responds to requests for network access by sending a dial tone. The distinguishing aspect of FX service is that the switching machine to which the telephone is connected is not located in the nearby telephone company central office but is in a distant office, as, for example, where a telephone located in Chicago is connected to a switching machine in New York City. Such an arrangement permits the user to make and receive calls in the distant city as though they were local calls, i.e., as if the subscriber had a local telephone in the distant city. For this reason, FX service is frequently used by such businesses as airlines and hotel reservation agents.
CCSA service offers a similar advantage to large subscribers who wish to link far flung branches or offices to each other via private telephone lines connected through switches in the local telephone company office. In essence, CCSA service allows a large subscriber to obtain a personal mini version of the nationwide telephone network. The FTS line that connects federal government offices is one example of a CCSA-type service.
. In this opinion, Microwave Communications, Inc., as well as its successor corporations, are collectively referred to as MCI.
. The general service carriers argued that the entry of specialized common carriers into the telecommunications industry would be contrary to the public interest because telecommunications services could be provided more economically by a single supplier; because additional microwave systems would be duplicative and wasteful; and because specialized carriers without general service responsibilities would “cream-skim” the existing averaged rate structure by selectively competing only along the most profitable long distance routes, thus imposing a heavier rate burden on low density and local telephone users.
. Indeed, the district judge in this case characterized the decision as an “abomination” and “one of the worst examples of legal draftsmanship I have ever seen.” Tr. 3785.
. Circuit miles measure the total distance covered by all lines leased by customers in a given month. McGowan, Tr. 355-56.
. The dispute over local distribution areas related to the geographic boundaries within which AT & T was obligated to provide local facilities to MCI. See infra, pp. 1145-1147.
. Multipoint service involves a situation in which a customer has an AT & T private line between Cities A and B, and an MCI private line between Cities B and C. MCI claimed that it was entitled to an interconnection between its terminal and the AT & T terminal in City B so that the customer could obtain direct, MCI-provided service between City "A and City C. Revenues for the City A to City B segment would, of course, redound to AT & T. See infra, pp. 1147-1150.
. Before its 1974 clarification, the FCC had taken the position, in a brief filed in the Ninth Circuit, that the Specialized Common Carriers decision did not permit the offering of “switched” service (such as FX and CCSA service). This apparent contradiction in the FCC’s position was fully presented at trial.
. For a detailed description of Telpak and how it works, see American Trucking Ass’ns v. FCC, 377 F.2d 121, 124-27 (D.C.Cir.1966), cert. denied, 386 U.S. 943, 87 S.Ct. 973, 17 L.Ed.2d 874 (1967).
. AT & T, Revisions of Tariff F.C.C. No. 260 Private Line Services, Series 5000 (TELPAK), 64 F.C.C.2d 971, 983-89 (1977). This holding was affirmed by the Court of Appeals for the District of Columbia Circuit after the completion of the trial in the instant case. Aeronauti*1099cal Radio, Inc. v. FCC, 642 F.2d 1221, 1223 (D.C.Cir.1980), cert. denied, 451 U.S. 920, 101 S.Ct. 1998, 68 L.Ed.2d 311 (1981).
. AT & T had initiated studies to consider a deaveraged rate structure as early as 1970, shortly after the FCC’s approval of MCI’s Chicago-St. Louis line. The initial proposal resulting from those studies was a so-called “exception tariff’ which would have matched MCI’s rate over the Chicago-St. Louis route as soon as MCI commenced operations. Warned by its economic advisors that such an exception tariff could be perceived as violative of the antitrust laws, AT & T decided against this approach and opted, instead, for the development of a broadly deaveraged national rate structure. deButts, Tr. 4038-39.
. The revenue projections for the “undamaged” MCI were adjusted downward to account for some delays and uncertainties that MCI officials felt were not fairly attributable to AT & T. See Uhl, Tr. 3228-37; PX 1203.
. MCI assumed that the effects of AT & T’s allegedly unlawful conduct would be completely eliminated as of 1995.
. MCI countered this contention by arguing, inter alia, that AT & T’s Telpak rate computations did not include applicable termination charges and that they failed to account for the percentage of base capacity actually used by a customer (the “fill factor”). See infra, pp. 1165-1166.
. Mann-Elkins Act of 1910, ch. 309, § 7, 36 Stat. 539, 544 (1910).
. See Mann-Elkins Act of 1910, ch. 309, §§ 7, 12, 36 Stat. 539, 544, 551 (1910).
. Clayton Act, ch. 323, § 7, 38 Stat. 731 (1914) (current version at 15 U.S.C. § 18 (1976)).
. Compare Mann-Elkins Act of 1910, ch. 309, §§ 7, 12, 36 Stat. 539, 544, 551 (1910) with Communications Act of 1934, ch. 652, §§ 201, 202, 43 Stat. 1064, 1070 (1934) (current version at 47 U.S.C. §§ 201(b), 202(a) (1976)).
. We acknowledge the brief on this issue of the United States as amicus curiae.
. But see Essential Communications, 610 F.2d at 1120 (Communications Act intended for the benefit of customers, not competitors).
. But we believe that FCC regulation of AT & T’s rates may be more pervasive than Federal Power Commission (now Federal Energy Regulation Commission) regulation of the wholesale rates of electric utilities. Cf. Otter Tail Power Co. v. United States, 410 U.S. 366, 93 S.Ct. 1022, 35 L.Ed.2d 359 (1973).
. The FCC maintains, however, that when it has prescribed or specifically approved a tariff, its judgment must control. See United States v. AT & T, 461 F.Supp. at 1327 n. 39; cf. Jeffrey v. Southwestern Bell Tel. Co., 518 F.2d 1129 (5th Cir.1975) (state action immunity from antitrust laws granted where challenged rate had been approved by municipality after thorough hearings).
. See, e.g., AT & T, Charges, Regulations, Classifications and Practices For Voice Grade/Private Line Service (High Density—Low Density), 55 F.C.C.2d 224, 244 (1975) (Interim Decision); 58 F.C.C.2d 362, 364, 370 (1976) (Final Decision) (finding Hi-Lo tariff “unlawful” because AT & T had not submitted sufficient cost data to justify the tariff); AT & T, Revisions of Tariff FCC No. 260 Private Line Services, Series 5000 (TELPAK), 61 F.C.C.2d 587, 651-62 (1976) (overall rate levels for AT & T’s private line telephone service unlawful because not set in accordance with fully distributed cost methodology), aff’d in part, rev’d in part sub nom. Aeronautical Radio, Inc. v. FCC, 642 F.2d 1221 (D.C.Cir.1980), cert. denied, 451 U.S. 920, 101 S.Ct. 1998, 68 L.Ed.2d 311 (1981).
. AT & T’s claim that its tariff filings with state commissions are immune from antitrust liability under the Noerr-Pennington doctrine is analyzed separately infra, at pp. 1153-1158.
. See, e.g., Travelers Insurance Co. v. Blue Cross, 361 F.Supp. 774, 780 (W.D.Pa.1972), aff’d, 481 F.2d 80 (3d Cir.), cert. denied, 414 U.S. 1093, 94 S.Ct. 724, 38 L.Ed.2d 550 (1973) (company was not a monopoly since it lacked control over rate-making mechanism); Nankin Hospital v. Michigan Hospital Service, 361 F.Supp. 1199, 1209-10 & n. 33 (E.D.Mich.1973) (company did not possess monopoly power since rates were controlled and actively reviewed by state insurance commission). Cf. International Railways of Central America v. United Brands Co., 532 F.2d 231, 240 (2d Cir.), cert. denied, 429 U.S. 835, 97 S.Ct. 101, 50 L.Ed.2d 100 (1976) (consent decree which fixed freight rates removed ability of banana grower to coerce lower freight rates from railroad and thus negated finding of monopoly power).
. Although many cases make reference to Alcoa ’s innovative presumption, in the more than three decades since that case was decided, courts have consistently found monopolization only in circumstances where predatory or exclusionary conduct was proven. Watson & Brunner, supra at 590 n. 83; see Hanover Shoe, Inc. v. United Shoe Machine Corp., 392 U.S. 481, 485-86, 88 S.Ct. 2224, 2227, 20 L.Ed.2d 1231 (1968); United States v. United Shoe Machinery Corp., 110 F.Supp. 295, 343-44 (D.Mass.1953), aff'd per curiam, 347 U.S. 521, 74 S.Ct. 699, 98 L.Ed. 910 (1954). In most successfully prosecuted section 2 monopolization cases, predatory or exclusionary practices held to constitute a violation of section 1 of the Sherman Act have also been present. See, e.g., United States v. Grinnell, 384 U.S. 563, 86 S.Ct. 1698, 16 L.Ed.2d 778 (1966); United States v. Griffith, 334 U.S. 100, 106-07, 68 S.Ct. 941, 945, 92 L.Ed. 1236 (1948); American Tobacco Co. v. United States, 328 U.S. 781, 66 S.Ct. 1125, 90 L.Ed. 1575 (1946); United States v. Reading Co., 253 U.S. 26, 40 S.Ct. 425, 64 L.Ed. 760 (1920); cf. Berkey Photo Inc. v. Eastman Kodak Co., 603 F.2d 263, 273-76 (2d Cir.1979), cert. denied, 444 U.S. 1093, 100 S.Ct. 1061, 62 L.Ed.2d 783 (1980) (integrated monopolist’s failure to predisclose innovations and its ability to sell monopolized and competitive products as a system are not unlawful uses of monopoly power, but legitimate advantages of size and integration); Telex Corp. v. IBM, 510 F.2d 894, 927-28 (10th Cir.), cert. dismissed, 423 U.S. 802, 96 S.Ct. 8, 46 L.Ed.2d 244 (1975) (reversing finding of section 2 liability in the absence of predatory conduct).
. We reject MCI’s suggestion that our holding in Mishawaka is limited to the unique situation of an alleged price squeeze by an electric utility subject to both federal and state rate regulation.
. Moreover, the trial court assured the jury that
the Sherman Act allows a regulated firm with monopoly power to compete vigorously whenever it may be faced with competition. So long as the defendant’s competitive responses to plaintiffs were based on legitimate business decisions and on the merits of defendant’s services, defendant cannot be found to have unlawfully maintained a monopoly. This is so even if plaintiffs were hurt by the competition and defendant sought to retain as much of its business as possible.
App. 1199-1200.
. See Northern Pacific Ry. Co. v. United States. 356 U.S. 1, 4, 78 S.Ct. 514, 517, 2 L.Ed.2d 545 (1958).
. In the course of instructing the jury, Judge Grady noted that the term “average costs” had also been referred to at trial as “fully distributed costs” or “embedded costs,” and the term “marginal costs” had also been used interchangeably with “incremental” or “long-run incremental costs.” Tr. 11485. This arguably imprecise terminology, see infra, was repeated in the special verdict. App. 1207-1209.
. We do not mean to imply that direct probative evidence of a defendant’s intent is inadmissible. See infra, note 59.
. Early antitrust cases could define predatory pricing only in vague verbal formulations relating to predatory intent and ruinous competition. See Moore v. Mead’s Fine Bread Co., 348 U.S. 115, 118, 75 S.Ct. 148, 149, 99 L.Ed. 145 (1954); Forster Mfg. Co. v. FTC, 335 F.2d 47, 52-53 (1st Cir.1964), cert. denied, 380 U.S. 906, 85 S.Ct. 887, 13 L.Ed.2d 794 (1965); Porto Rico Am. Tobacco Co. v. American Tobacco Co., 30 F.2d 234, 236 (2d Cir.), cert. denied, 279 U.S. 858, 49 S.Ct. 353, 73 L.Ed. 999 (1929). Such vague definitions resulted in erratic application of the law as well as lengthy and complex inferences 'of intent, which were of little predictive or precedential value. See Brodley & Hay, Predatory Pricing: Competing Economic Theories and the Evaluation of Legal Standards, 66 Cornell L.Rev. 738, 765-67 (1981) [hereinafter cited as Brodley & Hay, Predatory Pricing]; Areeda & Turner, Predatory Pricing at 699.
. See, e.g., 3 P. Areeda & D. Turner, supra, at ¶711-15 (1978); R. Posner, supra, at 189; Areeda & Turner, Predatory Pricing at 709-13; Joskow & Klevorick, A Framework for Analyzing Predatory Pricing Policy, 89 Yale L.J. 213 (1979).
. It should also be noted that AT & T is a regulated public utility. A dominant purpose of public utility regulation is to deny AT & T profits attributable to its monopoly power. It would be anomalous indeed in this context to require AT & T, as a matter of antitrust law, to maximize its profits.
. See Scherer, Predatory Pricing and the Sherman Act: A Comment, 89 Harv.L.Rev. 868 (1976); Areeda & Turner, Scherer on Predatory Pricing: A Reply, 89 Harv.L.Rev. 891 (1976); Scherer, Some Last Words on Predatory Pricing, 89 Harv.L.Rev. 901 (1976); Williamson, Predatory Pricing: A Strategic and Welfare Analysis, 87 Yale L.J. 284 (1977); Areeda & Turner, Williamson on Predatory Pricing, 87 Yale L.J. 1337 (1978); Williamson, A Preliminary Response, 87 Yale L.J. 1353 (1978); Williamson, Williamson on Predatory Pricing II, 88 Yale L.J. 1183 (1979); Greer, A Critique of Areeda and Turner’s Standard for Predatory Practices, 24 Antitrust Bull. 233 (1979); Koller, When is Pricing Predatory?, 24 Antitrust Bull. 283 (1979). See also R. Posner, supra, at 184-196.
. Professor William Baumol has defined long-run incremental costs of product X as “total company cost minus what the total cost of the company would be in the absence of production of X, all divided by the quantity of X being produced.” Baumol, Quasi-Permanence of Price Reductions: A Policy for Prevention of Predatory Pricing, 89 Yale L.J. 1, 9 n. 26 (1979).
. “Average total cost” as we use it here refers to average total economic cost, as the term is employed by economists in predatory pricing analysis. See generally Brodley & Hay, Predatory Pricing. The term “average total costs” is also sometimes used in utility ratemaking to refer to the revenue required to meet all the accounting costs of an entire utility enterprise. See infra, note 52.
. No objection has been made in the instant case to the introduction of fully distributed cost evidence. Thus, nothing in this opinion should be construed as reflecting on the admissibility of fully distributed cost evidence. Under some circumstances — for example, the operation of a single product enterprise in a stable economy— average balance sheet costs may provide an acceptable proxy for LRIC. See R. Posner, supra, at 190. The use of FDC, as a proxy, in various contexts, is always open to examination.
. Thus, various FDC methods will normally produce quite different calculations of the cost of a product or service. In one electric utility rate case one witness testified to the existence of at least 29 different methods of apportioning costs among services. J. Bonbright, Principles of Public Utility Rates 351 (1961). The FCC itself has required AT & T to submit at least seven different FDC cost studies. In the instant case, cost studies using these seven different FDC methods were introduced at trial. Not surprisingly, each method yielded a significantly different cost profile.
A simple example helps to highlight the arbitrariness of FDC methodology. Imagine a railroad line that simultaneously transports three different products: gold, lead and feathers. If the railroad attempted to calculate, on a fully distributed cost basis, the cost'of shipping each of these products, it would reach radically different results depending on whether it allocated joint and common costs on the basis of the value, weight, or bulk of the respective commodities shipped.
. See generally, 1 A. Kahn, supra, at 150; J. Bonbright, supra; Aeronautical Radio, Inc. v. FCC, 642 F.2d 1221, 1236-47 (D.C.Cir.1980) (Wilkey, J., dissenting), cert. denied, 451 U.S. 920, 101 S.Ct. 1998, 68 L.Ed.2d 311 (1981).
. This decision path illustrates that LRIC analysis is not, as the dissent suggests, a solely theoretical view of the question, but is, in fact, a meaningful representation of the implicit issues considered in the making of business decisions in the real world.
. MCI has cited no economic authority beyond the testimony of its own expert, Dr. Melody, which supports fully distributed cost as a measure of average total cost. Similarly, MCI has cited no economic authority supporting the use of fully distributed cost in an antitrust context. But cf. Noll & Rivlin, Regulating Prices in Competitive Markets, 82 Yale L.J. 1426 (1973) (use of incremental cost methods in setting regulated prices may invite predatory pricing). See generally Melody, Comment, in New Dimensions in Public Utility Pricing 205 (H. Trebing ed. 1976) (discussing the problems of marginal cost information in regulatory contexts); Melody, The Marginal Utility of Marginal Analysis in Public Policy Formulation, 8 J.Econ. Issues 287 (1974).
. Unfortunately the terms “average total cost” or “average total costs” have been used ambiguously. As we use “average total cost” in this opinion, we mean average total economic cost, i.e. costs on a forward-looking basis. See generally 1 A. Kahn, supra, at 73-74, 130-33. For a particular product or service of a multiproduct business “average total cost” is defined differentially as the average long-run incremental cost of the product or service in question. The term “average total costs” has also been used in quite a different sense in a utility ratemaking setting to refer to the revenues required to meet all the accounting, historical or embedded costs of the entire utility enterprise. See Bonbright, supra, at 300. This latter usage of “average total costs” is not applicable here since we are concerned with the economic cost of a particular service of a multiservice business. Such a cost must be one which is caused by the particular service in question. Cf., Brodley & Hay, Predatory Pricing at 780-86; Cudahy and Malko, Electric Peak-Load Pricing: Madison Gas and Beyond, 1976 Wis.L.Rev. 47, 62.
. In Aeronautical Radio, the District of Columbia Circuit held that the FCC did not act arbitrarily and capriciously in adopting a form of fully distributed cost methodology for purposes of evaluating proposed telephone rates. Judge Wilkey dissented, arguing that FDC was irrational, arbitrary and capricious even as applied to regulatory ratemaking.
. Judge Wilkey goes on to draw the parallel between this example and AT & T’s pricing decisions:
The parallel with the AT & T situation is obvious. The cost for monopoly services is fixed, and will be the same whether the competitive services are added by AT & T or not. The cost for the monopoly services is the equivalent of the $125 which would have been the hotel charge for Judge X; the cost for the competitive services would be the *1119additional $25 which would be added if Mrs. X enjoyed the use of the same services. From the hotel’s point of view, the additional cost for Mrs. X’s presence is only the additional linens and food, and therefore the incremental cost for her presence is a relatively small amount compared to the basic cost of providing that one room and facilities for one person. (The hotel might calculate that the presence of Mrs. X would generate sales in the shops on the hotel premises and thus add a small amount to hotel revenues, and thus the hotel could encourage her presence by charging even less than the cost of the linens and food and still make a profit on the incremental services. Similar comparisons might be made to AT & T services.)
To apply fully distributed costs to the stay of the couple at the hotel is economic nonsense; it is unquestionably true that considered ab initio the cost of providing the room and food for the two persons can be divided equally, $75 apiece, but this bears no relation to the economic logic of the way to conduct a hotel business or to conduct a moot court board’s business either. What the moot court board was faced with from the start was paying the total expenses of Judge X, which amounted to $125 at the price charged by the hotel. Similarly, the price charged by the hotel for Judge X individually was what it cost to put one person in the room and provide meals and all services, with a reasonable profit.
Both the hotel and the moot court board should logically and sensibly run their business on an incremental cost basis, just as is advocated by AT & T and the Antitrust Division in our case. The basic cost of having Judge X come to the moot court is going to be $125 a day whether Mrs. X comes or not. The board is logically forced to pay this price. The advent of Mrs. X is something entirely within the control of Judge X; she can come or not, and if she does come, there is no moral or economic right of the moot court board to profit by the incremental service the hotel is providing for Mrs. X by reducing the cost allocated to Judge X’s presence, which logically still remains at $125 a day and is not either economically or equitably reducible to $75.
Similarly, the charges which the monopoly customers of AT & T have been paying and which have been previously determined as fair, based on the costs of AT & T in providing these services, should not necessarily be reduced simply because AT & T can inaugurate other services in the competitive market. The costs properly allocated to the new competitive services of AT & T are incremental costs, not fully distributed costs, because the costs of the monopoly services should remain the same irrespective of whether AT & T enters competitive markets or not. (Actually, as shown by the detailed economic exposition above, there is hope that the competitive services of AT & T, with incremental cost pricing, will definitely contribute to a lowering of costs for the monopoly service customers.)
642 F.2d at 1245-47 n. 52.
. We regard this case as being tried under a stipulation that a form of average total cost would be used to determine predation. Thus, we analyze whether LRIC or FDC is the most meaningful economic measure of average total cost. This analysis should not be construed as a rejection of the Areeda-Turner rule using average variable costs. We affirm our previous holding in Chillicothe that pricing below average variable cost is normally one of the most relevant indications of predatory pricing.
. The trial court in Hommel, relying on the Supreme Court’s decision in Utah Pie Co. v. Continental Baking Co., 386 U.S. 685, 87 S.Ct. 1326, 18 L.Ed.2d 406 (1967), had allowed the jury to hear evidence of defendant’s pricing below total cost. See 472 F.Supp. 793, 795-97 (W.D.Pa.1979). On this evidence the jury found for the plaintiff on a Robinson-Patman violation, but not on a section 2 Sherman Act violation. Hence, on appeal, the Third Circuit faced only the issue of the appropriate pricing standard in price discrimination cases. Without ruling on the question whether the standards for predatory pricing were the same for both price discrimination and monopolization, the Third Circuit reversed the district court and entered judgment for the defendant, in part because of the absence of any evidence that prices were below average variable cost. 659 F.2d at 350-53.
. An important, but presently theoretical, issue not directly before this court is the propriety of using short-run marginal cost (as opposed to some measure of average total cost) in predatory pricing cases involving industries with high entry barriers. Several courts have suggested that exceptions to the Areeda-Turner rules may be appropriate where entry barriers are high — one of the circumstances in which true predatory pricing is more likely to occur. See, e.g., International Air Industries v. American Excelsior Co., 517 F.2d at 724; Hanson v. Shell Oil Co., 541 F.2d 1352, 1358 n. 5 (9th Cir.1976), cert. denied, 429 U.S. 1074, 97 S.Ct. 813, 50 L.Ed.2d 792 (1977); cf. Northeastern Telephone, 651 F.2d at 89 (barriers to entry into business telephone equipment market “relatively low”).
There is some evidence that barriers to entry may be high in the long distance telecommunications field. There is also evidence, however, that the development of microwave technology has significantly lowered those barriers. See Note, Recent Federal Actions Affecting Long Distance Telecommunications: A Survey of Issues Affecting the Microwave Specialized Common Carriers Industry, 43 Geo.Wash.L.Rev. 878, 894 (1975). Because both parties here have argued for measures of average total cost, we do not reach this question except to note that one of the principal barriers to entry in the telecommunications industry is the need for *1122FCC permission to enter the field. Since the FCC has extensive powers to open up the telecommunications industry to new competition and to “fine tune” the permissible competitive prices by regulation, any such barriers to entry may not be of overriding concern in an antitrust context.
. How practically to compute LRIC (including the possible use of proxies, where appropriate, cf. Brodley & Hay, Predatory Pricing at 780-86) *1123is a matter which we believe to be quite manageable and capable of development on an ongoing basis.
. We do not intend to imply that in all cases and in all circumstances we would only examine the price-cost relationship of a product or service. Our test merely suggests that a judge or jury may not infer predatory intent unless price is below long-run incremental cost. Thus, we agree, at least in principle, with Judge Wood’s advocacy of the use of non-economic (or less rigorous economic) evidence in some cases. Considering, however, among other things, the extent of regulatory control over entry and prices in the present case, and the highly ambiguous nature of the non-economic evidence which has been submitted, we think the price-cost relationship must be determinative. But, of course, some future case may admit more scope for “other factors.” Chilli-cothe, 615 F.2d at 432. In any event a strong presumption of lawfulness must attach when price is shown in a case like this one to be above an appropriately derived measure of long-run incremental cost. MCI has offered no credible direct evidence of intent that, in our view, directly rebuts this presumption.
. The implication of MCI’s theory would be that a multiservice firm must earn a rate of return for each service at least equal to its overall cost of capital for the firm. Such a requirement is illogical since a firm’s overall cost of capital is based on the level of risk in investing in the Grm and not in an individual service faced with particularized risks and competitive conditions. Also, to the extent all the services face competition, a demand that in the aggregate they earn the overall cost of capital suggests that, to the degree some services exceed this figure, others will fall short and thus arbitrarily appear “predatory.” In any event, rate of return calculations must be based on a host of arbitrary apportionments of plant and expenses.
. We recognize, of course, that under the consent decree approved in United States v. AT & T, 552 F.Supp. 131 (1982), Bell’s local operating companies will no longer be corporately linked .to AT & T’s long distance telephone service.
. Df course, quite apart from antitrust pricing standards in this regulated industry, the regulatory agencies can evaluate competitive prices by whatever standards they deem economically or socially desirable, including FDC. See Aeronautical Radio, Inc. v. FCC, 642 F.2d 1221, 1222 (D.C.Cir.1980), cert. denied, 451 U.S. 920, 101 S.Ct. 1998, 68 L.Ed.2d 311 (1981).
In this connection, although there may be some merit to the courts’ fashioning an antitrust rule of liability addressing limit pricing (or other “strategic” practices) as discussed by the dissent, administration might be difficult. See infra, pp. 1181-1182. Further, Prof. Baumol’s proposal of a quasi-permanent pricing approach (and other like proposals) may have promise if apparent problems of administration can be solved. See Baumol, Quasi-Permanence of Price Reductions: A Policy for Prevention of Predatory Pricing, 89 Yale L.J. 1 (1979). In any event, no evidence was presented based on such theories in this case.
. MCI belatedly argues that it was not required to prove that the Hi-D circuits were priced below cost but that Hi-Lo as a whole was priced below cost. This argument defies logic as well as MCI’s proof. The Hi-D circuits were the only portion of the long distance market in which AT & T lowered its price. The Lo-D and short haul portions of the market were subject to substantial price increases. Further, Dr. Melody’s testimony bolsters the notion that only the price cuts for the Hi-D circuits are relevant for purposes of determining predatory pricing. Dr. Melody stated that this decrease in the high density rates was AT & T’s competitive response to MCI and, therefore, the only relevant price to be examined. Tr. 2617.
. These were expressed in the studies as a ratio of net operating earnings to net investment.
. Dr. Melody also testified that the “revenue deficiency” shown in his chart was “greater after Hi-Lo went into effect.” Tr. 10477. Considered in the context of the other deficiencies in the evidence and the many variables involved, we do not consider this observation as evidence that Hi-Lo or Hi-D was “below cost.”
. It appears from our review of the record that Dr. Melody used the plant and expense allocations produced by AT & T’s FDC “Method 1” as applied to the various years. See supra, note 47, for a discussion of FDC methodologies. Thus AT & T’s “net operating earnings” under that method turns out for each of the years (with the possible exception of 1974) to be equal to the “revenue available” in Dr. Melody’s chart. Dr. Melody did not testify that he in fact followed “Method 1” or why he selected that method if in fact he did. Of course, neither Method 1 nor any other method was espoused, or had its probative value attested to, by Mr. Johnston, who merely stated that these FDC studies were required by the FCC.
. Dr. Melody attempted to address this problem by testifying that “Mr. Johnson (sic) [in his testimony] indicated that the other services were ... providing a profit.” Tr. 10477. What Mr. Johnston actually said was that the other services “were making a significant contribution to the earnings of the Bell System.” Tr. 6868. This statement might mean that the other services were contributing on an incremental cost basis or earning a positive rate of return on a fully distributed cost basis. The statement says nothing about whether the rate of return of the other services was greater or less than the overall cost of capital of the Bell System, which would seem to be the relevant consideration in terms of Dr. Melody’s chart.
. Referring to his exhibit (PX 3915) Dr. Melody said his “deficiency” calculations were not intended to be “exact or precise” and that he was looking for a “benchmark indicator.” Tr. 10476. MCI’s counsel referred to the numbers as “ballpark figures.” Tr. 10477.
. AT & T is correct in arguing that the jury instructions in this case failed to reflect this standard. Judge Grady instructed the jury that it could find unlawful pre-announcement if AT & T’s February 26, 1973 request for special permission constituted “the announcement of a price reduction by a firm with monopoly power a long time before it intends to put the reduction into effect,” App. at 1205, and if the announcement “was done not for legitimate reasons, but for the purpose of maintaining a monopoly.” Id. These instructions make no mention of deception or misleading conduct.
Because we conclude that there is insufficient evidence to support a finding of unlawful pre-announcement under the proper legal standard, we need not remand for a new trial on this issue.
. At one time, according to these memoranda, AT & T considered requesting an extension of the effective date of the tariff from the FCC. AT & T, however, in fact made no such request and, therefore, the memoranda have little significance.
. Judge Richey, in Southern Pacific Communications Co. v. AT & T, 556 F.Supp. 825 (D.D.C. 1982), analyzed the identical set of circumstances and attributed the entire delay to regulatory requirements. At 965-966.
. AT & T contends in its brief that the entire pre-announcement of Hi-Lo is immunized from antitrust scrutiny by the application of the Noerr-Pennington doctrine, discussed infra pp. 1153-1158. Inasmuch as we have concluded that the pre-announcement cannot be considered a violation of the antitrust laws we do not need to reach the question of immunity.
. In addition to the fact that the circuits have already been paid for, a Telpak customer wishing to utilize a previously unused circuit must pay additional terminal charges.
. In addition, the bulk marketing of Telpak made it comparable to the user-owned microwave systems against which it was designed to compete.
. The only difference is that AT & T may physically transmit the signal over whatever circuits are available when the call is placed, a matter which is irrelevant to the customer and without competitive significance.
. In this connection, it is somewhat unclear from MCI’s brief exactly what “fictional” characteristics of Telpak routing MCI complains of. Our interpretation of MCI’s argument differs slightly from that of Judge Richey of the District Court for the District of Columbia who considered a similar argument by a different specialized carrier in Southern Pacific Communications Co. v. AT & T, 556 F.Supp. 825 (D.D.C. 1982). Judge Richey characterized the complaint before him as alleging a situation where a customer having a Telpak bundle between New York City and Washington, D.C. could have calls from New York suburban offices to Washington, D.C. suburban points billed under the same Telpak circuit. Judge Richey upheld such an arrangement, stating that AT & T had done no more than recreate the same communications system that the customer would have in its own hypothetical private microwave system. To the extent MCI complains of the same or an analogous problem, we think the answer is the same as that provided by Judge Richey.
. A switched network is one in which a customer’s telephone is linked to one of the many local AT & T central offices located in each exchange area. When the caller picks up his or her telephone receiver, a switching machine sends a dial tone that permits the caller access to switching machines in other central offices in the exchange area. Long distance service involves this same process, except that upon dialing the proper area code and telephone number, the caller is shifted from the local to a long distance toll switching machine, which forwards the call. See supra, at notes 8 and 9.
. FX (“foreign exchange”) service involves the connection of a subscriber’s telephone to a switching machine in a distant, rather than a local, telephone company office. CCSA (“common control switching arrangement”) service is used by large subscribers to link far-flung offices to each other by private lines connected through switches in the local telephone company’s various offices. See supra, at note 10.
To provide both FX and CCSA, MCI would need “local loop” interconnections between an MCI terminal and the switch in a nearby AT & T central office. MCI would pick up the local switched call and send it long distance via MCI microwave relays to the customer called.
. A related argument made by AT & T is that, since the Third Circuit concluded that there was room for “legitimate dispute” over the meaning of the Specialized Common Carriers decision, MCI was collaterally estopped from denying that such a legitimate dispute existed, and therefore the district court should have directed a verdict in favor of AT & T on the FX/CCSA issue. Both AT & T and MCI were parties to the suit that gave rise to the Third Circuit’s decision which dealt with the interconnections at issue in the trial that spawned this appeal.
AT & T’s estoppel argument cannot prevail. The Third Circuit concluded that there was legitimate room for dispute over whether or not certain kinds of interconnection were ordered. The court did not say that the interconnections were not ordered, see Bell Tel. Co. v. FCC, 503 F.2d 1250, 1262-63 (3d Cir.1974), cert. denied, 422 U.S. 1026, 95 S.Ct. 2620, 45 L.Ed.2d 684 (1975), nor did it specifically find that AT & T believed or reasonably could have believed that denial of interconnection was justified under the Communications Act or the FCC decision. The court decided only that Specialized Common Carriers was sufficiently unclear to warrant deferral to the FCC under the doctrine of primary jurisdiction. There is no identity of issues, and thus the court’s “finding” does not bind MCI here. Of greater significance is the FCC’s resolution of the interconnection question against AT & T.
AT & T alternatively contends that in light of the Third Circuit case, Judge Grady improperly refused to give a proposed instruction that MCI’s interconnection denial charge arose out of a legitimate dispute. We note again that the Third Circuit stated that there was room for legitimate dispute, not that AT & T in fact relied on that interpretation as a basis for its interconnection denials. Judge Grady properly rejected the instruction, since it was for the jury to determine whether the dispute over the meaning of Specialized Common Carriers was legitimate. This is especially important here, where MCI’s case included evidence from which it could be inferred that AT & T denied interconnections regardless of the meaning of the Specialized Common Carriers decision, and thus, even though the language in that decision could have given rise to a legitimate dispute, in this instance it did not. AT & T’s tendered instruction was potentially misleading as to the significance of the Third Circuit decision.
. AT & T asserts that the FCC’s pre-Execunet opinion in the matter is due special deference because it embodies the view of the administrative agency responsible for enforcing the relevant statute. To begin with, the Execunet decision holds that the FCC’s view on this issue was inconsistent with the agency’s past practice, a situation in which the agency’s views are due no special deference. See United States v. Shreveport Grain & Elevator Co., 287 U.S. 77, 84, 53 S.Ct. 42, 44, 77 L.Ed. 175 (1932). See generally, W. Hurst, Statutes In Court 161 (1970). Moreover, Judge Grady permitted AT & T to argue that the FCC’s expressions of its incorrect views made AT & T’s reliance on them proper. This gave AT & T ample opportunity to lay out for the jury the historical context and all the facts and circumstances known to the parties at the time, consistent with the district court’s instructions.
. This is so notwithstanding AT & T’s argument that the court in Execunet recognized that other cases had not foreshadowed its decision. 561 F.2d at 377-78 n. 59. What the D.C. Circuit meant in this respect was that different arguments had been put before the courts that had previously dealt with the interconnection controversy. That the result was not foreshadowed does not undermine the court’s conclusion that the FCC departed from established doctrine when it limited MCI in the manner involved in that case. Indeed, as AT & T itself acknowledges, the Execunet court’s premise is that there is no limitation on a grant of service without a ruling to that effect by the FCC. The court’s holding that the FCC in Specialized Common Carriers did not make an affirmative determination that the public interest required such a limitation was clearly foreshadowed by earlier cases and is merely the latest application of established principles in this interconnection controversy.
. See, e.g., Chevron Oil Co. v. Huson, 404 U.S. 97, 92 S.Ct. 349, 30 L.Ed.2d 296 (1971) (earlier holding on applicability of state statute of limitations would eliminate present plaintiff’s right to institute suit, even though filed before earlier holding).
. The jury was instructed that the period of liability for AT & T’s acts extended between 1969 and mid-1975. The court told the jury at the outset of this portion of the instruction that the Execunet decision was rendered in 1977. Thus, AT & T’s contention that the district court failed to instruct on the relevant time frame of the case is contradicted by the record.
. With respect to the interconnection controversy, AT & T specifically requested an instruction explaining in great detail the pertinent regulatory statutes, but the district court rejected it and instead simply instructed that liability could not be found if AT & T had acted pursuant to a good faith interpretation of regulatory policy. In the interconnection controversy, the inquiry focused on whether AT & T in fact had acted on the basis of a perceived regulatory policy. Whether AT & T’s purported belief was correct was not at issue since the relevant policies actually required the interconnections. Particularly in that instance the district court’s more general instruction was sufficient.
. AT & T challenges the admissibility of one document containing such statements. That question is dealt with infra, at 1143.
. A similar inference of intent could be drawn from tariffs AT & T filed, which by their terms required AT & T to refuse FX and CCSA interconnections. When considered in the context of the comments noted above, the tariffs could well give rise to a legitimate inference by the jury that AT & T sought every avenue of obstruction to competition, while anticipating that its general denial of interconnections would prove unsupportable in the long-run. See infra, at pp. 1153-1158, the discussion of those tariffs.
. MCI undercut some of this testimony on cross-examination. One AT & T witness admitted that his view of the interconnection requirements of the Specialized Common Carriers decision was formed without his having read the opinion. Another witness could not articulate precisely what aspects of the decision gave rise to his view that the decision limited MCI to “point-to-point” service. Other AT & T witnesses were closely examined to expose their biases or lack of knowledge about the specifics of the decision. The jury was entitled to consider all this in according weight to the testimony of these witnesses and those whom MCI called to testify to their opposite conclusions about the meaning of the case.
. That portion of the opinion describes the options approved “with respect to local loop service,” 29 F.C.C.2d at 938, as including (a) provision of interconnection arrangements on “reasonable terms and conditions,” or (b) new carriers’ construction of “their own independent local facilities to provide end-to-end service” without the need for interconnection. Id. at 940. There is no limiting language about interconnections aside from the reference to “reasonable terms and conditions.”
. This is so even acknowledging the Third Circuit’s comment that the FCC’s 1971 opinion was not a model of clarity. The evidence was sufficient for the jury to infer that AT & T seized on whatever ambiguity existed in the opinion to mask the true anticompetitive animus that guided its decisions.
. Indeed, AT & T never objected on hearsay grounds to the use of the opinion, but rather on the basis that the order was irrelevant or else unreliable because it was purportedly obtained by MCI’s fraudulent representations to the FCC. In its Reply Brief, AT & T directs our attention to a portion of the transcript where it says it made an objection on hearsay grounds to the admission of the 1974 decision. We see nothing remotely resembling such an objection in that or any other relevant section of the transcript. Rather, the cited portion of the transcript contains an objection on relevancy grounds that “[t]he document has nothing to do with the interpretation of” the Specialized Common Carriers decision.
. AT & T claims that the opinion is not probative of the FCC’s state of mind, saying that only two of the five Commissioners sitting at the time joined in the opinion’s full reasoning. Actually three of the five formed the majority as to the relevant portions of the opinion. One of those three dissented from a single paragraph concerning an issue not raised here. AT & T’s citation of Assure Competitive Transp., Inc. v. United States, 629 F.2d 467 (7th Cir.1980), cert. denied, 449 U.S. 1124, 101 S.Ct. 941, 67 L.Ed.2d 110 (1981), is inapposite, since that case involved the validity of agency actions in the absence of a quorum. Here, by contrast, a majority of the FCC Commissioners sitting approved the decision in relevant part, and AT & T raises no issue about the validity of that action on the basis that a quorum was lacking. Cf. Marks v. United States, 430 U.S. 188, 97 S.Ct. 990, 51 L.Ed.2d 260 (1977) (where no majority, plurality opinion states the holding of a case).
. AT & T notes that the 1974 decision was submitted to the jury with the prejudicial portions highlighted. Since AT & T did not object to the highlighting after having an opportunity to inspect the exhibit prior to its submission, and itself highlighted portions of its own exhibits, we see no impropriety in the jury’s receiving the marked document.
. Similarly, the district court properly exercised its discretion in admitting the 1973 injunction issued by the United States District Court for the Eastern District of Pennsylvania, the Third Circuit’s decision upholding the FCC’s 1974 cease and desist order and the Supreme Court’s denial of certiorari to review that decision. All three were probative of the reasonableness of AT & T’s assertions of innocent motive.
. The federal case AT & T cites for a contrary rule is based on local rules of Texas courts, as applied in that diversity action. Lasiter v. Washington National Insurance Co., 412 F.2d 594 (5th Cir. 1969).
. Thus, AT & T’s argument that MCI could not have been harmed because it did not have facilities in service during part of the time these obstructive events occurred is not convincing. The harm was one that would have effect once MCI actually went into service, as well as at the time of AT & T’s actions.
. We express no opinion on the overall relationship between the doctrines of essential facilities and tying and do not mean to imply that *1145a tying violation will always flow from any denial of an essential facility by a dominant firm.
. Neither party questions the correctness of Judge Grady’s instruction that MCI was not entitled to interconnections for geographically unlimited local facilities. MCI contended only that the boundaries as drawn unreasonably restricted MCI’s ability to serve its customers. For example, MCI claimed it was unreasonable for AT & T to restrict the Washington, D.C. LDA in such a way as to preclude necessary interconnections extending to Rockville, Maryland, a densely populated and commercial suburb only 13 miles from the city. AT & T argues that there is no evidence that the LDA designations were unduly restrictive or constituted an improper attempt to keep MCI at a competitive disadvantage and to exclude it from the market. There was conflicting evidence as to whether the LDA’s were unreasonably restrictive. This question thus became appropriate for resolution by the jury, and in light of the testimony and arguments concerning the restrictiveness of the LDAs, Judge Grady adequately instructed the jury to determine the reasonableness of the LDAs.
The MCI witness whose testimony AT & T says supports the view that MCI sought geographically unlimited interconnections was referring only to his desire for MCI to receive reasonable interconnections for local areas. His statement that “[tjhere should have been no restriction on MCI’s ability to serve its customers” was linked directly to his expressions of concern over unreasonable limits on the LDAs. The witness went on carefully to distinguish between MCI’s requests for broader limits on local interconnections, and its desire for national “interexchange” facilities. On appeal, AT & T ignores the effective distinction made by the witness to its counsel — and the jury — at trial.
. AT & T’s argument that the district court failed to provide guidance to the jury on what constituted an “essential facility” is meritless. Judge Grady carefully instructed the jury on the elements of the essential facilities doctrine (including the need for the jury to find that MCI could not reasonably duplicate the facility) and specifically stated that MCI contended the facilities of AT & T’s local operating companies were “essential”, since without them, MCI could not provide service to its customers. Moreover, the district court instructed the jury that the essential facilities doctrine is applicable where “a business holds a monopoly of some essential facility that other businesses *1147need in order to compete .... ” Since the word “essential” is a term of ordinary meaning, and since the instruction explained that the facilities involved must be those that a firm needs in order to compete, the jury was given adequate guidance. Cf. Kocher v. Creston Transfer Co., 166 F.2d 680 (3d Cir.1948) (where enigmatic term used, jury may not be left to speculate on its meaning).
. AT & T’s Reply Brief on this issue cites an MCI memorandum detailing the construction of MCI facilities in South Chicago and Hammond, Indiana. The memorandum appears to refer to these facilities as part of the long distance portion of MCI’s planned service between Chicago and Cleveland, making them different in character from the local interconnections MCI sought in the LDA controversy.
. AT & T argues that the essential facilities concept is not applicable here because there has been no joint refusal to deal. Where a monopolist controls essential services, however, its refusal to allow potential competitors to use those services gives rise to antitrust liability where the purpose of the denial is to restrain competition, even if the monopolist is the only one that controls the facility. Otter Tail Power Co. v. United States, 410 U.S. 366, 93 S.Ct. 1022, 35 L.Ed.2d 359 (1973); Official Airline Guides, Inc. v. Federal Trade Commission, 630 F.2d 920, 927-28 (2d Cir.1980), cert. denied, 450 U.S. 917, 101 S.Ct. 1362, 67 L.Ed.2d 343 (1981). Otherwise, a monopolist would be able unreasonably to choke off all competition, yet escape sanctions simply because it was the only one in a position to do so. The thrust of Official Airline Guides and Almeda Mall, Inc. v. Houston Lighting & Power Co., 615 F.2d 343 (5th Cir.), cert. denied, 449 U.S. 870, 101 S.Ct. 208, 66 L.Ed.2d 90 (1980), is not, as AT & T contends, that a joint refusal to deal is necessary in order to establish the applicability of the essential facilities concept. Rather, the' point made in those cases is that in the absence of competition between a potential seller and a putative buyer, there is no room to apply the essential facilities doctrine. Here, in contrast, AT & T and MCI are in direct competition as providers of long distance service. The interconnections that MCI seeks from AT & T in the area beyond the restrictive LDAs are by definition a necessary component for MCI to reach its customers. The jury was entitled to find that absent a greater geographical range than AT & T was willing to concede, MCI would be placed at a severe competitive disadvantage. Since MCI showed that it had potential customers outside the restrictive LDAs, the jury had an ample basis to conclude that ÁT & T intended to undercut MCI’s competitive potential. Indeed, when AT & T provided the temporary interconnections pursuant to the 1973 injunction, several went to MCI customers beyond at least one LDA.
. It should be made clear that multipoint service, as requested by MCI, contemplated that AT & T would be entitled to all revenue generated by the use of AT & T lines.
. On appeal, MCI argues that the interconnections it sought for multipoint service were “purely local and do not involve provision of interstate facilities by AT & T to MCI.” Appel-lee’s Brief at 83. We find this argument to be somewhat disingenuous. While the interconnections in question involved, in a technical sense, physical facilities that were “local,” the purpose of these interconnections was to allow MCI to sell a long distance service package which included access to cities to which MCI had, as yet, not built its own facilities.
. AT & T’s theory on appeal is that the jury instruction for the multipoint claim conflicted with the instruction under the LDA claim. We find no conflict in these two instructions. The instruction on the LDA claim, when read in context, tracks the application of the essential facilities doctrine to the LDA controversy. The district court’s reference to “geographically unlimited local facilities” in that instruction applied only to the LDA claim.
. Instruction 32 reads, in full:
MCI claims that AT & T denied it interconnection for multipoint service. According to the evidence, multipoint interconnection involved a situation where a customer ordered an AT & T private line between City A and City B, and an MCI private line between City B and City C, and MCI sought an interconnection between its terminal and the AT & T terminal so that the customer could obtain service between City A and City C. To prevail upon its claim of denials of multipoint service interconnections, MCI must prove that AT & T unreasonably denied those interconnections with the intent of maintaining a monopoly in the relevant market rather than for legitimate business reasons.
App. 1203.
. Because MCI failed to request a jury instruction which would have allowed a finding of liability based upon this theory, MCI may not now claim as error or be entitled to a remand upon this inadequacy in the instruction actually given. See Fed.R.Civ.P. 51.
. AT & T, Offer of Facilities for Use by Other Common Carriers, 52 F.C.C.2d 727 (1975). Neither the agreement itself nor any of its terms was in evidence.
. We also reject AT & T’s argument that MCI’s injury from the inappropriate or inefficient interconnections was never specified at trial and so must be de minimis. For the reasons stated in our discussion of MCI’s damages, infra, p. 1168, MCI must be allowed to present its proof of damage without having to tightly compartmentalize its proof and without having to specify an exact dollar amount for each unlawful act of AT & T.
. According to the definition stipulated to by the parties, “clean interface” is a demarcation or method of connecting Bell services or facilities with non-Bell services or facilities in a manner which clearly delineates the end-to-end responsibilities of the carrier providing the services, obviating the necessity for joint installation, maintenance or testing.
. As the parties stipulated, signaling is the method of telling the telephone equipment that communication is being made or has stopped, such as making the telephone ring or stop ringing. E & M is one of three possible signaling systems.
. A connection in a 66-type block is formed by pushing wires down between small pins. There are two points of connection. In comparison, a 300-type connector is mounted on a frame, has a square pin and provides 32 points of connection.
. MCI’s witness on this issue, who testified from personal experience with the reporting procedures, did not join the company until 1973.
; The instruction reads as follows:
For plaintiffs to prevail on their claim that they were provided with inefficient or otherwise inappropriate equipment and procedures for interconnections, plaintiffs must establish that defendant knowingly furnished inefficient or inappropriate services or equipment with the intent of maintaining a monopoly in the relevant market. The basic controversy here concerns the equipment used by the Bell operating companies to interconnect with plaintiffs, including such things as connector blocks and equipment interfaces, the various kinds of signaling used by the Bell operating companies, the configuration of certain interconnections, such as those for Central Office Centrex Service, the provision of engineering information, and the proce*1153dures for coordination, installation, testing, and repairs.
App. 1203.
. AT & T also asserts that the instruction permitted the jury to find against AT & T on the theory that it was under an obligation to make its best and most up-to-date equipment and procedures available to MCI. Quoting from Berkey Photo, Inc. v. Eastman Kodak Co., 603 F.2d 263 (2d Cir.1979), cert. denied, 444 U.S. 1093, 100 S.Ct. 1061, 62 L.Ed.2d 783 (1980), AT & T notes that “if a firm that has engaged in the risks and expense of research and development was required in all circumstances to share with its rival the benefits of these endeavors, this incentive on which the proper functioning of our competitive economy rests would very likely be vitiated.” Id. at 281. We find no hint of the theory to which AT & T objects in the instruction that mentioned only efficient and appropriate equipment and procedures, not the “most” efficient or appropriate ones. The only theory contained in the instruction is that MCI was entitled to equipment and procedures that did not act as a hindrance. The cases AT & T relies on for the contrary conclusion are inapposite. MCI did not ask AT & T for advance knowledge of its new products, unlike the plaintiffs in Berkey Photo; nor did MCI suggest that AT & T abandon a technical innovation in favor of an already existing alternative as in Memorex Corp. v. IBM Corp., 636 F.2d 1188 (9th Cir.1980), cert. denied, 452 U.S. 972, 101 S.Ct. 3126, 69 L.Ed.2d 983 (1981).
. This discussion of the soundness of the theoretical underpinning for allowing antitrust claims based upon a single sham claim or lawsuit may be unnecessary under the facts of this case. AT & T filed the same tariff for interconnection charges with 49 individual state commissions. This may be regarded as a pattern of baseless claims within the language of California Motor Transport. Judge Grady in denying AT & T’s motion to dismiss regarded AT & T’s filings as constituting just such a series of “baseless, repetitive claims.” MCI Communications Corp. v. AT & T, 462 F.Supp. 1072, 1103 (N.D.Ill.1978).
We also note what is at least a theoretical question: whether Noerr-Pennington would apply to each and every filing with state regulators. Since Noerr-Pennington is designed to protect the right to petition the government to take some action, Noerr-Pennington might not apply if a tariff filing is only a pro forma publication perhaps required by law and not an exercise of the right to petition the government. The record does not indicate whether each state regarded the filing of the tariff as an application for approval or merely as a notification (formal or otherwise). We do not reach this issue given our conclusion that these filings, even if “petitions” to which Noerr-Pen-nington applied, were, in any event, “sham.”
. At .the time Bell filed its interconnection tariffs with the state utility commissions, no court had determined that the FCC had exclusive jurisdiction over the tariffs. The tariffs, of course, covered interconnections between interstate long distance carriers (regulated by the FCC) and local telephone operating companies (regulated by state agencies). Filings, such as those involved here, could presumably have been made with state commissions for informational purposes, or merely on request of a commission, whether or not the FCC had exclusive jurisdiction.
It was not until 1977 that a court directly addressed the question: what agency had jurisdiction over FX and CCSA interconnection. At that time the District of Columbia Circuit held that the FCC had exclusive jurisdiction over such interconnections. California v. FCC, 567 F.2d 84 (D.C.Cir.1977), cert. denied, 434 U.S. 1010, 98 S.Ct. 721, 54 L.Ed.2d 753 (1978).
. Bad faith is defined as:
The opposite of “good faith,” generally implying or involving actual or constructive fraud, or a design to mislead or deceive another, or a neglect or refusal to fulfill some duty or some contractual obligation, not prompted by an honest mistake as to one’s rights or duties, but by some interested or sinister motive.
Black’s Law Dictionary (4th ed. 1968).
. The business plan itself was not introduced at trial.
. Because of the deficiency of MCI’s revenue assumption we do not consider the validity of the assumptions concerning the size or financing of MCI’s system. With regard to MCl’s assumption that it had a right to receive local interconnections from AT & T at the same rate as Western Union, it must be noted that the jury rejected this claim and that, in fact, MCI has paid a higher charge than it assumed in its lost profits study.
. We also note that since the service terminal charge is a flat fee, the pro rata service charge depends on the length of the circuit. MCI makes its pricing calculation on the basis of a 306 mile circuit. When a 500 mile circuit is chosen for comparison the pro rata service termination charge is reduced and, according to MCI’s own figures, Telpak’s price drops to $.82, which is below MCI’s assumed average revenue rate.
. Although we recognize that MCI was the victim of certain unlawful acts of AT & T it should be noted that MCI never in fact earned anything close to $.85 per circuit mile prior to the introduction of Execunet. In 1975 MCI had never earned more than $.63 per circuit mile. While AT & T’s unlawful actions undoubtedly had some effect on MCI, it appears unlikely that the impact accounts for realized revenues so far below the projected revenue figure.
. In its petition for rehearing, AT & T raised an argument that the intent evidence presented on the predatory pricing claims, which we reject, was inextricably interrelated with the intent evidence applicable to the findings on the interconnection claims which we approve. And, as a result, AT & T contends that the jury’s finding on predatory pricing “necessarily tainted” its findings on the interconnection claims; hence those claims should be remanded for a new trial on liability. But we think this argument relies on unfounded speculation about the jury’s mental processes. On the face of things, because reductions in price, unlike the denial of interconnections, may be pro-competitive as well as anticompetitive, we think the “intent” relevant to pricing is not necessarily closely linked to the “intent” relevant to refusals of interconnection. Further, the jury itself exonerated AT & T on the important Telpak pricing claim while finding monopolistic intent in the denials of interconnection. Particularly in the context of this unusual case, we believe we should balance the possibility of spillover among claims against the prejudice to the prevailing party in the overturning of the jury’s findings. The jury’s determination that Hi-Lo was predatory did not so clearly taint its ability to decide the interconnection claims that justice would be served by overturning the jury verdicts on these interconnection claims.
. Some have argued that the bifurcation of trials in civil antitrust cases must be approached with trepidation. Response of Carolina, Inc. v. Leasco Response, Inc., 537 F.2d 1307 (5th Cir. 1976). In a private treble damage action impact or “fact of damage” is a necessary element of proving liability. However, “fact of damage” and “amount of damage” are distinct concepts. As such, separate trial of these issues is beyond constitutional challenge even when the partial new trial is to be heard by a separate jury. Gasoline Products Co. v. Champlin Refining Co., 283 U.S. 494, 51 S.Ct. 513, 75 L.Ed. 1188 (1931).
. AT & T argued in its petition for rehearing that the jury’s answers to special verdict questions numbered 5(a), (c), (f), (h), (j), and (k) do not define the conduct found unlawful by the first jury with sufficient clarity to separate the issue of damages from the issue of liability so as to permit a new trial limited to the question of damages alone. Gasoline Products Co. v. Champlin Refining Co., 283 U.S. 494, 500 (1931). As indicated, we believe that the district court and counsel can, on the basis of the record and in accordance with the procedures prescribed in this opinion, sufficiently identify the conduct found unlawful by the first jury to separate the issues relevant to damages for trial.
. We briefly note an additional issue which relates to the calculation of damages. At trial, MCI argued that the damages it suffered must be adjusted upward to reflect the taxes which the company would pay on the award. In its original presentation to the jury MCI asked for over $900 million so that, after taxes, it would receive approximately $452 million, its projected lost profits.
Tax issues received very scant attention in the district court, yet result in an approximate doubling of the recoverable damages if MCI is correct in its treatment of tax effects. It is not possible for this court to review the tax aspects of this case on the record before us. Rather than require additional briefs on an issue not fully dealt with below, we simply note the existence of the issue and leave its resolution to the proceeding on remand. In directing the district court’s attention to these issues, we particularly note MCI’s contention that the relevant tax rate was 49.78% as well as the provisions of 26 U.S.C. § 186 (1980), which suggest that a portion of antitrust damages may be subject to an offsetting deduction if they have not been previously deducted as unrecovered losses.
. With these guidelines in mind, we note that the district court might wish to appoint a special master under the terms of Rule 53 of the Federal Rules of Civil Procedure to aid the district court in the resolution of this case. We leave this entirely to the discretion of the district court, with the further knowledge that references to a special master are to be the exception rather than the rule, and that the problem here may argue as much for as against such a master. However, the language of Rule 53 permits reference to a master in a jury trial where “the issues are complicated” or in a non-jury case involving “difficult computation of damages.” Fed.R.Civ.P. 53(b). Reference is particularly appropriate to resolve issues “so technical or esoteric as to be outside ordinary judicial competence.” Comment, Masters and Magistrates in the Federal Courts, 88 Harv.L. Rev. 779, 795 (1975). Reference on the limited issue of the computation of lost profits seems to be within the standards announced for the use of special masters. See La Buy v. Howes Leather Co., 352 U.S. 249, 77 S.Ct. 309, 1 L.Ed.2d 290 (1957). See generally Kaufman, Masters in Federal Court: Rule 53, 58 Colum.L. Rev. 452 (1958); Note, Reference of the Big Case Under Federal Rule 53(B): A New Meaning for the “Exceptional Condition” Standard, 65 Yale L.J. 1057, 1065 (1956). The question of appointing a special master is, of course, wholly within the district court’s discretion.
If the district court chooses to use a master, the reference to the master should be strictly limited to the calculation of lost profits. The use of an impartial and knowledgeable master might be helpful in determining the validity of assumptions used in lost profits studies as well as the accuracy and reliability of the costs and revenues used in the calculation of damages. Any evidence of actual damages can be presented directly to the fact finder at trial since such evidence does not present the sort of extraordinary complexity suggesting reference.
The computation of damages in an antitrust case where liability has already been established is the type of issue that has often been referred to a master. District courts within this circuit have utilized this procedure in previous antitrust suits where liability has been established. See, e.g., Locklin v. Day-Glo Color Corp., 429 F.2d 873 (7th Cir.1970), cert. denied, 400 U.S. 1020, 91 S.Ct. 582, 584, 27 L.Ed.2d 632 (1971). In Locklin a special master was used to calculate lost profits following the district court’s judgment on the issue of liability in an antitrust suit. See also Arthur Murray, Inc. v. Oliver, 364 F.2d 28 (8th Cir.1966) (affirming reference for analysis of books and records in connection with antitrust award); Connecticut Importing Co. v. Frankfort Distilleries, Inc., 42 F.Supp. 225 (D.Conn.1940) (special master appointed to calculate lost profits relating to an unlawful boycott of plaintiff’s business); Eastern Fireproofing Co. v. United States Gypsum Co., 50 F.R.D. 140 (D.Mass.1970). Again, we emphasize that any reference is within the discretion of the district court.
. Federal Rule of Civil Procedure 16 provides in relevant part:
In any action, the court may in its discretion direct the attorneys for the parties to appear before it for a conference to consider
(1) The simplification of the issues;
The court shall make an order which recites the action taken at the conference, the amendments allowed to the pleadings, and the agreements made by the parties as to any of the matters considered and which limits the issues for trial to those not disposed of by admissions or agreements of counsel; and such order when entered controls the subsequent course of the action, unless modified at the trial to prevent manifest injustice.
. THE COURT: What I have tried to do here by asking you to stipulate and agree is to accomplish the substance of a pretrial order. Those pretrial orders, where you list the things you disagree about — well, that is nonsense. I have seen those things that go on for pages.
MR. HANLEY: You try the case on paper once and then you have to come back and try again.
THE COURT: If there is anything in there that we have not covered that could be covered, we ought to do it. But right offhand, I cannot think of it, and I am reluctant to impose upon anybody unnecessary paper work at this point.
MR. SAUNDERS: I agree with that.
THE COURT: We do not need it either. Let’s forget the pretrial order but go over it — I have not seen one in a long time. I have forgotten some of the parts of it, but if you see something in there that you think would be helpful, you can get together with each other.
Tr. 29-30.
. The following exchange took place:
MR. SAUNDERS: Judge are we going to do anything with that pretrial order that I gave to you?
THE COURT: I really think there is enough in it that would be controversial that it is not worth trying to reach an agreement on. The answer I think is no.
. We emphasize the value of a pretrial order, in a proper case, in facilitating the orderly conduct of a lengthy and complex trial. We suggest, subject of course to Judge Grady’s exercise of discretion, that the full procedures of Fed.R.Civ.P. 16 be utilized on remand.
. Referring to the list submitted by AT & T, the district court noted:
It is almost an understatement to say that defendants’ approach to this case is grandiose. Without intending to pass upon any evidence questions at this time, and recognizing that I asked the parties to be very brief in describing the proposed subject matter of the testimony, it does appear to me that much of defendants’ proposed testimony would be cumulative and that some of it would be irrelevant.
MCI Communications Corp. v. AT & T, 85 F.R.D. 28, 30 (N.D.Ill.1979). The record indicates that much of AT & T’s proposed testimony would in fact have been repetitive. AT & T’s list submitted November 14, 1979, for example, names no fewer than a dozen witnesses selected to present testimony on the installation of interconnection services for MCI, seven more who were to testify on the subject of repair services for MCI, and at least six more who were to testify concerning provision of other services to MCI. AT & T also predicted that it would require between twenty-two and forty-six days to cross-examine MCI’s witnesses who, by MCI’s estimate, could be directly examined in twenty-six days.
. AT & T cites Padovani as authority for reversal. There, the district court issued a preclusion order forbidding the plaintiff to introduce evidence of lay witnesses other than the plaintiff and his wife, any expert witnesses, any medical exhibits except three named, any evidence of damages with four exceptions and any evidence on the issue of liability either in negligence or based on breach of warranty. Finding that the order could only result in a judgment for the defendant, the Second Circuit granted plaintiff’s petition for a writ of mandamus. The court held that “[i]n no event at this pretrial stage should witnesses be excluded because of mere numbers, without reference to the relevancy of their testimony.” 293 F.2d at 550 (emphasis added). In the case before us, the lists and summaries submitted by the parties referred to the relevancy of the testimony, and the district court found that some of AT & T’s testimony would be irrelevant as well as needlessly cumulative.
. The National Commission for Review of Antitrust Laws and Procedures recommended in its report the imposition of time limits on the length of trial in extraordinarily complex antitrust litigation. The Commission wrote:
*1172It is desirable to establish time limits for each major phase of a complex case, so that within each fixed period the litigants are motivated to exercise self-discipline and creative choices between alternatives.
Time limits for length of trial, as imposed recently in a large antitrust case, have been rarely used. The power of judges to cut off cumulative, redundant presentations of proof may provide authority for the use of overall limits on trial presentations. As long as the limitations established are realistic and fair, and the judge prevents delaying tactics by hostile witnesses, we believe that trial time limits would also be an appropriate means of expediting litigation.
National Commission for Review of Antitrust Laws and Procedures, Report to the President and the Attorney General, 80 F.R.D. 509, 535-36 (1979) (footnotes omitted). See Lacey, Proposed Techniques for Streamlining Trial of Complex Antitrust Cases: Pro and Con, 48 Antitrust L.J. 487, 492 (1979); 1 J. Moore, Moore’s Federal Practice, Pt. II. (Manual For Complex Litigation) § 4.57 (1982).
. In this regard the following comment of AT & T’s counsel to the district court is of interest, but not determinative of the issue:
We are very pleased with the amount of time we have been given to deal with this kind of defense. We have no problems about that at all, and I am not trying to say that we are cutting it down [on the presentation of the defense] because you are leaning on us. You are not leaning on us.
Tr. 6554-55.
. In affirming the adequacy of the instructions, we note Judge Grady’s remarks during the instructions conference:
I hope, incidentally, that anyone who, at a later time, is critical of these instructions, will take a look at what I had to work with by way of the original submissions made by the parties a couple of weeks ago, by way of the plaintiffs’ submissions and the defendants’ submissions, and at least I may win somebody’s sympathy, if not their approval.