Full opinion text
Opinion for the Court filed PER CURIAM. PER CURIAM: As part of the 1982 consent decree that severed the seven Regional Bell Operating Companies (“BOCs”) from AT & T, the parties agreed that the BOCs, which inherited AT & T’s local exchange monopoly, would be prohibited from providing interex-change (long distance) or information services, manufacturing telephone equipment, and participating in any non-telecommunications industry. The district judge retained jurisdiction over the ease, and the Department of Justice (“DOJ”) pledged to report to the court every three years as to the continuing need for these “line of business” restrictions. In the first such “Triennial Review,” after considering the DOJ’s report, as well as the comments of the other parties and dozens of other individuals and organizations, the district judge issued two opinions lifting the restriction against BOC participation in non-telecommunication businesses, modifying the restriction against their entering the information services market, and leaving intact the interexchange and manufacturing restrictions. See United States v. Western Elec. Co., 673 F.Supp. 525 (D.D.C.1987); United States v. Western Elec. Co., 714 F.Supp. 1 (D.D.C.1988). With the exception of the district judge’s ruling dealing with information services — which we reverse and remand — we affirm. I. A. The 1982 Consent Decree In 1974, the DOJ filed this antitrust suit against AT & T. After seven years of pretrial proceedings, the case was tried in district court for eleven months but did not culminate in a verdict. Instead, the parties submitted a proposed consent decree to the court for review according to the “public interest” standard prescribed by the Antitrust Procedures and Penalties Act, 15 U.S.C. § 16(b)-(h) (“Tunney Act”). After extensive Tunney Act proceedings, and after the parties agreed to certain modifications added by the district judge, the district court approved the decree. See United States v. American Tel. & Tel. Co., 552 F.Supp. 131 (D.D.C.1982), aff'd sub nom. Maryland v. United States, 460 U.S. 1001, 103 S.Ct. 1240, 75 L.Ed.2d 472 (1983). Although the district court made no explicit findings of liability in the course of the Tunney Act proceedings, it did examine whether the evidence was sufficient to warrant antitrust relief. See 552 F.Supp. at 161. The evidence indicated that AT & T or the “Bell System” was, at the time of the trial, a massive, vertically-integrated enterprise which enjoyed a monopoly in local exchange services, provided long distance service, designed and developed telephone equipment (through Bell Laboratories), and manufactured that equipment at its wholly-owned subsidiary, Western Electric. AT & T used its local exchange monopoly — the so-called “bottleneck” — in a number of ways to promote its own affiliated operations in the long distance and equipment fields. In the interexchange field, other providers such as MCI were dependent on AT & T’s local exchange facilities since there was no other way to reach the ultimate consumer. AT & T therefore had a strong incentive to provide more expensive or inferior quality local exchange access to its long distance competitors than it provided to itself. According to the DOJ, despite vigilant FCC attempts to prevent it, AT & T was able to discriminate against its interexchange competitors and, in that way, to stave off significant interexchange competition. See 552 F.Supp. at 160-63. In the equipment market, the BOCs purchased over eighty percent of the nation’s central office switches and transmission equipment and nearly always purchased that equipment from AT & T’s Western Electric affiliate, even when those products were more expensive or of lesser quality than equipment available from competing vendors. The BOCs and Bell Labs also preferred Western Electric over competitors by granting it early and otherwise advantageous access to technical data and other information about the BOCs’ equipment requirements. Finally, there was evidence that AT & T cross-subsidized its equipment prices offered by its Western Electric affiliate using its monopoly revenues from local exchange services, thereby enabling Western Electric to undersell its competitors while telephone consumers were effectively overcharged for their local telephone service. Again, all of this was apparently carried out notwithstanding the FCC’s best efforts to stop it. See 552 F.Supp. at 190-92. Under the consent decree, AT & T re-t lined its long distance and equipment manufacturing operations but agreed to divest itself of its local exchange monopoly, transferring those operations to the BOCs which were to become totally separate from AT & T. In turn, the BOCs were to be limited to the provision of local exchange services and precluded from participating in the markets for interexchange (long distance) services, equipment manufacturing, information services, and all other non-telecommunications businesses. See 552 F.Supp. at 227-28. The BOCs were, however, permitted to provide — but not manufacture — customer premises equipment and also to produce, publish, and distribute “Yellow Pages” directories. See 552 F.Supp. at 231. These line of business restrictions were premised on the notion that, because the BOCs still controlled the local exchange bottlenecks, there was a risk that they would engage in the same sort of anticompetitive abuses that AT & T had. See 552 F.Supp. at 187-91. Before approving the restrictions, the district judge scrutinized them carefully to ensure “that they will not actually limit competition by unnecessarily barring a competitor from a market.” 552 F.Supp. at 186. He rejected as overly simplistic the DOJ’s equation of the post-divestiture BOCs with the pre-divestiture Bell System based simply on the proposition that they both possessed a monopoly in local telecommunications. The separated BOCs, the district judge recognized, would be far more manageable monopolists for regulators to oversee than AT & T had been because they would be regional rather than national in scope, they would not be vertically integrated, and, if permitted to enter competitive markets, they would face “the most potent conceivable competitor: AT & T itself.” 552 F.Supp. at 187. He therefore ruled that the Tunney Act’s public interest standard would permit banning the BOCs from a market only if there was a substantial possibility that the BOCs would use monopoly power to impede competition in that market. 552 F.Supp. at 187. As part of that “public interest” analysis, the district judge undertook a two-part inquiry to determine for each market barred to the BOCs whether (1) the BOCs would actually have the incentive and opportunity to act anticompetitively and (2) whether the participation of the BOCs would contribute to the creation of a competitive market. See 552 F.Supp. at 187, 188-94. In addition, the district judge examined the effect of the restrictions on “important public policies.” 552 F.Supp. at 187. The district judge ultimately determined that the proposed restrictions were indeed warranted. The line of business restrictions were not meant necessarily to be permanent, however. The district judge retained jurisdiction and insisted that a mechanism be inserted into the decree for removing them at a later date. Rejecting the DOJ’s view that mere monopoly power in local exchange services warranted the restrictions, the district judge thought that removal of the restrictions should be governed by a standard that allows a petitioning BOC to prove that there is no substantial possibility that the BOC could use its monopoly power to impede competition in the relevant market. See 552 F.Supp. at 195. The district judge therefore drafted, and the parties agreed to, section VIII(C) of the decree, which reads: The restrictions imposed upon the separated BOCs by virtue of section 11(D) [the line of business restrictions] shall be removed upon a showing by the petitioning BOC that there is no substantial possibility that it could use its monopoly power to impede competition in the market it seeks to enter. 552 F.Supp. at 231. This removal standard was intended to supplant the “test usually applied to a contested modification of a consent decree,” set out in United States v. Swift & Co. that asks whether “ ‘unforeseen conditions’ ” make modification appropriate. See 552 F.Supp. at 195 n. 266 (quoting United States v. Swift & Co., 286 U.S. 106, 119, 52 S.Ct. 460, 464, 76 L.Ed. 999 (1932)). Whether section VIII(C) would also apply to uncontested modifications is a question upon which the district court was silent in 1982. B. The Triennial Review When the decree was entered, the DOJ pledged to report to the court on the third anniversary of divestiture and every three years thereafter on the continuing need for the line of business restrictions. Since divestiture was not actually accomplished until 1984, the first such Triennial Review was held in 1987. The DOJ hired an independent consultant, Peter Huber, to conduct in-depth research on the telecommunications industry as a whole as well as on each of the relevant sub-markets. The Huber Report provided the factual basis for the DOJ’s preliminary submission, filed with the district court in February 1987, in which it recommended the complete removal of the manufacturing, non-telecommunications, and information restrictions as well as the modification of the interexchange restriction. Quite clearly, the DOJ’s position — which conceded the continued existence of the BOCs’ local exchange monopoly — represented a significant change from its position when the decree was entered that mere monopoly power in local exchange services necessitated the restrictions. After studying the comments of the parties and intervenors on its recommendations, the DOJ formally moved the court— apparently under section VIII(C) of the decree — for the removal of all the line of business restrictions, with the exception of the interexchange restriction. The DOJ was evidently persuaded to alter its position on interexchange services, and it instead asked the court to leave the restriction intact but to grant waivers as soon as local regulation in a given area is lifted. The Department believes that the BOCs’ bottleneck monopolies persist primarily because of local regulation which, if removed, would allow potentially competitive access alternatives to be made available by new technology. The seven BOCs also filed motions under section VIII(C) of the decree asking for complete removal of all the line of business restrictions. The district court held proceedings in which interested persons were invited to comment and respond to the report and the motions. With respect to the non-telecommunications and information services restrictions, all of the parties to the original decree — AT & T, the BOCs, and the DOJ— as well as the FCC agreed that the restrictions should be removed. AT & T opposed any modification of the other restrictions, thus making those BOC motions undeniably “contested.” Not surprisingly, numerous existing participants in the markets that the BOCs sought to enter intervened and vigorously opposed each of the proposed modifications. And, as noted above, the DOJ opposed the complete removal of the interexchange restriction. After discussing the standard for removal of restrictions under section VIII(C) of the decree, the district judge determined that the BOCs still possessed a bottleneck monopoly over local exchange service. See 673 F.Supp. at 536-40. The court then analyzed each line of business restriction to decide whether the BOCs had nevertheless met their burden under section VIII(C) to warrant removal. In reviewing the restriction concerning nontelecommunications businesses, the court noted that it had routinely reviewed and granted requests to waive this restriction since the decree became operational in 1984. Despite losing the safeguards that the waiver process afforded by virtue of the conditions imposed whenever a waiver was granted, the district court removed the restriction entirely because potential competitors did not actively oppose the removal and because cross-subsidization is more difficult in enterprises unrelated to telecommunications. See 673 F.Supp. at 599. In addition, the court asserted that lifting the restriction would eliminate a significant burden on BOCs’ business planning and free the court from unnecessary and detailed oversight of BOC decisions. See id. The district judge left largely intact the decree’s so-called “core” restrictions— those regarding interexchange services, manufacturing, and information services. The manufacturing and interexchange restrictions remained completely unchanged since the district court rejected arguments that circumstances had changed since the issuance of the decree so as to justify the BOCs’ entrance into those markets under the standard imposed by section VIII(C). The district judge did grant partial relief to the BOCs on the information services provision but rejected the request of the DOJ and the BOCs to remove the restriction entirely for much the same reasons as he left the manufacturing and interexchange restrictions in place. The court asserted that information services are vulnerable even to slight manipulation and discrimination in access or transmission quality, thereby making it especially easy for the BOCs to use their bottleneck monopolies anticompetitively if they entered the market. See 673 F.Supp. at 566. Nevertheless, the information restriction was lifted insofar as it prevented the BOCs from providing transmission of information services generated by others. See 673 F.Supp. at 587-97. The district judge described the economic and social advantages of making information services more widely available and noted that the telephone system was perhaps the only means of accomplishing that goal, because it uniquely offers providers a means “to reach large, dispersed audiences over reasonably priced, interactive facilities.” 673 F.Supp. at 564. So long as the BOCs do not compete with the companies that generate the information services, the district judge reasoned, they would have an incentive to afford the widest, fastest, and highest quality access and transmission to all information providers, thereby maximizing their own revenues. This determination was further explained and substantially reaffirmed by the district judge following a separate proceeding held to work out the details of the information services restriction. See 714 F.Supp. at 1. All of these rulings have been appealed. II. A. Standard of Review We have repeatedly held that the construction of a consent decree (indeed of this particular consent decree) is subject to de novo appellate review. See United States v. Western Elec. Co., 894 F.2d 1387, 1390 (D.C.Cir.1990); United States v. Western Elec. Co., 846 F.2d 1422, 1427 (D.C.Cir.), cert. denied, — U.S. -, 109 S.Ct. 306, 102 L.Ed.2d 325 (1988); United States v. Western Elec. Co., 797 F.2d 1082, 1089 (D.C.Cir.1986), cert. denied, 480 U.S. 922, 107 S.Ct. 1384, 94 L.Ed.2d 698 (1987); see also United States v. Armour & Co., 402 U.S. 673, 681-82, 91 S.Ct. 1752, 1757-58, 29 L.Ed.2d 256 (1971). “We read the Decree essentially as we would a contract. See United States v. ITT Continental Baking Co., 420 U.S. 223, 236-37 [95 S.Ct. 926, 934-35, 43 L.Ed.2d 148] (1975); United States v. Western Elec. Co., 797 F.2d 1082, 1089 (D.C.Cir.1986) (“Western Elec. II”), cert. denied, 480 U.S. 922 [107 S.Ct. 1384, 94 L.Ed.2d 698] (1987). Thus, while the meaning of the Decree’s terms ‘must be discerned within its four corners,’ United States v. Armour & Co., 402 U.S. 673, 682 [91 S.Ct. 1752, 1757, 29 L.Ed.2d 256] (1971), our inquiry is guided by conventional ‘aids to construction,’ including ‘the circumstances surrounding the formation of the consent order [and] any technical meaning words used may have had to the parties ... ’ ITT Continental Baking Co., 420 U.S. at 238 [95 S.Ct. at 935].” United States v. Western Elec. Co., 894 F.2d at 1390. Appellees nevertheless urge us to employ an “abuse of discretion” standard of review, arguing that what we are reviewing here is the district judge’s decision not to modify the consent decree — subject to review only for abuse of discretion, see System Fed’n No. 91 v. Wright, 364 U.S. 642, 647-48, 81 S.Ct. 368, 371-72, 5 L.Ed.2d 349 (1961) — rather than his interpretation of the decree. We think that is an inaccurate description of this proceeding. At least insofar as the district court was faced with motions brought under section VIII(C), it was not asked by any appellant to modify the decree; it was asked to apply it. To the limited extent that that process might require the district judge to find facts (for instance, to determine whether the BOCs still possess monopoly power over local exchange services), we would review under a clearly erroneous standard. Cf. City of Las Vegas v. Lujan, 891 F.2d 927, 931 (D.C.Cir.1989). But the legal standard for lifting the line of business restrictions pursuant to a BOC motion is explicitly provided by section VIII(C) of the decree and therefore the district judge need not “balance [any] imponderables,” Wright, 364 U.S. at 648, 81 S.Ct. at 371, in deciding whether to remove them as he might in a case seeking modification. Indeed, the district court enjoys no equitable discretion at all in applying section YIII(C); if the petitioning BOC makes the required showing, the district court “shall” remove the restriction. Therefore, aside from fact-finding, we owe no deference to the district court’s decisions under section VIII(C). And to the extent that we review the district court’s conclusions about the scope of the applicability of section VIII(C) (as opposed to section VII) for modifications, that is clearly a pure question of law subject to plenary review. Further, we reject the suggestion — apparently embraced by other circuits, see, e.g., Keith v. Volpe, 784 F.2d 1457, 1461 (9th Cir.1986) — that this particular district judge’s interpretations should be afforded some “special” deference because he drafted the pivotal provision of the decree, section VIII(C), and because he has had enormous experience overseeing the case and the decree since its inception. In addition to our discomfort with the concept that the degree of deference we afford should depend even in part on the identity of a district judge hearing the case below, we also note that appellate courts do not normally defer to anyone else’s non-contemporaneous interpretations of the Constitution, statutes, cases, or contracts — whether or not the interpreter was also the drafter of the language at issue. When we defer to agency interpretations of their own ambiguous statutes, Chevron U.S.A. Inc. v. NRDC, 467 U.S. 837, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984), or of contracts, National Fuel Gas Supply Corp. v. FERC, 811 F.2d 1563 (D.C.Cir.), cert. denied, 484 U.S. 869, 108 S.Ct. 200, 98 L.Ed.2d 151 (1987), we do so on the assumption that Congress delegated to departments or agencies the reconciliation of agency policies implicated in that function — institutional concerns not present when we review district court opinions. B. Applicability of Section VIII(C) Except for some BOC petitions regarding information services, the BOCs and the DOJ brought, and the district court analyzed, all of the motions to lift the line of business restriction under the standard set out in section VIII(C) of the decree. It is unclear to us, however, that section VIII(C) applies at all to some of those motions. In the first place, section VIII(C) refers to “a showing by the petitioning BOC,” and thus on its face contemplates only petitions brought by BOCs. The DOJ’s hard-line position in favor of the restrictions in 1982 further indicates that when the parties agreed to the decree, none of them contemplated that the DOJ would seek to invoke section VIII(C). We do not see, therefore, how the DOJ can petition for removal of restrictions under section VIII(C). That is not to suggest that if section VIII(C) is not available to the DOJ no avenue for requesting modification would be open to it. Section VII explicitly provides for modifications of the decree, and the district court would possess equitable power to modify the decree even if section VII were not included, see United States v. Swift & Co., 286 U.S. 106, 114-15, 52 S.Ct. 460, 462-63, 76 L.Ed. 999 (1932). Furthermore, as a party to the decree, the DOJ’s position on any BOC petition would be considered by the district court in much the same way as is an intervenor’s views— whether it supports or opposes the petition. Strictly speaking, however, only the BOCs can be petitioners under section VIII(C). We also believe that section VIII(C) does not apply to proposed modifications that are agreed to by all the parties to the decree, such as the motions for removal of the restrictions on BOC entry into the information services field. Section VTII(C) speaks of a “showing by the petitioning BOC,” thereby indicating that it applies only to contested motions for removal — and this reading comports with the intent of the parties as expressed to the district court in 1982. See infra Part III.C.l. As we explain more fully in Part III of this opinion, uncontested motions for modification— those involving the information services and non-telecommunications businesses restrictions — should be treated by the district judge under section VII of the decree, and should be approved so long as the modifications satisfy the “public interest” standard embodied in the Tunney Act. Thus, while both the district judge and the parties treated the removal of the line of business restrictions as though it were governed entirely by section VIII(C), that section was really the appropriate standard only for the BOC petitions for removal of the manufacturing and interexchange restrictions — the former opposed by AT & T and the latter by both AT & T and the DOJ. C. The VIII(C) Standard The proper construction and application of section VIII(C) is nonetheless critical in reviewing the district court’s decision not to remove the manufacturing and interex-ehange restrictions. There is no dispute, at least not on appeal, that the BOCs still possess their bottleneck monopolies in local exchange services. Despite certain technological innovations, only a minute percentage of telephone users can bypass the local exchange carriers for any of their calls. See 673 F.Supp. at 536-40. The question under -VIII(C), then, is' whether any petitioning BOC has made a showing “that there is no substantial possibility that it could use its monopoly power to impede competition in the market it seeks to enter.” Given the enormous legal resources expended on the issue, it is hardly surprising that the parties hotly dispute the meaning of the quoted phrase and the proper scope of the district judge’s inquiry in deciding whether the standard has been met. According to both the DOJ and the BOCs, the district court’s analysis suffered from several flaws: (1) misconstruing the actual terms of section VIII(C) quoted above; (2) failing to accord deference to the recommendations and opinions of the DOJ and FCC; and (3) taking into account allegedly irrelevant factors while ignoring or discounting critical changes in the industry since the decree. We begin with a close parsing of section VIII(C)’s terms. Section VIII(C) requires a BOC to show that there is no “substantial possibility that it could use its monopoly power to impede competition.” According to the DOJ and the BOCs, the district court altered the decree by implicitly equating the phrase “substantial possibility” with a mere theoretical possibility. Since the BOCs concede that they could always theoretically use their local exchange monopolies to impede competition, they claim that this putative misreading was tantamount to a ruling that retention of the local exchange monopolies precludes relief from the interexchange and manufacturing restrictions under section VIII(C). They base this argument on the district court’s statements to the effect that the local monopolies “continue to provide the same basis for anticompetitive activity as they did prior to the Bell System breakup.” 673 F.Supp. at 543. While we do not read the district court’s opinion, as appellants do, to have amended the decree, see, e.g., 673 F.Supp. at 536 n. 42 (explicitly rejecting the contention that “the restrictions are justified by the mere fact that a monopoly exists in an area”), the importance of the word “substantial” should not be minimized. The ultimate burden under section VIII(C) remains on the petitioning BOC, but the requirement that the possibility of using its monopoly power to impede competition be “substantial” relieves the BOC of the essentially impossible task of proving that there is absolutely no way for it to use its monopoly power to impede competition. For example, the district judge’s speculation that the BOCs could impede competition by way of illegal (and perhaps criminal) collusion to divide markets among them according to territory, see 673 F.Supp. at 558, would, in the absence of supporting evidence, seem to qualify only as a theoretical possibility. The parties also differ markedly concerning what precisely is meant by section VIII(C)’s ambiguous phrase “impede competition in the market it seeks to enter.” According to the DOJ, a BOC cannot impede competition in a given market unless it has market power — the ability to restrict output and/or raise prices. AT & T argues that the district court’s 1982 opinion equated impeding competition with “leveraging” monopoly power, something that AT & T claims a BOC can do so long as its local exchange monopoly is also an “essential facility” for the market it seeks to enter. Whatever it means to “leverage” one’s monopoly power, the DOJ is surely correct that no damage to competition— through “leverage” or otherwise — can occur unless the BOCs can exercise market power. Cf. General Leaseways, Inc. v. National Truck Leasing Ass’n, 744 F.2d 588, 596 (7th Cir.1984). To be sure, it may be difficult to decide whether the BOCs would have such power if they were allowed to enter a market. Moreover, it may be necessary to refine the analysis to deal with markets in which self-dealing bias is a risk, such as the production of central office switches and transmission equipment. In those markets, a BOC might be able effectively to raise prices (disguised as costs in the local exchange market) or restrict output — thereby impeding competition — in the segment it controls or forecloses. See infra Part III.B. The district court, however, was apparently concerned with the possibility that BOC entry into new markets would disadvantage or destroy small and innovative firms in those markets. See, e.g., 673 F.Supp. at 561 (castigating the DOJ for its indifference to the possible destruction of “many high-quality firms producing high-quality goods that have emerged since divestiture”). New entry or increased competition in any market typically hurts and sometimes even destroys existing competitors. A court’s solicitude for those firms- — ostensibly in an effort to foster competition — may well come at the expense of competition. Cf. Cargill, Inc. v. Monfort, 479 U.S. 104, 115, 107 S.Ct. 484, 492, 93 L.Ed.2d 427 (1986). Accordingly, unless the entering BOC will have the ability to raise prices or restrict output in the market it seeks to enter, there can be no substantial possibility that it could use its monopoly power to “impede competition.” And while there may be some complexities in defining precise boundaries of the relevant market, one thing that is clear from section VIII(C) is that it is the “market [the BOC] seeks to enter” that matters, and not the local exchange market. For the most part, then, the district court should decide motions under section VIII(C) without regard to the effect BOC entry into new markets will have on local service ratepayers. Concern for the ratepayers’ welfare is primarily the responsibility of the FCC and state regulators, not the district court. Appellees make much of DOJ’s prior position before the district court when the decree was entered when the Department urged the court to consider the interests of ratepayers in its evaluation and implementation of the decree. The DOJ concedes the shift, and the only explanation we are given is its statements in two footnotes of its brief that it now believes, contrary to its stance in 1982, that line of business restrictions should not be used— indeed, cannot be used under section VIII(C) — to protect ratepayers of local exchange services rather than solely to protect competition in unregulated markets. While this may have been the DOJ’s contention at the time, we see no clear evidence that ratepayer protection was part of the “contemporaneous understandings of [the decree’s] purposes,” 846 F.2d at 1427. And, in any event, we believe the text of the decree generally forecloses the goal of ratepayer protection by the use of the words “the market [the BOC] seeks to enter.” In that regard, to the extent that the district court’s consideration of cross-subsidization focused on the danger that the BOCs would overcharge local ratepayers, it was misconceived. See, e.g., 673 F.Supp. at 557, 572. Cross-subsidization is relevant under VIII(C) insofar as it may be used to price below cost in the competitive market, and thereby unfairly to acquire power and impede competition in that market. Still, the impact on the local exchange market of allowing BOC entry into a new market might well be relevant under section VIII(C) if the BOC is likely to be its own primary customer in the entered market, as with production of central office switches and transmission equipment. In that case, cross-subsidization or cost misallocation that allowed a BOC to pass on its (inflated) equipment costs to the local ratepayers would likely be the primary manifestation of market power and might constitute an impeding of competition in the entered market. Cf. 3 P. Areeda & D. Turner, Antitrust Law, 11726 (1978). Appellants also fault the district court for failing to give deference to the views of the FCC and the Justice Department with respect to the BOC petitions under section VIII(C). The Justice Department’s interpretation of the law is not normally given deference in a civil or criminal case; in federal courts, departments’ and agencies’ legal views are deferred to only when they make a determination (either quasi-legislative or quasi-judicial) that has independent legal significance — as opposed to when they act in a prosecutorial role. See Michigan Citizens for an Indep. Press v. Thornburgh, 868 F.2d 1285 (D.C.Cir.), aff'd by an equally divided Court, — U.S. -, 110 S.Ct. 398, 107 L.Ed.2d 277 (1989). Still it must not be forgotten that the Justice Department has the “principal responsibility for enforcing the Sherman Act.” Broadcast Music, Inc. v. Columbia Broadcasting Sys., Inc., 441 U.S. 1, 14, 99 S.Ct. 1551, 1559, 60 L.Ed.2d 1 (1979). Therefore, although we see no doctrinal basis for the district court to defer to the DOJ’s interpretation of the decree or its views about antitrust law, it is to be expected that the district court would seriously consider the Department’s economic analysis and predictions of market behavior. Indeed, it would seem that that is precisely why the district judge required the Department to report to the court every three years concerning the continuing need for the restrictions imposed by the decree. See 552 F.Supp. at 195. Economic analysis and market predictions are not an exact science. Antitrust scholars and courts have changed their views somewhat over the last fifty years concerning the interrelationship of the antitrust laws and market behavior. Compare, e.g., United States v. Arnold, Schwinn & Co., 388 U.S. 365, 87 S.Ct. 1856, 18 L.Ed.2d 1249 (1967) with Continental T.V., Inc. v. GTE Sylvania, Inc., 433 U.S. 36, 97 S.Ct. 2549, 53 L.Ed.2d 568 (1977) (overruling Schwinn after determining that its per se rule against vertical nonprice restraints was not economically sound). Also compare Sullivan, Economics and More Humanistic Disciplines: What are the Sources of Wisdom for Antitrust?, 125 U.Pa.L.Rev. 1214 (1977) (arguing that economic efficiency is not the sole goal of the antitrust laws) with 1 P. Aree-da & D. Turner, Antitrust Law ¶¶ 103-113 (1978) and R. Bork, The Antitrust Paradox 50-89 (1978) (agreeing that courts should treat antitrust laws as designed solely to advance consumer welfare and efficiency). Consequently, we recognize that the DOJ may change its views — to incorporate different policy concerns — over time. That is not to say that we do not have any sympathy for the district court’s attitude toward the DOJ’s position changes in this Triennial Review. As we noted above, the DOJ in 1982 wanted the line of business restrictions to remain in place unless and until the BOCs lost their local exchange monopolies. With little warning or explanation, the DOJ completely altered its stance and is now generally hostile toward the restrictions. In the absence of a complete explanation of how and why the DOJ’s position had changed, the district judge was understandably uneasy about relying on the DOJ in this first Triennial Review since to do so would be to undo much of the decree after only three years’ time. The FCC’s claim to deference is perhaps even more puzzling to analyze. The FCC argues that the district court acted improperly when it evaluated the effectiveness of the FCC’s regulatory scheme since that is solely the function of a court of appeals pursuant to direct review as provided by the Communications Act. The problem with this argument is that a court of appeals reviews FCC regulations only if they are challenged and only to ensure that they are not arbitrary and capricious. The district court below, however, was obliged to determine ultimately whether the FCC’s regulations would effectively prevent the BOCs from using their monopoly power to impede competition in the markets they sought to enter. Cf. 846 F.2d at 1433. The very premise of this case was that the FCC could not effectively control AT & T. We think it would therefore have been an abdication of judicial responsibility for the district court to assume that the FCC’s regulations would be effective merely because they had not been found to be arbitrary and capricious. On the other hand, we recognize the institutional anomaly presented by a district judge placed in the position of evaluating the effectiveness of a federal agency’s regulatory program. The DOJ, which brought this action based on its view that the FCC was incapable of preventing AT & T’s abuse of monopoly power, would appear to be in a better position, both institutionally and practically, to evaluate the FCC’s regulatory effectiveness. And, therefore, we would expect the district court to consider the Department’s comparative advantage in performing that task. However, two of the principal FCC regulations that bear on this appeal, the Joint Cost and the Computer III rules, had not been finally implemented at the time this case was submitted to the district court. Thus the DOJ’s assessments of those regulations that are in the record of this appeal are necessarily speculative. See, e.g., infra notes 16, 17, 21. Rather than hazard our own, necessarily under-informed, appraisal of how these rules have performed in the three years since this case was presented to the district court, we think it is more prudent to await the DOJ’s assessment in subsequent Triennial Reviews. Appellants further argue that the district court erred in its application of section VIII(C) by considering factors it should have ignored while ignoring some that it should have considered. Most importantly, appellants take issue with the minimal significance that the district court placed on the break-up of AT & T’s local exchange monopoly into seven separate BOCs in deciding whether to lift the line of business restrictions. The decree was premised on the notion that the BOCs would have both the incentive and the ability to use their local exchange monopoly to impede competition in these markets, therefore necessitating the line of business restrictions. See 552 F.Supp. at 187. Under section VIII(C), therefore, the BOCs must establish that something is different now from the time when the decree was entered so that they can no longer use their monopoly power to impede competition. Obviously, if all conditions and assumptions remain the same now as when the decree was entered, no relief can be due under section VIII(C). While we reject the BOCs overly loose reading of the restrictions under section VIII(C), we also reject the appellees’ overly rigid interpretation of those restrictions. The appellees appear to insist that the standard under section VIII(C) requires a finding of some unforeseen changed circumstances as an ingredient of the petitioners’ showing necessary to justify removal of a line-of-business restriction. This construction finds no support in the language of the decree. Rather, the decree provides that a petitioning BOC is entitled to relief under section VIII(C) so long as it can prove that its ability to impede competition is no longer present. It would make no difference whether the circumstances leading to that conclusion were entirely foreseen or even if they could not be discretely identified. Indeed, even if the economic assumptions or market predictions which governed the decree turned out to be in part wrong, the BOCs might thereby be entitled to relief under section VIII(C). To be sure, as the district court noted, the mere existence of seven BOCs in place of the prior unified Bell System is not by itself a significant factor. Not only was it the very product of the decree, the seven continue to exercise monopoly power in the local exchange market. However, the other conditions in the various telecommunications markets wrought by divestiture and the behavior of the BOCs since divestiture are appropriately considered under section VIII(C), even if they were foreseen by the parties or incorporated into the fabric of the decree. These are not limited to obvious changes such as new competitors that did not exist or were not stable at the time of the decree. The court also may properly consider the manner in which the seven BOCs behave competitively against each other, AT & T, and other firms, as well as the way in which the various markets have evolved since the decree. It was feared, for instance, at the time the decree was entered that the BOCs would favor one another and AT & T over unrelated firms, a concern that now appears unfounded. Also appropriately considered in the section VIII(C) calculus is the asserted existence of “benchmarks” for comparing BOC performance. According to appellants and the FCC, these benchmarks would make it far easier to regulate the BOCs than the old Bell System if the BOCs were permitted to enter other markets. That the “possibility of the existence of benchmarks was necessarily included in the decree assumption which imposed the restrictions,” 673 F.Supp. at 547, makes them no less significant under section VIII(C). Indeed, the greater ease with which the FCC can regulate the BOCs merely because they have a monopoly in only one geographic portion of one of the markets controlled by AT & T prior to the decree — although clearly contemplated (and therefore foreseen) at the time of the decree, see 552 F.Supp. at 187 — is properly considered under section VIII(C). Of course, the district court still legitimately imposes on the petitioning BOCs the burden of making the requisite showing. Finally, we consider appellants’ claim that the district court strayed beyond the competitive analysis mandated by section VIII(C) when it considered the impact of removing the restrictions on various public policies, including the welfare of local ratepayers, innovation in the manufacturing market, the goal of universal telephone service, first amendment values, and the United States’ position in international trade. The district court explained its discussion of these factors by noting that “the same standards may be applied in proceedings addressing continued viability of the restrictions as were used in determining whether the restrictions were to be imposed in the first place.” 673 F.Supp. at 583. We disagree. When the district court entered the decree in 1982, it decided — as it was required to do under the Tunney Act— whether the decree was in the public interest, Whatever the substance of the “public interest” standard, it is surely more far-ranging than the section VIII(C) standard. When a BOC petitions under section YIII(C) for the removal of a line of business restriction, section VIII(C) itself defines the limits of the district judge’s inquiry. If a BOC makes the showing called for by section VIII(C), the district judge may not, for example, deny the motion because of the possible impact on the United States’ balance of trade, or for any other reason not related to the antitrust laws. III. We turn next to the district judge’s rulings on the BOC motions to remove the interexchange and manufacturing restrictions. As we noted above, the BOC motions to remove these two restrictions were the only motions properly analyzed under section VIII(C) of the decree. We therefore need not separately consider the DOJ’s motions, which were brought explicitly pursuant to section VIII(C), to the extent that those motions are different in substance from the BOC motions. Despite our disagreements with the district judge concerning the matters described above, we agree that the BOCs did not satisfy their burden in this Triennial Review of showing that there was no substantial possibility that they could use their monopoly power to impede competition in these markets. On appeal, the DOJ and the BOCs focus their criticism on the district court’s analysis and the standard it applied. As such, they ask us to remand for reconsideration under the “legally correct VIII(C) standard.” Since we conclude that the district court reached the correct results on the motions properly considered under VIII(C), there obviously is no reason to remand those motions to the district court. A. Interexchange Services The BOCs, supported by the FCC but not the DOJ, argue that the interexchange restriction should be removed in spite of the conceded persistence of their local exchange monopoly — upon which interexchange carriers rely for access to ultimate consumers — because changed circumstances since the decree would prevent them from impeding competition if they were permitted to enter the market. The DOJ opposes the BOCs’ petition. The district court noted that the interex-change market is currently competitive. Even though AT & T still retains a lion’s share of the market, there are hundreds of long-distance carriers in the United States, eight of them serving twenty-five or more states. See 673 F.Supp. at 550. This apparently undisputed fact leads the district court to the curious observation that “the entry of the [BOCs] into that market is not necessary to give it vitality.” Id. Section VIII(C) does not put the BOCs in the “Catch-22” that the district judge seems to imply by that statement — that the restriction will not be lifted if the market to be entered is not competitive because the BOC will grab and wield market power nor will it be lifted if the market to be entered is highly competitive because the BOCs’ presence is not “necessary.” If the market to be entered is sluggish or concentrated, that might be an argument in favor of allowing BOC entry, although it might also make it easier for a BOC to acquire market power. But if the BOC wants to enter a competitive market, that is a powerful reason to grant the section VIII(C) petition since there is less of a danger that the BOC will be able to seize and wield market power. Nevertheless, we agree with the district court’s conclusion regarding the changes in the interexchange market. The crux of the BOCs’ argument is that equal access for all interexchange carriers has been achieved and cross-subsidization eliminated, thus closing off the primary way for the BOCs to acquire market power anticompetitively. This has allegedly come about primarily through FCC regulation that was made more effective by the fragmentation of the Bell System’s local exchange monopoly into seven BOCs. As we indicated above, even though the mere existence of the seven BOCs is not a significant consideration under section VIII(C), the break-up is critically important insofar as the FCC has been able to adapt its regulations to the more manageable task of overseeing seven regional monopolies instead of one national, vertically-integrated one. See 552 F.Supp. at 187. On this point, we think the DOJ’s assessment of the FCC’s regulations is entitled to significant weight. According to the DOJ, those regulations, combined with the BOCs’ own equal access plans have “eliminated most of the anticompetitive advantages AT & T formerly enjoyed.” At the time this case was before the district court, however, the DOJ asserted that “the FCC’s equal access rules are based on the assumption that the BOCs will not provide interexehange services. Thus the FCC has not yet developed rules that would apply the nondiscrimination and cost separation principles of the Computer III and Joint Cost proceedings to BOC provision of interexehange services.” And until those regulations are adjusted to take account of BOC entry into the interexehange market — entry which would, of course, provide an incentive to deny equal access and to cross-subsidize if possible — the DOJ represents that equal access and proper cost allocation cannot be assured. The DOJ also points out that violations of the equal access policy are extremely difficult to detect and remedy, see 846 F.2d at 1424-25, thereby underscoring the danger of allowing entry before the FCC’s regulations are designed to deal with the problem. Finally, the DOJ warns that the BOCs will have an easier time acquiring market power in the interexehange market than in other markets because many of the firms that began providing interexehange services after the decree have not yet become stable in the highly capital-intensive field. Since the BOCs raise no serious opposition on appeal to any of these points made by the DOJ and since we have no other reason to doubt the DOJ’s assessments, we affirm the district judge’s conclusion that the BOCs failed to show that there was no substantial possibility that they could use their monopoly power to impede competition in the interexehange market. B. Manufacturing The “manufacturing restriction,” imposed by section 11(D)(2) of the decree and amended by section YIII(A) is really more properly conceived of as two restrictions. First, it forbids the BOCs from manufacturing (but permits them to provide) customer premises equipment (“CPE”), which “includes equipment employed on the premises of anyone other than a carrier that is utilized to originate, route, or terminate telecommunications.” 673 F.Supp. at 552 n. 116. Second, it prohibits the BOCs from manufacturing or providing telecommunications equipment, that is, equipment other than CPE used by a carrier to provide telecommunications services. See id. The BOCs petitioned, with DOJ support, for the complete removal of the manufacturing restriction, and the district court left the restriction intact. 1. Telecommunications Equipment The BOCs and the DOJ argue that market changes since the decree and regulatory adaptations to the post-divestiture market warrant removal of the telecommunications equipment restriction. The Justice Department further divides the telecommunications equipment market into separate markets for central office switches and for transmission equipment (primarily metal cable). The DOJ makes the significant concession that any BOC that chooses to manufacture central office switches, either unilaterally or through a joint venture, will buy all (or nearly all) of its requirements from the affiliated producer — thereby foreclosing a certain portion of the market, whether or not there are economies to be gained from such integration. Nevertheless, the BOCs will not impede competition (although competitors will surely be hurt) in the switch market, it is argued, because large economies of scale would prevent any BOC from remaining in the market solely to sell to itself. Even the largest BOC buyer of switches in 1985 purchased only 1.4 million lines of new switching capacity, about 17 percent of the U.S. market. Only two other BOCs purchased over one million switches and two BOCs bought less than 500,000. Huber Report at 14.8. The Huber Report estimates that switch producers must sell upwards of 1.5 million switches per year to survive and many more to be profitable. Huber Report at 14.15-14.16. Of the three major U.S. switch producers, two (AT & T and Northern Telecom) sold over 4.5 million switches in the U.S. in 1985, and perhaps twice that many worldwide. The third (GTE) sold only about 1.5 million switches in 1985 and is losing money in its equipment businesses. Huber Report, 14.8, 14.-14-14.15. Assuming that there are no joint ventures among BOCs (which the DOJ admits would alter its assessment), the DOJ argues that only a few BOCs would enter the switch market and those that did could not afford to produce idiosyncratic (or overpriced) switches since they would have to attract at least some significant number of nonaffiliated buyers who, of course, can choose among many producers. The DOJ further concedes, however, that the image they convey of the efficient BOC producer of switches is somewhat clouded by the danger of anticompetitive interconnection discrimination and of cross-subsidization. The risk of interconnection discrimination, by which a BOC would design its switches in a way that would favor the BOC’s self-produced equipment over rival manufacturers, has allegedly been “substantially decreased” by more effective regulatory control — especially the availability of benchmark comparisons among the BOCs. Cross-subsidization is a “plausible concern” in the switch market, the DOJ tells us, because of the need to attract a large market share, the extensive shared costs with local exchange services (especially research and development) (“R & D”), and traditional difficulty encountered by regulators in discovering cost misalloca-tions. Indeed, according to the DOJ, a BOC “might not have to produce efficiently to attract its own operating companies as buyers if regulators did not prevent recovery of excessive switch costs.” And while the risk of cross-subsidization cannot be eliminated completely, FCC regulation — especially the availability of benchmarks to enforce effective accounting rules — would “significantly mitigate” it. Finally,' the DOJ maintains that a small amount of cross-subsidization would not impede competition in the entered market; rather its primary effect would be to raise the price of local exchange service, a problem that the DOJ suggests is to be handled by regulators and is irrelevant to this proceeding. The DOJ’s assessment of the transmission equipment market is substantially similar to that for central office switches. In the transmission equipment market, as was true of central office switches, any BOC would purchase all or most of its own equipment from its own manufacturing affiliate; not all BOCs would manufacture every type of transmission equipment. There is some danger of discrimination and of cross-subsidization, but, due in large part to the availability of benchmark comparisons, that risk is substantially less than it was prior to divestiture. Among the salient differences is that the market has supported competition even though the BOCs have already been allowed to provide transmission equipment in the form of CPE, and therefore already possesses an incentive to discriminate in interconnection. In addition, cross-subsidization is allegedly less probable than in the switch market because R & D costs, normally the prime source of cross-subsidization, are so low. Even if we did accept the DOJ’s market forecasts and regulatory assessments wholesale, it would not suffice to compel removal of the telecommunications equipment restriction under the section VIII(C) standard. As we discussed above in Part 11(C), the possibility of self-dealing bias in the telecommunications equipment market poses dangers to competition that do not exist in the other markets the BOCs seek to enter. The DOJ’s submissions provide little solace against those dangers. As we noted above, the DOJ “assumes that any BOC that manufactures equipment would purchase substantially all of its requirement from its affiliate,” presumably regardless of price or quality. While the BOCs and the DOJ contend that not all BOCs will produce each type of equipment and therefore dispute the district judge’s conclusion that the BOCs would foreclose 70% of the telecommunications equipment market, there seems to be no dispute that some substantial portion (5-15%) of the equipment market will be foreclosed. Such foreclosure, if combined with cross-subsidization, would appear to allow the BOCs, in effect, to raise prices (and therefore exercise a form of market power) in the foreclosed sectors of the equipment market by disguising inflated equipment prices as costs in the local exchange market. See supra Part II.C. To be sure, the DOJ advised the district court that FCC regulation would substantially reduce the risk of cross-subsidization. In its own reports on the two principal telecommunications equipment submarkets, the DOJ nevertheless concedes that the BOCs would possess both the incentive and the ability to cross-subsidize, at least somewhat. The DOJ and the BOCs nowhere explain, however, why any significant amount of cross-subsidization that, in practical terms, enables the BOCs to charge higher prices for the equipment it produces would not be akin to an exercise of market power that would impede competition in the telecommunications equipment market. At least in this first Triennial Review, it is not enough for the BOCs (independently or through the DOJ) to show that a significant number of stable competitors will be able to survive BOC entry. We think that the BOCs cannot meet their burden under section VIII(C) without also explaining why foreclosure combined with cross-subsidization does not itself pose a “substantial possibility that the BOC could use its monopoly power” in the telecommunications equipment market. 2. Customer Premises Equipment The DOJ subdivides the CPE market into the private branch exchange (“PBX”) market and the terminal equipment market. While it analyzed those sub-markets separately, the DOJ believes — for reasons similar to those offered for the telecommunications equipment market— that BOC entry into both would not threaten competition. The PBX market is described as “moderately concentrated”— with three major and many minor producers — but nevertheless competitive. The DOJ points out that the BOCs have been permitted to provide PBXs (as well as all other forms of CPE) since the decree, and that they therefore have always had the incentive to provide discriminatory interconnection. If allowed to produce PBXs, that discrimination could manifest itself through BOC manipulation of its local network or through discriminatory dissemination of network information. We are assured, however, that the BOCs’ ability to discriminate will be blocked, as it has been until now, by numerous obstacles, especially FCC regulation. Similarly, any danger of cross-subsidization in the PBX market is dismissed by the DOJ as negligible. The DOJ’s argument with respect to the terminal equipment market is based again on the premise that no BOC could successfully discriminate against its rivals, either by way of interconnection or by failing to provide critical network information. Their ability to discriminate is asserted to be undercut somewhat by the advent of PBX use, as well as cellular and paging systems for which the BOCs do not even provide the interconnection with the terminal equipment and which all contribute to the prevalent use of standards in interconnection. Residual risk of discrimination, as well as the risk of cross-subsidization, are adequately prevented by FCC regulation. While we certainly have some reservations concerning the DOJ’s assessment of the CPE market, we are inclined to think that the question is much closer than it was for telecommunications equipment. Since the BOCs purchase only a minute percentage of the nation’s CPE output, there is no risk of the combined cross-subsidization and foreclosure that is so crucial to our decision on telecommunications equipment. Indeed, on appeal, the BOCs and the DOJ complain primarily that when the district judge discussed foreclosure, he did not differentiate between CPE and the various types of telecommunications equipment that the BOCs do purchase themselves in great quantity. While we agree that different competitive concerns should motivate decisions regarding CPE than those regarding telecommunications equipment, we do not perceive that as a reason to upset the district court’s decision as it applies to CPE manufacturing. The burden under section VIII(C) is not on the district judge; it is on the petitioning BOCs to show that they can enter certain markets without raising a substantial possibility that they could use their monopoly power to impede competition in those markets. In this Triennial Review, the BOCs petitioned for complete removal of the manufacturing restriction, and the DOJ explicitly urged the district judge not to distinguish between the two types of equipment in making his decision, because line-drawing between them is so difficult. Given the risks to competition we identified in the telecommunications equipment market, the district court understandably did not allow the entire manufacturing restriction to be removed in this first Triennial Review, and the motions were properly denied. C. Information Services The BOCs and the DOJ also appeal from the district court’s decision not to lift the restriction on “information services” under section 11(D)(1), as that restriction is applied to the generation of information. The district court found that there had “been no significant, relevant change in” market conditions justifying removal of this restraint under section VIII(C). 673 F.Supp. at 564. Noting that neither the DOJ nor AT & T opposed lifting the information services ban, the BOCs contend that the district court should have reviewed this question under a more flexible “public interest” standard pursuant to section VII. See supra note 6. We agree, and hold that the district court erred in applying section VIII(C) to the uncontested motion to remove the line-of-business restriction on information services. And because we are unable to say that the district court would have reached the same result had it applied the proper legal standard, we are constrained to remand the case for further consideration of the BOCs’ motion to remove the information services ban in its entirety. 1. The Applicability of Section VII to Uncontested Motions to Modify a Line-of-Business Restriction Under the Decree It is well established that a less demanding standard of review applies to an uncontested motion to modify a consent decree than applies to a contested one. See generally Note, Modifications of Antitrust Consent Decrees: Over a Double Barrel, 84 Mich.L.Rev