dissenting:
The AT&T Modified Final Judgment dismantled the old Bell System. See United States v. AT&T, 552 F.Supp. 131, 226-34 (D.D.C.1982) (“MFJ Opinion”), aff'd without opinion sub nom. Maryland v. United States, 460 U.S. 1001, 103 S.Ct. 1240, 75 L.Ed.2d 472 (1983). Section 11(D) of the decree bars one set of thé resulting offspring — the Bell Operating Companies or BOCs — from entering into various lines of business, such' as provision of interexchange telecommunications services and manufacture of “telecommunications products” (subject to- various exceptions). Id. at 227-28. Section 11(D) explicitly applies the restriction to the BOCs’ affiliates: “[N]o BOC shall, directly or through any affiliated enterprise” provide the forbidden goods and services. Id. The majority here concludes that, as a result of this reference to affiliation, the § 11(D) restriction bans “all arrangements, contractual or otherwise, in which the BOCs have a direct and continuing share in the revenues of entities engaged in prohibited businesses.” Maj.Op. at 232 (emphasis in original). The direct result of the ruling is to scotch the transaction that gave rise to this litigation, under which a BOC, Ameritech, advanced money enabling David Systems to apply an innova--tive technology to a telecommunications product, with, David Systems in return obliged to pay Ameritech a share of its revenues from sales of the product.
The indirect results are far more serious. In adopting this sweeping idea of affiliation, the court looses the concept from its core meaning of control, characteristically exercised through some kind of ownership. The court substitutes an amorphous notion that seiems likely to obstruct a wide range of BOC efforts to advance entry into telecommunications by firms that, by any ordinary standard, would be viewed as independent. Although the court’s interpretation has modest support in some parol evidence, it has none *238in the language of the decree itself and is sharply contradicted by the behavior of the parties to the MFJ — especially that of AT&T, which now appears as the interpretation’s major champion. Ironically, the court’s construction of § 11(D) will — in the name of competition — actually inhibit competition in the lines of business from which the BOCs were excluded; it will thereby facilitate AT&T’s . overwhelming dominance in some of those lines — represented, for example, by its 49% of sales of central office switches. See Appellants’ Brief at 34 n. 40 (citing NATA, 1992 Telecommunications Market Review & Forecast 78 (1992)).
This dissent will first consider some specific clues to the meaning of “affiliated enterprise” in § 11(D) of the consent decree: the language of § 'II(D) in relation to other references to affiliation in the MFJ, uses of the concept of affiliation in other antitrust consent decrees of substantially the MFJ’s vintage, parol evidence as to the parties’ intent át the time the decree was agreed to, and the later behavior of the parties. Then it will consider whether the majority’s expansive interpretation can be justified in terms of the anti-competitive risks that drove adoption of the decree. ‘
Decree references to affiliation. Section II(D)’s use of the term “affiliated entity” is not the decree’s only reference to affiliation. Section IV(A) defines as an affiliate of AT&T “any organization or entity ... that is under direct or indirect common ownership with or control by AT&T or is owned or controlled by another affiliate.” MFJ Opinion, 552 F.Supp. at 228. It goes on to say, “For the purposes of this paragraph, the terms ‘ownership’ and ‘owned’ mean a direct or indirect equity interest (or the equivalent thereof) of more than fifty (50) percent of an entity.” Id. While directed specifically to AT&T itself, the definition suggests that drafters of the decree thought of affiliation in conventional terms — control, typically via ownership.
The decree also defines the BOCs so as to explicitly encompass affiliates under the usual understanding of the term; § IV(C) states that “the BOCs mean [any corporation listed on an attached appendix] and any entity directly or indirectly owned or controlled by a BOC or affiliated through substantial common ownership.” Id. The definition indicates an intent to reach entities either owned or controlled by BOCs, or affiliated with the BOC through a common parent with substantial ownership in both entities. Again, affiliation appears to depend on ownership or control.
As the panel properly notes, however, § 11(D) also refers to affiliation. Thus, although “BOCs” by definition under § IV(C) include affiliates under the usual understanding of the term, § 11(D) explicitly applies the ban not only to BOCs but also to any “affiliated enterprise”. The majority, invoking the maxim that contracts should be interpreted so as to give meaning to every provision, concludes that the § 11(D) reference must somehow encompass more than § IV(C). Maj.Op. at 230.
Like so many constructional maxims, this one seems questionable. An alternative view is that the seemingly redundant definition arose’ either from neglect, or, more likely in the intensely lawyered atmosphere surrounding this decree, to make assurance doubly sure. Cf. Fort Stewart Schools v. FLRA, 495 U.S. 641, 646, 110 S.Ct. 2043, 2047, 109 L.Ed.2d 659 (1990) (drafting redundancy may have been inserted “out of an abundance of caution — a drafting imprecision venerable enough to have left its mark on legal Latin (ex abundanti cautela)”).
The decree’s pervasive tendency to repeat references to affiliation supports the abundance-of-caution view. In referring to AT&T in §§ 1(A) 1, 11(B), VIII(B) and VIII(D), the decree in each case adds mention of its “affiliates”, which is redundant because § IV(B) has already defined AT&T as including its affiliates. In addition, §§ III and V refer to the defendants, the BOCs and their affiliates, again redundantly in light of the definitions of AT&T and the BOCs. Under the majority’s interpretive method, evidently, supplementary meanings must be found for each of these seemingly redundant usages. While the pattern convinces me that the drafters emphatically intended to cover affiliates, I see no support for the view that they meant *239to send the courts off to hunt for special meanings for every repetitious use.
References to affiliate in other antitrust consent decrees. The question of affiliation is obviously posed in antitrust settlements generally, so we may look to such consent decrees for some sign of their meaning in that context. In a sample of 70 consent decrees from July 1980 to the present,1'23 of which were published prior to the AT&T Modified Final Judgment (Aug. 24,1992), the term “affiliate” generally appears in two places in the competitive impact statements and final judgments: (1) the definition of a shorthand company name, which usually adds a term such as “and any ... affiliate thereof’ to the full name of the defendant, see, e.g., U.S. v. Merck & Co., Inc., 45 Fed. Reg. 60,044, 60,045 (Sept. 11,1980); and (2) a boilerplate “applicability” clause, which is identical in virtually all of the searched decrees:
The provisions of this Final Judgment applicable to Reveo shall also apply to each of its officers, directors, agents, employees, affiliates, subsidiaries, successors and assigns, and to all other persons in active concert or participation with any of them who receive actual notice of this' Final Judgment by personal service or otherwise.
See, e.g., United States v. Reveo D.S., Inc., and Zale Corp., Proposed Final Judgment and Competitive Impact Statement, 46 Fed. Reg. 13,418, 13,419 (Feb. 20, 1981).
While these decrees typically do not bother to define “affiliates”, the applicability clause — which lumps affiliates in with “officers, directors, agents, employees, ..., subsidiaries, successors and assigns” — suggests a narrow meaning. Affiliates are grouped with entities that the defendant (or a parent) is entitled to commit to the consent decree by virtue of ownership or control, in contrast to “all other persons in active concert or participation”, seemingly a catch-all for persons who may actively assist the defendant in violating the decree.
Only- five of the consent decrees reviewed explicitly define affiliation. All óf these definitions include some form of ownership or substantial control. In United States v. Gillette Co., Proposed Final Judgment and Competitive Impact Statement, 55 Fed.Reg. 12,-567,12,571 (April 4,1990), the decree defined an “ ‘Affiliate’ of a legal entity” as “a person controlled, directly or indirectly by a common parent of that legal entity.” A decree in the telecommunications context uses a similar definition:
“Affiliate” or “affiliates” means any organization or entity in which, directly or indirectly, the named person has control or substantial ownership. For purposes hereof, “substantial ownership” means a direct or indirect equity interest (or the equivalent thereof) of fifty percent (50%) or more of an entity. Any parent company of a named person shall also be deemed its affiliate.
United States v. Pacific Telesis Group, Proposed Final Judgment and Competitive Impact Statement, 51 Fed.Reg. 9277, 9278 (March 18, 1986). What “the equivalent thereof’ means in the above definition is not clear, but it seems to require an interest somehow comparable to the specified equity share — 50%.
Another telecommunications decree, United States v. GTE Corp., Proposed Final Judgment and Competitive Impact Statement, 48 Fed.Reg. 22,020, 22,021 (May 16, 1983), published nine months after the MFJ Opinion, states that affiliate “means any organization or entity in which, directly or indirectly, GTE has any ownership or equity interest or control.” GTE also defines “BOC” as including the corporations’ “successors and assigns, and any entity directly or indirectly owned or controlled by a BOC or affiliated through substantial common ownership.” Id.
One decree spells out not only the concept of affiliation but the idea of “control” embedded therein:
“Affiliate” means, with respect to any Person, any other Person directly or indirectly *240controlling, controlled by or under common control with such Person ... For purposes of this definition ‘control’ (including ‘controlling’, ‘controlled by’ and ‘under common control with’) shall mean the possession, directly or indirectly, of the power to direct or cause the direction of the management and policies of .a Person, whether through the ownership of voting securities or by contract or otherwise.
United States v. Alcan Aluminum Ltd., Proposed Final Judgment and Competitive Impact Statement, 49 Fed.Reg. 40,454, 40,459 (Oct. 16,1984). Though broad, the definition of control here seems to depend on legal or contractual power over the other firm’s management and policies. There is no hint that it would reach links that merely afford one firm an incentive to influence or favor the other.
Finally, one decree defines “subsidiary” and “affiliate” together, with the former defined as “a company of which the parent owns more than 50% of capital stock”, while the latter implies “an entity of which the defendant has more than 50% nonstock ownership interest or has less than 50% interest and exercises or has the right to exercise control.” United States v. Hercules Incorporated, Proposed Final Judgment and Competitive Impact Statement, 45 Fed.Reg. 85,-840, 85,841 (Dec. 30, 1980).
The decrees, then, are united by a theme of genuine control. They represent the product' of the Antitrust Division’s lawyers, many of them in the period in which those same lawyers negotiated the MFJ. If those lawyers wanted some drastically broader coverage, one would expect them to say so.
Pre- and post-decree statements by the Department of Justice. The majority places considerable weight on certain assertions by the Department of Justice — before and after adoption of the decree — as to the reach of the áffiliation concept of § II(D). See Maj. Op. at 230-31. First a general caution. Par-ol evidence may be used to clarify an ambiguity, and I am ready to assume arguendo that
§ II(D)’s redundant reference to affiliation creates an ambiguity. But there is nothing within the decree, or within standard usage of the affiliation concept in antitrust decrees, remotely suggesting a stretch beyond the twin ideas of ownership and control. While I can well imagine parol evidence helping us to work' out the exact line within the zone framed by those basic ideas, I doubt the validity of its use to substitute a completely new concept. Cf. Tataranowicz v. Sullivan, 959 F.2d 268, 276 (D.C.Cir.1992) (when statute contains ambiguity that is readily resolved, the ambiguity’s force is spent and does not create ambiguity on every other interpretive issue).
In any event, the pre-decree history yields a Justice Department reference to possible capacity-sharing arrangements between the future BOCs .and interexchange carriers and information providers:
First, to the extent that, as a practical matter, such [capacity-sharing] agreements amount to a joint venture with the sharing enterprise, or otherwise give the BOC a stake in its financial success, e.g., payments on a per-unit-of-traffic basis (exclusive of tariffed access charges), the modification’s prohibition against the BOC’s reintegration into interexchange or information services markets would be violated.
Response to Public Comments on Proposed Modification of Final Judgment, 47 Fed.Reg. 23,320, 23,347 (May 27, 1982) (emphasis added) (cited at Maj.Op. at 230-31).
The remark seems strikingly discursive. It refers first to joint ventures, suggesting a focus on highly integrated activities. Then it suggests that any BOC “stake in [the other firm’s] financial success” would be fatal — a test that if taken seriously would doom a whole range of relations with suppliers and customers whose financial collapse would’ injure a BOC. Then it refers to “payments on a per-unit-of-traffic basis”, a phrase possibly suggestive of some revenue-sharing arrangement, but without much precision.2 Taken as *241a whole, the passage does not seem carefully worded, although it is, to be sure, vaguely suggestive of a very expansive interpretation of § 11(D).
In assigning weight to this remark, one must also look at the overall context. The Department was, of course, the initiator of the underlying antitrust suit, and presumably a prime mover behind the line-of-business restrictions; that it should have sought to work expansive notions into the parol evidence is not surprising. The significance (if any) of its pre-decree statements depends not so much on Justice’s assertions but on AT&T’s apparent failure to respond (appellants have not called our attention to any responses).
Yet the significance even of AT&T’s silence is questionable. First, we do not have a grasp of the total volume of contentions floated at the time; other claims may have seemed more egregious and therefore more demanding of refutation. Second, who was around to do any rebutting? The BOCs were mere embryos within AT&T. To the extent that AT&T managers may have anticipated more of a future with the new AT&T than with the BOCs, they would not have been inclined to attack understandings that would protect the new AT&T from later competition. At a minimum, the AT&T management faced, as we have noted before, “a significant inherent conflict of interest” with respect to these restrictions. See United States v. Western Electric Co., 969 F.2d 1231, 1239 (D.C.Cir.1992). With the context borne in mind, the Justice Department remark seems wholly inadequate to overthrow conventional understandings of the phrase the parties actually used.
The majority then turns to an assertion of the Department of Justice made November 10, 1983, after the decree was adopted (August 24, 1982) and even after the Supreme Court had affirmed the judgment (February 28, 1983). The remark broadly attacks a division of revenues arrangement between a BOC and AT&T, an arrangement proposed by AT&T and resisted by the BOC in question, Bell Atlantic. See 'Maj.Op. at 230-31. It is hard to know what to make of this, but since the BOC was objecting to something AT&T wished to foist upon it, the context is hardly one where BOCs can be said to have “acquiesced” in a particular DOJ claim. At most, we seem to have evidence of the prac-. tice of the parties under the contract — evidence overwhelmed by AT&T’s own conduct in matters far closer to home, namely its agreement to pay royalties to BOCs as a method of reimbursement for development funding. I now turn to that conduct.
AT&T Conduct. AT&T concedes that since adoption of the MFJ there have been what it calls “a few instances” in which it agreed to undertake development work for a BOC if the BOC paid the expenses, in exchange for a reduced purchase price or royalties on sales to other buyers. See J.A. 558-59. So far as appears, these were all entered into without any request for waivers under § VIII(C). One gets some idea of what “a few” means to AT&T when one goes on to read that there were six such agreements (evidently a trivial fraction in AT&T’s eyes) in which the royalty pay-out was not capped at the initial development expenses. Id. at 559.
AT&T leans heavily on this distinction— the cap on royalties — to dismiss the significance of its own conduct. See id. But even for the capped agreements, the BOC must have had — until the cap was reached — the financial stake in sales of the product that AT&T now claims is absolutely forbidden under the MFJ. Further, even if we zeroed in only on the six uncapped transactions, six seems like a substantial run of the parties’ course of conduct. “‘[S]how me what the parties did under the contract and I will show you what the contract means.’ ” Thompson v. Fairleigh, 300 Ky. 144, 187 S.W.2d 812, 816 (1945) (quoting unidentified English judge) (cited in E. Allan Farnsworth, Contracts § 7.13, n. 28 at 535 (2d ed. 1990)). These actions of AT&T, flatly contradicting its current claims, seem far more convincing than a sprawling- sentence once uttered by *242the Department of Justice to address a hypothetical transaction marginally relevant to the present issue.
Finally, AT&T urges us to disregard the six uncapped transactions on the basis of a claim that “after AT&T’s management and counsel became aware of the form of these agreements, a memorandum was sent to AT&T sales personnel in December, 1987,” directing that all such future arrangements should be of the capped variety. J.A. 559. Again, the cap only diminishes the scale of AT&T’s violation of the principle that AT&T and the majority find in the decree. Further, the implicit excuse that AT&T’s “management and counsel” were wholly unaware of these transactions is no help. If the contracts were binding (and there is no assertion to the contrary), they must have been entered into by persons with adequate authority. They are acts of AT&T.
* * *
Although there seems only the weakest formal ease for reading “affiliated enterprise” to encompass any firm sharing revenue with a BOC, it makes sense to see whether such arrangements so clearly impinge on the pro-competitive purposes of the MFJ that the wrench of the language deserves serious consideration.
The majority identifies three ways in which a royalty arrangement of the kind agreed upon by Ameritech and David Systems might imperil the antitrust objectives of the decree. As a result of the agreement the BOC could (1) grant the funded manufacturer privileged access to its technical requirements or adopt preferential standards; (2) engage in cross-subsidization, paying inflated prices on its own purchases, thus enabling the funded manufacturer to “undersell its competitors and gain power in other markets”; and (3) buy from the funded manufacturer even if the price/quality relationship of the manufacturer’s product was inferior to its competitors’. Maj.Op. at 232-33.
Of course it is true that a BOC could act in these ways with respect to a funded manufacturer. That alone is surely not enough — a BOC could act that way towards any firm. The question would seem to be whether the funding/royalty relationship is likely to create such strong incentives to engage in this behavior, and with such serious likelihood of anticompetitive impacts, that we should regard the funding relationship as substantially akin to garden-variety affiliation.
Let us first take cross-subsidization, the most concrete of the hazards, and, in fact, the template for the other two. The feared result — “gain[ing] power in other markets” — of course cannot inflict an antitrust injury unless the BOC and funded company can overcome all the conventional hazards to successful predatory pricing; the prospects of driving competitors out, and the hurdles to any new entry, must be so great that the present discounted value of the hypothetical future monopoly overcharges exceeds the present discounted value of the guaranteed upfront losses. See Brook Group v. Brown & Williamson Tobacco Corp., — U.S.-,- -, 113 S.Ct. 2578, 2588-89, 125 L.Ed.2d 168 (1993). As a result, “predatory pricing schemes are rarely tried, and even more rarely successful”. Id. at --, 113 S.Ct. at 2589 (citations omitted). We must assume, of course, that a BOC may be able to pass the cost of overpayments to its “affiliate” forward to customers as a cost of business under its regulated rates, see National Rural Telecom Ass’n v. FCC, 988 F.2d 174,178, 179-80 (D.C.Cir.1993), and that therefore the BOC could fund the predation at a lower real cost than could a firm in an unregulated market — could fund it with, as it were, “free” money. Thus, the price-regulated environment makes the prospect of cross-subsidization and predation far more plausible than normally.
The question, though, is how conducting such an operation through a funded independent manufacturer, which simply owes the BOC a royalty on sales of the funded product, compares with doing so through a genuine affiliate. The answer is that it is a rather feeble substitute. Although overpayments to the funded firm may be “free” to the BOC, the money is by no means free to the funded entity. Once in the fundee’s hands, the money is its own, so that investing it in a scheme of predation is just as costly for the fundee as for any firm in any ordinary market not subject to price regulation. The absence of *243either BOC ownership or control, or the ownership or control of both entities by a com- ■ mon parent, thus severs precisely the link that made cross-subsidization and predation more serious risks in this economic environment.
Of course the BOC could seek to enter into side agreements with the fundee, committing it to use the funds for predation. But the absence of control, which we must assume under the majority analysis, clearly increases .the hazards of such a conspiracy; the number of people necessarily brought in increases, and there are at least two chains of command to be silenced rather than one. The district court recognized this distinction at the time it approved the decree: .“Anti-competitive activities undertaken by two separate corporations rather than by two components of the same corporation are likely to be far more difficult to accomplish because of increased problems of coordination and the greater possibility of detection.” MFJ Opinion, 552 F.Supp. at 191. Further, as a competitor can probably survive in these fields only by selling to several operating companies,3 the funded entity will likely have many customers other than the funding BOC, making it harder for the BOC to inflict effective punishment for “cheating”. .
The two other concerns identified by the majority suffer from the same basic problem. Preferring the fundee despite an inferior price/quality relation seems just an intricate way of overpaying it. Again, the absence of ownership or control of the fundee, or ownership of both entities by a common parent, deprives .the BOC of the ability to assure that these overpayments, perhaps “free” from its perspective but surely not from that of the fundee, will be applied in accordance with its purpose rather than the fundee’s. Similarly, favoring the fundee with advance technical information or with discriminatory technical standards also appears to be simply a complicated way of shifting value to the fundee at the expense of the BOC (or, by the assumption we are indulging here, the BOC’s customers). Without control over the fun-dee’s use of the profits that derive from this advantage, the BOC is in a weak position to achieve its goals.
Again, I cannot argue that the funding- and-royalty arrangement is absolutely without antitrust risk; probably nothing is without such risk. What seems plain to me is that whatever risks exist in that context are trivial compared with those that the drafters assumed applicable to entry by a BOC into the forbidden lines of business via a conventional affiliate.
In fact, funding/royalty arrangements are likely to enhance competition in telecommunications products by providing a new source of funding for smaller companies with innovative ideas. BOCs have a comparative advantage in judging the prospects for investments in research and development of products complementary, to their business, and an obvious interest in ensuring that such innovation occurs. They thus can diminish the imperfection of financial markets due to normal lenders’ lack of information about the market and the technology. The funding/royalty arrangement increases the likelihood of such financial assistance, for it enables the BOC to commit capital in a form that entitles it to share in the high returns on very successful projects, just as a wildcatter arranges to share in the rare success among exploratory oil and gas wells. Similarly, just as a wildcatter assembles leases in the area of intended exploration so as to capture as much as possible of the value of the information that a successful well will yield (and to prevent free riding by others), so a BOC taking substantial risk on a new technology would want to diminish free-riding by other buyers, which is precisely what the royalty arrangement permits.
The BOC’s investment, to be sure, carries a marginal anticompetitive risk. A BOC may *244persist in dealing with a funded firm longer than it would with a totally disconnected one; as with any lender, the hope that a little more indulgence will save the project will weigh against the advantages of cutting its losses. This is a far cry, however, from the deliberate cross-subsidization and predation, which, as I argued above, is pertinent in the case of a genuine affiliate but is rendered highly unlikely here by the fundee’s independent interest in any revenues once it has received them from the BOC.
The majority’s methodology is somewhat unclear to me. Once it abandons what it correctly identifies as the “usual” understanding of affiliation, see Maj.Op. at 230-31, it turns, unguided by any contract language, to parol evidence and postdecree assertions of the Department of Justice. These include, as we have seen, direct or indirect references .(1) to revenue-sharing with a firm in a forbidden line of business and (2) to arrangements giving a BOC “any stake” in the success of such a firm. The majority does not commit itself as between these two formulae. If revenue-sharing is forbidden because, it involves a BOC “stake” in the funded enterprise, or because of the anticompetitive hazards sketched by the majority, then any loan is equally forbidden, and a variety of long-term arrangements such as requirements contracts are at risk. A decision embracing so radical an interpretation should confront its implications. On the other hand, if revenue-sharing is singled out from other arrangements by which a BOC might have a stake in a funded entity’s success in a forbidden line of business, then we should learn just what the analytic distinction is.
For purposes of this case, all that is needed is that we recognize that § II(D)’s reference to “affiliated entity” is within the conventional range of affiliation, and thus requires ownership or control by the BOC, or the ownership of both entities by a common parent. As the David Systems transaction is outside that range, we need not settle the narrow dispute as to whether § II(D)’s reference to affiliation simply incorporates § IV(C)’s definition or calls for a slightly broader one such as that advanced by the Department of Justice4, which appears here in support of Ameritech’s view that the transaction is outside § 11(D).
This case is full of ironies. The first, of course, is that AT&T is hereby enabled to' use the line-of-business restrictions, adopted in the name of pro-competitive purposes, to stifle competition from small firms that might enter the telecommunications products markets as a result of BOC funding, thus protecting, for example, its 49%' share of sales of central office switches. Further, it is only because of AT&T’s role as a successor party to the MFJ that it is able to press its claim here. United States v. Western Electric Co., 969 F.2d 1231, 1237-41 (D.C.Cir.1992). As AT&T would gain in higher prices from any reduction in competition, unless it were driven from the field, it would have “antitrust standing” only if it were able to make a showing that the feared predation has some likelihood of eradicating AT&T itself as a competitor, see Matsushita Elec. Industrial Co. v. Zenith Radio Corp., 475 U.S. 574, 588-93, 106 S.Ct. 1348, 1356-59, 89 L.Ed.2d 538 (1986), a laughably implausible scenario.
A second irony is that it is now seriously asserted in some quarters that the nature of the long distance and local telephone markets may be precisely the opposite of what was assumed at the adoption of the MFJ. That assumption, of course, was that the technology of the long distance market would be radio, that its unit costs would not decline with volume, and that it therefore would be competitive; the assumption for local exchanges was that their technology would be wire, that unit costs would decline throughout the relevant market, and that therefore it would be a natural monopoly. To the extent that glass fiber is replacing radio for long distance, and cellular radio emerges as the optimal technology in the local loop, these *245premises are reversed. See Peter W. Huber, “Telephones, Competition, and the Candice-Coated Monopoly”, Regulation (1993 No. 2) 34-43. Of course no record has been made as to any such matters, and no attempt to amend the decree is before us. Such contentions tend, however, to undermine the majority’s implicit assumption that there is some overwhelming economic need to reach out and expand the decree’s line-of-business restrictions.
In any event, our job is to construe the agreement as written. As I can find.no basis at all for conceiving “affiliated enterprises” to include funded royalty payors, either in the decree’s language, the tradition, of antitrust consent decrees, the skimpy parol evidence, the practice of the parties, .or the overall purposes of the MFJ, I respectfully dissent.
. LEXIS Library: Genfed, File: Fedreg, Search Request: "competitive impact statement” and affiliat!
. The AT&T brief goes on to quote a remark that the decree prevents "the reemergence of any division of revenues procedures to supplant the exchange access tariff provisions." Competitive Impact Statement in Connection with Proposed Modification of Final Judgment, 47 Fed.Reg. 7170, 7178 (Feb. 17, 1982) (cited in AT&T Br. at 11-12). Because of the statement's focus on *241supplanting access pursuant to tariffs, it evidently rests on § 11(A) of the MFJ, requiring that such access be “on an unbundled, tariffed basis”, see 552 F.Supp. at 227, and the majority does not rely on it.
. See Peter Huber, The Geodesic Network: 1987 Report on Competition in the Telephone Industry, 14.8 (1987) (reporting very large economies of scale in development of switches, with prospect the 18 firms manufacturing digital switches worldwide in 1984 will likely to fall to fewer than a dozen); see also Comments of David Systems, Inc., in Support of Ameritech’s Revised Request for a Waiver to Allow the Receipt of Royalties On Third-Party Sales of Telecommunications Products, 5 (DOJ June 30, 1988) (Joint Appendix 408) ("no one buyer has a large enough share of the market to make it economically attractive to produce a product for that one buyer alone”).
. The Department of Justice urges us. to read "affiliated enterprise" in § 11(D) to include entities in which a BOC has a more than.de minimis equity interest (5% or more) or exercises substantial management control. Brief for Appellant United States of America at 2.