Graphic Products Distributors, Inc. v. Itek Corporation, Individually and D/B/A Itek Graphic Products, Defendants

This is an antitrust case involving non-price, vertical restraints on trade. The defendant, Itek Corporation (Itek), appeals from a jury verdict in favor of plaintiff Graphic Products Distributors, Inc., (GPD) in a suit brought under, inter alia, Section 1 of the Sherman Act, 15 U.S.C. § 1 (1976), and Section 4 of the Clayton Act, 15 U.S.C. § 15 (1976 & Supp. V 1981). Itek contends on appeal that the district court should have granted it a directed verdict or judgment notwithstanding the verdict (n.o.v.) because (1) the evidence was insufficient to establish that its distribution system was an unreasonable restraint of trade; and (2) the evidence was insufficient to establish the amount of GPD’s damages. It also contends that the district court should have granted it a new trial because the jury instructions misstated the applicable law. We affirm.

*1564 I.

Although we conclude that the evidence was sufficient to create jury questions regarding both liability and damages, we do so not without some reluctance. The evi-dentiary record is neither as extensive nor as precise as one would expect in a well-tried antitrust case. Most important, the history of Itek and evidence of its need to utilize the challenged vertical restraints to compete in the marketplace were not fully presented to the district court.

Itek’s Graphic Products Division manufactures graphic equipment and supplies for the national graphic arts market. Its product line includes offset platemaking machines (platemakers), duplicators, camera processors, and microfilm equipment. 1 Itek sells supplies to accompany these products, and provides the needed servicing as well. It had annual revenues of some $85,000,000 in 1977, the only year for which we have this data. Approximately one-third of those sales were in equipment, the rest in service and supplies. The record does not indicate what percentage of this revenue derived from each of Itek’s product lines.

Prior to 1975, Itek distributed its equipment and supplies exclusively through its own sales organization. It had twenty-two direct sales or branch offices, concentrated in major urban areas with large potential markets. In order to increase sales in the areas outside those major urban centers already relatively well-covered by the branch offices, Itek decided to switch to a dual distribution system in the period 1975-76.

Under this dual system, Itek confined its branch offices’ direct sales activity to within a 50-mile radius of the cities in which they were located. Using marketing areas designated by the Business Equipment Manufacturers Association (BEMA), Itek divided up the rest of the country and sought independent distributors for these remaining areas. BEMA areas assigned to a distributor did not overlap with those of any other distributor or any branch office. At the time of this litigation, there were approximately 30 such independent distributors.

We emphasize that the record does not indicate the extent of Itek’s sales activities in the areas outside the major urban centers before or after the institution of the dual distribution system. The only evidence bearing on this point indicates that, prior to the implementation of the distributorship program, the branch office personnel would make infrequent trips to these areas. Nor does the record reveal the extent to which Itek’s competitors were selling in these secondary markets. ' We do know, however, that all of Itek’s previous efforts to appoint distributors had been with outside graphic equipment dealers. We do not know precisely what these existing graphic equipment dealers were selling, i.e., whether they were selling Itek-type products.

Other than for the territory ultimately assigned to GPD, the record does not disclose the extent of the potential market in these outlying areas for Itek’s (or its competitors’) products; nor does it reveal the extent to which that potential had been exploited. With respect to GPD’s assigned territory, the record indicates that Itek had realized only 10-13 percent of the potential market for platemakers, and only 1-2 percent of the potential market for other graphic products. 2

Pursuant to this distributorship program, in 1975 an Itek representative approached Anthony Zatzos, a long-time Itek salesman, and asked if he would be interested in a distributorship. After several months of *1565 negotiation, Zatzos formed GPD, and in July 1975 it received a distributorship covering seven BEMA areas in Georgia and South Carolina. The distributorship agreement between Itek and GPD provided as follows:

1. GRANT OF DISTRIBUTORSHIP

[Itek] hereby grants to [GPD] and [GPD] accepts, a non-exclusive, non-transferrable Distributorship to purchase for resale and to service the [Itek] products and equipment specified in the attached Schedule “A” hereinafter referred to as “Products” in the area hereinafter referred to as the “Territory.” [GPD] represents that [it] is actively engaged in the business of Graphic Equipment and supplies, Sales & Services and maintains 2 outside fulltime salesmen who regularly call on prospects and customers in the Territory hereinafter defined and maintains 2 outside fulltime Technical Service personnel in the same area.

GPD began operations in September 1975, with Zatzos as its president and two other former Itek employees as the other principals; it distributed the full line of previously described Itek products in the seven BEMA areas.

GPD made 90% of its sales within its assigned area, but 10% of its sales occurred within the territory of Itek’s Atlanta branch office. The Atlanta branch manager complained to Itek’s management about these sales, as well as a sale GPD made in Columbus, Georgia, to a customer within the territory of an Alabama-based distributor. Pursuant to a clause of the distributorship agreement providing for termination at will by either party upon 90 days written notice, Itek notified GPD on June 10, 1976 of its intent to terminate the agreement.

GPD brought this suit in May 1977, alleging violations of federal and state antitrust law. The complaint alleged that pursuant to a contract, combination or conspiracy to restrain trade, Itek compelled its distributors, as a condition of doing business with them, to enter into agreements restricting them to reselling its products only within a designated geographical territory, and only to customers within that territory. The complaint further alleged that Itek terminated GPD pursuant to its conspiracy to maintain these territorial and customer restrictions. GPD also asserted that these trade restraints were for the purpose of fixing prices.

In narrowing the issues for trial, the district court granted Itek’s motion for summary judgment on GPD’s price-fixing claim and on Itek’s permissive counterclaim for a debt. Neither of these rulings are presented to us for review. The court also rejected GPD’s claim that Itek’s dual distribution system was a horizontal restraint of trade 3 and should be considered under the per se rule. Again, this finding is not before us for review.

The ease was tried before a jury in May 1981. Much, if not most, of the testimony at trial was devoted to a dispute about the reasons for GPD’s termination. At the close of GPD’s ease and at the close of all the evidence, Itek moved for a directed verdict. The court denied these motions. The jury returned a special verdict pursuant to Fed.R.Civ.P. 49(a). The jury specifically found: (1) Itek entered into a conspiracy, contract or combination with individuals or companies other than GPD the purpose of which was to restrain interstate trade or commerce; (2) the territorial restraints placed upon GPD by Itek were unreasonable and without valid business justification; (3) the motivating reason for the termination of GPD was GPD’s violation of the territorial restrictions; (4) GPD was injured in fact as the direct and proxi *1566 mate result of the restraint placed upon it by Itek; (5) GPD suffered actual damage as a sole consequence of such injury. The jury awarded GPD $200,000 in damages. 4

Itek moved for judgment n.o.v. and, alternatively, for a new trial on the grounds that: (1) GPD failed to show the challenged restraint had an anticompetitive impact in the relevant market; and (2) GPD failed to prove the amount of its damages. 5 The district court summarily denied the motion and, pursuant to Section 4 of the Clayton Act, 15 U.S.C. § 15, tripled the jury’s damages award and granted reasonable attorney’s fees.

II.

At the outset, we must emphasize the limited character of our review. Itek’s challenge to the court’s failure to grant it a directed verdict or judgment n.o.v. is based on the insufficiency of the evidence to establish: (1) that the alleged restraints were unreasonable; and (2) the amount of GPD’s net economic loss. In this posture, we consider all of the evidence in the light and with all reasonable inferences most favorable to the party opposed to the motion. “If the facts and inferences point so strongly and overwhelmingly in favor of one party that the Court believes that reasonable men could not arrive at a contrary verdict, granting of the motions [for directed verdict or judgment n.o.v.] is proper.” Boeing Co. v. Shipman, 411 F.2d 365, 374 (5th Cir.1969). 6 Thus, “[w]e must uphold the district court’s judgment unless ‘there is a complete absence of substantial probative facts to support the jury’s verdict.’ ” Copper Liquor, Inc. v. Adolph Coors Co., 624 F.2d 575, 579-80 (5th Cir.1980) (quoting Lavender v. Kurn, 327 U.S. 645, 653, 66 S.Ct. 740, 744, 90 L.Ed. 916 (1946)).

Section 1 of the Sherman Act literally prohibits every contract, combination or conspiracy in restraint of trade, but this has long been confined to only such concerted activity as unreasonably restrains trade. See, e.g., Standard Oil Co. v. United States, 221 U.S. 1, 31 S.Ct. 502, 55 L.Ed. 619 (1911). 7 Itek’s basic argument is that the evidence was insufficient to show that its dual distribution system, complete with territorial restraints, was an unreasonable restraint of trade. We analyze this contention within the framework set out by the Supreme Court in Continental T.V., Inc. v. GTE Sylvania Inc., 433 U.S. 36, 97 S.Ct. 2549, 53 L.Ed.2d 568 (1977).

In Sylvania, the High Court overruled United States v. Arnold, Schwinn & Co., 388 U.S. 365, 87 S.Ct. 1856, 18 L.Ed.2d 1249 (1967). Schwinn involved a complex, three- *1567 level distribution system with a variety of territorial and customer restraints. The Schwinn Court began by stating “we are remitted to an appraisal of the market impact of these [restrictive] practices[,]” id at 373, 87 S.Ct. at 1862, but it failed to use economic analysis in distinguishing between those restrictive practices it condemned under the per se rule and those it found worthy of “rule of reason” treatment. Instead, it elevated to prime importance a purely legal distinction between “the situation where the manufacturer parts with title, dominion, or risk with respect to the article, and where he completely retains ownership and risk of loss.” Id at 378-79, 87 S.Ct. at 1865.

The Schwinn Court held that where a manufacturer sells products to a wholesale distributor subject to territorial or customer restrictions upon resale, or places similar restraints upon the retailers to which the goods are sold, its conduct is per se unreasonable. It did not analyze market impact; it merely concluded “[s]uch restraints are so obviously destructive of competition that their mere existence is enough.” Id at 379, 87 S.Ct. at 1865. On the other hand, where the manufacturer retains ownership and the risk of loss, the more flexible rule of reason would apply. The Court then upheld that portion of the lower court’s opinion which found territorial allocation and franchising of retailers reasonable where agency or consignment arrangements alone were involved, and title to the goods did not pass.

The Sylvania Court noted that the Schwinn opinion did not distinguish among the challenged restrictions on the basis of their respective competitive effects, but instead only on the basis of whether title to the goods had passed. 433 U.S. at 52, 97 S.Ct. at 2558. It pointed out that the opinion lacked “even an assertion ... that the competitive impact of vertical restrictions is significantly affected by the form of the transaction.” Id at 54, 97 S.Ct. at 2559. Because “an antitrust policy divorced from market considerations would lack any objective benchmarks[,]” the legality of particular vertical restrictions must henceforth be judged by their competitive impact. Id at 53, n. 21, 97 S.Ct. at 2559 n. 21. Finding that the form of the transaction by which a manufacturer conveys its products to its dealer does not affect the competitive impact of the vertical restrictions, the Court abandoned that basis for the per se rule of Schwinn.

Turning to an analysis of the market impact of vertical restrictions, the High Court noted that they have a “potential for a simultaneous reduction of intrabrand competition and stimulation of interbrand competition.” 8 433 U.S. at 51-52, 97 S.Ct. at 2558. Without dwelling on their potential for reducing intrabrand competition, the Court discussed three of the ways that vertical restrictions can be wielded by a manufacturer to enhance its ability to compete interbrand: (1) they can be used by new manufacturers or manufacturers' entering new markets to induce competent and aggressive retailers to invest large amounts of capital and labor in the distribution of products unknown to the consumer (the “market access” rationale); id at 55, 97 S.Ct. at 2560; (2) they can be used by established manufacturers to induce retailers to spend money on promotional or servicing activities necessary to market their goods efficiently. These services are likely to increase sales and therefore it would appear to be in the self-interest of each retailer to provide them. However, individual retailers might fail to provide them in optimal amounts if other dealers could take a “free ride” on these services by not providing them, selling at a lower price, and at the same time enjoying the competitive benefits to the brand flowing from the providing dealer’s expenditures (the “dealer services-free-rider” rationale); id; (3) they can be used to ensure product quality and safety, in accordance with products liability *1568 and consumer warranty law. Id. at 55 n. 23, 97 S.Ct. at 2560 n. 23. 9

Finding substantial authority supporting their economic utility, the Court concluded that the per se rule against vertical restraints stated in Schwinn must be overruled and the rule of reason employed instead. Since the Court of Appeals in Sylvania had distinguished Schwinn (a course the Supreme Court rejected in favor of overruling it), the High Court had no occasion to apply the rule of reason to the facts before it. It emphasized, however, that although particular applications of vertical restrictions might justify per se treatment, any future “departure from the rule-of-reason standard must be based upon demonstrable economic effect rather than — as in Schwinn —upon formalistic line drawing.” 433 U.S. at 58-59, 97 S.Ct. at 2562.

Sylvania places the competitive effects of particular vertical restraints at the center of the analysis under the rule of reason. Accord Muenster Butane, Inc. v. Stewart Co., 651 F.2d 292, 295, 296 (5th Cir.1981); Red Diamond Supply, Inc. v. Liquid Carbonic Corp., 637 F.2d 1001, 1005 (5th Cir. Unit A), cert. denied, 454 U.S. 827, 102 S.Ct. 119, 70 L.Ed.2d 102 (1981). However, it provides us with little guidance as to how the rule should be applied. 10 We have narrowed the broad-ranging inquiry called for by the rule of reason by insisting, at the threshold, that a plaintiff attacking vertical restrictions establish the market power of the defendant:

[a] requirement that plaintiff prove market power in this case would have saved the litigants and the courts much expense. [Defendant] had no market power in [the relevant geographic market]. The market was highly competitive. Whatever vertical restraints [defendant] imposed on its dealers, their effect could not have been to raise the price consumers paid for [the relevant product].”

*1569 Muenster Butane, 651 F.2d at 298. 11

In order to establish the defendant’s market power, a plaintiff must first offer proof of “a well-defined relevant market upon which the challenged anticompetitive actions would have had a substantial impact.” Cornwell Quality Tools Co. v. C.T.S. Co., 446 F.2d 825, 829 (9th Cir.1971), cert. denied, 404 U.S. 1049, 92 S.Ct. 715, 716, 30 L.Ed.2d 740 (1972). The relevant product and geographic market is a question of fact, and findings concerning the market should be overturned on appeal only if clearly erroneous or where there is no evidence to support the finding below. Associated Radio Service Co. v. Page Airways, Inc., 624 F.2d 1342, 1348-49 (5th Cir.1980), cert. denied, 450 U.S. 1030, 101 S.Ct. 1740, 68 L.Ed.2d 226 (1981).

In this case, the court did not submit the issue of the relevant product and geographic market to the jury. Under Fed.R. Civ.P. 49(a), if the court omits any issue of fact raised by the pleadings or by the evidence, the court may make a finding itself or, if it fails to do so, it shall be deemed to have made a finding consistent with the judgment on the special verdict.

Concerning the relevant geographic market, we observe that GPD did not challenge its distributorship termination alone as illegal; it challenged its termination pursuant to a nationwide distribution system that it claimed unreasonably restrained trade. Under this system, Itek divided the nation into exclusive territories, and there was testimony indicating nationwide compliance with these restrictions. Moreover, Itek enforced a policy against competitive bidding on governmental orders nationally. Finally, much of the testimony concerning Itek’s share of the market referred to its national market share, and GPD’s principal witness, Zatzos, testified about the national competitive effects of Itek’s alleged product superiority. There was ample evidence to support a finding that the relevant geographic market was national.

The record is not so unequivocal with respect to the relevant product market. GPD clearly tried this case to the jury on the theory that the relevant product market was platemakers, despite the fact that the record indicated that Itek, and GPD, also sold duplicators, camera processors, and microfilm equipment. 12 All the market share data pertained exclusively to platemakers. Zatzos testified that Itek’s primary product line consisted of five models of automated platemakers; he emphasized this point throughout his testimony. He added that Itek sold imported duplicators, microfilm equipment and camera processors.

Itek never challenged this characterization of the relevant product market. 13 Don Malagamba, Itek’s Vice President for United States Sales Operations, testified at length for the defendant but did not challenge Zatzos’ claim that platemakers were Itek’s major product line. His testimony largely reiterated that of Zatzos, stating *1570 that Itek sold platemakers, duplicators, and camera processors. If data concerning the portions of Itek’s total sales deriving from each product and its “after-markets” for supplies and service would have refuted GPD’s claims, Malagamba surely would have enjoyed ready access to it. Malagam-ba gave no such testimony. Considering the evidence in the light and with all reasonable inferences favoring GPD, as we must do in our present posture, we believe the evidence supported a finding that the relevant product market was platemakers. We now consider Itek’s market power in the relevant market.

Market power is the ability to raise price significantly above the competitive level without losing all of one’s business. Valley Liquors, Inc. v. Renfield Importers, Ltd., 678 F.2d 742, 745 (7th Cir.1982). Market share is frequently used in litigation as a surrogate for market power for two reasons. First, market power is conceptually difficult to define in any given case. Second, its measurement requires sophisticated econometric analysis. Therefore, market power is not well suited to presentation in an adversary proceeding. See Id. Moreover, market share directly relates to the effectiveness of interbrand competition in minimizing the anticompetitive effects of a restraint on intrabrand competition. Vertical Restrictions Limiting Intrabrand Competition 63 (ABA Antitrust Section Monograph No. 2 1977) (hereinafter cited as ABA Monograph); see Sylvania, 433 U.S. at 52 n. 19, 97 S.Ct. at 2558 n. 19.

Itek enjoyed significant market power in the national platemaker market in 1975-76 when it was implementing the challenged dual distribution system. Zatzos testified that Itek enjoyed 70 to 75 percent of the national platemaker market during that period. Malagamba agreed that, despite some local variations, Itek had “on the average” approximately 70 percent of the national market. Samuel Reeder, another Itek Vice President, conceded that its strongest competitor in the platemaker field — A.B. Dick — had only 10 percent of the market, and several other firms divided the remaining 20 percent. The evidence indicated, moreover, that Itek designed its platemakers so that the supplies of its competitors could not be used therewith; thus, Itek equipment purchasers were locked into buying Itek supplies. 14 Finally, John Fountain, a principal of GPD specializing in the service area, indicated that there was little or no interbrand competition in the servicing of platemakers because of warranty prohibitions and the inability to service another company’s machines.

In addition to establishing Itek’s large market share, GPD introduced substantial evidence pertaining to another major criterion of market power: product differentiation. 15 Zatzos and James Boswell, the third GPD principal, both testified that the Itek platemaker was a superior product to any then on the market. Zatzos stated “[t]here were other manufacturers that sold a piece of equipment that was designed to do the same type of job, but the operation of the equipment, the design of the equipment just wasn’t even a factor in the market place.” Joseph Crabtree, former Itek Atlanta District Manager, corroborated this testimony. Reeder confirmed that Itek was more often successful than not in writing the specifications for machinery on which governments would accept bids, thereby ensuring their success in the market for *1571 government orders. 16 We are convinced that GPD satisfied its threshold burden of showing Itek had substantial market power when it imposed vertical restraints on its distributors. 17

Having crossed this threshold, GPD was required to take its “first step”: “show [an] anticompetitive effect, either in the intrabrand or interbrand markets.” H & B Equipment Co. v. International Harvester Co., 577 F.2d 239, 246 (5th Cir.1978). We emphasize, however, that demonstrating an anticompetitive effect is but the first step of what must be a systematic comparison of the negative effects of the restraint on intrabrand competition and interbrand competition, if any, with any alleged positive effects on interbrand competition stemming from the restraints. This is the critical analysis required by Sylvania. Muenster Butane, 651 F.2d at 296. 18 Accord Donald B. Rice Tire Co. v. Michelin Tire Corp., 483 F.Supp. 750, 760 (D.Md.1980), aff’d, 638 F.2d 15 (4th Cir.1981).

We note first that a vertical restraint on trade, almost by definition, involves some reduction in intrabrand competition. When a manufacturer restricts a dealer to selling only within a certain territory, or only to certain customers, or only from certain locations, it is necessarily restraining intrab-rand competition. However, this may or may not have a negative effect on the welfare of the consumer. “[A]bstract lessening of intrabrand competition is not enough [to establish a cause of action for violation of the antitrust laws].” Aladdin Oil Co. v. Texaco, Inc., 603 F.2d 1107, 1116 (5th Cir.1979). The effects of a restraint of intrabrand competition on consumer welfare cannot be viewed in isolation from the interbrand market structure. 19 A restriction of intrabrand competition may — depending on the interbrand market structure — either enhance or diminish overall competition, and hence consumer welfare. As one commentator puts it, “the real issue remains whether the loss of intrabrand competition itself injures or benefits potential consumers of the brand in question. Those consumers are injured by the restraint if, without obtaining more services, they are denied intrabrand choices that are sources of consumer welfare, but are benefited [notwithstanding the restraint] if dealer services increase.” Gerhart, The “Competitive Advantages” Explanation For Intrabrand Restraints: An Antitrust Analysis, 1981 Duke L.J. 417, 439. Moreover, if *1572 enhanced dealer services to the consumer result from the restraint, interbrand competition should be sharpened. 20

The second implication of the Sylvania analysis is that even if a negative effect on consumer welfare and competition can be shown to flow from a restriction of intrab-rand competition, the court must still look to any possible pro-competitive effects on the interbrand market stemming from the *1573 intrabrand restriction. The Sylvania court stressed increased market access for the manufacturer, increased dealer services to the consumer, and product safety and quality as examples, but clearly this list was intended to be illustrative rather than exhaustive. There are a wide variety of pro-competitive rationales for vertical restrictions on intrabrand competition. 21

To require an antitrust plaintiff to conjure up every possible pro-competitive rationale for a vertical restraint, and prove their inapplicability to the restraint in question, would be to impose an insurmountable burden. It would, in effect, convert the rule of Sylvania — that pro-competitive interbrand effects of vertical restraints must be considered — into a rule of per se legality for intrabrand restraints. Sylvania clearly does not extend that far, and we decline to so extend it on the facts of the case before us. 22

However, “[t]he burden of proving unreasonable effects [in a rule of reason case] rests with the antitrust plaintiff.” Kentucky Fried Chicken Corp. v. Diversified Packaging Corp., 549 F.2d 368, 380 (5th Cir.1977). Therefore, GPD, after crossing the threshold of showing Itek’s market power, was required to establish that the interbrand market structure was such that intrabrand competition was a critical source of competitive pressure on price, and hence of consumer welfare. GPD was also required to show that the nature and effects of the restraint were such as to be “substantially adverse” to market competition. United States v. Arnold, Schwinn & Co., 388 U.S. at 375, 87 S.Ct. at 1863, overruled on other grounds, Sylvania, 433 U.S. at 57-58, 97 S.Ct. at 2561-62. 23 Of course, Itek was free to come forward with a showing that, notwithstanding its market power, the vertical restraints were reasonably necessary to achieve legitimate, pro-competitive purposes, and that the purpose and effect of the restraints were pro-competitive when viewed as a whole. 24 We now apply this analytical framework to the present case by analyzing first the effect of the intrabrand restraints on consumer welfare, in light of the interbrand market structure. We then analyze any possible pro-competitive effects on interbrand competition.

GPD presented considerable evidence bearing on Itek’s anti-competitive intent in imposing territorial restrictions. Evidence of intent is highly probative “not because a good intention will save an otherwise objectionable regulation or the reverse; but because knowledge of intent may help the court to interpret facts and to predict consequences.” Chicago Board of Trade v. United States, 246 U.S. 231, 238, 38 S.Ct. 242, 244, 62 L.Ed. 683 (1918).

Lee Hollingsworth, Itek’s former sales manager for special accounts, stated that Itek’s purpose in assigning exclusive territories to branch offices and distributors was to have no competition between them. In a memorandum to all distributors explaining *1574 a system of rules apportioning bids on state government orders between distributors and branches, Itek stated that their purpose was to avoid a distributor and an Itek branch office, two distributors or two branches competing with each other for the same order. The impetus for this anti-competitive bidding system was GPD’s “low-bidding” of Itek’s Atlanta branch office on a platemaker sale to the Georgia Department of Education. The regional manager wrote Don Malagamba stating that this territorial violation “could impact on direct sales with wide-spread adversity ... The dealer obviously was willing to risk all relationships with the Company in order to ‘grab’ the total order credit ... With increasing dealer activity throughout the South, such conflict with direct and indirect efforts must be avoided.”

GPD also emphasized at trial the “airtight” or absolute nature of the territorial restraint system. 25 Itek Vice President Reeder admitted that this marketing system, if followed, completely eliminated in-trabrand competition, and Itek did not contest this point at trial. Clearly, complete elimination of intrabrand competition differs qualitatively from merely lessening it, if intrabrand competition is of proven importance to consumer welfare in the circumstances of the case under consideration. 26 We now consider GPD’s evidence about the anticompetitive effects of the restrictions, bearing in mind the evidence concerning Itek’s market power.

GPD brought forth evidence that its “extra-territorial” activities in the area assigned to Itek’s Atlanta branch office were highly competitive with that office. Zatzos testified that GPD substantially underbid the Atlanta branch office on a contract to provide the Georgia Education Department with a platemaker, and that this caused consternation among Itek officials. 27 The jury found, and Itek does not contest on appeal, that territorial violations such as this were the reason for GPD’s termination. Zatzos testified that on at least two other occasions after GPD’s termination it underbid Itek in competition for orders and received the contract.

Zatzos also testified that GPD charged only $19.00 per hour for servicing Itek machines, in contrast to $32.00 or $32.50 hourly rates charged by Itek. He stated that Itek complained about this pricing of service during the time GPD retained its distributorship. 28 Given the absence of interbrand *1575 service competition, such intrabrand competition could have been important to consumer welfare. This possibility was borne out by the testimony of then-Itek President Preschlack: “[S]ervice was ten to twelve percent of the revenue generated by our field organization. It was the highest profit margin part of our sales, and it was a very important contributor of our profit.” Clearly, the jury could have inferred that elimination of GPD as a competitor would have eliminated competitive pressure on the price of service as well as on the price of equipment. 29 See Eiberger v. Sony Corp., 622 F.2d 1068, 1080 (2d Cir.1980).

This intrabrand competition existed in the context of Itek’s considerable interbrand market power. We have discussed the importance of market share in minimizing the anticompetitive effects of a restraint of intrabrand competition, supra at 1569-1570. The product differentiation achieved by Itek, owing to the superiority of its platemakers, also suggests that intrabrand competition was an important source of consumer welfare: “significant product differentiation increases somewhat the importance of intrabrand competition between distributors and increases correspondingly the required justification for abolishing it.” Sandura Co. v. FTC, 339 F.2d 847, 857 (6th Cir.1964); accord ABA Monograph, 64-65 (“The greater the product differentiation, the lesser the degree to which interbrand competition will be effective.”); Gerhart, supra, 1981 Duke L.J. at 442 (“One must determine whether a firm — because of product differentiation or some other advantage over its rivals — has a degree of discretionary power that makes consumer choice between outlets selling that firm’s product an important source of consumer welfare.”)

We are convinced that there was sufficient evidence for the jury to find that intrabrand competition, in this context, was an important source of competitive pressure on price, and tha