Opinion
MDL No. 369.
October 29, 1981.
Stephen R. Steinberg, John A. Lowe, Reavis McGrath, New York City, Marshall G. Mintz, Reavis McGrath, Henry Shields, Jr., Harold E. Kruth, Irell Manella, Los Angeles, Cal., Robert B. Pringle, Joseph A. Darrell, Thelen, Marrin, Johnson Bridges, San Francisco, Cal., for Data General Corp.
Jack E. Brown, Paul E. Eckstein, Brown Bain, P. A., Phoenix, Ariz., Russell L. Johnson, Pillsbury, Madison Sutro, San Francisco, Cal., for Fairchild Camera and Instrument Corp.
Jack C. Provine, M. Janice Smith, Miller, Starr Regalia, Oakland, Cal., for Digidyne Corp.
OPINION AND ORDER
In this exceedingly difficult and complex antitrust litigation involving the legality of certain tying arrangements utilized by defendant, Data General Corporation ("Data General"), in marketing its computer equipment, a jury rendered a verdict for plaintiffs, Fairchild Camera and Instrument Corporation and Digidyne Corporation, after a forty-five day trial limited to the single issue of whether Data General possessed sufficient economic power in the relevant market for the tying product (operating systems software) appreciably to restrain competition in the relevant market for the tied product (central processing units). Defendant has moved for judgment non obstante veredicto or, in the alternative, for a new trial, and plaintiffs have moved for an injunction and for a trial on the issue of damages. After a painstaking review of the voluminous evidence in this case, the Court grants defendant's motion for judgment non obstante veredicto or, in the alternative, for a new trial, and denies plaintiffs' motion for an injunction and for a trial on the issue of damages.
Fairchild Camera and Instrument Corporation and Digidyne Corporation are the only plaintiffs who actually proceeded to trial.
A total of eleven separate actions were brought before the Court in the course of this litigation. Three of these actions were originally filed in this District: Fairchild Camera Instrument Corp. v. Data General Corp., No. C-78-2418 WHO; SCI Systems v. Data General Corp., No. C-78-2417 WHO; Digidyne Corp. v. Data General Corp., No. C-78-1261 WHO. During 1979 and 1980, the Judicial Panel on Multidistrict Litigation transferred an additional eight actions for consolidated or coordinated pretrial proceedings: Bytronix v. Data General Corp., No. C-79-1101 WHO; Data General Corp. v. Ampex, No. C-79-1102 WHO; Data General Corp. v. Ampex, No. C-79-1103 WHO; Data Compass v. Data General Corp., No. C-79-1345 WHO; Ampex v. Data General Corp., No. C-79-3441 WHO; Data General Corp. v. Data National Corp., No. C-79-3442 WHO; Data General Corp. v. Ampex, No. C-79-3443 WHO; Keronix v. Data General Corp., No. C-80-1217 WHO.
The Court has requested the Judicial Panel on Multidistrict Litigation to issue an order to show cause why Data General Corp. v. Data National Corp. should not be remanded to the District of Massachusetts, in accordance with the procedure set forth in A. H. Robins Co., "Dalkon Shield" IUD Products Liability Litigation, 453 F. Supp. 108 (Judicial Panel on Multidistrict Litigation 1978). The remaining actions were settled prior to the commencement of the jury trial.
The Court bifurcated the trial and ordered the issue of liability to be tried first, with the trial on damages to follow soon thereafter. The scope of the liability trial was confined to the question of whether Data General possessed sufficient economic power in the market for the tying product appreciably to restrain trade in the market for the tied product.
The Court excluded from the present trial Data General's counterclaims, alleging that plaintiffs had misappropriated its trade secrets and proprietary information in the development of their NOVA emulators, and plaintiffs' claim that Data General has instituted "sham" litigation regarding its trade secrets, reserving these claims for trial following the trial of the antitrust issue. See the Court's Order of July 9, 1980, and the Court's decision on cross-motions for summary judgment, In re Data General Antitrust Litigation, 490 F. Supp. 1089 (N.D.Cal. 1980).
The jury rendered its verdict on June 9, 1981. On June 18, 1981, defendant filed its motion for JNOV, or in the alternative, for reconsideration of the Court's grant of partial summary judgment and for a new trial, or in the alternative, for certification for interlocutory appeal. On July 10, 1981, plaintiffs applied for an injunction.
On August 21, 1981, the Court denied defendant's motion for reconsideration of its decision granting partial summary judgment and for a new trial, denied defendant's motion for certification for interlocutory appeal, and took defendant's motion for JNOV and plaintiffs' application for injunction under submission. On September 2, 1981, the Court vacated the trial date for the damages phase, originally set for October 5, 1981. On September 25, 1981, plaintiffs moved for an early setting of the damage trial.
I A
Data General manufactures and markets various computer systems, including a central processing unit ("CPU") called the NOVA and a line of software, called RDOS, designed for use with the NOVA. Data General sells its systems in a general market found by the jury to consist of manufacturers of general purpose minicomputers and microprocessors. Its competitors in this market include, among others, DEC, IBM, Hewlett-Packard, Honeywell, Prime, Texas Instruments, Interdata, Wang, Computer Automation, General Automation, Intel, Zilog, Burroughs, and NCR. Its customers are primarily original equipment manufacturers ("OEMs"), who apply software to the hardware and resell the completed systems to end-users. Plaintiffs manufacture "NOVA emulators," which are CPUs modeled after the NOVA and which are compatible with RDOS, the operating software which executes the NOVA instruction set.
Other companies which manufacture such "NOVA emulators" include Ampex, Lear-Siegler, Keronix, SCI, Point Four, Bytronix, Nixdorf, and Beehive.
The roots of this lawsuit lie in Data General's refusal to license its software, RDOS, for use with the NOVA emulators, or with any CPU other than Data General's own NOVA. Data General makes its software available pursuant to a Program License Agreement which restricts the use of such software to CPUs designated by Data General, and with two exceptions not relevant here, Data General has designated only its own CPUs for use with its licensed software. Plaintiffs allege that these marketing practices are per se illegal under § 1 of the Sherman Act, 15 U.S.C. § 1, and § 3 of the Clayton Act, 15 U.S.C. § 14, which proscribe the utilization of tying arrangements affecting a not insubstantial amount of commerce by a manufacturer who possesses sufficient economic power in the market for the tying product (here, RDOS) appreciably to restrain competition in the market for the tied product (here, the NOVA).
Data General has approved the use of its software with CPUs manufactured by ROLM Corporation and Nippon Minicomputer Corporation, neither of whose computer products compete with products offered by Data General in the United States and Europe.
B
Upon cross-motions for summary judgment supported by affidavits, and referencing many of the 600,000 documents, 150 depositions, and hundreds of interrogatories and requests for admissions available at that time, the Court found that Data General did in fact tie the sale of its software, RDOS, to the sale of its CPUs. In Re Data General Corp. Antitrust Litigation, 490 F. Supp. 1089 (N.D.Cal. 1980) (hereinafter cited as In Re Data General). The Court also found, however, that the question of whether Data General possessed sufficient economic power in the market for operating software appreciably to restrain competition in the CPU market presented genuine issues of material fact appropriate for determination by a jury.
In order to make out a per se violation of § 1 of the Sherman Act and § 3 of the Clayton Act, plaintiffs must prove three critical elements. First, there must be two separate products, with the purchase of one (the tying product) conditioned upon the purchase of the other (the tied product). Second, the seller must possess sufficient economic power in the tying product market appreciably to restrain competition in the tied product market. Third, a not insubstantial amount of commerce in the tied product market must be affected. Fortner Enterprises, Inc. v. United States Steel Corp., 394 U.S. 495, 499, 89 S.Ct. 1252, 1256, 22 L.Ed.2d 495 (1969) (hereinafter cited as Fortner I) (quoting Northern Pacific Ry. Co. v. United States, 356 U.S. 1, 6, 78 S.Ct. 514, 518, 2 L.Ed.2d 545 (1958)).
Plaintiffs have relied in this case only on the per se rule, and have not attempted to show that Data General's tying arrangements create an unreasonable restraint of trade under the "rule of reason" analysis. See In re Data General Antitrust Litigation, supra n. 2, 490 F. Supp. at 1101.
The Ninth Circuit has indicated that the standards for a violation of § 1 of the Sherman Act and § 3 of the Clayton Act are "virtually identical." Moore v. Jas. H. Matthews Co., 550 F.2d 1207, 1214 (9th Cir. 1977). See also von Kalinowski, 9 Antitrust Laws and Trade Regulation § 64.05[2]; In re Data General Antitrust Litigation, supra n. 2, 490 F. Supp. at 1100.
In considering the parties' cross-motions for summary judgment, the Court addressed five issues: (1) whether there were two separate products, with the purchase of one (the tying product) conditioned upon the purchase of the other (the tied product); (2) whether Data General possessed sufficient economic power in the tying product market appreciably to restrain competition in the tied product market; (3) whether the tie-ins affected a not insubstantial amount of commerce; (4) whether plaintiffs were in fact damaged by the tie-ins; and (5) whether there were business justifications for the otherwise unlawful tie-ins.
First, despite defendant's vigorous argument to the contrary, the Court found that software and CPUs are separate and distinct products, notwithstanding the fact that neither the software nor the CPU can function without the other. The relevant inquiry is not whether the two items must be used together, but whether they must come from the same seller. Siegel v. Chicken Delight, Inc., 448 F.2d 43 (9th Cir. 1971), cert. denied, 405 U.S. 955, 92 S.Ct. 1172, 31 L.Ed.2d 232 (1972). Data General itself admits that it sells CPUs for use with non-Data General software, and its price lists contain separate prices for hardware and for software. The undisputed material facts establish that software and CPUs are "separate products" for purposes of finding a tie-in. In re Data General, supra, 490 F. Supp. at 1104-06. The Court next found that defendant's tying practices affect a "not insubstantial" amount of interstate commerce, based on standards set forth in Fortner I. Plaintiffs showed that Data General sold and shipped approximately 52,700 CPUs between 1970 and 1978, and that in 1977 alone its shipments were valued at $254,000,000. Moreover, defendant's CPUs are clearly in interstate commerce, inasmuch as they are currently installed throughout the United States and Europe. In re Data General, supra, 490 F. Supp. at 1116-17.
Addressing the "fact of injury" requirement, the Court found that plaintiffs demonstrated that they were actually damaged by defendant's tying practices, and that their injury was of an antitrust nature. Id. at 1117-19.
Finally, the Court found that the three business justifications which defendant asserted in support of its tying arrangement were not sufficient to permit the maintenance of an otherwise unlawful tie-in scheme. Id. at 1120-24.
As to the final issue, whether Data General possesses sufficient economic power in the tying product market appreciably to restrain competition in the tied product market, the Court found there were genuine issues of material fact precluding summary judgment and mandating a jury trial limited to this issue. Id. at 1111-16.
II A
The testimony at trial necessarily focused upon the definitions of the relevant markets for the tying product (operating systems software), and for the tied product (CPUs). In order to instruct the jury properly on this critical issue, the Court considered the contentions of the parties against the backdrop of the Supreme Court's most recent expressions on the subject contained in the Court's opinions in Fortner I and United States Steel Corp. v. Fortner Enterprises, Inc., 429 U.S. 610, 97 S.Ct. 861, 51 L.Ed.2d 80 (1977) (hereinafter cited as Fortner II).
The threshold legal issue presented is the degree of market analysis necessary to prove economic power in a tying case. Defendant has taken the position that in order to prove economic power plaintiffs must establish the relevant markets by proof similar to that required in a rule of reason case, including extensive analysis of such factors as substitutability, cross-elasticity of demand, and potential competition. Plaintiffs contend that proof of the relevant markets is unnecessary in a per se tying case in which economic power is predicated on the "uniqueness" of defendant's tying product, and in which plaintiffs need only show that the seller possessed economic power over an "appreciable number of buyers" within the tying product market. Fortner I, supra, 394 U.S. at 504, 89 S.Ct. at 1259.
An examination of the precedents requiring proof of the relevant tying and tied product markets as a prerequisite for proving economic power reveals that antitrust law has always treated tying arrangements harshly. Before the advent of the Fortner litigation, tying arrangements possessing certain critical elements were repeatedly held to be per se violations of both §§ 1, 2 of the Sherman Act, 15 U.S.C. § 1, 2, and § 3 of the Clayton Act, 15 U.S.C. § 14. See, e.g., Times-Picayune Publishing Co. v. United States, 345 U.S. 594, 73 S.Ct. 872, 97 L.Ed. 1277 (1953); Northern Pacific Ry. Co. v. United States, supra, 356 U.S. at 5-6, 78 S.Ct. at 518-519; United States v. Loew's, Inc., 371 U.S. 38, 83 S.Ct. 97, 9 L.Ed.2d 11 (1962); Siegel v. Chicken Delight, Inc., supra, 448 F.2d at 47. The rationale for subjecting tying arrangements to a rule of per se illegality was expressed by the Supreme Court in Northern Pacific, supra, 356 U.S. at 5-6, 78 S.Ct. at 518-519, as follows:
"[T]here are certain agreements or practices which because of their pernicious effect on competition and lack of any redeeming virtue are conclusively presumed to be unreasonable and therefore illegal without elaborate inquiry as to the precise harm they have caused or the business excuse for their use. This principle of per se unreasonableness not only makes the type of restraints which are proscribed by the Sherman Act more certain to the benefit of everyone concerned, but it also avoids the necessity for an incredibly complicated and prolonged economic investigation into the entire history of the industry involved, as well as related industries, in an effort to determine at large whether a particular restraint has been unreasonable — an inquiry so often wholly fruitless when undertaken.
* * * * * *
For our purposes a tying arrangement may be defined as an agreement by a party to sell one product but only on the condition that the buyer also purchases a different (or tied) product, or at least agrees that he will not purchase that product from any other supplier. Where such conditions are successfully exacted competition on the merits with respect to the tied product is inevitably curbed. Indeed `tying agreements serve hardly any purpose beyond the suppression of competition.' Standard Oil Co. of California [and Standard Stations] v. United States, 337 U.S. 293, 305-306 [ 69 S.Ct. 1051, 1058, 93 L.Ed. 1371]. They deny competitors free access to the market for the tied product, not because the party imposing the tying requirements has a better product or a lower price but because of his power or leverage in another market. At the same time buyers are forced to forego their free choice between competing products."
Since it is the use of "power or leverage in another market" which the antitrust laws proscribe, the Court reasoned that tie-ins are "unreasonable in and of themselves whenever a party has sufficient economic power with respect to the tying product to appreciably restrain free competition in the market for the tied product and a `not insubstantial' amount of interstate commerce is affected," and noted that "where the seller has no control or dominance over the tying product so that it does not represent an effectual weapon to pressure buyers into taking the tied item any restraint of trade attributable to such tying arrangements would obviously be insignificant at most." Id. at 6, 78 S.Ct. at 519.
A review of the case law prior to Fortner I and II reveals that, despite the economic power requirement stated in Northern Pacific, the courts did not require plaintiffs to submit detailed market information in order to prevail on per se tying claims. In United States v. Loew's, Inc., supra, 371 U.S. at 45, n. 4, 83 S.Ct. at 102, n. 4, the Supreme Court stated:
"Since the requisite economic power may be found on the basis of either uniqueness or consumer appeal, and since market dominance in the present context does not necessitate a demonstration of market power in the sense of § 2 of the Sherman Act, it should seldom be necessary in a tie-in sale case to embark upon a full-scale factual inquiry into the scope of the relevant market for the tying product and into the corollary problem of the seller's percentage share in that market. This is even more obviously true when the tying product is patented or copyrighted, in which case, as appears in greater detail below, sufficiency of economic power is presumed. Appellants' reliance on United States v. E. I. du Pont de Nemours Co., 351 U.S. 377, 76 S.Ct. 994, 100 L.Ed. 1264, is therefore misplaced."
Thus, the Court indicated that the intricate market analysis and detailed evidence of factors such as cross-elasticity of demand, substitutability, and potential competition appropriate in a monopolization case like du Pont was not required in the context of per se tying cases.
In two leading tie-in decision rendered during the past decade, however, the Court expressed its dissatisfaction with the evolution of the per se analysis and suggested that some degree of market analysis is necessary even in a per se case. In Fortner I and Fortner II the Court emphasized that the focus in a per se tying case must be on "whether the seller has the power, within the market for the tying product, to raise prices or to require purchasers to accept burdensome terms that could not be exacted in a completely competitive market. In short, the question is whether the seller has some advantage not shared by his competitors in the market for the tying product." Fortner II, supra, 429 U.S. at 620, 97 S.Ct. at 867. The Court went on to suggest three indicia of such economic power.
First, it may be shown that the seller occupies a dominant position in the tying product market. Id. at 620, 97 S.Ct. at 867. Second, it may be shown that the seller's product is sufficiently unique that he has "some advantage not shared by his competitors in the market for the tying product." Id. The Supreme Court has described such "uniqueness" in terms of a tying product's legal, physical, or economic characteristics that confer special advantages upon the seller. The Court in a footnote explained:
"Uniqueness confers economic power only when other competitors are in some way prevented from offering the distinctive product themselves. Such barriers may be legal, as in the case of patented and copyrighted products, e.g., International Salt, Loew's, or physical, as when the product is land, e.g., Northern Pacific. It is true that the barriers may also be economic, as when competitors are simply unable to produce the distinctive product profitably, but the uniqueness test in such situations is somewhat confusing since the real source of economic power is not the product itself but rather the seller's cost advantage in producing it." Fortner I, supra, 394 U.S. at 505 n. 2, 89 S.Ct. at 1259 n. 2.
Third, economic power may be shown where a substantial number of customers have accepted the tie-in and there are no explanations other than the seller's economic power for their willingness to do so. Fortner II, supra, 429 U.S. at 618 n. 10, 97 S.Ct. at 867 n. 10.
The cases shed little light on the question of what degree of market proof is required in a per se tying case in the wake of Fortner I and Fortner II To require market evidence identical to that appropriate in a rule of reason case would effectively eradicate the distinction between rule of reason cases and per se cases. However, if the jury is to make the determinations required by Fortner, definition of the relevant market and identification of competitors therein is necessary to determine whether there are exclusionary effects in the market for the tied product, and whether the seller has some advantage not shared by his competitors in the market for the tying product.
See Antitrust Law Developments, Antitrust Law Section of the ABA (2d Supp.) at 15. Cases subsequent to Fortner II have confirmed the necessity for defining the markets in which economic power is to be measured. See, e.g., Kingsport Motors v. Chrysler Motors Corp., 644 F.2d 566 (6th Cir. 1981); Reisner v. General Motors Corp., 511 F. Supp. 1167 (S.D.N.Y. 1981). The utility of these cases is limited by the fact that the courts state their findings of relevant markets without explaining the analysis leading to those findings.
Thus, the task of plaintiffs in this case was to introduce sufficient evidence of the tying product market to permit the jury to find that Data General's competitors were prevented from developing competitive software, and sufficient evidence of the tied product market to permit the jury to find an appreciable restraint of trade. Although plaintiffs were not required to present market evidence as extensive as that required in a rule of reason case, they could not recover on the alleged tie-ins unless they identified and proved the relevant markets for the tying and tied products.
See the Court's in limine ruling of March 20, 1981.
Defendant takes the position that the relevant market for the tied product consists of all general purpose minicomputers, microprocessors, and mainframe computers that utilize a CPU within the $300 to $20,000 price range, and that if the market is to be narrowed further, suggests that it include all general purpose minicomputers or at the very least all 16-bit machines. ( See defendant's proposed jury instructions on the relevant markets; defendant's motion for JNOV at 3.) Defendant's definition of the tying product market parallels its definition of the tied product market, and consists of that operating systems software which will run on the CPUs described above. Plaintiffs contend that the relevant submarket or market is much more limited. Initially plaintiffs seemed to favor a relevant tying market based on the principle of software "lock-in," which would include only those types of operating systems software compatible with applications programs written to run on Data General's software RDOS. Plaintiffs settled, however, on a market based on the principle of compatibility with the NOVA set, defining the tying market as those types of systems software which will run on CPUs utilizing the NOVA instruction set, and the tied market as those CPUs utilizing the NOVA instruction set. ( See plaintiffs' proposed jury instructions on the relevant markets; plaintiffs' response to defendant's motion for JNOV at 20-21.)
The number of bits in a computer determines the number of characters which can be manipulated at a time, and thus affects the speed at which data can be processed.
The focus in defining a relevant product market must be upon recognizing those firms and products which present relevant alternatives to the defendant's product, to the extent that they are "reasonably interchangeable" for the same or similar uses, and thus restrict the defendant's power to raise prices. See United States v. E. I. du Pont de Nemours Co., 351 U.S. 377, 76 S.Ct. 994, 100 L.Ed. 1264 (1956); Kaplan v. Burroughs Corp., 611 F.2d 286 (9th Cir. 1979), cert. denied, 447 U.S. 924, 100 S.Ct. 3016, 65 L.Ed.2d 1116 (1980).
Within this relevant market there may in some cases exist a well-defined submarket which can in itself constitute a tying product market for purposes of assessing economic power, and a corresponding tied product submarket for purposes of assessing restraint on trade. Whether such a submarket exists and how it should be defined will depend on the facts of the individual case, but the factors to be considered are set forth in an often-cited passage from Brown Shoe, Inc. v. United States, 370 U.S. 294, 325, 82 S.Ct. 1502, 1523, 8 L.Ed.2d 510 (1961):
"The outer boundaries of a product market are determined by the reasonable interchangeability of use or the cross-elasticity of demand between the product itself and substitutes for it. However, within this broad market, well-defined submarkets may exist which in themselves constitute product markets for antitrust purposes. United States v. E. I. du Pont de Nemours Co., 353 U.S. 586, 593-595, 77 S.Ct. 872, 877, 1 L.Ed.2d 1057 (1956). The boundaries of such a submarket may be determined by examining such practical indicia as industry or public recognition of the submarket as a separate economic entity, the product's peculiar characteristics and uses, unique production facilities, distinct customers, distinct prices, sensitivity to price changes, and specialized vendors."
B
The Court's instructions to the jury with respect to the definition of the relevant markets for the tying and tied products were based on the legal standards described above. First, the Court instructed the jury that in order to decide the economic power issue, it must determine relevant markets for the tying and tied products. Jury Instructions ("JI") 28. The Court next instructed the jury as to plaintiffs' and defendant's proposed definitions of the relevant markets. JI 29. Plaintiffs' define the tying product market as "all operating systems software which runs on CPUs utilizing the NOVA instruction set." Defendants contend that the markets or submarkets proposed by plaintiffs do not exist, and that the tied product market is "all general purpose minicomputers, microprocessors, and mainframe computers, regardless of what company makes them," and the tying product market consists of "all operating system software suitable for use with the CPUs just described." The instructions clearly stated that the jury, as the sole judge of the facts, was not limited to just plaintiffs' or defendant's view of the relevant markets or submarkets and was free to find that these markets or submarkets were constituted differently. JI 29.The Court next instructed the jury as to the criteria which should govern their definition of the relevant markets. JI 30. A relevant product market consists of "all products for which there is a reasonable degree of substitutability," and "products which are not reasonably interchangeable or substitutable should not be considered in the same tying and tied product market." The Court then explained that "within the general product market there may exist a submarket which in itself constitutes a product market for determining economic power in the tying product market or appreciable restraint in the tied product market." JI 31. The Court then set forth the factors enumerated in Brown Shoe, supra, noting that these factors are merely practical aids in identifying possible submarkets and are not compulsory bases for the jury's finding.
C
In conjunction with its general verdict that defendant "possessed sufficient economic power within the market for the tying product, the operating system software, to have appreciably restrained competition within the market for the tied product, the CPU, from 1977 to the present," the jury made six findings regarding the relevant markets in response to special jury questions. Question No. 1: "What is the appropriate market for the tying product?" The jury rejected the definitions proposed by plaintiffs and defendant, and defined the relevant market as "all general purpose minicomputers and microprocessors." Question No. 2: "Is there an appropriate submarket for the tying product?" The jury answered in the affirmative. Question No. 3: "What is the appropriate submarket for the tying product?" The jury found in accordance with plaintiffs' definition that the submarket is "operating software which run with CPUs utilizing the NOVA instruction set."
With respect to the tied product market, the jury questions and answers paralleled those given for the tying product market. The jury in Question No. 4 rejected both parties' proposed definitions of the relevant tied product market, and defined it as consisting of "all general purpose minicomputers and microprocessors." The jury found in response to Question No. 5 that an appropriate submarket for the tied product did exist and, in Question No. 6, accepted plaintiffs' definition of that market "CPUs utilizing the NOVA instruction set." Thus, in defining the general markets, the jury did not simply follow the definitions proposed by the parties or any instruction of the Court, but instead developed its own definitions of those markets.
D
Having instructed the jury as to the standards governing market definition, the Court next instructed the jury to determine whether defendant possesses "sufficient economic power within the tying product market appreciably to restrain competition in the tied product market." JI 32. The Court defined economic power as "the power to raise prices, or instead of raising prices, to impose burdensome terms, that could not be exacted in a completely competitive market with respect to any appreciable number of buyers within the market." The Court explained that such power could be proven by direct evidence of a seller's power to raise prices or impose burdensome terms, or by proof that "the seller's product is sufficiently unique that he has some advantage not shared by his competition in the market for the tying product." JI 33.
The Court further instructed the jury that uniqueness confers economic power only when other competitors are in some way prevented from offering the distinctive product themselves, noting that the question of whether competitors must be able to offer compatible software, i.e., software which will run with RDOS-based applications programs with little or no modification of those programs, or simply comparable software, i.e., software which is functionally equivalent to RDOS but which will not run with RDOS-based applications programs without substantial modifications, was for the jury `to decide. JI 36. The Court instructed the jury that a competitor may be prevented from offering the distinctive product by legal barriers, economic barriers, and/or physical barriers. JI 35. The Court explained that copyright protection is presumed to constitute such a legal barrier, and that the burden is on defendant to rebut that presumption by proving that its copyrights have not prevented others from developing the distinctive product themselves. JI 37, 38. Similarly, the Court explained that trade secrets may constitute a legal barrier only upon plaintiffs' proof that such trade secrets protection prevents others from offering the distinctive product themselves. JI 39, 40.
Turning to the question of economic barriers, the Court instructed the jury that software "lock-in" renders RDOS unique and confers economic power upon defendant only if the evidence shows that (1) some of defendant's customers are in fact "locked-in"; and (2) this "lock-in" phenomenon gives defendant an advantage not shared by its competitors in the market for the tying product. JI 42. The Court then instructed the jury that the alleged superiority of RDOS in terms of comprehensiveness, field testing, and other desirable qualities, could constitute an economic barrier rendering RDOS unique only upon a showing that others were prevented from offering software with the same distinctive qualities. JI 43. The Court explained that these alleged barriers could be considered individually or in combination as proof of uniqueness conferring economic power. JI 44. The Court also instructed the jury that economic power over an appreciable number of buyers in the market, even if such power is not complete over these buyers or over all other buyers in the market, is sufficient to satisfy the economic power test. JI 45.
Finally, the Court instructed the jury as to the meaning of appreciable restraint of competition in the market for the tied product: "The term `appreciably to restrain competition in the market for the tied product' means to interfere appreciably with the ordinary, usual, and freely competitive pricing or distribution system of the open tied product market." JI 46.
On the basis of the foregoing instructions, the jury found, in response to Question Nos. 7, 8, and 9 that "defendant had power to raise prices, or instead of raising prices, to impose burdensome terms that could not be obtained in a completely competitive market or submarket"; that "defendant's operating system software was sufficiently unique that defendant had some advantage not shared by its competitors within the appropriate operating system software market or submarket"; and that there was "an appreciable restraint within the tied product market or submarket."
The difficulty in interpreting the jury verdict arises from the fact that it is impossible to determine from the form of the general verdict and the special questions whether the findings of economic power and appreciable restraint refer to the market or to the submarket. Nonetheless, whether the jury based its findings on the market or on the submarket, the Court finds that the jury verdict cannot stand, because the evidence in this case does not support a finding of sufficient economic power appreciably to restrain competition in either of these markets.
III A
The standards for granting a judgment notwithstanding the verdict are the same as those for granting a directed verdict. Fountila v. Carter, 571 F.2d 487 (9th Cir. 1978). A court may grant a motion for JNOV only if, without in any sense passing upon the credibility of the witnesses, it finds that the evidence and its inferences, considered as a whole and viewed in the light most favorable to the nonmoving party, can support only one reasonable conclusion: that the moving party is entitled to JNOV. Inglis v. ITT Continental Baking Co., 652 F.2d 917, 930 (9th Cir. 1981); Murphy Tugboat v. Crowley, 467 F. Supp. 841, 848 (N.D.Cal. 1978), aff'd, 658 F.2d 1256 (9th Cir. 1981); Fount-Wip, Inc. v. Reddi-Whip, Inc., 568 F.2d 1296, 1300 (9th Cir. 1978). The Court is not free to weigh the evidence or to reach a result that it finds more reasonable, so long as the jury's verdict is supported by "substantial evidence." Inglis v. ITT Continental Baking Co., supra; Murphy Tugboat v. Crowley, supra; Continental Ore v. Union Carbide Carbon Corp., 370 U.S. 690, 696 n. 6, 82 S.Ct. 1404, 1409 n. 6, 8 L.Ed.2d 777; Butte Copper Zinc Co. v. Amerman, 157 F.2d 457, 458 (9th Cir. 1946).
The Supreme Court has defined the term "substantial evidence" as meaning "more than a mere scintilla. It means such relevant evidence as a reasonable mind might accept as adequate to support a conclusion." Consolidated Edison Co. v. NLRB, 305 U.S. 197, 229, 59 S.Ct. 206, 216, 83 L.Ed. 126 (1938). The Ninth Circuit recently held that "[i]n order to benefit from the favorable inferences available under either the directed verdict or j. n. o. v., a party must present substantial evidence, defined as evidence as a reasonable mind might accept as adequate to support a conclusion." Kaplan v. Burroughs Corp., supra, 611 F.2d at 290. The Ninth Circuit most recently applied the "substantial evidence" test in the context of a motion for JNOV in Inglis, a predatory pricing case. The district court had granted JNOV in favor of defendant on the grounds that plaintiff's evidence regarding defendant's cost-price relationship was legally insufficient to establish predatory pricing. The Court of Appeals reversed, finding that the district court's conclusion had been based at least in part on the Court's doubts as to the credibility of plaintiff's witness. It may be noted here that such weighing of the evidence plays no part in the Court's decision to grant JNOV in the instant case.
B
The jury's definition of the relevant tied product market as consisting of all general purpose minicomputers and microprocessors, and of the relevant tying product market as consisting of all the operating systems software which will run on such minicomputers and microprocessors, is the only reasonable definition on the basis of the evidence presented.
The evidence points to a broad, dynamic, highly competitive market in which numerous vendors of microprocessors, minicomputers, and mainframes compete to offer original equipment manufacturers (OEMs) all purpose computer solutions designed to meet a variety of end-user needs. Dr. Stiegler, economist and witness for the defendant, states that the market for all computer services which are supplied by Data General's CPUs and software must be much wider than only the minicomputers themselves because:
"Large minicomputers and shared mainframe computers are in full, direct competition to provide the identical or equivalent services to the potential customers. This is the fundamental market in which Data General competes." RT 6850, lines 2-8.
Computer vendors of different instruction sets are all offering OEMs general-purpose solutions for the OEMs to meet end-user needs for a broad variety of applications. The end-user can in turn choose from among the solutions offered by the OEMs of different manufacturers, not only in the minicomputer market, but also in the microprocessor and mainframe markets. Thus, the relevant market must also be defined by the competition to meet end-user needs, since it is the demand of end-users which determines what computer systems OEMs will buy and what systems vendors will sell.
This competition was recognized by plaintiffs' economist, Professor Stiegler, who, while reluctant to define a market, did acknowledge competition not only within the minicomputer market place, but between minicomputers and mainframes and minicomputers and microprocessors. RT 3625, lines 7-16, RT 6991-6993.
The testimony of numerous "locked-in" customers shows that they selected Data General only after evaluating the broad range of competitive hardware and software offerings. Customer testimony indicated that the desirable features of RDOS could be found in other operating systems, and identified several operating systems in particular as competitive with or superior to Data General's operating systems. The industry itself views the market as a general market for a broad range of computer services. Data General itself regards its competition as "the traditional minicomputer industry" and identifies a large number of competitors in that industry as restraining its pricing ability. Each of these competitors offer systems software for use with their CPUs' particular instruction set, and there are currently more than 100 different sets of operating system software available in the United States. Competitors, too, have testified as to the breadth of competitive offerings in the industry, and plaintiffs' own documents show that competitors do not view the market in terms of any particular vendor's instruction set. Thus, the Court finds that the jury's definition of the relevant tying and tied product markets as encompassing all general purpose minicomputers and microprocessors is clearly supported by the evidence in this case.
See, e.g., RT 620, line 25 — 621, line 13; RT 633, line 5 — 635, line 4; RT 844, line 2 — 845, line 9; RT 855, line 20 — 856, line 1; RT 916, line 16 — 918, line 24; RT 985, lines 7-16; RT 1060, lines 13-23; RT 1145, line 16 — 1147, line 5.
See, e.g., RT 807, lines 3-14; RT 2746, line 6 — 2751, line 25; RT 2368, line 25 — 2370, line 21; RT 2506, line 22 — 2207, line 2; RT 2370, lines 6-14; RT 1277, line 21 — 288, line 11; RT 842, line 23 — 846, line 21; RT 914-915, line 14; RT 2466, line 22 — 2467, line 9; RT 2505, line 23 — 2506, line 19.
See, e.g., RT 3936, 3937, 3939, 4307.
See, e.g., RT 2005, line 6 — 2006, line 19; RT 2034, line 19 — 2036, line 3; Exhibit Nos. 92, 314, 442, 524, 696, 699, 1261, 1569.
RT 406, lines 5-10.
Exhibits 00Q, OOS, OPD, OPE.
The Court also finds that no reasonable jury could find that the defendant possessed sufficient economic power in the relevant tying product market appreciably to restrain competition in the relevant tied product market as these markets have just been defined.
Plaintiffs seek to prove Data General's economic power by two methods: (1) by direct evidence of Data General's power over price, and (2) by evidence showing that RDOS is "unique" and that this uniqueness confers upon Data General an advantage over its competitors because they are unable to offer the distinctive product themselves. Plaintiffs' evidence on the economic power issue was directed toward the NOVA instruction set submarket, and plaintiffs have made no real attempt to prove that Data General possessed economic power over an appreciable number of buyers in the general market. The Court will nonetheless examine the evidence concerning defendant's economic power to the extent that it can support a finding of economic power in either of the relevant tying product markets found by the jury.
Plaintiffs' direct evidence of defendant's power over price rests essentially on the software lock-in theory discussed above. Plaintiffs contend that the relative price insensitivity of Data General's locked-in customers enables it to impose discriminatory prices or terms upon its old customers, or alternatively, to impose higher prices uniformly as to all customers even at the cost of losing some new business. The evidence unmistakably proves, however, that lock-in does not free Data General from the competitive restraints of the general market, and that Data General's prices are in fact competitive.
The degree of "lock-in" depends on a number of variables, particularly on the manner in which a customer writes its applications programs. The testimony indicates that the technology to develop conversion aids already exists. The witness Trevor, for example, testified that his company, Cold Spring Investment Corporation, has reduced future conversion difficulties by segregating all of the operating system calls into a block of routines independent of the applications program, and that this helped when converting an application program from RDOS to AOS. RT 1178, lines 16-23. The testimony indicates that conversion is in each case an individual business decision. As stated by the witness Bossett in his testimony on the feasibility of conversion:
"If the dollars that I have to invest and the time that I have to invest warrant the return, then conversion is the route to take." RT 1183, lines 3-19.
Witness Richard Bajackson testified that certain additional factors warrant consideration in deciding whether to convert. These factors are the resulting benefits in hardware capability, reliability, or servicing. RT 2907-2908.
The evidence shows that it is virtually impossible to calculate the degree of commitment to Data General software and the sensitivity to price of Data General's locked-in customers. Dr. McClure testified that "there are degrees of lock-in and some applications are more locked-in than other applications." RT 696, lines 22-25. In addition, plaintiff's expert, commissioned to survey OEMs on this issue, acknowledged that his findings "don't really help us to decide whether Data General OEMs think they are dependent on Data General software." RT 3412, lines 9-24.
The evidence also shows that Data General's locked-in customers are not sufficiently insulated from competition in the broad market to enable Data General to impose discriminatory prices or terms upon old customers. The evidence shows that OEMs are very price-sensitive, and that if Data General overcharged its old customers it would destroy its customer base because its OEMs would not be able to compete against OEMs of other vendors in the resale market.
Moreover, plaintiffs have offered no evidence to rebut Data General's contention that it actively seeks to attract new customers as well as to make sales for new applications and additional sales for use with existing applications. The fact that the continuing satisfaction of existing customers is an indispensable base for attracting new business, since decisions to purchase from a particular vendor are based largely on reports from existing customers, supports a finding that Data General's competitive position with regard to existing customers is no different than its position with regard to new customers. Plaintiffs' only evidence that Data General discriminates against old customers consists of speculation and testimony by some old Data General customers that they occasionally experienced delays in service or delivery. Such evidence is not sufficient to uphold a finding that Data General imposed discriminatory prices or terms upon its locked-in customers.
RT 812, line 19 — 813, line 2; RT 785, lines 5-7.
See, e.g., RT 1105, line 7; RT 841-842; RT 1655, line 11; RT 1656, lines 14-17.
Moreover, the overwhelming weight of the evidence shows that Data General's pricing is in fact competitive both in policy and result, and that competitive constraints on Data General's pricing practices prevented Data General from possessing or exercising legally cognizable power over price. Evidence contained in witness testimony and exhibits indicates that competitive constraints are of prime importance in Data General's pricing practices. Customer testimony shows that Data General's hardware and software packages are competitive in terms of functionality and price, and are sometimes functionally superior, to those offered by competitors. For example, witness David Clinton testified that his company made the decision to purchase the Data General hardware and software, as opposed to those of DEC, Hewlett-Packard, Varian, and others because:
See Exhibit 5; see also RT 53-55; RT 2380, line 20 — 2381, line 17; RT 2383, lines 3-15; RT 2385, lines 8-13; RT 2389, line 12 — 2390, line 1; RT 3088, line 16; RT 2504, lines 11-22; RT 2520, lines 3-16.
RT 919, line 14 — 920, line 2; RT 2621, line 20 — 2623, line 6; RT 2812, lines 5-14.
"We felt that the Data General NOVA offered more performance in terms of the computing capability per dollar of cost than these other machines represented." RT 635, lines 11-14.
Plaintiffs contend that Data General's January, 1980, announcement "unbundling" software prices from hardware prices demonstrates its power over price. The evidence, however, shows that the price "unbundling" was not a price "increase" in software, but rather a price "shift" which enabled existing customers to pay only for those software items they actually needed, and that the "unbundling" actually constituted a reduction in package price.
Exhibit 647; RT 3131, line 17 — 3132.
Plaintiffs further contend that Data General's practice of requiring its software licensees to purchase a minimum amount of its memory and peripheral products (the so-called "minimum equipment configuration," or "MEC") or to pay a program license charge for its software demonstrates Data General's economic power. The evidence, however, does not support plaintiffs' contention that the MEC is an example of Data General's ability to raise prices or to impose burdensome terms. Plaintiffs' single-item price comparisons as to certain peripherals included in MEC fly in the face of Fortner II as well as economic common sense: where a tying product is part of a larger package, price competitiveness can only be measured in terms of what the buyer paid and what he received for the package as a whole. Customer testimony establishes that the benefits of MEC included not only the products involved, but also the real and perceived advantages for buyers of system reliability and single source accountability where a complete configuration is obtained from a single vendor. In short, plaintiffs have not presented any substantial evidence to support their theory that software lock-in enables Data General to raise prices or to impose terms that could not be exacted in a completely competitive market.
RT 2301, line 13 — 2302, line 9.
See, e.g., RT 918, line 16 — 918, line 11; RT 922, line 4 — 925, line 4; RT 624, lines 10-18; RT 657, line 23 — 658, line 2.
Similarly, plaintiffs have failed to prove that RDOS is unique and that such uniqueness gives Data General an advantage over its competitors. Plaintiffs claim that RDOS is unique for two distinct reasons: (1) it is the most comprehensive, field-proven, and generally desirable software available; and (2) it is the only software compatible with RDOS-based applications programs.
Proof that RDOS is inherently superior to other software and thus has unusual appeal for consumers may render RDOS "unique" within the meaning of Fortner I and Fortner II to the extent that competitors are prevented from offering a comparable product themselves. However, plaintiffs have not carried their burden of proof on this issue. Looking at the evidence in the light most favorable to the plaintiffs, the Court will assume that customers viewed RDOS as uniquely desirable and that it in fact possesses various features which render it superior to other software, although there was substantial evidence indicating that other types of software are equally or more attractive and were often preferred by customers. Assuming that RDOS is in this sense "unique," however, plaintiffs have failed to prove that Data General's competitors were prevented from developing functionally equivalent software.
The evidence does not support plaintiffs' claims that legal barriers in the form of copyright notices and trade secrets protection, as well as economic barriers, deterred competitors from developing comparable software. As the Court noted in its earlier opinion, the existence of copyright notices generally creates a presumption of economic power, but does not conclusively prove such power. Moreover, the presumption of economic power may be inappropriate in the computer software context because copyright notices do not necessarily prevent others from copying the material embodiment of the source program.
In re Data General Antitrust Litigation, supra n. 2, 490 F. Supp. at 1113.
Id.
Plaintiffs' evidence on the copyright issue came almost exclusively from plaintiffs' expert witness Professor Latman, who explained the limited nature of the protection afforded by copyrights in the software context. RT 2087. No evidence was offered as to the actual effect of copyright notices on the development of comparable software. The testimony of plaintiffs' chief witness Professor Latman merely addressed the question of how copyright might affect the development of any compatible software. Professor Latman bases his testimony on the assumption that "Data General's customers must develop applications software that has to dovetail with Data General's software." RT 2085, lines 23-25. Yet, a competitor's software does not have to "dovetail" with RDOS-based applications programs because as previously stated it is comparability, not identicality, which is relevant insofar as uniqueness is predicated on the superior qualities of RDOS.
Nor does the evidence support a finding that Data General's trade secrets protection created a legal barrier preventing competitors from developing comparable software. Trade secret protection has never been held to create a presumption of economic power. The evidence pertaining to trade secrets came primarily from plaintiff's expert witness Roger Milgrim. It should be noted that Milgrim's testimony was not based on the specific facts of this case, since he had made no investigation as to Data General's particular policies or practices in this area. Milgrim testified that trade secrets protection can only arise where confidential information is disclosed pursuant to a confidential relationship. Thus, trade secret law:
"primarily protects the owner against wrongful use or disclosure of the trade secret by persons standing in a legal relationship to the owner, which relationship limits the right to use or disclose. The most typical such relationships arise out of a contract, such as an employment or a license agreement, or a so-called confidential relationship * * * such as that which exists between an employer and an employee." RT 2282-2283.
Since there is no claim of such a relationship between Data General and its competitors, Data General's licensing of software and claims of trade secret protection cannot prevent competitors from developing equivalent software. In other words:
"While an owner of a trade secret has extensive rights with respect to those standing in a contractual or confidential relationship to him, he * * * has no rights at all against independent third parties who develop similar or identical matter. * * * Thus, if XYZ first develops program `Dynamite' — a software program — retaining it and licensing it as a trade secret, the law of trade secrets will not prevent ABC — another company — from developing a program * * * serving the same function and meeting the same specifications, whether or not the second program is identical to the first * * *." RT 2282-2283.
Moreover, the testimony regarding trade secrets focused on the possible effects on the development of software compatible with RDOS-based applications programs, rather than comparable, functionally equivalent software.
Plaintiffs have also attempted to prove the existence of economic barriers which deter competitors from developing comparable software. Plaintiffs presented evidence showing the high costs and high risks associated with developing comparable software. RT 7228-7240. Under the standard established by Fortner II, however, such evidence is without significance unless it is based on comparative statistics; the critical question for purposes of assessing Data General's "advantage" over competitors is whether it is more difficult or expensive for competitors to develop such software initially. RT 812, line 19 — 813, line 2. The evidence shows that comparable software could be developed more easily today due to increased technology. In fact, plaintiffs' expert witness Dr. McClure admits that as a result of the increase in the available tool of technical knowledge about the software, it would not take as long to develop systems software comparable to the systems software that Data General offers in its NOVA line as it took Data General to do it in the first place. RT 747, lines 10-25. Furthermore, Fairchild's own development of NOVA compatible software, and testimony indicating that Fairchild could add to that software all of the desirable features of RDOS, confirms the absence of significant economic barriers. RT 7228-7240; Exhibit ETJ.
Finally, plaintiffs seek to prove that RDOS is unique because it is the only software that will run with RDOS-based applications programs, and that this software lock-in confers a competitive advantage upon Data General because its competitors are prevented from developing RDOS-compatible software by the legal and economic barriers discussed above. Looking at the evidence in the light most favorable to the plaintiffs, the Court will assume that these barriers prevent competitors from developing compatible software and that the costs of converting to a noncompatible software are substantial. The Court finds, however, that on the basis of the evidence presented, this type of "uniqueness" does not give Data General the type of competitive advantage and consequent economic power envisioned by Fortner II. In Fortner II, the Court cited with approval the following interpretation of uniqueness as an indicia of economic power: "Whenever there are some buyers who find a seller's product uniquely attractive, and are therefore willing to pay a premium above the price of its nearest substitute, the seller has the opportunity to impose a tie to some other good." Fortner II, 429 U.S. at 621, 97 S.Ct. at 868 (emphasis added). Thus, uniqueness is an indicia of economic power to the extent that the manufacturer of a unique product which competitors are prevented from offering may exact a premium, in the form of higher price or burdensome terms, which could not be exacted in a completely competitive market. As the Court demonstrated above, however, the evidence unmistakably shows that Data General's prices are competitive and that Data General cannot exploit the supposed advantages of software lock-in because lock-in does not free it from the price constraints of the general market. Thus, although uniqueness may in many cases support a valid inference of economic power, the record in this case overwhelmingly rebuts any such inference of economic power from the fact of lock-in.
Plaintiffs' failure to prove that Data General possesses economic power in the relevant tying product market necessarily means that Data General cannot exercise such power to restrain competition in the relevant tied product market. Quite apart from this failure of proof, however, the record clearly shows the absence of any appreciable restraint in the market for the tied product.
Plaintiffs' evidence of appreciable restraint of competition has been directed not at the general market described above, but at the effects of defendant's tying practices upon the NOVA emulator manufacturers. Plaintiffs' expert witness Professor Schmalensee plainly stated that his discussion of restraint of competition was directed to the NOVA emulator market, and that he has not concluded that Data General's licensing practices and MEC policy worked an appreciable restraint on the CPU manufacturers in the general market, such as DEC, Hewlett-Packard, IBM, Burroughs, or Computer Automation. RT 3606-3607. Indeed, plaintiffs have not even attempted to prove that Data General's practices have appreciably restrained trade in this market.
The evidence shows that this general market is characterized by a wide range of competitive hardware offerings, intense price competition, ease of entry, and rapid growth. Data General's competitors in this market include DEC, Hewlett-Packard, Texas Instruments, Burroughs, General Automation, IBM, Honeywell, and Microdata, as well as the manufacturers of "NOVA emulators," which include Fairchild, Digidyne, Lear-Siegler, Ampex, Keronix, SCI, Point Four, and Bytronix. The testimony further shows that there were numerous new entries into this market in the years following the introduction of RDOS, the Program Licensing Agreement, and the MEC. RT 3556. And, assuming that software lock-in appreciably restrained competition in the NOVA instruction set submarket, the portion of the broad market conceivably affected by this phenomenon is so small that it cannot be characterized as indicative of power "appreciable to restrain competition" in the general market.
C
The Court now turns to the evidence pertaining to the "NOVA instruction set" market or submarket found by the jury, finding that there is no substantial evidence justifying such a definition of the relevant market and, furthermore, that the evidence does not support a reasonable finding of economic power or restraint of competition even in this narrow market.
According to the market definition principles established in Brown Shoe, a "well-defined" submarket which in itself would constitute a "product market" for antitrust purposes may exist where certain factors insulate those competitors in the submarket from the competitive forces operating in the general market. Such a submarket may be delimited on the basis of practical factors such as "industry or public recognition of the submarket as a separate economic entity, the product's peculiar characteristics, and uses, unique production facilities, distinct customers, distinct prices, sensitivity to price changes and specialized vendors." Brown Shoe, Inc. v. United States, supra, 370 U.S. at 325, 82 S.Ct. at 1523.
The indicia suggested in Brown Shoe "are not to be used in a `talismanic fashion' * * rather, we must approach the problem pragmatically and keep in mind Brown Shoe's instruction that a submarket must be `economically significant.'" Kaplan v. Burroughs Corp., supra, 611 F.2d at 292.
Plaintiffs have offered no evidence, apart from mere opinion, demonstrating that the "NOVA instruction set" submarket is a "well-defined" submarket which is insulated from the competitive forces in the general market, or that identification of particular instruction sets provides a meaningful principle of market definition.
Plaintiffs' expert witness Professor Schmalensee defined the tying product market as "all system software that will execute the NOVA instruction set," on the grounds that "only such software can be used with the CPUs in the tied product market," which is in turn defined as "CPUs that execute the NOVA instruction set." The circularity of this reasoning is obvious. The only characteristic shared by the different types of software in this tying product market is their compatibility with NOVA CPUs and NOVA emulators. Virtually all of the evidence offered by plaintiffs in this case, however, has focused on the fact that the consideration foremost in the minds of purchasers is whether the software will be compatible with their existing applications programs. A Data General customer who has written application programs to run with RDOS can no more easily convert his program to run with another type of software within the NOVA instruction set submarket than he can convert it to run with a type of operating software outside of the NOVA instruction set submarket. Indeed, the testimony clearly indicates that plaintiffs'"principle" of submarket definition is simply to include those products and firms most directly aggrieved by defendant's tie and to exclude other competitive alternatives which do not seem to be adversely affected. Schmalensee, RT 3534-3540. This kind of evidence does not support a finding that the "NOVA instruction set" submarket is an economically significant market, under the principles of Brown Shoe.
RT 3534-3540.
Plaintiffs have emphasized the fact that the tying product market found by the jury was based not on the principle of lock-in, but rather on the principle of compatibility with the NOVA instruction set, and that the submarket was not limited to Data General's software, and included software manufactured by companies such as IPI (BLIS/COBOL), Point Four (IRIS), Dynamic Concepts (BITS), RMD and Associates (VMOS), Izhar Shy Industries (IOS), Minicomputer Systems (MICOS), and Fairchild (IMDOS). An examination of the record reveals that plaintiffs' evidence in this case has focused almost exclusively on the "locked-in" Data General customer, and on "lock-in" both as a source of economic power and as a principle of market definition. In explaining the justification for a "NOVA instruction set" submarket, for example, Professor Schmalensee stressed the fact that buyers who are "locked-in" to RDOS-based applications programs cannot economically switch to non-NOVA type CPUs because such CPUs could not run with RDOS. And at the close of the case, plaintiffs' counsel argued:
"But as so many of the customer witnesses told you, from their hearts, and expert witnesses as well explained, none of that matters at all to the appreciable number of customers who are locked in to the continued use of defendant's systems software. For them the reality is that the only available choices are in the market of those vendors who have created operating systems software that runs with NOVA instruction set machines. In that marketplace there's just no question that the operating system known as RDOS is the operating system not only of choice, but of absolute need." RT 7627, lines 4-15.
"Members of the jury, this is a case in which the uniqueness rises to the level of necessity for the substantial number of customers who have written their application programs to run with the defendant's operating system software.
Those, who have been referred to in this case, and have been referred to by everyone, as the committed or the locked-in customers, for them, there is just nothing else that can come close to providing an adequate substitute for defendant's operating system software." RT 7635, line 16 — 7636, line 1.
Under the "lock-in" theory, the market is defined not in terms of the competitive alternatives available to the potential customer when he makes his initial decision to purchase, but rather after the customer has already decided to purchase RDOS, and has thereafter become committed to that product. Such an approach to market definition is untenable, and submarkets based on customer commitment to a given product have been rejected by other courts as too narrow as a matter of law. Moreover, the Court notes that the evidence does not support a submarket composed of Data General's locked-in customers because of proof that: (1) new customers, and old customers with new applications, are not insulated from the competitive forces of the larger market, and a vendor's inability to discriminate between old and new customers insures old customers the benefits of competition in the larger market; and (2) old customers' competition at the resale level with OEMs offering other vendors' products insures that Data General's sales to all customers are directly affected by competition in the general market.
See, e.g., Kaplan v. Burroughs Corp., 611 F.2d 286, 293-95 (9th Cir. 1979), cert. denied, 447 U.S. 924, 100 S.Ct. 3016, 65 L.Ed.2d 1116 (1980); General Business Systems v. North American Phillips Corp., 1980 Trade Cases ¶ 63,607 (N.D.Cal. 1980); Reisner v. General Motors Corp., 511 F. Supp. 1167 (S.D.N.Y. 1981); ILC Peripherals v. IBM Corp., 458 F. Supp. 423, 427 (N.D.Cal. 1978), aff'd 636 F.2d 1188 (9th Cir. 1980); Telex Corp. v. IBM Corp., 510 F.2d 894, 917 (10th Cir. 1975), cert. denied, 423 U.S. 802, 96 S.Ct. 8, 46 L.Ed.2d 244 (1975).
Moreover, a reasonable jury could not find, on the basis of the evidence presented, that Data General had sufficient economic power in the tying product market appreciably to restrain trade in the tied product market, even if the relevant markets were defined in terms of the NOVA instruction set. Plaintiffs' evidence on this issue consists of testimony by Data General's competitors ( see, e.g., RT 1962-1963, 3983, 1372-1373) alleging that NOVA emulator manufacturers could have sold more CPUs if not for the existence of the tying arrangement; by Data General customers, who indicated that the licensing restrictions precluded them from using non-Data General CPUs ( see, e.g., RT 840, 950-954, 1084, 1092, 1932-1933); and by Professor Schmalensee's opinion that the tie appreciably restrained competition in the NOVA emulator market. RT 3534.
Evidence that Data General's competitors were unsatisfied with their sales of NOVA emulators, and conclusory statements that Data General's marketing practices were the cause of their alleged lack of success, simply cannot, without more, establish that defendant appreciably restrained competition by exercising its leverage in the software market. Fundamentally, plaintiffs have not presented evidence demonstrating that their alleged inability to compete with Data General's NOVAs is attributable to the software licensing restrictions rather than to their failure to meet the standards set by Data General with regard to such factors as the quality of the CPUs, the level of field service available, the ability to provide single-vendor accountability, and marketing strategies. In short, plaintiffs have not demonstrated that competitors provided viable alternatives for customers who ultimately purchased Data General's NOVA CPUs. Neither of the plaintiffs has presented evidence showing that customers would rather buy an emulator CPU and pay a reasonable software license charge to license Data General's operating systems software, than to buy a NOVA CPU and license Data General's software under its current terms, conditions, and prices. Indeed, one of plaintiffs' customer witnesses testified that if he had to pay $2,000 to license Data General's software, he would not purchase Fairchild hardware. RT 1277.
The Court emphasizes the point, explained in its discussion above of "lock-in"' as a source of economic power, that the time at which competition in a relevant market should be assessed is the point at which the customer makes his initial decision to purchase, and not after he has already made his decision and become committed to a particular product. Even in the so-called "NOVA instruction set" submarket, the customer has available to him a variety of competitive alternatives, and plaintiffs have failed to demonstrate that Data General's tying practices deprived such customers of the benefits of competition. Moreover, as noted in the discussion above, competitors were not prevented from developing the software necessary for them to offer a viable CPU product or from procuring such software from a third-party supplier; indeed, the evidence shows that non-Data General NOVA-compatible software has been used with SCI's and Digidyne's CPUs. RT 471-475, 1845-1846.
IV
In deciding to grant defendant's motion for judgment non obstante veredicto, the Court has considered all of the evidence presented in the light most favorable to the plaintiffs. Plaintiffs' arguments in support of the jury verdict may be briefly reiterated as follows: (1) evidence as to the relevant product market is not required in a per se tying case; (2) if a relevant market must be defined, the relevant market for the tying product consists of software which will execute the NOVA instruction set, and the relevant market for the tied product consists of CPUs which will execute the NOVA instruction set; (3) Data General possessed sufficient economic power within the relevant tying product market appreciably to restrain competition in the relevant tied product market because (a) software lock-in gives Data General power over price, by enabling it to impose discriminatory prices and terms upon its locked-in customers, or alternatively, to impose uniformly higher prices and burdensome terms upon all customers; and (b) RDOS is unique because it is the only software which can reasonably be used by Data General's locked-in customers, and competitors are prevented from offering the unique product, i.e., software which is comparable or functionally equivalent because of the barriers described above; and, finally, (4) Data General's power in the relevant tying market enables it appreciably to restrain competition in the relevant tied product market.
The Court finds a critical failure of proof as to three critical aspects of the jury verdict. First, the Court finds that, on the basis of the evidence presented in this case, the only reasonable definition of the relevant market for the tied product is the market for general purpose minicomputers and microprocessors. Second, the evidence cannot reasonably support a finding that Data General possessed economic power within the general market for the tying product, nor even that it possessed such power in the submarket of the NOVA-compatible software. Third, there is no evidence from which a reasonable jury could conclude that Data General's tying practices appreciably restrained competition in the general market for the tied product, or even that it imposed significant restraints upon competition in the submarket for NOVA emulators.
Accordingly, IT IS HEREBY ORDERED that:
1. Defendant's motion for judgment non obstante veredicto is granted and defendant's motion for a new trial is conditionally granted.
The Court may grant defendant's motion for a new trial conditionally, even if the verdict is supported by substantial evidence, if the verdict is contrary to the clear weight of the evidence or results in a miscarriage of justice. Murphy Tugboat v. Crowley, 467 F. Supp. 841 (N.D.Cal. 1978), aff'd, 658 F.2d 1256 (9th Cir. 1981); Inglis v. ITT Continental Baking Co., 652 F.2d 917 (9th Cir. 1981).
2. Plaintiff's application for an injunction and motion for a trial on the issue of damages are denied.
3. Defendant's motion to expunge certain statements from the record is denied.
4. All applications for further discovery on the damages issue are denied.5. No motion or motions for reconsideration will be entertained.
6. Judgment will be entered accordingly.