FOR THE SECOND CIRCUIT
FOR THE WESTERN DISTRICT OF NEW YORK
Ball Memorial Hospital, Inc. v. Mutual Hospital Insurance, 784 F.2d 1325 (7th Cir. 1986)
Board of Education v. Dowell, 498 U.S. 237, 247 (1991)
Broadway Delivery Corp. v. United Parcel Service of America, Inc., 651 F.2d 122 (2d Cir. 1981)
Eastman Kodak Co. v. Image Technical Services, Inc., 112 S. Ct. 2072 (1992)
Jefferson Parish Hospital District No. 2. v. Hyde, 466 U.S. 2 (1984)
NCAA v. Board of Regents of the University of Oklahoma, 468 U.S. 85 (1984)
New York State Association for Retarded Children, Inc. v. Carey, 706 F.2d 956 (2d Cir.), cert. denied, 464 U.S. 915 (1983).
Patterson v. Newspaper & Mail Deliverers' Union, 13 F.3d 33 (2d Cir. 1993)
Rufo v. Inmates of Suffolk County Jail, 112 S. Ct. 748 (1992)
State of New York v. Anheuser-Busch, Inc., 811 F. Supp. 848 (E.D.N.Y. 1993)
United States Anchor Mfg., Inc. v. Rule Industries, Inc., 7 F.3d 986 (11th Cir. 1993), cert.denied, 114 S. Ct. 2710 (1994)
United States v. Agri-Mark, Inc., 156 F.R.D. 87 (D. Vt. 1994)
United States v. E.I. du Pont de Nemours & Co., 366 U.S. 316 (1961)
United States v. Eastman Kodak Co., 230 Fed. 522 (1916), appeal dismissed, 255 U.S. 578 (1921)
United States v. Eastman Kodak Co., 226 Fed. 62 (W.D.N.Y. 1915)
United States v. Grinnell Corp., 384 U.S. 563 (1966)
United States v. Morton Plant Health Systems, M.D. Fla., Civ. No. 94-748 CIV-T-23E
United States v. Rockford Memorial Corp., 717 F. Supp. 1251 (N.D. Ill. 1989), aff'd, 898 F.2d 1278 (7th Cir.), cert. denied, 498 U.S. 920 (1990)
United States v. Swift & Co., 286 U.S. 106 (1932)
United States v. Swift & Co., 1982-1 Trade Cas. (CCH) ¶ 64,464 (N.D. Ill. 1981)
United States v. Topco Associates, Inc., 405 U.S. 596 (1972)
United States v. United Shoe Machinery Corp., 391 U.S. 244 (1968) 1
Statutes and Rules:
Section 2, 15 U.S.C. 2
Section 4, 15 U.S.C.
28 U.S.C 1291
F.R.Civ.P., Rule 60(b)(5)
Areeda, Phillip & Donald F. Turner, Antitrust Law (1978)
Areeda, Phillip & Herbert Hovenkamp, Antitrust Law (1993 Supplement)
Carlton, Dennis W. & Jeffrey M. Perloff, Modern Industrial Organization (2d ed. 1994)
Elzinga, Kenneth G. & Thomas F. Hogarty, "The Problem of Market Delineation in Antitrust Suits," 18 Antitrust Bulletin (1973)
Elzinga, Kenneth G. & Thomas F. Hogarty, "The Problem of Market Delineation Revisited: The Case of Coal," 23 Antitrust Bulletin (1978)
Landes, William M. & Richard A. Posner, "Market Power in Antitrust Cases," 94 Harv. L. Rev. 937 (1981)
New York Times, Aug. 16, 1994
Tirole, Jean, The Theory of Industrial Organization (1988) 20
U.S. Department of Justice & FTC, Horizontal Merger Guidelines (1992)
FOR THE SECOND CIRCUIT
FOR THE WESTERN DISTRICT OF NEW YORK
The United States appeals from a final order of the United States District Court for the Western District of New York (Hon. Michael A. Telesca), terminating two antitrust decrees. The decision (J.A. 11-97) is reported at 853 F. Supp. 1454.
The district court had jurisdiction pursuant to section 4 of the Sherman Act, 15 U.S.C. 4, and its continuing jurisdiction over its own decrees. It entered judgment on May 20, 1994. The United States filed a notice of appeal to this Court on July 18, 1994. This Court has jurisdiction pursuant to 28 U.S.C 1291.
1. Whether the district court properly allocated the burden of proof in concluding that Kodak no longer has market power with respect to the sale of amateur color negative film.
2. Whether the district court's conclusion that Kodak no longer has market power with respect to the sale of amateur color negative film is clearly erroneous, in light of the district court's findings of fact pointing to the likelihood that Kodak continues to exercise such power.
On May 20, 1993, Kodak filed a motion to modify or terminate antitrust decrees entered in 1921 and 1954 (J.A. 133), which the government opposed (J.A. 166). On May 20, 1994, after a nine-day evidentiary hearing, the district court terminated both decrees.
a. The 1921 Decree. George Eastman and his Eastman Kodak Co. pioneered amateur photography. They also monopolized it in violation of the Section 2 of the Sherman Act by buying competitors and imposing various forms of exclusive dealing contracts on retailers. United States v. Eastman Kodak Co., 226 Fed. 62 (W.D.N.Y. 1915). This successful government antitrust suit concluded in 1921 in a consent decree.(1) The 1921 decree barred Kodak from "preventing dealers * * * from freely selling goods produced by competitors" (Section VI), from hindering dealers in freely selling Kodak products (Section VII), and from selling "so-called fighting brands" or any product without the Kodak name on it (Section X) (J.A. 98).(2)
b. The 1954 Decree. Kodak began to market a color slide film called Kodachrome in the late 1930s, and a color print film, Kodacolor, by 1954 (J.A. 211-21). At that time, it had over 90% of the color film market. Since Kodak sold its color film only as a package deal with processing included in the price, it also had over 90% of the color photofinishing market (J.A. 220-21). The tying arrangement resulted in a government antitrust suit and a consent decree in 1954 (J.A. 109). Section V of the 1954 decree permanently enjoined Kodak from "[t]ying or otherwise connecting in any manner the sale of its color film to the processing thereof, or the processing of its color film to the sale thereof" (J.A. 114-15).(3)
a. Film. Five firms manufacture all the amateur color negative film sold in the United States: Kodak, Fuji, Konica, Agfa, and 3M (J.A. 42; 257, 303-04). Although "there is little, if any, difference in the quality of film manufactured by Kodak, Fuji, Konica, and Agfa" (J.A. 66; 266-67) in the United States, Kodak greatly outsells its rivals and commands a substantially higher price.
According to the court, Kodak accounts for about 75% of film sales in dollar terms, and about 67% of unit sales (J.A. 54). Worldwide, it accounts for 36% of sales (J.A. 257). As would be expected given Kodak's share in the United States, almost all 241,000 major film retailers, such as mass merchandisers (e.g., K Mart), food and drug stores, and camera specialty shops, carry Kodak film (J.A. 316, 561-62). By contrast, only about 71,000 outlets carry its nearest rival, Fuji (J.A. 561), although they include the stores that sell a majority of the film in the country (J.A. 67; 316, 571-72). Fuji's prices are about 10% lower than Kodak at the wholesale level (J.A. 563-64).(4) The other films are available at even fewer stores,(5) and their prices are much lower than Fuji's (J.A. 314-15, 456, 477-78).
Kodak can greatly outsell its rivals despite charging a higher price primarily because 50% of consumers in this country will buy only Kodak film regardless of price, and another 40% prefer Kodak (J.A. 62; 319-20). Another relevant factor is that Kodak provides rebates to dealers who sell extra (or only) Kodak film (J.A. 72-73).(6)
Kodak not only sells far more film here than its rivals and at higher prices, but those prices vastly exceed Kodak's marginal costs. Kodak's expert economist, Jerry Hausman, testified--and the district court agreed--that Kodak has an "own elasticity" of demand of approximately 2.(7) This means that if Kodak raised prices by 5%, it would lose 10% of its sales (J.A. 56; 367-68; 669). As the government's expert economist, Robert Masson, explained without contradiction, an own elasticity of 2 indicates that "fifty percent of Kodak's price is in margin above manufacturing costs" (J.A. 600). In other words Kodak's prices are twice its marginal costs.(8)
b. Photofinishing. "The markets for color film and color photofinishing in 1954 were indisputably controlled by Kodak" (J.A. 75). Kodak had over 90% of the amateur color negative film market in 1954 (J.A. 214). Kodak did the photofinishing on all of its own color film (J.A. 220-21), because it controlled the technology, and because its photofinishing was included in the cost of the film (J.A. 234).(9)
The 1954 antitrust decree introduced competition into the photofinishing industry, both by barring Kodak from tying its film and photofinishing sales, and by requiring Kodak to license the technology and provide technical assistance to other firms that desired to enter the business (J.A. 80; 221-32). Thus, by about 1968, when color film had captured half the market from black and white film (J.A. 308-09), Kodak was processing less than 5% of its own film (J.A. 225-26). Moreover, in 1977 the first minilab was installed in the United States (J.A. 247). The minilab does on-site photofinishing in about an hour. Because of their convenience these small labs expanded rapidly through the 1980s, and now account for about one-third of the photofinishing done in the United States (J.A. 251-52, 278; 664). Macrolabs (including both wholesale and captive(10) labs) have remained viable because they are somewhat less expensive per photo (J.A. 277-78, 526-28), but they have had to start providing faster service, and overnight wholesale service has become the norm (J.A. 254-55, 536-37, 546-47). While there has obviously been an interplay between the different types of labs, each has its own niche. Macrolabs cannot provide one-hour service, but minilab costs per print are higher, and they cannot handle the volume of work required by large retail customers (J.A. 243-45, 258-60). Thus, retailers, such as department stores, food stores, and drug stores, use macrolabs (J.A. 244-45).
Since 1986, Kodak has reclaimed a large market share in photofinishing by making several acquisitions. The most important of these was a joint venture to establish Qualex, Inc. (J.A. 227).(11) Qualex grew rapidly, largely by acquisition, to a nationwide chain of 65 labs that had 70% of the nation's wholesale macrolab photofinishing market (J.A. 237, 253; 728-29).(12) At present, three firms--Qualex, Konica, and Fuji--have 95% of the wholesale photofinishing business (J.A. 532-33).
The district court terminated both decrees in their entirety (J.A. 97). It stated that it was applying the standard for modifying decrees set forth in Rufo v. Inmates of Suffolk County Jail, 112 S. Ct. 748 (1992), which, in its view, allowed the court "to modify the decrees to fit changes in market conditions" (J.A. 35).(13)
a. The 1921 Decree. The court determined that Kodak no longer had market power, which it defined as the "power of controlling prices or unreasonably restricting competition," with respect to the sale of film (J.A. 50). In doing so it found the relevant geographic market to include not only the United States, but also Western Europe and Japan (ibid.). In that "world" market Kodak has only a 36% share, clearly not enough to infer market power (J.A. 51). Alternatively, even in a geographic market limited to the United States, the court found no market power, despite Kodak's share of 67% (by units) and 75% (by dollars). It held that "[p]rice elasticities are better measures of market power" than market share data (J.A. 55). It found that Kodak had an own elasticity such that if it raised the price 5%, it would lose 10% of sales (J.A. 56)--an own elasticity of 2.(14) It accepted Dr. Hausman's representation that this finding is incompatible with the possession of market power (J.A. 55-58). It found that, despite the equal quality of competing film, "Kodak is obtaining a 'premium' price for its products in some retail outlets" (J.A. 62). But it determined that "Kodak's price premium is not evidence of market power acquired illegally, but of the perceived quality difference that exists in the minds of consumers who are satisfied with Kodak products" (J.A. 68).
"Having found that Kodak does not possess market power in film" (J.A. 68-69), the court had little trouble concluding that the various decree restrictions should be removed.(15)
b. The 1954 Decree. The court stated that this decree was designed to dismantle Kodak's technological dominance of the color film photofinishing industry, by requiring Kodak to license and give technical information to competing photofinishers (J.A. 78-79). Finding that the decree had accomplished its essential purpose of creating a competitive photofinishing market (J.A. 80), and that neither Kodak nor its affiliate Qualex has power in that market (J.A. 92), it thought that allowing Kodak to bundle(16) film and photofinishing would be pro-competitive (J.A. 94).
This is the first time in memory that a district court has terminated an antitrust decree over the government's objection. It did so, we submit, because it misread Rufo and Patterson as substantially eviscerating the requirement that a defendant seeking termination of a decree prove that its purposes have been achieved. Thus, the district court held that Kodak had sufficiently shown that it lacks power in the amateur color film market, and so is entitled to termination, even though the court's own findings of fact strongly indicate that Kodak does have such market power.
This erroneous judgment will not only have an impact on the multibillion dollar film and photofinishing industries, which Kodak has long dominated and which affect millions of American consumers. It also poses a serious threat to the ability of federal enforcement agencies and the courts to protect the public interest through injunctions. There are nearly 1200 federal antitrust decrees in force, and they cover almost every major sector of the American economy. We recognize, of course, that some judgments merit termination or substantial modification, and we have agreed in recent years to the termination or substantial modification of numerous decrees. But the Antitrust Division has concluded with respect to many of the decrees it has reviewed in recent years that they continue to be necessary to the protection of competition.
The district court's judgment presents a significant risk that decrees will be terminated on the basis of minimal showings of changed conditions. It so reduces the burden of defendants convincingly to demonstrate their right to relief that in effect it places the burden on the government to relitigate any cases it has won or settled through consent decrees at the option of the defendants in order to preserve the remedies and safeguards it has obtained. If the district court's opinion is upheld, defendants will be encouraged to seek termination, and the government will be forced to prove de novo the threat to competition with respect to any decree any defendant chooses to challenge at any time. Since the resources of public enforcement agencies are limited, every dollar committed to decree termination litigation is a dollar taken away from initial investigation and prosecution of newly discovered antitrust offenses. And the easier it is to modify or terminate consent decrees, the less attractive they become as a means of settling litigation. Thus, it is exceptionally important to enforcement of the nation's antitrust laws that the Court reverse and remand.
THE DISTRICT COURT ERRED IN TERMINATING THE DECREES ON THE GROUND THAT KODAK LACKS MARKET POWER(17)
It is well settled that district courts may modify antitrust decrees providing for prospective relief, including consent decrees, in response to changed conditions. United States v. Swift & Co., 286 U.S. 106, 114-15 (1932); Rule 60(b)(5), F.R.Civ.P. The courts have consistently emphasized, however, that this power must be exercised with restraint. Thus, in Swift, the Court rejected the private defendants' bid to be released from a decree: conditions had not changed sufficiently to eliminate the threat to competition. In so doing, it enunciated a very demanding standard, requiring a "clear showing of grievous wrong evoked by new and unforeseen conditions." 286 U.S. at 119-20.
The Supreme Court has subsequently explained that Swift's standard was not designed for rote application in every case. In United States v. United Shoe Machinery Corp., 391 U.S. 244 (1968), the Court, explaining Swift, stated that a decree "may not be changed in the interest of the defendants if the purposes of the litigation as incorporated in the decree (the elimination of monopoly and restrictive practices) have not been fully achieved." 391 U.S. at 248 (emphasis in original).
In Rufo v. Inmates of the Suffolk County Jail, 112 S. Ct. 748, 758 (1992), where state and local government officials sought modification of a decree governing conditions in a correctional institution, the Supreme Court repeated that "the 'grievous wrong' language of Swift was not intended to take on a talismanic quality, warding off virtually all efforts to modify consent decrees." The Court concluded that a "flexible approach is often essential to achieving the goals of reform litigation." 112 S. Ct. at 758.(18) This Court recently applied the "flexible approach" described in Rufo in affirming the termination of a civil rights decree that established a comprehensive affirmative action program (including detailed rules governing promotions, transfers, hiring, and back pay) for a private institution. Patterson v. Newspaper & Mail Deliverers' Union, 13 F.3d 33, 38 (2d Cir. 1993).
The district court in this case concluded that the Rufo/Patterson standard applied and that this should significantly ease Kodak's burden (J.A. 34-35).(19) Neither Rufo nor Patterson, however, countenances the undermining of antitrust decrees. Those cases dealt with decrees that inherently demanded judicial flexibility, because of their detailed requirements and those requirements' inevitable intrusions on the efficient working of public, or quasi-public, institutions. But while they plainly teach that the stern Swift formulation is not of "talismanic quality," their flexible standard is also not meant to be a talisman, indiscriminately easing the burden of any defendant seeking termination of a decree.
Rufo, after all, did not address or hint at any change in the principle, established in Swift and explained in United Shoe, that an antitrust decree "may not be changed in the interests of the defendants" if the purposes of the decree have not been "fully achieved." Indeed, the Court in Rufo cited United Shoe for its explanation of Swift. 112 S. Ct. at 758. This is hardly surprising, since the Supreme Court only a year earlier had expressly reaffirmed the United Shoe test in Board of Education v. Dowell, 498 U.S. 237, 247 (1991).(20) Moreover, this Court in Patterson described the United Shoe standard as an example of the more flexible standard it was applying, and it affirmed termination of the decree only upon a finding that its purposes had been achieved. 13 F.3d at 39.
Since under United Shoe an antitrust decree may be modified in the defendants' favor only if they prove that its purposes have been "fully achieved," a fortiori the termination of an antitrust decree requires such demanding proof. The entry of an antitrust decree is a matter of high public importance. United States v. E.I. du Pont de Nemours & Co., 366 U.S. 316, 323 (1961). It is for that reason that the Supreme Court and this Court have underscored the importance of the public interest in determining whether to terminate (or modify) antitrust decrees. Protection of the public interest in competition and respect for the prior judgment of the court are the reasons that Kodak, bound by the 1921 decree after violating the antitrust laws, must be held strictly to its burden of proving that it is entitled to be free of the decree's restrictions. And, as we shall show, this is a burden that Kodak failed to sustain.
The fundamental question for the district court was whether Kodak proved that it no longer has market power in film. In answering that question, the district court made several key findings of fact which, especially when added to important pieces of undisputed evidence, cannot be reconciled with an ultimate conclusion that Kodak carried its burden of establishing that it lacks market power. We believe that the district court's conclusion as to market power flowed directly from its legally erroneous failure to hold Kodak strictly to its burden. And, in any event, an ultimate conclusion as to market power that is incompatible with the court's own supporting findings of fact and the undisputed evidence constitutes clear error.
The court's own findings establish, first, that Kodak sells film no better than its rivals' at a higher price. That Kodak film sells at a premium is obvious from the difference between Kodak's share of U.S. film sales measured in units (67%) and measured in dollar volume (75%). It reflects a Kodak price premium over its nearest rival, Fuji, of at least 4.5% at the retail level and at the more relevant wholesale level of 10%. Kodak has even higher premiums over other competing brands. Second, Kodak's 67%-75% share of U.S. film sales is only slightly below the 75%-80% share found by Judge Hazel in 1915 when he held that Kodak had monopolized film. Third, Fuji, despite selling film of equal quality at 10% under Kodak's price, has been unable to garner more than 10% of U.S. sales. Finally, Kodak faces a demand elasticity of 2, which indicates that it is pricing at twice its marginal cost. All this is possible because 50% of consumers will not buy any other brand of film regardless of price, and another 40% prefer Kodak film.
We submit that these findings and undisputed facts are sufficient for this Court to hold that Kodak in fact does have market power, order the reinstatement of the decrees, and thus obviate further district court proceedings in this protracted matter. At the very least, a reversal and remand is necessary for the district court to assess the evidence under a correct legal standard.
1. Market power is "the ability to raise prices above those that would be charged in a competitive market." NCAA v. Board of Regents of the University of Oklahoma, 468 U.S. 85, 109 n.38 (1984). Monopoly power is a significant degree of market power. Eastman Kodak Co. v. Image Technical Services, Inc., 112 S. Ct. 2072, 2090 (1992). It is at the heart of this case and most antitrust cases, because the ability to act anticompetitively depends on the possession of market power. If Kodak still has such power, it can exercise it to the detriment of consumers, and there is abundant reason to maintain the decrees.
Traditionally courts determine the existence of market power inferentially: they define a relevant market and "infer [market power] from the predominant share of the market." United States v. Grinnell Corp., 384 U.S. 563, 571 (1966). But, since the ultimate inquiry is power over price, courts have increasingly addressed that critical subject directly. As Judge Easterbrook wrote for the Seventh Circuit: "Market share is just a way of estimating market power, which is the ultimate consideration. When there are better ways to estimate market power, the court should use them." Ball Memorial Hospital, Inc. v. Mutual Hospital Insurance, 784 F.2d 1325, 1336 (7th Cir. 1986).(21)
The district court properly focused on market power as a threshold matter. It quoted a sound definition of such power (J.A. 55).(22)
It followed Judge Easterbrook's advice in Ball Memorial Hospital, and that of William M. Landes & Richard A. Posner, "Market Power in Antitrust Cases," 94 Harv. L. Rev. 937, 950 (1981), to use better measures of market power than market share statistics (J.A. 55). It sensibly held that "[p]rice elasticities are better measures of market power" (ibid.). And finally, it reasonably relied on the testimony of Kodak's expert economist, Jerry Hausman, that Kodak's own elasticity of demand is 2 (J.A. 55-56).
What the court failed to appreciate, however, was that this evidence, especially when backed by other undisputed testimony and by generally accepted principles of economics, flatly contradicts Kodak's position. It unequivocally shows Kodak's failure to prove that it no longer exercises market power. We submit, moreover, that this Court can take it (and other findings to be discussed later) as affirmative proof that Kodak still is exercising market power as it admittedly did in 1921 and 1954.
2. The economist's term "own elasticity of demand" expresses the change in quantity of goods a firm will sell in response to a change in the price it charges (J.A. 367). Thus an elasticity of 2 means that a price increase by Kodak would produce a quantity decrease (or a price decrease by Kodak would produce a quantity increase) twice the size of the price change in percentage terms. Or, to use Dr. Hausman's own example, if Kodak raised its current prices five percent, it would suffer a ten percent decrease in sales (J.A. 367-68).
An own elasticity of demand of 2 in itself tells us something important about Kodak's power over price. As economists agree, when a firm is charging a profit-maximizing price, "if the elasticity of demand is 2, price is twice marginal cost." Dennis W. Carlton & Jeffrey M. Perloff, Modern Industrial Organization 137 (2d ed. 1994).(23) At trial, the government's expert, Dr. Masson, stressed this point (J.A. 599-603), and Dr. Hausman did not disagree.(24) Dr. Masson also stressed--and Dr. Hausman did not disagree--that this large an excess of price over marginal cost is generally a strong indicator of market power. (J.A. 600-03). Indeed, as an authoritative antitrust law treatise states: "The degree of market power is measured by the excess of the profit-maximizing price over short-run marginal cost." 2 Phillip Areeda & Donald F. Turner, Antitrust Law 337 (1978).(25)
Thus Kodak, which had the burden of proving that it does not have market power, instead by its evidence on own elasticity of demand submitted strong proof that it does have market power. Dr. Hausman recognized the predicament caused by his testimony about own elasticity of demand and attempted to escape from it by arguing that Kodak's "fixed costs are enormous" and so suggesting that the difference between price and short-run marginal cost is not the appropriate measure of market power in this case (J.A. 392-93). But Dr. Hausman neither had nor claimed any expertise as to Kodak's actual fixed costs. Aside from a second-hand recitation of an undocumented claim of R&D costs of "8 or 9 percent" (J.A. 305, 393), he offered no figures at all. More significantly Kodak, which has these figures both precisely and readily available, flatly declined to produce them. Indeed, when, during the course of the hearing the government asked Kodak to disclose its profit margins--which would have settled definitively the issue of the relationship between its prices and cost--Kodak refused (J.A. 323-24). The obvious inference to be drawn is that the evidence would have been embarrassing to Kodak. Nonetheless, the district court upheld its refusal (J.A. 325-26).
This ruling was typical of the district court's misunderstanding of the burden of proof with which Kodak undertook the litigation and which Kodak greatly increased with its own price elasticity evidence during the hearing. The court never focused on the fact that embracing Dr. Hausman's testimony on own demand elasticity cut the ground out from its own ultimate conclusion that Kodak lacks market power. Moreover, this was not the only finding that showed Kodak's market power.
3. The district court's findings that Kodak sells film that is no better than its rivals' at a substantial price premium, while maintaining an enormous share of U.S. sales, further demonstrated Kodak's continuing market power.
According to the district court, "Kodak's competitors now manage to produce film of equal quality . . ." (J.A. 66). But while Kodak film is not any better than its rivals', it charges substantially higher prices than its competitors. This is obvious from the district court's finding that Kodak sells 67% of the film in the U.S. but garners 75% of the revenue (J.A. 54). And the court found that Kodak enjoys a retail price premium at food and drug stores of 4.5% (J.A. 59, 62). More importantly, since Kodak and its rivals sell at wholesale not at retail, Kodak, according to uncontradicted testimony, sells to U.S. dealers at a premium ranging from 10% above Fuji to at least 20% above 3M's Scotch brand (J.A. 477, 563-64; 718).(26)
Despite this significant price disparity, Kodak nonetheless continues to maintain a 67%-75% share in the U.S.--not greatly different from its 75%-80% share in 1915, when the district court found it in violation of Section 2 of the Sherman Act. United States v. Eastman Kodak Co., 226 Fed. 62, 79 (1915). Moreover, Kodak dwarfs its nearest rival, Fuji, which has a mere 10% of U.S. sales (J.A. 690-91).(27) Since Fuji's strategy is "to undercut whatever price Kodak is charging for its film" (J.A. 60) and sells to retailers at 10% less than Kodak, it is all the more significant that Kodak nevertheless maintains a dominance of 67% to 10% over Fuji in U.S. sales.
This ability to maintain both a price premium and an enormous market share without a quality difference is the essence of proof of market power. Cf. United States Anchor Mfg., Inc. v. Rule Industries, Inc., 7 F.3d 986, 997-98 (11th Cir. 1993), cert. denied, 114 S. Ct. 2710 (1994). Indeed, even the district court in its discussion of the price premium seemed not seriously to doubt that this is so. Rather than denying that the price premium was evidence of market power at all, it concluded that "Kodak's price premium is not evidence of market power acquired illegally, but of the perceived quality difference that exists in the minds of consumers" (J.A. 68). Once again, the district court failed to appreciate the burden the law placed on Kodak to terminate the decrees.
We readily admit that Kodak enjoys strong consumer loyalty. As Kodak said--and the court agreed: "50 percent of consumers will only buy Kodak film, while another 40 percent of consumers prefer Kodak film, but are willing to purchase another brand of film" (J.A. 62). Indeed, it is precisely this brand loyalty which enables Kodak to retain market power. If large numbers of consumers did not think Kodak film was of better quality (whether or not it actually is), Kodak would not be able to charge more than its rivals and still maintain an immense market share, nor for that matter would it have an own demand elasticity of 2. Since market power necessarily is the direct consequence of consumer preference, their coincidence hardly provides a basis for terminating a decree designed to counter the effects of just such market power. Put differently, the important point is not the reason for market power when a consent decree is at issue; it is the effect of that market power on price and business behavior in the market.
Likewise, the court's determination that Kodak's current market power was not "acquired illegally" (J.A. 68) as a matter of law is insufficient to warrant termination of the antitrust judgments.(28) This is not an antitrust enforcement action, in which the government has the burden of proving illegal conduct. The government brought such an action eighty years ago, proved in the district court that Kodak illegally acquired and maintained market power, and Kodak, rather than exercise its right to obtain Supreme Court review, chose to settle with a decree. In agreeing to the 1921 decree, it agreed to be bound by restrictions which as a matter of law are not to be lifted until the purposes of the decree are fully achieved. Those purposes are the protection of the public from Kodak's market power. Kodak at all times in this proceeding had the burden of showing that it no longer has market power and hence that those protections are now unnecessary. The district court, by excusing Kodak's current market power as not acquired illegally, not only misapplied the governing legal standard but put consumers in jeopardy as well.
c. The court's failure to appreciate the significance of its finding on own elasticity of demand and successful maintenance of a price premium also explains its finding that the United States is not a relevant market. The issue in this case is whether Kodak can exercise market power in the United States, to the detriment of American consumers. The purpose of defining markets in antitrust cases is to assess the ability of a firm to exercise market power. Thus, a relevant geographic market is the area in which it would be possible to exercise such power. If, as the district court's findings and undisputed evidence indicate, Kodak can exercise market power in the United States, then the United States is the relevant market for purposes of this case. Whether it would be more appropriate to define a broader market in another context for another purpose is beside the point.
The district court, however, relying on market delineation tests proposed by Landes and Posner and by Elzinga and Hogarty (J.A. 43-44)(29) concluded that the market is worldwide because foreign manufacturers sell significant amounts of film in the United States. In other words, the court found that Kodak could not exercise significant power over price and output in the United States because competitive pressure from foreign manufacturers would prevent Kodak from maintaining supracompetitive prices in the United States.
It is certainly true that foreign competition should be taken into account in defining markets, and that the prospect of foreign firms increasing their sales into the United States may sometimes prevent American firms from maintaining prices above competitive levels. But the Landes and Posner and Elzinga-Hogarty tests do not justify ignoring the reality of Kodak's ability to exercise market power in the United States. Indeed, Landes and Posner themselves note that when evidence of demand elasticities is available to measure market power directly, "no market share criterion of market power is either necessary or appropriate." Landes & Posner at 953.
If a firm can discriminate against purchasers in one geographic area--profitably charging them higher prices than it could profitably charge elsewhere--it may be able to exercise market power in that limited area even if it lacks such power elsewhere.(30) Thus, it is generally acknowledged that tests such as the Elzinga-Hogarty test overstate the size of a geographic market if a firm is engaged in price discrimination. See, e.g., United States v. Rockford Memorial Corp., 717 F. Supp. 1251, 1267 n.12 (N.D. Ill. 1989), aff'd, 898 F.2d 1278 (7th Cir.)(opinion by Posner, J.), cert. denied, 498 U.S. 920 (1990). See also, Phillip Areeda & Herbert Hovenkamp, Antitrust Law (1993 Supplement) at 595-96.
The record here amply demonstrates that Kodak can, and does, engage in such price discrimination, taking advantage of the consumer preference that it enjoys in the United States, but not in most of the rest of the world.(31) Kodak data for l993 showed its average wholesale prices lower in Europe (where its market share is 43%) than in the United States, and lower in Japan (where its market share is 6%) than in Europe (J.A. 707, 718). Moreover, further uncontradicted testimony (mostly from Kodak witnesses) established that over several years Kodak wholesale prices have been higher in the United States than in Japan and Europe (J.A. 335, 359, 490).
The district court's response to this evidence of price discrimination was to suggest that the government had not carried the burden of persuasion (J.A. 61). It raised a series of objections to the government's evidence, which fail to blunt the clear--and hardly surprising--point that Kodak can charge higher prices in a country where 50% of consumers will buy only Kodak and another 40% strongly prefer it (J.A. 62, citing Kodak's own studies).(32) And, in any event, the burden of persuasion does not rest with the government in this case. It was Kodak's burden to prove that foreign competition limited its power to exercise market power in the United States. The court thought that the 1993 Kodak pricing data submitted by the government was too limited in time adequately to compare Kodak's pricing at home and abroad (J.A. 61). But Kodak chose not to submit its data for other years, and the court should have drawn the reasonable adverse inference from its failure to do so.
The district court's decision to terminate the 1921 decree was premised on its determination that Kodak no longer has market power, or at least not market power acquired illegally. Since there is ample evidence that Kodak still has market power, this Court should reverse with instructions to reinstate the 1921 decree.(33)
The Court should also reverse and direct the district court to reinstate the 1954 decree. The disputed provision, Section V, prohibits the tying or bundling of photofinishing to film. Since Kodak has market power in film, any tying it did of photofinish-ing to film would be a per se violation of the Sherman Act. Jefferson Parish Hospital District No. 2. v. Hyde, 466 U.S. 2 (1984). There is no conceivable reason to remove the ban as to such flagrantly anticompetitive conduct. Bundling of film and photofinishing is not unlawful per se for Kodak, but there is good reason to fear the consequences in the already Kodak-dominated film market of letting Kodak bundle. Kodak admitted that one of its immediate goals in bundling is to improve its bargaining position with retailers (J.A. 294-97). Kodak, of course, already has market power over the film it sells to those retailers, and giving it more power will strengthen its film monopoly and make it harder for its rivals ever to improve their competitive position.(34) It is hard to imagine any competitive benefit from bundling that will outweigh this competitive harm, and Kodak certainly has not shown any.
The federal antitrust laws are the "Magna Carta of free enterprise . . . as important to the preservation of economic freedom and our free-enterprise system as the Bill of Rights is to the protection of our fundamental personal freedoms." United States v. Topco Associates, Inc., 405 U.S. 596, 610 (1972). Antitrust enforcement is thus a matter of major public importance. And, as the Supreme Court has emphasized, in antitrust cases an effective remedy "is crucial. For the suit has been a futile exercise if the Government proves a violation but fails to secure a remedy adequate to redress it." United States v. E.I. du Pont de Nemours & Co., 366 U.S. 316, 323 (1961).
We recognize, of course, that decrees sometimes outlive their usefulness, and the government freely consents to decree modification or termination where such relief seems appropriate. Over the last fifteen years, over two hundred antitrust consent decrees have been modified or terminated with the government's consent.(35) But the district court's decision, replete as it is with findings attesting to Kodak's continuing market power, stands as a statement that defendants seeking decree termination need not, in reality, prove that a decree's purposes have been accomplished, and indeed that the government in effect must prosecute an antitrust case de novo. The message will not go unnoticed by antitrust defendants past and future. If this decision is affirmed, it is reasonable to assume that many more defendants will seek to follow Kodak's example. Cf. United States v. Agri-Mark, Inc., 156 F.R.D. 87, 88 (D. Vt. 1994) (citing decision below). An inevitable result will be the erroneous termination of competitively important antitrust decrees.
Moreover, the decision below will endanger the consent decree settlement process that provides the public the benefits of antitrust enforcement while minimizing the cost of litigation. Until now, the government has had the ability to avoid protracted litigation in monopolization cases, such as Kodak and Swift, by settling for a decree that strictly regulates conduct for the future rather than pursuing the more draconian remedies of divestiture or dissolution. But if such decrees are easily subject to termination or substantial modification at the defendant's behest, perhaps within a very short time of their entry, the government will have greater incentives to insist on relief that cannot easily be modified in the future.
Not only would this distortion of the government's incentives lead to more litigation, but it could rob the public of the benefits to be gained by innovative decrees carefully tailored to address competitive problems. In May of this year, for example, the United States sued to prevent the merger of two large hospitals that dominated their market for inpatient services in Florida. United States v. Morton Plant Health Systems, M.D. Fla., Civ. No. 94-748 CIV-T-23E. Rather than insisting on "all or nothing" relief, the government agreed to a highly innovative settlement which simultaneously protects competition where it is threatened and encourages money-saving efficiencies where competition is not threatened.(36) But those benefits would be in jeopardy if the hospitals, on the sort of slim showing countenanced by the district court in this case, could have it modified to eliminate the restrictions on conduct threatening competition. Faced with this sort of risk, the Antitrust Division would have little incentive to enter such innovative decrees, and a perverse incentive to litigate to the hilt, despite the availability of a consent decree that solves the immediate competitive problem. Such distorted incentives ill serve the government, defendants, the public interest, or sound and reasonable enforcement of the antitrust laws.
The district court's erroneous ruling threatens serious harm to sound enforcement of the antitrust laws, and therefore we urge the Court to reverse and remand with instructions to reinstate the 1921 and 1954 decrees.
CERTIFICATE OF SERVICE
I, Robert J. Wiggers, a member of the bar of this Court, hereby certify that on this 29th day of November, 1994, I caused copies of the foregoing BRIEF FOR THE UNITED STATES in final form to be served by first class mail upon:
1. The district court's opinion finding that Kodak had violated Section 2 of the Sherman Act is reported as United States v. Eastman Kodak Co., 226 Fed. 62 (W.D.N.Y. 1915). The court entered a decree the following year. United States v. Eastman Kodak Co., 230 Fed. 522 (1916). While Kodak's appeal to the Supreme Court was pending, the parties reached a settlement subsequently embodied in the 1921 decree. The appeal was dismissed. 255 U.S. 578 (1921).
2. The decree also required Kodak to divest--as it did--several acquired firms.
3. The decree also included certain affirmative requirements, which have now expired, for Kodak to license its photofinishing processes and to provide technical assistance to any person seeking to establish a photofinishing business (J.A. 116-21).
4. At the retail level, Kodak testified to a 4.5% price premium over Fuji, ranging from 1% at mass merchandisers to perhaps 7% to 8% at food and drug chains (J.A. 59; 373-75, 403-04; 670).
5. At most 20,000 outlets carry 3M's film, and 10,000 carry Konica's (J.A. 488a, 519-20). Polaroid brand conventional film, made by 3M and Konica (J.A. 464, 470, 540-44), is available in stores accounting for only about 30% of U.S. film sales (J.A. 447-48, 454-55).
6. Kodak has both a volume incentive program ("VIP") (J.A. 339-43, 449-51, 506, 552-54), and explicit exclusivity arrangements (J.A. 451, 482, 501-13, 557-60).
7. Hausman measured demand elasticity for Kodak's 100 ASA 35 mm film, using Nielsen data for food stores in five cities (J.A. 367-68, 401-02; 669). His underlying data, however, also showed that in those stores Kodak film was already priced at a substantial premium, perhaps 7% to 8% (J.A. 404). Kodak's sales share for that particular type of film was 78% (J.A. 596-97; 726), and its share of all film sales in those stores was 80% in units and 83% in dollars (J.A. 726).
8. Hausman suggested that some of this difference was due to high fixed costs, but showed no personal knowledge of Kodak's fixed costs (J.A. 392-93), and Kodak, with the court's approval, refused to disclose its profit margin (J.A. 324-26). The district court opinion does not mention the subject.
9. The customer or retail dealer mailed the exposed film to Kodak for processing, and the prints were returned by mail in two to three weeks (J.A. 219-20). Kodak did the photofinishing of color film in large laboratories, supervised by engineers, due to the sensitivity of the process (J.A. 217-19). It refused, however, to process film produced by any other company, because its equipment could be contaminated by different chemicals they used (J.A. 220-21).
10. A "captive" lab is one owned by a retailer, such as Wal-Mart, to do its own work (J.A. 243). Very few retailers produce sufficient volume to make captive labs worthwhile at current levels of scale efficiencies (J.A. 286-88, 290-91). A "wholesale" lab is one that provides photofinishing services for a retailer by contract (J.A. 228). There are also mail order labs that provide relatively inexpensive service directly to consumers, but they are much slower than the others and have been rapidly losing market share, except for rural areas where the others are not convenient (J.A. 607-08; 664).
11. The joint venture was with the Actava Group (formerly Fuqua Industries) (J.A. 577-78). There was some question at the hearing regarding the extent of Kodak's control over Qualex (J.A. 238-39). After the district court entered its decision, however, Kodak bought out Actava for $150 million, and became the sole owner of Qualex. See "Kodak Buys All of Qualex," New York Times, Aug. 16, 1994, at D12.
12. Qualex has also acquired control of Lerner Processing Labs, the fifth largest wholesale photofinisher (J.A. 288-89; 730). It has rapidly become the second largest minilab operator, and is expanding those operations exponentially (J.A. 290-90a; 731).
13. It found support for this view in Patterson v. Newspaper & Mail Deliverers' Union, 13 F.3d 33 (2d Cir. 1993).
14. The court did not mention the undisputed economic inference of this: that Kodak prices film at double its marginal cost.
15. The government has never contended that the decree should be maintained if Kodak lacks market power.
16. As noted above, the decree prohibited not only tying, i.e., conditioning sales of film on the purchase of photo-finishing, but also bundling, i.e., offering film with or without photofinishing, or the use of coupons. In terminating the decree, the court eliminated the ban on tying as well as on bundling.
17. Our argument that the district court misunderstood and misapplied the legal standard governing termination of antitrust decrees presents an issue of law reviewable by this Court de novo. Our argument that the court erroneously found that Kodak no longer has market power is subject to appellate review under the clear error standard.
18. It cited this Court's seminal decision in New York State Association for Retarded Children, Inc. v. Carey, 706 F.2d 956 (2d Cir.), cert. denied, 464 U.S. 915 (1983).
19. The district court concluded both that Rufo establishes a new standard generally applicable to decree modifications and terminations (J.A. 31-34), and that the "changes sought in these antitrust decrees fit the description given by Judge Newman in Patterson, because Kodak seeks pervasive change in long-established practices affecting a large number of people, and seeks the changes to vindicate significant rights of a public nature, i.e., the consumer benefits from increased competition which Kodak claims is stifled by both decrees" (J.A. 35). It would be perverse, however, if a defendant were held to a lower burden because it sought to change long established rules of substantial public importance.
20. Dowell involved termination of a decree. Rufo, by contrast, involved a modification that the Court believed would make the decree more workable, not a termination, and so the Court did not need to determine whether the decree's purposes had already been fulfilled.
21. This Court some years earlier had indicated a receptivity to this kind of approach. Broadway Delivery Corp. v. United Parcel Service of America, Inc., 651 F.2d 122, 130 (2d Cir. 1981). That case dealt, of course, with a traditional antitrust suit where the plaintiff has the burden of proof. In this case the burden was on Kodak to show that it no longer has market power in any market governed by the decrees.
22. It cited the definition from State of New York v. Anheuser-Busch, Inc., 811 F. Supp. 848, 873 (E.D.N.Y. 1993): "market power is the ability to raise prices and maintain such prices above competitive levels."
23. Profit maximization implies a direct relationship between the excess of price over short-run marginal cost for a particular firm and its own elasticity of demand. See, e.g., Landes & Posner at 940. This relationship holds in the cases of monopolies, dominant firms, and firms selling differentiated products and pricing without cooperation with rivals.
The inference from the estimated own elasticity is that Kodak is already exercising significant market power by charging prices substantially above competitive levels. Whether Kodak could profitably raise prices from prevailing levels is irrelevant. Kodak's own elasticity of demand at the competitive price and quantity would have been far lower than that at prevailing prices.
24. 24 He could hardly do so, since three individuals he recognized as authorities in the field (J.A. 362) say the same thing. See Jean Tirole, The Theory of Industrial Organization 66 (1988); Landes & Posner, supra, 94 Harv. L. Rev. at 940.
25. 25 The explanation for this principle is that perfect competition drives price down to short-run marginal cost, and the further a market deviates from the competitive model toward monopoly the greater the difference between marginal cost and price.
26. 26 The district court never mentioned the subject of the wholesale premium.
27. 27 Agfa, Konica, 3M, and Polaroid have yet smaller shares of the market. The total share for all four is roughly 20% in units and 10% in dollars (J.A. 690-91).
28. 28 The United States in a sort of "fruit of the poisonous tree" argument had claimed that much of Kodak's immense, current reputational advantage with consumers was the result of its earlier illegal activities. The court disagreed. But we had no obligation to prove this, for under the proper legal standard it is Kodak's burden to show that it no longer has market power--whatever its source.
29. 29 Kenneth G. Elzinga & Thomas F. Hogarty, "The Problem of Market Delineation in Antitrust Suits," 18 Antitrust Bulletin 45 (1973); Kenneth G. Elzinga & Thomas F. Hogarty, "The Problem of Market Delineation Revisited: The Case of Coal," 23 Antitrust Bulletin 1 (1978); Landes & Posner, supra.
30. 30 Markets defined on this basis are known as price discrimination markets. See, e.g., U.S. Department of Justice & FTC, Horizontal Merger Guidelines § 1.22 (1992)
31. 31 In Japan, for example, Fuji is the overwhelming favorite, with some 70% of sales (J.A. 67 n.17). Kodak is third (also trailing Konica) with only 6% of sales (J.A. 491; 707). Not surprisingly, Fuji enjoys a price premium in Japan (J.A. 329).
32. 32 The court criticized the government's 1993 pricing data, citing an "entirely different" distribution system in Japan from America (J.A. 61). But both Kodak and its rivals use that distribution system and so should be affected equally by it. Moreover, the systems' inefficiencies should be reflected in higher retail prices, not lower wholesale prices, if wholesale prices are related to costs. The fact that Kodak's prices in France are almost as high as in the United States ignores the fact that Kodak's market share in France is its highest in Europe (J.A. 359), and its prices are lower elsewhere in Europe where its market share is smaller (J.A. 396-97).
33. 33 Kodak did not claim that it was entitled to termination or major modification of the 1921 decree if it still has market power. Indeed, the provisions remain important protection against anticompetitive use of that market power. Section VI prohibits voluntary exclusive dealing arrangements, which Kodak would find a relatively cheap way to exclude competitors. By conditioning price reductions on exclusivity, Kodak, with its huge advantage in sales volumes and profit margins, could make it prohibitively expensive for any of its rivals to make comparable offers, and by excluding competitors more than make up the cost of the discount through its own increased volume. Section X protects the private label film market. Since Kodak is obviously concerned that its entry into that market might cannibalize its highly lucrative "Kodak" label business (see, e.g., J.A. 327, 328), it is hard to see any reason it would want to enter other than to drive out of this actively competitive market its smallest member, 3M, which has a 4% market share (J.A. 51 n.10) and which could use its film making facilities for other lines of business (J.A. 493). Finally, as to Section VII, which bans non-price vertical restraints, until Kodak indicates--as it has not--just what marketing practices it wants to implement and why they are lawful under the Rule of Reason, there is no basis for eliminating the provision.
34. 34 It should be kept in mind that Kodak, through its Qualex subsidiary, already dominates the wholesale segment of the photofinishing market, which serves retailers (p. 7, supra).
35. 35 These include the 1920 Swift decree. United States v. Swift & Co., 1982-1 Trade Cas. (CCH) ¶ 64,464 (N.D. Ill. 1981).
36. 36 The hospitals agreed to undertake instead a new joint venture to perform specified medical services as to which there is competition in the market generally, as well as to do administrative work jointly.