Times Mirror Magazines, Inc. v. Las Vegas Sports News, L.L.C., D/B/A Las Vegas Sporting News

BARRY, Circuit Judge,

Dissenting.

How famous a mark must be before it can be afforded protection under the Federal Trademark Dilution Act (“FTDA”) is a question of first impression in this Circuit, a question which implicates the expansion of trademark rights under the Lanham Act and one which has received much judicial attention elsewhere since the passage of the FTDA. The correct answer to this question is of critical importance in order that an appropriate balance between free competition and property rights be maintained.1 The majority holds that the District Court did not err in finding that “The Sporting News” mark was sufficiently famous to merit protection under the FTDA. Because I conclude that Times Mirror has not shown and, in my view, cannot show that it is likely to satisfy the threshold fame requirement, I respectfully dissent.

I.

Fame means FAME

The FTDA offers little guidance as to what is required to find a mark famous. Thus, courts must rely on legislative history and, where helpful, look to dilution theory which has developed over years of judicial interpretation of state anti-dilution statutes. Dilution theory in the United States emanated from a 1927 Harvard Law Review article which posited that protection against dilution would fill the gap in trademark law left by infringement theory which only provided protection when a junior user applied a deceptively similar mark to similar goods to confuse a competitor’s consumers about the source of the goods. See Frank I. Schechter, The Rational Basis of Trademark Protection, 40 Harv. L.Rev. 813, 825 (1927). In the absence of a dilution cause of action, trademark law was unable to protect mark owners from the unauthorized use of a deceptively similar mark placed upon dissimilar or non-competing goods. Before passage of the FTDA, state anti-dilution statutes attempted to fill this gap in trademark law. However, because truly famous marks — and “truly” will become the operative word here — are ordinarily used on a nationwide basis but only half of the states provided remedies for dilution, Congress recognized the need for a federal anti-dilution statute. See H.R.Rep. No. 104-374, at 4 (1995), reprinted in 1995 U.S.C.C.A.N. 1029. Accordingly, the FTDA was passed to fill the gap and “to bring uniformity and consistency to the protection of famous marks.” See id. at 3.

Historically, the Lanham Act has attempted to balance the two competing goals of protecting consumers and protecting a trademark owner’s investment. The FTDA, however, is concerned only with the latter:

It does not have those twin public policy goals of the laws of trademark infringement[.] As a result there may be a kind of judicial restraint about the new law. The perception may be that it does not carry any compelling need to protect the public, and that it benefits only a coterie of American business elite, not the general public.

Jerome Gilson, 2 Trademark Protection & Practice (1999) § 5.12[1][e] at 5-272 to 5-273 (hereinafter “Gilson”). Moreover, *171there can be little doubt that Congress sought to protect only a select and narrow class of truly famous and well-recognized marks. “Without such a requirement, an anti-dilution statute becomes a rogue- law that turns every trademark, no matter how weak, into an anti-competitive weapon.” 4 J. Thomas McCarthy, McCarthy on Trademarks and Unfair Competition, (4th ed. 1999) § 24:108 at 24-210 (hereinafter “McCarthy”). “To save the dilution doctrine from abuse by plaintiffs whose marks are not famous and distinctive, a large neon sign should be placed adjacent wherever the doctrine resides, reading: ‘The Dilution Rule: Only Strong Marks Need Apply.’ ” 4 McCarthy § 24:108 at 24r-209; see also 2 Gilson § 5.12[l][b] at 5-260 (referring to class of trademarks protected by FTDA as “Supermarks”).

The legislative history of the Act is crystal clear that Congress intended courts to be highly selective in determining which marks are famous and accorded those truly famous marks an unprecedented degree of protection. A 1987 Report of the Trademark Review Commission of the United States Trademark Association (USTA) emphasized how limited this universe should be: “We believe that a limited category of trademarks, those which are truly famous and registered, are deserving of national protection from dilution[.] We therefore urge the adoption of a highly selective federal dilution statute[.]” Trademark Review Commission, Report & Recommendations, 77 Trademark Rep. 375, 455 (1987).2 The Report of the Senate Judiciary Committee on the precursor to the FTDA stated that the 1988 bill

creates a highly selective federal cause of action to protect federally registered marks that are truly famous from dilution of the distinctive quality of the mark. The provision is specifically intended to address a narrow category of famous registered trademarks where the unauthorized use by others, on dissimilar products for which the trademark is not registered, dilutes the distinctiveness of the famous work[.]
Section 43(c) of the Act is to be applied selectively and is intended to provide protection only to those marks which are both truly distinctive and famous, and therefore most likely to be'adversely affected by dilution. To protect these 'special marks, and to ensure that the bill does not supplant the current protection of trademarks based on the likelihood of confusion, the committee amended the legislation to place greater emphasis on the factors the courts must weigh in determining whether a mark possesses a sufficient level of fame and distinctive quality to qualify for federal protection from dilution.

S.Rep. No. 100-515 (reproduced in 6 McCarthy App. A5, at 41-42), 1988 U.S.C.C.A.N. 5577, at 5604(emphasis added). Examples of truly famous marks cited in a House Report on the FTDA included “Buick”, “Dupont”, and “Kodak”. See H.R.Rep. No. 104-374, at 4 (1995), reprinted in 1995 U.S.C.C.A.N; 1029, 1031. In a nutshell, the legislative history amply supports the conclusion that the FTDA should be restricted to a narrow category of marks,' ensuring that it does not swallow infringement law by allowing mark owners to end-run a likelihood of confusion analysis which they fear — or, indeed, know— they cannot win.

Despite Congressional intent that the FTDA protect only a narrow category of truly famous marks, some early judicial interpretations of the Act granted dilution protection after engaging in only a cursory analysis (or no analysis at all) of the fame of the mark. See, e.g., Gazette Newspapers, Inc. v. New Paper, Inc., 934 F.Supp. 688, 696-97 (D.Md.1996)(no separate anal*172ysis of fame); Hasbro, Inc. v. Internet Entertainment Group, Ltd., 1996 WL 84853 (W.D.Wash.1996)(entering preliminary injunction on dilution claim without discussing fame). Little if any analysis, of course, would be required to find marks such as “Buick”, “Dupont” or “Kodak” truly famous or, in the context of sports with which we deal here, that the mark “New York Yankees” is so famous that even non-sports fans are well aware of it. If, however, marks which are not such household names can be protected by the Act — and, in my view, that is a big “if — those marks must be subjected to a rigorous analysis. As one commentator has noted with some alarm:”

[Cjourts thus far have shown little inclination to limit protection to the truly famous marks envisioned by the drafters of the [FTDA]. Instead, the courts, when they acknowledge the fame requirement at all, simply state a mark’s fame in eonclusory terms without attention to the eight fame factors. Unless courts strictly adhere to the admittedly vague dictates of the federal dilution statute, federal dilution protection will surely give rise to a broad regime of trademark rights in gross.

Robert N. Klieger, Trademark Dilution: The Whittling Away of the Rational Basis for Trademark Protection, 58 U. Pitt. L.Rev. 789, 68 (Summer 1997).

This concern has not gone unnoticed, and courts are now taking pains to emphasize the rigor of the fame requirement. For example, the Ninth Circuit recently set itself apart from the expansive interpretations of the FTDA by other courts by vacating a permanent injunction after finding that plaintiffs trademarks were not sufficiently famous for FTDA protection and remanding with instructions to enter summary judgment for defendant. See Avery Dennison Corp. v. Sumpton, 189 F.3d 868 (9th Cir.1999). There, Avery Dennison, the seller of office supplies and industrial fasteners, sued an Internet email business which offered “vanity” e-mail addresses, alleging that defendant’s maintenance of the domain name registrations and <dennison.net> diluted Avery Denni-son’s trademarks. The “Avery” mark had been in continuous use since the 1930s and had been registered since 1963. The “Dennison” mark had been in continuous use since the late 1800s and registered since 1908. See Avery Dennison, 189 F.3d at 873. Avery Dennison’s annual advertising expenditures exceeded $5 million, and its annual sales reached $3 billion (although no evidence indicated what percentage of these dollar figures applied exclusively to the “Avery” or “Dennison” trademarks as opposed to the company’s other marks). See id. Avery Dennison also maintained its own website.

After reviewing dilution theory and the legislative intent behind the FTDA, the Court emphasized the role of the fame requirement in “reinstating the balance” in the Lanham Act to avoid “over-protecting trademarks, at the expense of potential non-infringing uses.” Id. at 875. Despite the fact that the registered marks had acquired distinctiveness and that four of the eight statutory fame factors favored a finding that the marks were famous, the Ninth Circuit held that, as a matter of law, Avery Dennison had failed to meet its burden of proving fame for two reasons. Id. at 876-77. First, while recognizing that fame in a “specialized market segment” might be adequate if the “diluting uses are directed narrowly at the same market segment,” the Court noted that Avery Dennison provided no evidence of customer overlap or that defendant’s customers possessed any degree of recognition of plaintiffs marks. Id. at 877-78. Second, widespread third-party use of the names “Avery” and “Dennison” undermined the famousness of the marks. Id. at 878. Thus, the Court held that the marks were not entitled to protection under the FTDA. See also I.P. Lund Trading ApS v. Kohler Co., 163 F.3d 27, 46 & 49 (1st Cir.1998)(finding mark not famous and noting “mark [must] be truly promi*173nent and renowned”; “courts should be discriminating and selective in categorizing a mark as famous”); G. Kip Edwards, Developments in Dilution Law, 579 PLI/ Pat 209, 217 (Nov.-Dec.1999)(noting that many early judicial interpretations of the FTDA neglected to heed Congressional intent that the Act be applied sparingly to truly famous marks and mistakenly granted protection to marks that were “famous” only within a specialized market niche (citing Gazette Neivspapers), but noting that the “pendulum may be swinging back toward protection only of truly famous marks” (citing Avery Dennison)).

If one heeds the legislative history, it is simply beyond the pale to find “The Sporting News” mark to be another “Buick”, “Dupont”, or “Kodak”, names which have long been associated in the public’s eye with a particular company or a particular product and which immediately strike one as being truly famous. Stated somewhat differently, to find that “The Sporting News” mark meets the fame requirement, thus entitling it to the extraordinary protection the FTDA provides — a nationwide injunction — would defeat Congress’s deliberate design.

II.

Insufficient evidence of fame

The majority, after paying lip service to the fame requirement, holds that the District Court did not err by concluding that “The Sporting News” mark is famous in the sports periodicals market, a “niche market”. I disagree. For starters, the legislative history does not mention much less embrace a so-called “niche market” theory of fame.3 Beyond that, the niche market theory risks lowering the bar for trademark protection unless it is applied prudently to cases which clearly call for such an analysis, and this is not one. For one thing, The Sporting News is directed at the general public via subscriptions and at newsstands, thereby begging the question: is not the general public the appropriate universe for assessing the fame of the mark? But even if fame exclusively within the sports periodicals market would be enough to establish fame under the FTDA, the paltry evidence here does not permit any such finding Moreover, I take issue with the rather cursory discussion by the majority, and by the District Court, of the eight statutory factors for fame.

The fundamental problem with the majority’s application of the niche market theory, sometimes known as the “big fish in a small pond theory,” to the facts of this case is that it is hard to conceive of any consumer goods or services that are not in a narrow market of some type, be it luxury cars, cameras, or sporting publications.

Courts approving the “big fish in a small pond” theory of trademark dilution fail to recognize that it threatens to overrun trademark infringement law. Trademark infringement law permits similar, or even identical, marks to coexist on non-competing goods. If even a locally famous mark can preclude all other *174marks in every channel of trade, then conceivably every trademark can be used to create a monopoly in a word or symbol — a proposition clearly contrary to the intent and practice of trademark law. It is possible to find virtually any mark to be “famous” within some market, depending on how narrowly that market is defined.

Courtland L. Reiehman, State and Federal Trademark Dilution, 17 Franchise L.J. 111, 133 (Spring 1998).

If marks can be “famous” within some market, depending on how narrowly that market is defined, then the FTDA will surely devour infringement law. Indeed, the unauthorized use of a mark in the same or a similar market is precisely what good old-fashioned infringement principles have traditionally been there to remedy once actual confusion or likelihood of confusion has been shown, and there is simply no need for dilution principles. Can one imagine a clearer case for application of those principles than if one were to begin manufacturing automobiles and calling those automobiles “Buick”? Similarly, if the parties here operate within the sports periodicals market, then this case, at least in my view, is a garden variety infringement case, and the complaint alleges just that.

Congress was quite clear, however, that the FTDA was not designed for situations in which ordinary infringement law provided a remedy but, rather, for those situations in which a truly famous mark on dissimilar products deserves, but cannot receive, protection under infringement law — those situations in which, for example, no one would ever confuse that truly famous mark with the goods or services to which it has been wrongly attached. Congress was explicit as to where protection was warranted: “DUPONT shoes, BUICK aspirin, and KODAK pianos.” H.R.Rep. No. 104-374, at 4 (1995), reprinted in 1995 U.S.C.C.A.N. at 1030. The extensive relief the FTDA authorizes, which gives the owner of the famous mark a virtual monopoly by precluding all others from using the mark “regardless of the presence or absence of ... competition between the owner of the famous mark and other parties, or likelihood of confusion,” is itself something federal trademark law had not before seen, and surely was not meant to be accorded to any marginally “famous” mark. 15 U.S.C. § 1127. It follows inexorably that if a mark is famous in the general public, it is also famous in its niche market and, in such a case, dilution and infringement theories need-not be mutually exclusive. Before, however, a Court categorically adopts the theory that a mark that is not generally renowned, but famous only in its niche market, is entitled to protection under the FTDA, the evidence of fame should be rigorously examined. Had such an examination been performed here, only one conclusion could have been reached: the evidence of fame is woefully lacking.

A. Factor (F)

The FTDA lists eight non-exclusive statutory factors for fame which a court may but is not required to consider. The “may” is important because it would make little sense to require that a mark which immediately strikes one as truly famous— again, “Buick”, “Dupont”, or “Kodak” — be analyzed for fame in accordance with these factors, although certainly such an analysis would confirm the immediate impression of fame. The less truly famous a mark is, however, the more rigorous the analysis of the statutory factors must be in light of the evidence of record.

It is Factor (F), “the degree of recognition of the mark in the trading areas and channels of trade used by the marks’ owner and the person against whom the injunction is sought,” which gives .the FTDA’s fame requirement its “teeth.” As the majority notes, however, the District Court did not explicitly address this factor — and neither did the majority.

The guidance which Factor (F) affords to courts is somewhat 'murky. The Act, *175for example, does not define “channels of trade,” although presumably that phrase means the chain of distribution of the goods featuring the mark in question, i.e. the route by which the goods travel to the ultimate consumer — here, from the publisher to the reader. Importantly, although Factor (F) focuses the analysis on the channels of trade in which the parties operate, it does not dictate the conclusion that fame solely within those channels of trade is enough for protection under the FTDA.

According to the legislative history, a finding of fame “requires substantial renown or fame within both the trading area of the mark and the trading area of the other party to the dilution suit.” S.Rep. No. 100-515 (reproduced in 6 McCarthy App. A5, at 43), 1988 U.S.C.C.A.N. at 5605(emphasis added); see also 77 Trademark Rep. at 461 (advocating that a mark “should be well known to a substantial portion of the relevant purchasers of the goods or services.”)(emphasis added). This, I note, is a higher standard than the “appreciable number of persons” standard applied in an infringement action in which a plaintiff may prevail only if it shows that an appreciable number of ordinarily prudent purchasers of the type of product in question are likely to become confused as to the source of the goods by the defendant’s use of the mark. See 77 Trademark Rep. at 461; 4 McCarthy § at 163.

The crux of any discussion of Factor (F) is whether Times Mirror is likely to prove that its mark is recognized by a substantial portion of LVSN’s potential consumers. Again, the FTDA does not quantify the requisite “degree of recognition”. Consequently, some commentators have called for a clear percentage cut-off for consumer recognition. McCarthy, for example, recommends that a plaintiffs mark must be known by more than 50% of the defendant’s potential customers in order to be considered “famous”. See 4 McCarthy § 24:92 at 24-164; see also Xuan-Thao N. Nguyen, The New Wild West: Measuring and Proving Fame and Dilution Under the Federal Trademark Dilution Act, 63 Albany L.Rev. 201, 233 (1999)(advocating a 40% rate of recognition among defendant’s potential customers in a nationwide survey). While I am not wed to any specific minimum percentage for consumer recognition, I do take issue with the fact that Times Mirror has been granted a preliminary injunction without offering any evidence whatsoever of consumer recognition in LVSN’s channel of trade.4

In the absence of evidence indicating that consumers in LVSN’s channel of trade recognize Times Mirror’s mark, it was wrong, in my view, for the District Court and the majority to conclude that *176the publications share a common market and that the mark is famous within that market. The Sporting News, moreover, is available at newsstands to members of the general public, just as Buick automobiles and Kodak film are available to the general public. Accordingly, to be entitled to a preliminary injunction, it was incumbent upon Times Mirror to demonstrate that it is likely to succeed in proving that its mark is truly famous among members of the general public and, although this should follow almost automatically, that its mark is recognized by a substantial portion of LVSN’s consumers — those who like to gamble, who read gambling publications, or who frequent casinos. Such a showing is typically achieved through a properly-conducted recognition survey.5 Certainly, Times Mirror had, ample time and notice (sixteen months passed between its discovery of LVSN’s title and the preliminary injunction hearing) to conduct a recognition survey of its mark and/or a survey to determine whether its mark would be affected by the presence of LVSN’s title in the marketplace.6

Assuming, arguendo, that a showing of fame only within the sports publication market suffices for protection under the FTDA, Congress’s intent to reserve dilution protection for a select and narrow category of truly famous marks cannot be glossed over, as the majority has done, by an unsupported finding that “The Sporting News” mark is famous within its niche and recognized by a significant portion of Las Vegas Sporting News readers. “A preliminary injunction may not be based on facts not presented at a hearing, or not presented through affidavits, deposition testimony, or other documents, about the particular situation [ ] of the moving part[y].” Adams v. Freedom Forge Corp., 204 F.3d 475, 487 (3d Cir.2000). Times Mirror has simply not come forward with any evidence of “the degree of recognition of the mark in the trading areas and channels of trade used by the marks’ owner and the person against whom the injunction is sought.” 15 U.S.C. § 1125(c)(1)(F). This failure weighs formidably against any conclusion that Times Mirror is likely to succeed on its dilution claim.

B. The remaining factors

Examination of the remaining statutory factors underscores the inadequacy of the evidence offered by Times Mirror. Factor (A), the degree of inherent or acquired distinctiveness of the mark, encompasses more than simply whether “The Sporting News” mark has inherent distinctiveness or has acquired distinctiveness through secondary meaning, as it must to be eligible for protection under the Lanham Act. This factor suggests that the degree of the mark’s distinctiveness is relevant to the fame inquiry. As discussed below with regard to Factor (G), the degree of a mark’s distinctiveness is weakened by third party use of the mark and by the descriptive nature of the mark. Therefore, while this factor favors Times Mirror, it does so only slightly.

*177Factor (B), the duration and extent of the use of the mark, means more than simply the length of time “The Sporting News” mark has been in use, but also the breadth of its distribution. ■ Times Mirror did not introduce evidence of The Sporting News’s sales figures either in toto or broken down by source, i.e. newsstand, subscription, advertising, or Internet, relying only on its weekly circulation of half a million copies in Canada and the United States. It cannot be seriously argued that this weekly circulation is not small relative to other major publications, including sports magazines, and not small period given the population of those countries.7 Thus, despite over one hundred years of publication of an inexpensive product distributed in countries in which there is a huge interest in, and concomitant market for, anything to do with sports, the relatively limited extent of The Sporting News’s circulation certainly does not compel the conclusion that the mark has generated a mental association among consumers sufficient to support a finding of fame.

Factor (C) addresses how widely and frequently a mark has been advertised or publicized which, in turn, suggests the public’s familiarity with the mark. See 4 McCarthy § 24:92. Times Mirror presented evidence that it advertises primarily by direct mail, but also on television and “occasionally” on the radio in “selected markets”. It did not, however, provide evidence of its annual advertising expenses,, nor did it detail where, when, or how the mark has been advertised. Moreover, the unadorned fact that Times Mirror has an Internet website, a fact the majority noted, is of little significance because there is no evidence regarding the extent of sales or advertising on the Internet, nor is there any evidence regarding, for example, the number of “hits” received from visitors to the website which would assist in determining the degree of consumer recognition of the mark.8

The FTDA, I note, does not specify quantitative measurements for Factors (B) and (C), such as a basic minimum for sales or advertising. When, however, the evidence supporting these factors in this case is compared to that in dilution cases in which the mark has been deemed “famous”, Times Mirror’s evidence of sales revenue and advertising expenditures falls short.9

*178Factor (G) clearly favors LVSN. Factor (G) takes into account the possibility that third party use of the mark or elements of the mark has already diluted the mark’s strength, thereby rendering the mark less famous. See 77 Trademark Rep. at 461 (“Third party uses of the same or similar marks are relevant in determining the fame and distinctiveness of the mark, since the mark must be in substantially exclusive use. If a mark is in widespread use, it may not be famous for the goods or services of one business.”).10

The words “sporting” and “news” are commonplace words in our vocabulary appearing on many items, not only publications. The majority does not acknowledge that at least six other publications use the word “sporting” in their titles: Grays Sporting Journal, . Southern Sporting Journal, Sporting Thoughts, The Sporting Scene, The Sporting Life and Sporting Green. LVSN’s use of the word “sporting” in its title describes the magazine’s content. “Sporting” is defined, in this record, as “involving betting or gambling as sporting men. Involving or inducing the taking of risk as a sporting proposition.” In Viacom, Inc. v. Ingram Enters., 141 F.3d 886 (8th Cir.1998), the Eighth Circuit held that while the trademark “Blockbuster” for a chain of video stores was strong for purposes of an infringement analysis, the mark’s strength was not necessarily sufficient to sustain a claim for dilution-by-blurring due to the ordinariness of the word “Blockbuster”. See 141 F.3d at 891-92. The court noted that “the fact that Viacom is seeking a complete monopoly on the use of a rather common word with multiple meanings would make us hesitate to uphold summary judgment on its dilution-by-blurring claim.” Id. (citing 3 McCarthy § 24:114 at 24-208) (dilution “is a potent legal tool, which must be carefully used' as a scalpel, not a sledgehammer.”). Third party use of the commonplace elements of Times Mirror’s mark weakens its fame. Factor (G), therefore, strongly favors LVSN.

The legislative history indicates that the eight factors should be weighed independently “and it is the cumulative effect of these considerations which will determine whether a mark qualifies for federal protection from dilution.” S. Rep. 100-515 (reproduced in 6 McCarthy App. A5, at 42), 1988 U.S.C.C.A.N. at 5605. Moreover, the factors should be interpreted flexibly “so that their relative weight in any given case can be balanced.” Hershey Foods Corp. v. Mars, Inc., 998 F.Supp. 500, 504 (M.D.Pa.1998). In Hershey, for example, the District Court found that even though *179six of the eight enumerated statutory factors favored Hershey (inherent distinctiveness, degree of consumer recognition, duration and extent of use, advertising and publicity, geographical extent of trading area and widespread distribution channels), Hershey’s trade dress was unlikely to meet the statute’s stringent fame requirement because of third party use of the same aspects of the trade dress and because it was not registered. Here, Factors (B), (D) and (H) favor Times Mirror because “The Sporting News” is a registered mark used continuously for over 100 years nationwide. However, there is little evidence going to Factor (B)’s extent of sales. Times Mirror has offered little or no evidence going to factors (C), (E) and (F). Factor (G) weighs strongly against Times Mirror. Finally, because third party use and the descriptive nature of the mark tend to weaken its distinctiveness, Factor (A) only slightly favors Times Mirror.

In my view, the District Court failed to sufficiently evaluate the mark — it did not consider several of the statutory factors, nor did it qualitatively weigh those factors it did consider. Even if a mark not immediately recognizable by the general public can, nonetheless, meet the fame requirement of the FTDA, and I do not believe it can, at this preliminary stage, based upon the inadequate record before us, I cannot agree that Times Mirror is likely to succeed in proving the fame of its mark.11

III.

Conclusion

The FTDA grants broad discretion to the federal courts and, as one commentator has remarked, “it is up to the judiciary to apply such potent laws with care and common sense lest they damage the competitive systems they are designed to enhance.” 4 McCarthy § 24:114 at 24-222. Lax interpretation of FTDA requirements forecasts easier lawsuits for trademark owners who will use a dilution cause of action as a “tack-on” to an infringement claim in the event that likelihood of confusion cannot be shown, and even when, as perhaps here, it can. See Klieger, Trademark Dilution, 58 U. Pitt. L.Rev. at 64 (“It may just be a matter of time before dilution eclipses confusion as the gravamen of most federal trademark actions and trademark rights in gross displace consumer protection as the defining feature of *180United States Trademark law.”); I.P. Lund Trading ApS v. Kohler Co., 163 F.3d 27, 48 (1st Cir.1998)(“Dilution laws are intended to address specific harms; they are not intended to serve as mere fallback protection for trademark owners unable to prove trademark infringement.”).

Naturally, when a court rules on a motion for a preliminary injunction, it makes an initial judgment based on an incomplete factual record; its findings of fact and conclusions of law are subject to revision based on additional discovery. The stakes are, nonetheless, high; here, for example, had the injunction not been stayed with the consent of the parties, lowering the bar for the fame requirement would have forced LVSN to alter its publication at great cost or cease publishing altogether.12

Yet again, we have recently stressed our respect for the extraordinary nature of the preliminary injunction power and the fact that “the use of judicial power to arrange relationships prior to a full determination on the merits is a weighty matter” to be reserved for extraordinary situations. Adams v. Freedom Forge Corp., 204 F.3d 475, 487 (3d Cir.2000). This is surely not such a situation. , I would vacate the preliminary injunction.

. Commentators have warned that expansion of a dilution cause of action could harm competition. See, e.g., Mark A. Lemley, The Modern Lanham Act and the Death of Common Sense, 108 Yale L.J. 1687 (May 1999); Glynn S. Lunney, Jr., Trademark Monopolies, 48 Emory L.J. 367 (Spring 1999); William Marrolet-ti, Dilution, Confusion, or Delusion? The Need for a Clear International Standard to Detennine Trademark Dilution, 25 Brook. J. Int’l L. 659 (1999); Robert N. Klieger, Trademark Dilution: The Whittling Away of the Rational Basis for Trademark Protection, 58 U. Pitt. L.Rev. 789 (Summer 1997).

. In 1988, legislation based on the Commission's proposal for a dilution statute limited only to "famous marks” was approved by the Senate, but did not survive in the House of Representatives. Subsequently, the Commission’s proposal became, in large part, the basis for the FTDA. See 4 McCarthy § 24:87.

. The only reference in the legislative history that even comes close to suggesting a so-called "niche-market” theory is found in a discussion of the degree of a mark's recognition, where the Trademark Review Commission noted that dilution might occur with respect to one universe of consumers, but not necessarily to another. "For example, if a mark is famous at the industrial level but not at the consumer level, protection may be appropriate at the industrial level but not at the consumer level.” 77 Trademark Rep. at 461. This reasoning may provide the basis for the application of a niche market theory within a specialized industrial niche. See, e.g., Syndicate Sales, Inc. v. Hampshire Paper Corp., 192 F.3d 633, 640 (7th Cir.1999)(finding niche market fame sufficient under circumstances in which both parties operated within the narrow wholesale market for plastic baskets used for funeral floral bouquets); Avery Dennison, 189 F.3d at 877-78 (involving markets for office products, industrial fasteners and email addresses); Teletech Customer Care Management (California), Inc. v. Tele-Tech Co., Inc., 977 F.Supp. 1407 (C.D.Cal.1997)(involving "teleservicing industry” where both parties provided services for large, corporate clients).

. The majority averts its gaze from what little evidence does exist relating to either party’s channel of trade. The majority of LVSN copies (approximately 22,000 out of 42,000 in circulation) are made available at no charge in Nevada. Many others are available at no charge at casinos in Mississippi, Louisiana, Atlantic City, New Jersey and Foxwood, Connecticut. Only a small percentage of copies is sold at newsstands. Of. the approximately 10,000 to 11,000 copies sent to a couple of hundred newsstands in a handful of states, only approximately 1,500 are actually sold; the rest are returned to the publisher. By contrast, The Sporting News is available nationwide through subscriptions and at newsstands.

In addition, the record does not evidence much if any overlap in advertising revenues or readership. Because LVSN is given away at casinos, it survives primarily on its advertising revenues. LVSN's advertisers tend to be "casinos, paging companies, [and] handicappers.” Times Mirror, 1999 WL 124416, *2. The Sporting News, on the other hand, advertises, inter alia, sports memorabilia, collectibles, commemorative collections, apparel, sporting equipment, tobacco products and automobiles. LVSN's competitors for advertising dollars and readership are not sports magazines, but gambling publications, such as Gaming Today, Atlantic City Magazine and Casino Player. The Sporting News reports on the six major spectator sports; LVSN reports not only on sports gambling, but also on gambling on horse racing, car racing, roulette, craps, blackjack, slots — and presidential elections.

. See, e.g., Star Markets, Ltd. v. Texaco, Inc., 950 F.Supp. 1030, 1033 & 1035 (D.Haw.1996)(secondary meaning survey found 75% of respondents associated mark “Star” with plaintiff’s grocery stores; recognition survey found that over 96% of respondents recalled plaintiff's mark when asked to name any grocery store); Ringling Bros.-Barnum & Bailey Combined Shows, Inc. v. Utah Div. Of Travel Development, 955 F.Supp. 605, 612 n. 4 (E.D.Va.1997)(40% of respondents to recognition survey associated phrase "Greatest Show on Earth” with plaintiffs circus), aff'd, 170 F.3d 449 (4th Cir.1999), cert. denied. - U.S. -. 120 S.Ct. 286. 145 L.Ed.2d 239 (1999); Hershey Foods Corp. v. Mars, Inc., 998 F.Supp. 500, 517 (M.D.Pa.1998) (94% of respondents recognized orange, brown and yellow packaging of non-labeled peanut butter candy as Reese’s brand).

. Between 1997 and 1998, The Sporting News spent $500,000 to study the market's perception of its title. See "In Brief: The More, The Merrier," Media Daily, Mar. 2, 1998. The results of this study were not, however, introduced into evidence.

. See Bill Wallace, "Web Hits Becomes Baseball's New Statistic, Knight-Ridder Tribune Business News”, Feb. 22, 2000, available at 2000 WL 14920170 (comparing distribution rates of sports publications — 3.2 million weekly distribution of Sports Illustrated, for example — and noting “second-tier” Sporting News's half million "static circulation” rate.)

. The District Court's fame analysis appears to have added an additional factor to the eight statutory factors in emphasizing that "Times Mirror has spent millions of dollars improving the magazine.” Times Mirror, 1999 WL 124416, *5. Large expenditures aimed at retooling a product do not contribute to establishing fame unless it can be shown that those efforts were effective among the relevant group of consumers. While it is possible that a company's investment in its product may result in the heightened fame of its mark, it is far from clear whether that has occurred here.

. See, for example: Eli Lilly & Co. v. Natural Answers, Inc., 86 F.Supp.2d 834, 849 (S.D.Ind. 2000)(findmg "Prozac” famous; $12 billion in sales over twelve year period and "massive” unsolicited publicity rendering mark part of the "popular lexicon”); Planet Hollywood (Region IV), Inc. v. Hollywood Casino Corp., 80 F.Supp.2d 815, 840 (N.D.Ill.1999)(finding "Planet Hollywood” mark famous and noting annual sales of more than $195 million in merchandise bearing mark); NBA Properties v. Untertainment Records, L.L.C., No. 99-2933, 1999 WL 335147, *7 (S.D.N.Y. May 26, 1999)(finding NBA logo famous; logo appeared on $1.6 billion of merchandise over three year period and mark was widely promoted); Nabisco, Inc. v. PF Brands, Inc., 191 F.3d 208 (2d Cir.1999) (finding Pepperidge Farm Goldfish crackers famous; $120 million three year marketing campaign and $200 million annual net sales); Ringling Bros., 955 F.Supp. at 609 (finding slogan "Greatest Show on Earth” famous; over $103 million in annual sales derived from goods bearing slogan and over $19 million in annual advertising expenditures); American Exp. Co. v. CFK, Inc., 947 F.Supp. *178310, 312 (E.D.Mich.1996)(finding slogan “Don’t Leave Home Without It” famous; over $600 million in marketing expenditures 'over six years).

. See, e.g., Avery Dennison Corp. v. Sumpton, 189 F.3d 868 (9th Cir.1999); Carnival Corp. v. SeaEscape Casino Cruises, Inc., 74 F.Supp.2d 1261, 1271 (S.D.Fla.1999)("the word ‘fun’ is used by many other businesses in the travel, gaming, and entertainment industries ... cut[ting] against Carnival’s dilution claim”); Michael Caruso & Co., Inc. v. Estefan Enterprises, Inc., 994 F.Supp. 1454, 1463 (S.D.Fla.1998)(extensive third party use of word "bongo” undermines inherent distinctiveness of mark), aff'd without opinion, 166 F.3d 353 (11th Cir.1998); Hershey, 998 F.Supp. at 517 (finding trade dress not sufficiently famous and noting several examples of third party’s trade dress in food industry similar to plaintiff's color combination and lettering); Sports Authority v. Abercrombie & Fitch, Inc., 965 F.Supp. 925, 941 (E.D.Mich.1997)(third-party use of "authority,” whether or not in the relevant market, diminishes any distinctive 'or famous aspects of mark rendering it "not so famous as to deserve protection” under the FTDA); Trustees of Columbia University v. Columbia/HCA Healthcare Corp., 964 F.Supp. 733, 744 & 750 (S.D.N.Y.1997)(fame of mark "Columbia” for healthcare services "has been seriously undermined by third party use of the same or similar marks” both within the health care industry and in other industries); Star Markets, 950 F.Supp. at 1035 (noting multiple third party uses of "Star” and "Star Markets” in food industry and unrelated industries); Golden Bear Int’l, Inc. v. Bear U.S.A., Inc., 969 F.Supp. 742, 749 (N.D.Ga.1996)(third parties extensively used both the word "bear” and a bear design in connection with the sale of sporting goods and clothes).

. My disagreement with the majority rests primarily on my conclusion that Times Mirror has not come close to satisfying the threshold requirement of fame to qualify for protection under the FTDA. It goes without saying, therefore, that I would also disagree that Times Mirror was likely to prevail on its dilution claim. One observation: the majority holds that the District Court did not err in applying what have become known as the “Sweet factors” to determine whether LVSN's use blurred and, therefore, diluted, Times Mirror’s mark for “The Sporting News”. In addition to the Sweet factors, which have been roundly criticized, the majority appears to have adopted the multiple factor test articulated in Nabisco, Inc. v. PF Brands, Inc., 191 F.3d 208 (2d Cir.1999), which includes the factor of likelihood of confusion, a factor which is required in a standard infringement analysis, but not in a dilution analysis. While I have no difficulty with adopting an appropriate list of factors for consideration, I note that we are the only Circuit to have considered the applicability of the Sweet factors — and the Nabisco factors— which has not articulated a specific critique or rejected some or all of the factors. We should do so as well.

The majority also holds that the District Court did not err in finding that irreparable injury may be shown even in the absence of actual economic harm, presumably siding with the Second Circuit and rejecting the Fourth Circuit’s position on the issue. Compare Ringling Bros.-Barnum & Bailey Combined Shows, Inc. v. Utah Div. of Travel Development, 170 F.3d 449, 461 (4th Cir.1999), cert. denied, - U.S. -, 120 S.Ct. 286, 145 L.Ed.2d 239 (1999), with Nabisco, 191 F.3d at 223-24. I agree, and note only that it would be well-nigh impossible for a widely sold product such as Kodak to show that its sales have been impacted by a diluting use of its mark. Indeed, Kodak’s sales might well be increasing even as the distinctiveness of its truly famous mark is being whittled away by an unauthorized user. See S.Rep. No. 100-515, at 108 (noting that distinctive quality of a mark "could be materially reduced during a period of rising sales”).

. A District Court's review of the merits of a dilution claim at the preliminary injunction stage may also be significant because, as at least one court has held, the cause of action is essentially equitable in nature and may not provide a right to a jury trial. See Ringling Bros., 955 F.Supp. 598, 605 (E.D.Va.1997), aff'd on other grounds, 170 F.3d 449 (4th Cir.1999)(reserving constitutional issue for another day), cert. denied, - U.S. -, 120 S.Ct. 286, 145 L.Ed.2d 239 (1999); see also 25 U.S.C. §§ 1116(a), 1117(a), 1118; 2 Gilson § 5.12[1][c][vii].