|
|
AND CONCLUSIONS OF LAW Dated: July 19, 2002 COUNSEL FOR PLAINTIFF
COLM F. CONNOLLY
Paulette K. Nash
William E. Berlin
REDACTED
As set forth in the following Proposed Findings of Fact and Conclusions of Law, Dentsply has unlawfully maintained its monopoly power in the prefabricated artificial tooth market in the United States in violation of Section 2 of the Sherman Act, 15 U.S.C. § 2, and unreasonably restrained competition under Section 1 of the Sherman Act, 15 U.S.C. § 1, and Section 3 of the Clayton Act, 15 U.S.C. § 14. Since at least 1976, if not before, Dentsply has terminated, or threatened to terminate, dealers selling Dentsply's Trubyte teeth if they add a competing brand. In 1993, this exclusionary policy was memorialized as Dealer Criterion 6, one of the two practices challenged in this case. The other is Dentsply's agreements with new dealers to drop some, or all, competing tooth brands in order to obtain Dentsply's Trubyte line in the first place. Proposed Findings of Fact ("PFF") ¶¶ 37-43. Through this exclusionary conduct, Dentsply has foreclosed its closest competitors from approximately 80% of the laboratory dealer outlets in the country. PFF ¶¶ 242-48. A network of dental lab dealers is necessary in order to compete effectively in the market. PFF ¶¶ 61-167. Dentsply's conduct has maintained its monopoly power, reduced competition, and increased prices. Indeed, the record shows, and Dentsply concedes, that in the absence of this challenged conduct:
Given these conceded effects, Dentsply bears a heavy burden to come forward with a nonpretextual, procompetitive rationale for its conduct. It has not done so here. The theoretical justification offered by Dentsply's expert, Professor Howard Marvel, is a pretext for the real reason that is expressed quite clearly in contemporaneous evidence: to "block competitive distribution points"; "tie-up dealers"; prevent competitors from gaining a "foothold" in the U.S. market; and to [REDACTED TEXT] posed by dealers selling the teeth of Dentsply's primary competitors. PFF ¶¶ 231-41. The United States' expert, Dr. David Reitman has spent five years studying this industry and reviewing the extensive evidentiary record from this case and the investigation leading up to it — including over 100 deposition transcripts, 20-25 interviews of market participants, 10 boxes of documents, and various types of market data, as well as visiting facilities of firms at each level of distribution. PFF ¶ 285. He has concluded that the substantial anticompetitive effects of Dentsply's conduct outweigh any possible procompetitive benefits. PFF ¶ 329. The net effect on competition and consumers is clear. Competition has been reduced, and consumers have had fewer choices and paid higher prices. PFF ¶¶ 263-84. The evidence on five key issues in the case -- monopoly power, foreclosure, agreement, effects, and Dentsply's justifications -- is summarized below. Monopoly power. Dentsply cannot seriously dispute that its market share in the artificial tooth market is approximately 75%-80%, a share sufficiently high to give rise to a presumption of monopoly power. Market share surveys commissioned by Sam Thumim, Dentsply's own Manager of Market Research who testified during the government's case and was not cross examined by Dentsply, show that its share has been roughly 80% for at least the past 10 years. PFF ¶¶ 169-72. Dentsply has also controlled price in the market. By successfully excluding its closest competitors from the dealers necessary to compete effectively, it has been able to charge monopoly prices. According to William Turner, Senior Product Manager for Trubyte's tooth products for almost 10 years, and Steve Jenson, the current General Manager of the Trubyte Division, Dentsply's high prices have acted as a "price umbrella," under which other companies compete in the marketplace. PFF ¶¶ 173-78. A particularly telling example of Dentsply's monopoly power occurred during a several-year period in the 1990's, prior to Dentsply's introduction of its Portrait tooth line. Its Bioform premium teeth were a poor match to the popular Vita shade guide and, apart from their shading, these teeth were considered aesthetically inferior in other respects to the teeth made by both Vita and Ivoclar. Dentsply's teeth were also significantly higher priced. Yet, it was still able to maintain its market share during this period. When labs received a prescription for a Vita-shaded tooth, they substituted the higher priced, poorly-matched Dentsply tooth in 72% of the cases. Not surprisingly, several dealers tried to add the Vita tooth line in response to the requests of their lab customers. Yet Dentsply consistently enforced Dealer Criterion 6, and successfully blocked Vita from each of these dealers. PFF ¶¶ 191-211. At trial, Dentsply touted the introduction of its Portrait teeth as evidence that it is not a monopolist. Yet the evidence shows that Dentsply was receiving complaints, from both dealers and labs, about the poor aesthetics and shading of its teeth for at least a few years. It was late in addressing the problem. During the time in which Dentsply was working to develop its Portrait line, it could have competed on the merits by permitting dealers to decide freely what tooth lines they should carry. Instead, it threatened those dealers with the loss of their Trubyte tooth business, convincing them not to add the aesthetically superior and less expensive teeth requested by their lab customers. PFF ¶¶ 191-211. Dentsply's monopoly power is protected by the high barrier to successful entry and expansion created by its exclusionary conduct. Dentsply's primary competitors, Vita and Ivoclar, have competed in the market for over 20 years. Despite substantial efforts, their market shares are stuck in the mid-single digits. Dentsply's conduct has completely excluded at least one firm from the market, delayed the entry of another, and significantly limited the success of two firms, Heraeus Kulzer and Davis Schottlander, that have just recently entered. PFF ¶¶ 212-29. Dentsply has tried to make much of the unimpressive entry of these two firms, probably because it is the only evidence of entry it could find. Yet the entrants' experience confirms that it is virtually impossible for a new entrant to find dealer distribution in the face of Dentsply's exclusive dealing agreements, and that it is difficult to achieve any kind of success without dealer distribution. If actions speak louder than words, then Dentsply's lack of any reaction -- in terms of its pricing or otherwise — to this recent entry speaks volumes. Dentsply's monopoly power is far from being eroded by either of these entrants. PFF ¶¶ 216-19.(1) Foreclosure. Dentsply's exclusive dealing agreements have done what they were intended to do — "tie-up" the "key dealers" selling Trubyte teeth. Dr. Reitman's analysis shows that Dentsply has foreclosed its competitors from approximately 80% of the laboratory dealer outlets in the country. PFF ¶¶ 242-48. Dentsply did not offer into evidence their own foreclosure rate analysis. Instead, they argued that because direct sales to labs are possible, Dr. Reitman should not have focused on dealer outlets when measuring foreclosure. Dr. Reitman's methodology was proper given the abundant evidence that dealer distribution is necessary for a tooth supplier to be an effective competitor in this market. That evidence comes from numerous sources: from the dealers, who described the valuable services they provide; from the labs, in the form of Dr. Jerry Wind's survey of dental laboratories; and from the rival tooth manufacturers, that have tried repeatedly to develop a dealer network. Most significantly, that evidence comes also from Dentsply, which has a diverse dealer network of national, regional and local dealers collectively maintaining Trubyte tooth stocks in over 100 locations throughout the country. [REDACTED TEXT] Dealers are important in this industry, to Dentsply and to its competitors alike. That is why Dentsply imposed and has consistently enforced Dealer Criterion 6, and that is why it is defending it in this litigation. PFF ¶¶ 61-167. Agreement. When Dentsply has recognized new dealers, it has explicitly required the dealer to agree to drop some or all competing tooth brands -- particularly, those sold by its closest competitors, Vita and Ivoclar. Dentsply cannot seriously dispute that these are "agreements" for purposes of Section 1 of the Sherman Act and Section 3 of the Clayton Act. PFF ¶¶ 293. Dealer Criterion 6 also constitutes an agreement. It is not a unilateral policy that is announced and automatically enforced. Dentsply has engaged in negotiations — at times, lengthy negotiations — with dealers to persuade them not to add competitive brands. It has coerced dealers to agree not only by threatening them with the loss of the Trubyte tooth business, but with the loss of other product lines as well. In the case of Frink Dental, Dentsply sent three high-level executives, including its chief executive officer, to Illinois to try to talk Frink's owner, Tom Cavanaugh out of adding the Ivoclar line. When he did so anyway, Dentsply terminated him not only as a tooth dealer but as a Trubyte merchandise dealer as well. Dentsply then threatened him with the loss of other business, until Mr. Cavanaugh finally relented and dropped the Ivoclar line. Dentsply has done what it takes to make sure its dealers agree not to add competing tooth brands. PFF ¶¶ 287-92. Effects. Dentsply's monopoly power and the very high foreclosure rate create a powerful presumption of anticompetitive effects. Here, there is also ample direct evidence of harm to competition and consumers. Dentsply's concession that it would lose sales and market share without Dealer Criterion 6 means that consumer preferences in today's market are being frustrated. Some labs that want to buy the teeth of Vita, Ivoclar, and others, are not buying them because those teeth are not available through the same dealers selling Trubyte teeth. PFF ¶¶ 265-74. There is no dispute that prices will fall in the absence of Dentsply exclusive dealing. Dentsply and its expert Professor Marvel concede that. The testimony of Dentsply's competitors showed that they price more aggressively in areas in which they have better dealer distribution, and would lower their prices even further if they developed a better dealer network. Dr. Reitman's analysis of the Wind Survey data confirmed and quantified this price effect, concluding that premium tooth prices would fall by approximately 4%-5% in just three months. Additional, long-term decreases are likely as well. PFF ¶¶ 275-80. Consumers will benefit not only from these lower prices but from the greater competition that will result if Dentsply's conduct is enjoined. On a level-playing field, both Dentsply and its rivals will compete harder. The competitors testified that they will promote their products more vigorously. Likewise, Dentsply will try harder to regain its lost market share, with any number of procompetitive tactics — such as more research and development, more sales and marketing expenditures, or a bigger sales force. PFF ¶¶ 282. Dentsply's business justifications. Professor Marvel's free riding justification is pretextual and not supported by the facts. As noted above, the real reason Dentsply uses exclusive dealing is to exclude its competitors from competing effectively. Professor Marvel claims that Dentsply uses exclusive dealing, and charges higher prices, so it can advertise its teeth more. Yet the evidence shows that Dentsply will engage in more, not less, promotion if it is no longer able to practice exclusive dealing. PFF ¶¶ 373-77. The evidence also undermines certain necessary conditions that must apply for Professor Marvel's free riding theory to work. One is that dealers must steer lab orders for Dentsply teeth to other brands through the use of what Professor Marvel calls "bait and switch" tactics. There is no evidence that dealers in this industry do that. PFF ¶¶ 343-49. Professor Marvel's theory is just that -- a theory of an expert retained in litigation. It is not grounded in, or supported by, the facts of this case. Dr. Reitman, however, who began investigating the issue of procompetitive justifications five years ago, consistently cited the detailed factual support for his conclusions that the procompetitive benefits, if there are any, are negligible and outweighed by the significant anticompetitive harm that has been demonstrated. PFF ¶¶ 328-29.
1. The relevant product market for purposes of this case is the sale of prefabricated artificial teeth in the United States. GX 445 at 6-8.(2) 2. Artificial teeth today are manufactured in either porcelain or plastic. In order to match the different characteristics of a person's mouth, they are made in thousands of different shades, sizes and shapes. Teeth are made in different grades of quality, commonly known as "premium," "mid-line," "economy," and "sub-economy." Weinstock Tr. 81-84; Reitman Tr. 1479-80. 3. Artificial teeth are manufactured for use in dentures. A denture is a removable prosthetic device comprised of artificial teeth fixed in an acrylic or other base material to replace some or all of a person's natural teeth. D.I. 368, Exh. 1, Stipulation 11. 4. "Removable" appliances are ones that patients can remove from their mouth themselves, clean them and place them back in. Ryan Tr. 1206. This can include either full and partial dentures. Weinstock Tr. 85. 5. "Fixed" appliances, by contrast, include crowns, bridges, and implants. A crown is a single, individual tooth restoration. A bridge is a restoration of at least three units bridging a gap of at least one missing tooth. An implant case is where a device is actually screwed into the bone. Weinstock Tr. 85-86. 6. The term "combination case" refers to the use of both fixed and removable appliances. Ryan Tr. 1208. 7. Dental laboratories purchase almost all of the artificial teeth sold in the United States and use the teeth to make dentures. Labs buy artificial teeth on cards containing six (for anteriors) or eight (for posteriors) teeth. A full denture, i.e., one that replaces all natural teeth, requires 28 teeth from a total of four tooth cards. When fabricating a partial denture, a dental lab may only use a portion of the teeth on a card. The remaining teeth are the tooth cards known as "broken sets." D.I. 368, Exh. 1, Stipulations 13-16. 8. Labs fabricate dentures according to the prescription, impression and any other information provided to the lab by the dentist. Weinstock Tr. 81; Ryan Tr. 1211-17 (describing process by which denture is made). The process of fabricating a new denture involves "a rather large number of steps," and any given denture case goes back and forth between the lab and the dentist several times. Ryan Tr. 1211. 9. The market demands that this process move quickly as possible. When a patient is waiting for a new denture to be fabricated, or an existing one repaired, they want the work done as quickly as possible. Reitman Tr. 1480-81 ("there is someone sitting there waiting, a patient waiting for their denture to come back, not wanting a lengthy process"); Weinstock Tr. 90 ("they are anticipating something that fits better, chews better, gets rid of pain. They want it as fast as possible"); Ryan Tr. 1225 (denture patients often in pain or undergoing major life event requiring denture work); Armstrong Tr. 2369 (patients value fast service particularly when losing anterior teeth). 10. There are four distinctive characteristics of artificial teeth that bear on how they are distributed: (1) teeth are available in thousands of different mould and shade combinations; (2) they are ordered frequently, generally daily; (3) labs, dentists and patients value quick turnaround times in obtaining teeth; and (4) the handling of restocking and returns of teeth is a very labor-intensive effort. Reitman Tr. 1479-81; GX. 364-A.(3) 11. Dentsply International, Inc. is a Delaware for-profit corporation headquartered in York, Pennsylvania. Dentsply transacts business in, and is found within, the District of Delaware within the meaning of 15 U.S.C. § 22. Dentsply's business activities that are the subject of this lawsuit are within the flow of, and substantially affect, interstate trade and commerce. D.I. 368, Exh. 1, Stipulations 1, 4-5. 12. Dentsply's artificial teeth are developed, designed, sold, and marketed by its Trubyte Division, located in York, Pennsylvania. Dentsply manufactures artificial teeth in the premium (under the names "Portrait," "TruBlend," "Bioblend" and "Bioform"), mid-range ("Biotone") and economy ("New Hue" and "Classic") segments. D.I. 368, Exh. 1, Stipulations 8-9; Jenson Tr. 2108, 2116-17. Dentsply does not compete in the sub-economy tooth segment. Jenson Tr. 2250-51. 13. Dentsply sells its artificial teeth exclusively to independent dealers. Dentsply does not own the dealers it has authorized to distribute Trubyte teeth. D.I. 368, Exh. 1, Stipulations 17-18. 14. Dentsply has been the dominant tooth manufacturer in the United States market for a very long time. Its current market share is between 75% and 80%. Reitman Tr. 1471-72; [REDACTED TEXT] 15. In 2001, Dentsply's gross tooth sales to dealers were [REDACTED TEXT] Net sales, taking into account broken sets and other tooth returns, totaled [REDACTED TEXT] DX 1650; Jenson Tr. 2253. Dentsply's Trubyte Division also sells lab merchandise products. Teeth, however, represent approximately [REDACTED TEXT] of the division's revenue. Jenson Tr. 2255-56.
16. Ivoclar Vivadent, Inc., headquartered in Liechtenstein, is a manufacturer and marketer of dental restorative materials, including artificial teeth. Ganley Tr. 982-83. Its U.S. subsidiary, Ivoclar Vivadent, Inc. ("Ivoclar"), is based in Amherst, New York and is responsible for marketing Ivoclar teeth in the United States market. Ganley Tr. 982-83. Its president is Robert Ganley. He has been involved in the sale of Ivoclar teeth in the United States market since 1986. Ganley Tr. 983. 17. Ivoclar sells a number of different lines of artificial teeth. Among its premium plastic teeth are the Antaris and Postaris teeth, which were introduced by Ivoclar in the 1990s and made of a material that is more resistant to wear. Ganley Tr. 984, 1013. 18. Ivoclar is one of Dentsply's two primary competitors in the tooth market. Clark Tr. 2683-84; Miles Tr. 3461, 3494; Jenson Tr. 2249-50. Although it is Dentsply's closest competitor in terms of tooth sales, Ivoclar's market share, at 5%, is 15 times smaller than Dentsply's market share. Reitman Tr. 1472.
19. Vita Zahnfabrik is a German manufacturer of artificial teeth. Whitehill Tr. 221. Vita Zahnfabrik sells teeth in the United States through an affiliated importer and distributor named Vident. Whitehill Tr. 288-89. Vident is a closely held California corporation owned, in part, by the same family that owns Vita Zahnfabrik. Whitehill Tr. 222. Vident's president is Wayne Whitehill, who has been involved in the sale of Vita teeth since they were first imported into the United States market in the 1970's. Whitehill Tr. 221, 223. 20. Vident sells both porcelain and plastic (or "resin") teeth in the United States. The brand name of the resin teeth is "Vitapan." Whitehill Tr. 225. 21. Vident has been the entity responsible for marketing the Vita Classical Shade Guide in the United States market. Whitehill Tr. 231-32. A shade guide is used by dentists to match the shade of an artificial tooth (or crown, bridge, etc.) with the shade of a patient's natural dentition. Whitehill Tr. 230-31. The Vita Classical Shade Guide is the most popular shade guide in the market, used by approximately [REDACTED TEXT] of the dentists in the United States. Whitehill Tr. 231-32. 22. Vita, through its importer Vident, is the other primary competitor to Dentsply in the United States tooth market. Clark Tr. 2683-84; Miles Tr. 3461, 3494; Jenson Tr. 2249-50. Vita's market share is approximately 2%-3%. Reitman Tr. 1472; [REDACTED TEXT]
23. Myerson LLC is a tooth manufacturer based in Chicago, Illinois selling premium (Myerson, Universal, Swissedent), economy (Kenson), and midline teeth. At one time, Myerson was a free-standing division within the Austenal Corporation. In January 2002, Dentsply acquired Austenal, and Myerson became a wholly separate company. Myerson's president and chief operating officer is James Swartout, who has been with the company (and, before that, Austenal) since 1994. Swartout Tr. 1291-95. 24. Myerson teeth have been sold in the United States market since the company was founded in Cambridge, Massachusetts in 1917. Dr. Myerson was a Professor of Dentistry at Harvard Dental School, and hand carved almost all of Myerson's teeth. Myerson was a pioneer in cross-linked resin technology and in the move from using porcelain to plastic to manufacture artificial teeth. Swartout Tr. 1293, 1295. 25. Myerson's market share is approximately 3%. Reitman Tr. 1472.
26. American Tooth Industries ("ATI") sells teeth under the Imperial and Justi brand names. ATI's share is roughly 2%. Reitman Tr. 1472. 27. Universal is a diminishing competitor in the market. Jenson Tr. 2250 (level of activity "certainly trending downward"). In the fall of 2001, it sold some of its tooth lines to Myerson. Swartout Tr. 1295, 1340-42. Its market share is currently between 1% and 2%. Reitman Tr. 1472. 28. Heraeus Kulzer began selling its mid-line teeth, called "Jeldent Basic," in the United States market two years ago. Becker Tr. at 1817-1818. Its market share is about 1%. Reitman Tr. 1472; Marvel Tr. 3726. 29. Davis Schottlander & Davis Ltd. is an English company that sells a premium, Vita-shaded tooth under the brand name "Enigma." It is distributed in the United States by Dillon Company, Inc, which is also referred to as Leach & Dillon. Dillon Tr. at 4079-80, 4086, 4088. The market share of Enigma teeth is negligible. Reitman Tr. 1472
30. Dental laboratory dealers, like the ones to which Dentsply sells its teeth, are dealers carrying the full range of products that dental labs use. Weinstock Tr. 101-02; Reitman Tr. 1482-83. These products can include artificial teeth, metals, porcelains, teeth, acrylics, waxes, and anything else necessary to fabricate fixed or removable restorations. Weinstock Tr. 93. 31. Lab dealers that sell teeth vary in the size and scope of their operations. In general, there are three main types of tooth dealers — national, regional, and local. (a) National tooth dealers, such as Zahn Dental and Patterson Dental, sell teeth nationwide through a network of tooth stock inventories scattered throughout the country. Whitehill Tr. 244-45. (b) Regional tooth dealers are those that are particularly strong in certain regions of the country and have multiple tooth stocks scattered throughout the states in which they sell. Whitehill Tr. 245. (c) Local, specialty tooth dealers typically operate within a single state or single city. They almost always have just one tooth stock. They are much smaller organizations than national or regional dealers, carry a narrower range of products, and have fewer resources such as catalogues and sales representatives. Whitehill Tr. 245-46; Reitman Tr. 1512. 32. Due to the thousands of mould and shade combinations of artificial teeth, most tooth dealers carry large inventories of teeth. Weinstock Tr. 82; Reitman Tr. 1481. A dealer's "tooth counter" is a separate part of a laboratory dealer dedicated almost entirely to handling teeth. Weinstock Tr. 104-05; Reitman 1481-82. Tooth counters are extremely labor-intensive operations, requiring the employment of friendly, detail-oriented customer service personnel. Weinstock Tr. 126-27; Reitman Tr. 1479.
33. There are approximately 16,000 labs that perform fixed and/or removable work in the United States. Weinstock Tr. 86. Of these, approximately 7,000 fabricate dentures. Weinstock Tr. 86; Jenson Tr. 2247; Reitman Tr. 1484. 34. The 7,000 labs that fabricate dentures are a very heterogeneous group. Reitman Tr. 1484; Jenson Tr. 2247 ("highly fragmented" in size). (a) The large labs are those employing 25 or more denture technicians. There are only approximately 500 labs of this size (or only 7% of the total) in the country. Weinstock Tr. 88. (b) The mid-size labs employ between four and 25 technicians. There are approximately 700-800 mid-size labs (or 11% of the total) in the country. Weinstock Tr. 88. (c) The remaining 82% are small labs, defined as labs employing four or fewer technicians. Weinstock Tr. 88. [REDACTED TEXT] See also Jenson Tr. 2247 (approximately 80% of labs employ 1-5 people); Mariacher Tr. 2895 (75% of labs employ 3-5 people); [REDACTED TEXT] 35. Denture labs compete with each other on the basis of price and service. Weinstock Tr. 89. Patients and dentists value fast service, particularly in the case of lost or damaged dentures. Weinstock Tr. 89; Reitman Tr. 1480-81. 36. Labs are the relevant consumer for prefabricated artificial teeth because they choose the brand of tooth used in a denture in the majority of cases. Pohl Tr. 1911; Reitman Tr. 3931-32. Ryan Tr. 1210, 1215-16, 1220-22 (dentists usually provide only a shade); Armstrong Tr. 2332-33 (only 10% of dentist prescriptions specify tooth brand); Jenson Tr. 2141 (same).
37. This case challenges Dentsply's exclusionary conduct that has unlawfully maintained its monopoly and substantially restrained competition in the artificial tooth market. Under scrutiny here are two aspects of Dentsply's exclusionary policies: (i) its agreements with dealers that they will lose their Trubyte tooth business if they add a competing brand of teeth; and (ii) its agreements with new dealers to drop some, or all, competing tooth brands in order to obtain the Trubyte line in the first place. Reitman Tr. 1521-23.
38. Since at least 1976, if not before, Dentsply has terminated, or threatened to terminate, dealers selling Trubyte teeth if they add a competing tooth brand. Reitman Tr. 1521-22. In 1988, Dentsply terminated Frink Dental of Elk Grove, Illinois as both a tooth and merchandise dealer when it began selling Ivoclar teeth. Cavanaugh Tr. 700-01; Brennan Tr. 1720. 39. In publishing its "Dealer Criteria" in February 1993, Dentsply expressly stated its refusal to deal with dealers that added its rivals' tooth lines. The exclusionary policy that had been applied to Frink Dental was memorialized in Dealer Criterion 6, which reads, "[i]n order to effectively promote Dentsply/York products, dealers that are recognized as authorized distributors may not add further tooth lines to their product offering." GX 31. Dentsply permitted dealers to keep selling any competing brands, commonly called "grandfathered" brands, they were carrying as of the date Dealer Criterion 6 was formally announced. Weinstock Tr. 41; Reitman Tr. 3942. 40. The express purpose of Dealer Criterion 6 is to "block competitive distribution points" and "[t]ie up dealers." GX 171 at DPLY-A 004360; Clark Tr. 2608 (GX 171 a "reiteration" of Dealer Criterion 6). 41. In at least the past 15 years, no dealer has agreed to walk away from its Trubyte tooth business to take on a competitive line. Reitman Tr. 1514-15; Jenson Tr. 2287; Clark Tr. 2631; Pohl Tr. 1907.
42. On several occasions, Dentsply has required dealers to drop some, or all, competing tooth brands in order to obtain the Trubyte tooth line in the first place. For example, in 1992, Dentsply recognized Jan Dental in exchange for its agreement to stop selling Vita, Kenson, Dentorium, and Justi teeth. GX 24, 26; Pohl Tr. 1908-1910. Two years later, Dentsply authorized Darby as a Trubyte tooth dealer upon Darby's agreement not to add the Vita tooth line. GX 82 at DS 015663; Clark Tr. 2636. 43. The agreements have had an anticompetitive purpose as well. E.g., GX 26 at DS 016474 ("[o]pening Jan with teeth will increase our presence within the laboratory market and eliminate several competitors"); GX 86 at DS 015805 (DTS recognized to "fully eliminate the competitive threat they pose by representing Vita and Ivoclar in three of four regions [in which DTS operated]").
44. Testimony and documents introduced at trial establish that Dentsply has for at least the past 14 years coerced dealers to agree to drop competing brands from, or not add them to, their product offerings. These incidents began at least as early as 1988 and include at least 8 separate incidents since this case was filed in January 1999. As a result, 12 separate brands of teeth have been excluded, and counting each time a single brand has been excluded from a single dealer, there have been a total of 25 incidents where a rival brand has been excluded from a dealer carrying Trubyte teeth. These incidents involve a total of 11 separate dealers, located throughout the United States from New York to California.
45. Frink Dental, already an effective dealer for other Ivoclar products, took on the Ivoclar tooth line in 1988 in response to customer requests for Ivoclar teeth. Ganley Tr. 992-93, Cavanaugh Tr. 724. After Frink began carrying Ivoclar teeth, Mr. Cavanaugh of Frink was told in a letter from Bob Brennan, then Trubyte General Manager, and in a meeting with Mr. Brennan and Mr. Borgelt, then Dentsply's president, that Frink had an agreement with Dentsply to sell its teeth exclusively. Either Frink had to cease selling Ivoclar immediately or Dentsply would stop selling Frink all Trubyte products, including teeth and merchandise. Cavanaugh Tr. 692-99, 727-28. Borgelt explained Dentsply's position by stating "we cannot let Ivoclar get a foothold in the United States. This is our most highly profitable product out of our Dentsply division." Cavanaugh Tr. 695. When Mr. Cavanaugh decided to keep the Ivoclar line, Dentsply terminated Frink for all Trubyte products, both teeth and merchandise, Cavanaugh Tr. 699-701, because Dentsply "wanted to make a strong point." Brennan Tr. 1720. Other dealers provided teeth and other Trubyte products to Frink at cost, because they felt Dentsply was trying to exercise too much control. Cavanaugh Tr. 701-05. Over time, Dentsply tracked down all but one of the dealers and threatened to cut them off if they continued to supply Frink, and as a result of these threats, these dealers stopped supplying Frink. Cavanaugh Tr. 705-07.(4) After consulting with his sales force, Mr. Cavanaugh gave up the Ivoclar tooth line, and when he told Mr. Brennan he was dropping Ivoclar and returning to Dentsply Mr. Brennan told him he would be re-established as a Trubyte dealer from the day Frink stopped selling Ivoclar. Cavanaugh Tr. 712-14, Ganley Tr. 1105. See also Reitman Tr. 1523. Moreover, after Frink was reestablished as a Trubyte dealer, Cavanaugh considered carrying other lines. Although he didn't do so because of what happened after he added Ivoclar, he would have added Myerson if he could have. Cavanaugh Tr. 717-18.
46. In 1988, Norman Weinstock of Zahn Dental met with Kevin Dillon, then the president of Ivoclar, and discussed the possibility of Zahn taking on the Ivoclar tooth line and becoming Ivoclar's national dealer, with Frink Dental servicing the Midwest. Weinstock initially felt Zahn would take on the line. Weinstock Tr. 149-51. However, Bob Brennan, then head of the Dentsply tooth division, and Gordon Hagler, then Sales Manager, told Weinstock about the Dentsply policy later embodied in Dealer Criterion 6, and stated that if Zahn took on the line of Ivoclar teeth, Dentsply would not allow Zahn to distribute Dentsply teeth. In several telephone calls and personal meetings over a two-month period, including a very heated discussion with Hagler, Weinstock learned that Dentsply would not tolerate Zahn taking on a competing product line. Dentsply management explained that they felt they had to protect their business and were not going to allow Zahn or anybody else to take on competing premium lines of teeth. Weinstock Tr. 149-52. After these discussions -- and after he saw how unfavorably Ivoclar's $1.2 million in projected U.S. sales compared to Zahn's annual Dentsply tooth sales of around $8 million -- Weinstock decided to keep his Dentsply line rather than replace it with the much smaller amount of Ivoclar sales. Consequently, Weinstock told Mr. Dillon about Dentsply's threat, that Zahn couldn't afford to lose the Dentsply line, and that Zahn therefore would not carry Ivoclar teeth. Weinstock Tr. 152-53; Ganley Tr. 1001; Reitman Tr. 1524.
47. In October 1992, Dentsply opened Jan Dental Supply Company as a Trubyte tooth dealer, after Jan agreed to stop selling Vita, Kenson, Dentorium and Justi teeth. GX 24, 26; Reitman Tr. 1525. Dentsply opened Jan, in part, to "eliminate several competitors." GX 26 at DS 016474. Vident believed that before it was cut off Jan Dental had done "a wonderful job" as a Vita tooth dealer. Whitehill Tr. 263-64.
48. In the early 1990's Atlanta Dental considered adding Vita teeth to its product offering after Betsy Harris, manager of Atlanta Dental's tooth department, received requests from current and potential customers asking whether she could carry Vita teeth. Harris Tr. 599-600, 615. Ms. Harris understood that these customers were interested in buying Vita teeth from Atlanta Dental rather than from Vident in California because Atlanta Dental sold them locally. Harris Tr. 599-600. Ms. Harris had initial discussions with Vident about taking on the Vita line, and they planned further discussions, after Ms. Harris had a chance to review Vident product information and a sample contract. Harris Tr. 601-03; GX 296. Ms. Harris later met with Vident representatives, and they decided to draw up a contract for Atlanta Dental to acquire a $30,000 stock of Vita teeth. Harris Tr. 603-04. Ms. Harris then talked to Bill Yacola of Dentsply to find out what the consequences would be if Atlanta Dental put in a competitive line -- in particular, because of her experience with Frink, Ms. Harris was concerned that she ran the risk of losing her Dentsply line. Harris Tr. 606-07. After checking with others at Dentsply, Mr. Yacola replied that if Atlanta Dental took on Vita teeth Atlanta Dental would no longer be able to sell Dentsply teeth. Harris Tr. 607-08. Atlanta Dental decided not to put in the Vita line, in order to avoid jeopardizing their Dentsply business. Harris Tr. 608-10. At that time their sales revenue for artificial teeth was a million dollars a year, and Dentsply teeth comprised 90% of that revenue. Harris Tr. 615. "I had no way of knowing what our Vita sales would be at that time, so losing that much business was -- this is my livelihood, this is what I do, and I didn't want to jeopardize my company or myself in that way." Harris Tr. 616. Atlanta Dental would have added Vita teeth to its product line in 1993 if it could have done so without jeopardizing its Dentsply business. Harris Tr. 617.(5)
49. In late 1993 or early 1994, Pearson took on a consignment of Vita teeth at its southern California facility. Kashfian Tr. 1386. The stock consisted of one tooth cabinet and was worth "a few thousand dollars." Id. at 1388. Pearson also included the Vita teeth in its catalogue. Kashfian Tr. 1404. After a Dentsply sales representative noticed the Vita consignment stock, the Dentsply Regional Manager, Dave Louda, told Pearson that it was not supposed to carry competing tooth lines and that continuing to carry Vita would jeopardize Pearson's ability to be a Trubyte dealer. Kashfian Tr. 1386-87. Because Pearson was "doing a tremendous amount of business with the Trubyte division," and because Vita teeth were "not as popular as Trubyte," Pearson agreed to return the tooth consignment to Vita. Id.; Reitman Tr. 1524. Moreover, at the same time that Pearson took on Vita teeth, it considered adding Meyerson teeth. Kashfian Tr. 1388-89. When Pearson asked its Dentsply sales representative what the consequences would be of taking on Meyerson, it was informed that it would be "the same scenario as Vita," and Pearson did not take on the Meyerson tooth stock. Id. at 1389.
50. DLDS sought to add Universal and Vita teeth in 1994 to fulfill customer demand. Vetrano Tr. 1423-24. However, a week after DLDS introduced the teeth to its customers, Dentsply informed DLDS that if it carried the teeth it would lose the entire Trubyte line of teeth and merchandise. Id. at 1426-27. As a result, DLDS did not take on the Universal and Vita teeth. Id.; GX 58; GX 66; Reitman Tr. 1524.
51. Darby Dental had lost its Trubyte tooth line when it purchased a company called Nordent, which sold Nordent's house brands of competitive teeth. Nordhauser Tr. 4102-03, 4112-13.(6) Darby Dental wanted to regain the Trubyte tooth line, and Mr. Nordhauser tried several times to negotiate with Dentsply, but Dentsply demanded that Darby give up all its teeth other than Dentsply in order to get the Dentsply line back. Nordhauser Tr. 4117, 4120. At one point Mr. Nordhauser offered to give up selling a number of brands of teeth, including, inter alia, Justi, Myerson, and Kenson, but Dentsply still refused to re-open Darby as a Trubyte tooth dealer. Nordhauser Tr. 4119-21, 4125-27, GX 434. When Darby made plans to take on the Vita tooth line, however, Dentsply finally agreed to reinstate Darby. Nordhauser Tr. 4127-32. Dentsply and Darby agreed that Darby would regain the Dentsply tooth line in return for agreeing to the following conditions:
Nordhauser Tr. 4131-33; GX 82. See also GX 434; Clark Tr. 2636 (Dentsply did recognize Darby as a tooth dealer, and Darby did not take on the Vita tooth line).
52. Dental Technicians Supply ("DTS") was a laboratory dealer with locations in New Hyde Park, NY, Kansas City, MO, and Denver CO, and Orlando, FL. Raths Tr. 1144-45; Underwood Tr. 3377.(7) DTS had been a Trubyte tooth and merchandise dealer since the mid- 1980's, but in 1990 Dentsply terminated DTS as a tooth dealer. Raths Tr. 1147-48; Underwood Tr. 3405. In approximately 1991 DTS began selling Vita and Ivoclar teeth at all its locations, and between 1991 and 1995 sales of Vita and Ivoclar teeth increased at all DTS locations. Raths Tr. at 1149-56; Underwood Tr. 3400; Whitehill Tr. 259-60; Ganley Tr. 1002; GX-19. In subsequent negotiations over those years regarding reinstating DTS as a Trubyte tooth dealer, Dentsply consistently maintained that, as a condition of reinstatement, DTS would be required to stop selling Vita and Ivoclar teeth. Raths Tr. 1157; Underwood Tr. 3407-09. DTS finally reached an agreement with Dentsply at a meeting in York, Pennsylvania in June 1995. Raths Tr. 1157-59. Under that agreement, in return for being reinstated as a Trubyte tooth dealer DTS dropped Vita and Ivoclar teeth from the Kansas City, Denver, and Orlando locations, and in New York, DTS agreed to remove the Ivoclar line and to limit its Vita sales to existing customers in the Northeast. Raths Tr. 1159-65; GX 93 at DPLY-A 18372-79; GX 158 at DS 015783-91; Reitman Tr. 1522-23, 1525. One of Dentsply's "considerations" in recognizing DTS was that it would "fully eliminate the competitive threat that [DTS locations] pose by representing Vita and Ivoclar in three of four regions." GX 86 at DS 015805-06.
53. In the spring of 2000, Marcus Dental, a Trubyte dealer in Minneapolis, had taken on the Kenson tooth line because of an out of stock problem with Trubyte teeth. Jenson Tr. 2291. For several months during 2000, Trubyte was having problems supplying teeth to dealers. Jenson Tr. 2292. The service problems started in the spring but continued into October 2000, and in August 2000 Trubyte's success rate for fulfilling one-day shipments dropped to an all-time low of 80.5% (Dentsply's goal was 97%). Jenson Tr. 2292-93 (level dropped more than three points in one month alone, dropping from 83.7% in July to 80.5% in August). That rate of order fulfillment by Dentsply caused concern among Trubyte dealers such as Marcus. Jenson Tr. 2292. Due to these problems with Dentsply's service levels, some of Marcus' lab customers had switched to Kenson teeth and Marcus was trying to retain its customers by selling them Kenson teeth. Jenson Tr. 2291-92. Dentsply, however, enforced Dealer Criterion 6 against Marcus, and Marcus returned the Kenson teeth to Myerson. Jenson Tr. 2293; Swartout 1314-15; Reitman Tr. 1525.
54. In mid-2000, when Leach & Dillon agreed to take on the Enigma tooth line after Mr. Dillon failed to recruit another dealer, [REDACTED TEXT] Jenson Tr. 2295-96 [REDACTED TEXT] See also Jenson Tr. 2296-97 (Dentsply informed Zahn that it viewed this proposal as a violation of Dealer Criterion 6).
55. In his November 2000 monthly report to his superior, Mr. Roos, Mr. Jenson reported that Thomson Dental, a Trubyte tooth dealer, was exploring competitive tooth lines. Jenson Tr. 2297. Mr. Uthus, Trubyte's Director of Sales, quickly squelched Thompson's effort, explaining Dentsply's "agreement" with Thompson on competitive teeth and faxing Thompson a copy of the dealer criteria. Jenson Tr. 2297-98.
56. Dentsply put a quick stop to another dealer's consideration of adding a rival line in Fall 2000, when Patterson inquired about carrying competitive tooth lines. Jenson Tr. 2298. Mr. Jenson told Mr. Easty of Patterson that the Dentsply dealer criteria would be enforced. Jenson Tr. 2298.
57. [REDACTED TEXT] (9) Last fall, Horst Becker of Heraeus Kulzer also spoke with Zahn's president and its director of marketing about Zahn carrying Heraeus Kulzer teeth. Becker Tr. 1818-20.
58. In 2001 Patterson bought a Trubyte tooth dealer in L.A. named Guggenheim which also carried Kenson teeth. Patterson itself did not carry Kenson teeth, and so Trubyte asked Patterson to comply with Dealer Criterion 6 and drop the competing tooth lines from the Guggenheim locations. Jenson Tr. 2289-90; Reitman Tr. 1524-25. Patterson complied and dropped the Kenson teeth that Guggenheim had been selling. Jenson Tr. 2291; Reitman Tr. 1524-25.
59. When Darby acquired DTS in 1998, the DTS facilities in Kansas City and Denver did not carry Vita teeth; DTS carried Vita teeth only in New York and sold them only to customers in New York. Nordhauser Tr. 4101, 4104-05; Jenson Tr. 2288 (prior to DTS becoming a Trubyte dealer, Dentsply had agreed that DTS could keep its New York Vita stock). See GX 158 at DS 015783. When Darby acquired DTS, Dentsply "made it very clear . . . that we cannot promote or give to the rest of our customers Vita teeth. We can only, for a very short period of time, sell it to the customers we have." Nordhauser Tr. 4106, 4135-36; GX 130 at DARBY 001120Y-21Y. As a result, Mr. Nordhauser agreed that Darby would not sell Vita teeth to any other Darby or DTS division,(10) and that Darby would sell Vita only through the New York office to the "few" Vita teeth customers that DTS already had. Nordhauser Tr. 4104-07, 4136-37. Dentsply also insisted, "as you can see by the letter," that Darby ultimately "would have to terminate the sale of Vita teeth." Nordhauser Tr. 4107, 4135-37; GX 130 at DARBY 001120Y-21Y; Jenson Tr. 2289 (when Darby bought DTS Trubyte asked Darby to comply with Dealer Criterion 6 and get rid of the Vita line). Although Darby had wanted to keep the Vita tooth line in New York, after this transitional period Darby ultimately complied with Dealer Criterion 6 in 2001, and now DTS and Darby do not sell Vita teeth. Jenson Tr. 2289-91.(11)
60. [REDACTED TEXT]
61. The "general norm in the [artificial tooth] business is distribution of teeth through dental dealers. It's the expectation of the laboratory to purchase and receive products that way and it's also the expectation of the laboratory to receive the level of service provided by a dental dealer." Ganley Tr. 1007. As Kevin Dillon, Ivoclar's former president and the current distributor of Enigma teeth, stated, "[i]f you don't have distribution with the dealer network, you don't have distribution." Dillon Tr. 4081.
62. While the 7,000 dental labs in this country are quite heterogeneous, and the strength of each lab's preference for any particular dealer service will vary, most labs find at least some benefit from the services offered by dental lab dealers. Reitman Tr. 1484-85. Zahn Dental alone sells at least some teeth to 3,000 to 4,000 labs each year. Weinstock Tr. 168-69. Some of these services are unique to local dealers, while others are provided by all dealers regardless of their location. Reitman Tr. 1485. All of the numerous services that dealers provide to dental labs indirectly benefit manufacturers. Weinstock Tr. 135.
63. Dealers maintain tooth stocks, which provide labs with a local inventory of teeth. Miles Tr. 3493; Weinstock Tr. 106; Reitman Tr. 1494; GX 364-C. Labs greatly value having this local access to a wide array of teeth. Reitman Tr. 1541 (Wind Survey showed that second-most important attribute of dealer was providing local stock of teeth); Miles Tr. 3493; Weinstock Tr. 106; Dillon Tr. 4094. Small labs, in particular, regard a local dealer inventory as a surrogate for its own, in-house inventory. Reitman Tr. 1495. [REDACTED TEXT] 64. The record is replete with specific, real-world examples showing that labs value the ability to purchase teeth from local stocks. (a) When Zahn opened its Indianapolis tooth counter, it dramatically increased its tooth business in that local area. Before it did so, it had only a couple of Indiana tooth customers, and its tooth sales were under $50,000 per year. Weinstock Tr. 108, 166. [REDACTED TEXT] (b) Zahn's experience in Florida was the same. [REDACTED TEXT] Prior to that, its sales had been between $300,000 to $400,000. Weinstock Tr. 109, 166. (c) According to Patterson Dental's president, "[w]e seem to do well only where we had a tooth stock and significant tooth stock. In areas where we did not, we had a difficult time to do business." Wiltz Tr. 3822. When Patterson tried to consolidate five tooth stocks in the Northeast into one location in Lancaster, Pennsylvania, its tooth sales declined in the areas where tooth stocks were removed. A year later, Patterson concluded that this consolidation effort was a failure. Wiltz Tr. 3821-23. After it returned tooth stocks in the Pittsburgh and New York areas to the previous levels, it regained the market share it had lost. Wiltz Tr. 3824-25. (d) Atlanta Dental's tooth sales in the Birmingham, Alabama area declined considerably after it consolidated most of that stock with its main stock in Atlanta (due to the death of the local tooth counter specialist there). It still maintains a skeleton stock in Birmingham for the convenience of customers who want to pick up their tooth orders themselves. Harris Tr. 586-87, 640-41. (e) Sidney Nordhauser of Darby Dental testified that his company does not compete with Hendon Dental, another dealer selling Trubyte teeth, "because they are in Texas. We are not." Nordhauser Tr. 4145. "We don't have a tooth counter in Texas or anywhere near there we can get teeth to them." Nordhauser Tr. 4145. Although Darby sells teeth in Texas when asked, Mr. Nordhauser believes Darby "can't compete" with a local tooth counter in selling teeth. Nordhauser Tr. 4145. (f) Like many other dealers, Accu Bite Dental has found that a high percentage of its tooth business comes from the area in which it has a tooth stock. Accu Bite's one tooth stock is in Livonia, Michigan, a western suburb of Detroit. Desautel Tr. 2426. In 1999, 80% of Accu Bite's tooth sales were made to labs located in southeast Michigan. Desautel Tr. 2462. (g) Dentsply's competitors have found that their tooth sales have been higher in the few areas in which they have been able to obtain dealer distribution. Whitehill Tr. 251 (Vident's sales "substantially higher" in limited areas served by small, specialty dealers selling Vita teeth); Swartout Tr. 1317 (Myerson has had sales growth of 20% per year in Connecticut, where it has dealer distribution).
65. Proximity to a dealer increases the delivery options available to a lab and decreases shipping charges. Reitman Tr. 1495; Turner Tr. 423-24; Swartout Tr. 1309. Because ground transportation is less expensive than air service delivery, Turner Tr. 424, a lab can lower its shipping charges by buying teeth from a dealer close enough to provide next-day service by UPS ground service. Reitman Tr. 1496; Turner Tr. 423-24. Some dealers have arranged very favorable deals with delivery services, and are able to pass those savings on to their lab customers. [REDACTED TEXT] Obst Tr. 2717 (DSG labs receive free overnight delivery from Zahn); Ryan Tr. 1238-39 (DLDS pays entire shipping charge for teeth purchased by Sonshine Dental Lab).
66. The use of a local dealer is perhaps most critical when labs need teeth on the same day they order them. The need for same-day pick or delivery of teeth is fairly common in the industry. [REDACTED TEXT] Desautel Tr. 2463 (in 1999, "a lot of" walk-up business at Accu Bite's Detroit tooth counter; labs would "call in and about 15 minutes later they will come in and pick it up"); Kashfian Tr. 1379-80 (10%-15% of Pearson Dental's overall tooth orders; 40% at Southern California location); Harris Tr. 662 (at Atlanta Dental, 5-10 customers per day); Nordhauser Tr. 4142-43 (perhaps five customers in New York, "but some of the other facilities have a steady walk-in business"); Coykendall Tr. 3312, 3332 (50% of Hopkins Dental Lab's business is repairs, where lab needs a ready stock of teeth to complete job in same day). 67. Same day service is particularly important for emergency cases, such as in repair cases where a denture cracks (or a tooth chips off) and the lab needs to replace the tooth with one that matches the shade and brand precisely. Reitman Tr. 1496-97; Armstrong Tr. 2369 (patients value fast service particularly when losing anterior teeth). See also Becker Tr. 1831 (even non-emergency cases can lead to emergency tooth orders: labs often run out of stock in the middle of a denture case and need a "replacement of materials basically in the next hour or so to finish their case, and so that means a fast delivery and fast service is, to many labs, essential"). One important way in which labs compete against each other is by providing fast and reliable service in such situations. Weinstock Tr. 89-90; Reitman Tr. 1497. 68. Labs prefer to buy tooth brands that would be available on the same day, if necessary, even though the vast majority of their cases do not require same-day delivery or pickup. Weinstock Tr. 133 ("perception is, if you are not local, many people are afraid to deal with you . . . The perception in the dental lab business is basically reality"); Reitman Tr. 1500. Repairs usually require a lab to obtain the same brand of tooth originally used in the denture. Because labs want to be able to obtain teeth on the same day for repair work, the availability of a tooth brand through a dealer is a significant factor influencing the lab's initial choice of a tooth brand to use in its business. Reitman Tr. 1499-1500. 69. Dentsply itself recognizes the importance of having tooth counters located in proximity to dental labs to be able to provide same-day service. In a September 15, 1993, letter John Weiland, Senior Vice President of Dentsply's North American Group, explained to Zahn's Norman Weinstock about opening a new Dentsply dealer less than ten miles away from an existing Zahn tooth counter in Dania, Florida: The York Division's decision to open J&S Dental Supply is based on a market demand that cannot be fulfilled from Dania, Florida. Namely, we believe certain customers in the Miami area need a local presence to be able to drive over and pick up teeth in rapid turnaround situations. GX 44; Weinstock Tr. 114-16.
70. "Relationships are very important in this industry." Reitman Tr. 1491. In selling teeth to denture labs, a dealer's relationship is one of the three biggest things. Weinstock Tr. 129. Judd Ryan selected DLDS as his dealer because he has "got a great relationship with them [and] we've been dealing with them for a long time." Ryan Tr. 1237. Similarly, Ralph Langer of Langer Dental Arts likes to buy teeth from DLDS because of the close relationship he has developed with that dealer:
I go where I feel most comfortable. In other words, where people recognize me and have a genuine interest in me because I'm spending some dollars. Well, that's the case with DLDS. They are very, very nice people, people that we've done business with for ten-plus years. People that we know we can relate to and they can relate to us. In fact, it goes so far, our DLDS rep has a picture of my wife and I in Hawaii on her place and we have one in our office. Langer Tr. 3279-80. See also Vetrano Tr. 1418 (relationships between labs and DLDS sales reps are "our lifeline" and "very valuable"); Obst Tr. 2748 (increased purchases from Zahn, in part, due to "exceptional service and relationship that has been built up" with Zahn). Moreover, local tooth counters can help build these customer relationships, because they give dealers places to hold programs and get closer to local dental lab associations. Weinstock Tr. 107.
71. Dealer sales representatives visit dental labs and manage their inventory of teeth by determining which tooth cards have been used, placing orders to replace those that have been used, and taking back broken sets. Reitman Tr. 1500; Becker Tr. 1820; Swartout Tr. 1309; Nordhauser 3421-22. As Horst Becker of Heraeus Kulzer testified, dealers can provide a "far higher" level of service than manufacturers because dealer reps regularly call on labs, count teeth and restock inventory. Becker Tr. at 1820. 72. Atlanta Dental has one lab specialist whose primary function is to call on labs and service their tooth stocks. Harris Tr. 644-45. The majority of Atlanta Dental's lab customers that maintain an inventory on site choose to have Atlanta manage it. Harris Tr. 648-49. 73. This inventory management function is particularly valuable when a lab's tooth stock is consigned from a manufacturer or dealer. In that case, dealer sales representatives can monitor the stock to ensure that the lab pays for the inventory that is being used. Weinstock Tr. 132; Ganley Tr. 1010-11; Reitman Tr. 1492-93. Zahn employs not only its own sales force, but many of the Sullivan-Schein operatory sales persons, to visit dental labs and monitor consignment stocks. Reitman Tr. 1493. See also Obst Tr. 2743 ("a lot" of inventory in DSG labs is on consignment).
74. Prompt, accurate, and reliable delivery is the essence of what it means to be a dealer. When labs place an order, they want to get it quickly, they want to make sure what they order is what they receive, and they want to receive that service day in and day out. Reitman Tr. 1486; Weinstock Tr. 129-30. A direct-selling supplier can try to provide that same service, but delivery is far less of a focus for a manufacturer. Dealers, on the other hand, specialize in distribution. Reitman Tr. 1487; [REDACTED TEXT] Brennan Tr. 1706 ("distributors did distribution" better than Dentsply). 75. Trubyte General Manager Steven Jenson also acknowledged that Dentsply's dealer network, at least in part, makes Trubyte teeth readily available to labs. Jenson Tr. 2267. Indeed, while at Trubyte, Mr. Jenson directed the Trubyte sales staff to reinforce to dentists and labs the service advantage of a dealer network versus a manufacturer's direct tooth sales. Jenson Tr. 2268.
76. Labs benefit from reducing the number of vendors with which they must deal. Reitman Tr. 1487 (explaining that term "one-stop" shopping does not literally mean buying all supplies from one source). These benefits include fewer accounts, fewer invoices to pay, and the savings of time, effort and cost. Weinstock Tr. 102-03; Reitman Tr. 1487-89. As Judd Ryan of the Sonshine Dental Lab in Bear, Delaware testified, the time it takes for his staff to make phone calls or send faxes to different vendors is "important. Time is money." Ryan Tr. 1286. See also Weinstock Tr. 102 (Zahn believes labs benefit by reducing number of invoices and statements, which are costly to process); Swartout Tr. 1308-09 (many labs are small businesses and reducing vendors "makes their business simpler, fewer bills to pay"); Nordhauser Tr. 4104 (small labs "don't have a person to do the buying, so if they can get everything from one salesman coming in or one telemarketer or something, it's a great advantage, it is obvious"). 77. The large lab dealers selling Trubyte teeth offer these "one-stop" shopping benefits. Zahn Dental offers an incredibly broad array of products for sale — over 25,000 tooth choices and 8,500 merchandise and equipment options. GX 160; Weinstock Tr. 102. Under the direction of Norman Weinstock, Zahn's Chairman, one-stop shopping is "something that we have been selling and selling hard for a number of years." Weinstock Tr. 102. Similarly, Betsy Harris of Atlanta Dental testified that "[a] lot of our customers . . . like to do one-stop shopping. They like to be able to make one phone call and get everything that they need rather than call around to two or three different places. And their time is more well spent doing their job than shopping on the phone." Harris Tr. 617. See also Vetrano Tr. 1414. 78. By consolidating purchases with a single dealer, labs can take advantage of volume purchase discounts such as those offered under Zahn's VIP program. Weinstock Tr. 147; Reitman Tr. 1488. For example, Dental Services Group ("DSG") labs increased their purchases from Zahn "a lot" as a result of Zahn's volume purchasing program. Obst Tr. 2747. DSG labs do not have a volume discount arrangement with any other dealer, preferring instead to consolidate more of its purchases through Zahn. Obst Tr. 2749.
79. [REDACTED TEXT] Weinstock Tr. 81 (30% of the teeth sold by Zahn are returned to it). This percentage is likely to increase in the future because partial dentures are becoming more and more common. Clark Tr. 2498. Taking back returns of broken sets is a service provided by dealers and valued by dental labs. Reitman Tr. 1489; Swartout Tr. 1309; Weinstock Tr. 81. It is also less costly for a lab to return several brands of teeth to a single dealer, rather than to several different locations. Reitman Tr. 1489.
80. Dealers inform labs of new products in the tooth and dental market, as well as other markets. Weinstock Tr. 130-31; Reitman Tr. 1490; Vetrano Tr. 1417-18. Labs are able to make more informed choices about products by receiving information from more than one source. Reitman Tr. 1490-91. 81. Through their tooth counter specialists, dealers also provide advice to labs on tooth and mould selection. Dealers are experienced in handling teeth and are familiar with different shades and moulds, and are therefore able to assist labs when they have questions. Reitman Tr. 1491.
82. Because dental lab dealers are the preferred tooth distribution channel for great numbers of dental labs, the main benefit that dealers provide to tooth manufacturers is that dealers make their teeth available through the channel that customers prefer. Reitman 1501-02. Moreover, a number of additional benefits accrue directly to tooth suppliers by being able to sell through dealers. Weinstock Tr. 134-37; Reitman Tr. 1502-03; GX 364-D.(12)
83. Dealers carry a substantial amount of a manufacturer's inventory, which reduces the capital costs incurred by the supplier. Reitman Tr. 1503; Becker Tr. 1821-22; Weinstock Tr. 134; Swartout Tr. 1307-08. Dealers carrying Trubyte teeth maintain approximately 100 tooth stocks across the United States. Reitman Tr. 1483-84; GX 364-B; Jenson Tr. 2267. Given that just one tooth stock, Atlanta Dental's, contains $145,000 worth of Trubyte teeth, Harris Tr. 617, the total amount of Trubyte tooth inventory owned and maintained by dealers is in the millions.
84. Manufacturers benefit significantly by having dealers handle the accounts receivable function. Reitman Tr. 1503; Kashfian Tr. 1383; Turner Tr. 450. Direct-selling manufacturers incur the substantial costs of billing, invoicing, and collecting debts from thousands of dental labs. Weinstock Tr. 134-35; Reitman Tr. 1503. For example, Dentsply currently sells to only 23 dealers. If it tried to sell directly to the 7,000 labs that fabricate dentures, it would incur a huge increase in billing, invoicing and collections costs. Reitman Tr. 1503; Weinstock Tr. 134 (Zahn "takes a lot of the cost away from the manufacturers, in that we collect the money"). This function is particularly important because the dental laboratory industry has not had the greatest of reputations over the years, of being fast payers or good credit risks. Weinstock Tr. 93. As a result, dealers like Zahn are able to provide extended credit terms to labs that are "way beyond what any manufacturer would really accept." Weinstock Tr. 134. See also Vetrano Tr. 1423 (discussing dealers' willingness to extend credit terms for labs "when things are hard").
85. Although each manufacturer employs its own sales representatives to promote its tooth products, dealer sales reps add another "voice in the marketplace." Reitman Tr. 1504; Weinstock Tr. 99-100, 139. Often dealer personnel serve a wider array of customers than just those served by a particular tooth supplier's representatives. Reitman Tr. 1504; Desautel Tr. 2469-70 (there aren't enough Dentsply sales reps to canvass all of Accu Bite's accounts, and Accu Bite reps have a greater reach into the labs than Dentsply). See also Becker Tr. 1820 (dealers "multiply" the size of a manufacturer's sales force); Swartout Tr. 1307 (dealers are important conduit for supplier's promotional message).
86. Similarly, dealers provide additional advertising vehicles such as dealer catalogues and invoice stuffers. Reitman Tr. 1504; Kashfian Tr. 1382; Desautel Tr. 2424-25; Weinstock Tr. 99-100. Again, having additional advertising channels and more ways of reaching customers, particularly customers that may not be current customers of the supplier, is a benefit to suppliers provided by dealers. Swartout Tr. 1307; Kashfian Tr. 1382; Reitman Tr. 1504.
87. Co-traveling, or "detailing," occurs when sales representatives from both the supplier and dealer travel together to a dental lab. Harris Tr. 635; Reitman Tr. 1505. This benefits the manufacturer because the dealer sales representatives have more frequent access to the dental lab, and probably has a stronger relationship with the lab. Reitman Tr. 1505. Dealers also benefit because the supplier's sales representative is better able to talk about the specific products. Therefore, co-traveling helps to increase sales and benefits both the dealer and the supplier. Reitman Tr. 1505.
88. Dealers refer new lab customers to supplier representatives. Reitman Tr. 1506. When a new lab enters the business, it needs various equipment, benches, lathes, grinders, flasks, trays and other items that are typically purchased from a dental lab dealer. As a result, dealers know about these new customers and can provide valuable leads to its suppliers, leads that the suppliers would not have if they were selling teeth directly. Reitman Tr. 1506. 89. Dealers employ tooth counter specialists who, along with dealer sales reps, have the daily, ongoing contact with the lab customer. Miles Tr. 3489. These relationships enable dealers to help market teeth, particularly based on their sales representatives' intimate knowledge of dental labs. Weinstock Tr. 113, 135. These established relationships are important to a supplier such as Dentsply because it does not have a close relationship with most of the labs selling Trubyte teeth. Miles Tr. 3489. Indeed, such relationships have enabled Zahn to help grow Dentsply's business. Weinstock Tr. 113, 135.
90. Dealers can assist suppliers in generating incremental business by promoting the manufacturer's product and providing these other services. Reitman Tr. 1506; Weinstock Tr. 99-100 (Zahn sales force generates new customers for denture products). Mr. Desautel of Accu Bite, one of Dentsply's own witnesses, testified that Accu Bite brings in new customers and new business for Dentsply. Desautel Tr. 2471. Even Dentsply agrees. See Reitman 1506, GX 31 (Dealer Criterion 5 anticipates that new dealers can bring incremental business to Dentsply, since it requires that they do so in order to be recognized as a Trubyte dealer).
91. Given all of the benefits that dealers provide to both tooth suppliers and dental labs, it is not surprising that Dentsply greatly values its diverse network of dental lab dealers. It benefits from the roughly 100 dealer stocks of Trubyte teeth that are maintained throughout the country. Reitman Tr. 1483-84; GX 364-B; Jenson Tr. 2267. Trubyte tooth stocks are located in essentially all of the major metropolitan areas in the United States, far exceeding the tooth stock network of any of its competitors. Jenson Tr. 2267-68. And Dentsply would like even more tooth stocks than it has today. Miles Tr. 3493 ("I could sell more teeth"). 92. Dentsply has valued not only its large, national dealers such as Zahn, but its smaller, more local dealers as well. As Chris Clark said, "I valued all my dealers, absolutely ... I'm a cash cow business in no growth. I need all of my dealer friends." Clark Tr. 2648-49. [REDACTED TEXT] 93. One reason why Dentsply values its smaller dealers is that they command a very high share of the Trubyte teeth sold in their local areas. For example, in 1998, Bernie McNickle of Reeve/Burkhardt Dental Supply in Oklahoma City wrote to Dentsply out of a concern that he was losing a substantial amount of business to larger, out-of-state dealers such as Zahn. Clark Tr. 2647-48. In responding to McNickle's concern, Dentsply's Chris Clark wrote that Trubyte tooth dealers with local tooth stocks in Oklahoma accounted for 80% of Trubyte sales in the state in 1997. "While [this percentage] may be down from ten years ago," Clark wrote, "I believe it underscores the value of full-service dealers like Reeve/Burkhardt to the Trubyte tooth business." GX 121 at DPLY-A 023489; Clark Tr. 2647-49. 94. Another reason Dentsply has valued its local dealers is that they provide important service benefits to lab customers. While larger, more mail-order-oriented dealers might offer a price break, as Dentsply's Robert Brennan wrote in a letter to another dealer, Dentsply "believe[s] that if a local dealer is successful at selling and providing service, a few dollars in savings from mail order will be more than offset. Customers will return for excellent service. Price can never replace service over the long haul." GX 30. 95. Dentsply has recognized that labs often need same-day delivery or pickup of teeth for their emergency cases. In 1993, Dentsply recognized a new dealer in Miami, Florida, J&S Dental Company, that was located no more than 10 miles from an existing Zahn tooth counter. Zahn's Norman Weinstock raised his concerns about this new dealer with Dentsply's senior management. Weinstock Tr. 113-117. In a September 15, 1993 letter, a Senior Vice President of Dentsply stated that: The York Division's decision to open J&S Dental Supply is based on a market demand that cannot be fulfilled from Dania, Florida. Namely, we believe certain customers in the Miami area need a local presence to be able to drive over and pick up teeth in rapid turnaround situations. GX 44; Weinstock Tr. 116-118. 96. Dentsply has encouraged dealers selling Trubyte teeth to open more tooth counters and has tried to slow down any efforts by dealers to consolidate tooth stocks. Dentsply encouraged Zahn to open tooth counters both in North Carolina and Texas when Healthco, another dealer selling Trubyte teeth, was on the brink of insolvency because otherwise "it would have left a tremendous gap in the southeastern United States." Weinstock Tr. 111. In 1996, when Patterson Dental was considering tooth stock consolidation, Dentsply suggested that Patterson conduct market research before doing so in order to slow down the consolidation process. Clark Tr. 2657-58 ("I was trying to buy time"). See also DX 41 at DS 024281 (noting that Patterson "ruined a good thing" and "lost a ton" by consolidating tooth stocks in late 1970's). 97. Dealers and the support they provide have been very important to Dentsply's tooth business. Miles Tr. 3489; Turner Tr. 450 (dealer support "very important"); Brennan Tr. 1706 (dealers, not Dentsply, do distribution well). In refusing to sell directly to its largest lab customers, Dentsply has done more than merely state that it is committed to its current distribution system of selling to dealers -- it has specifically told labs that "Dentsply simply cannot provide adequate service to our lab customers absent our dealer network." DX 653 at DS 005171.
98. Dr. Yoram (Jerry) Wind designed and conducted a survey of dental laboratories in order to produce a representative sample of labs and obtain data from that sample that could be used to establish the relative importance of brand, distribution and pricing in laboratories' purchasing decisions for prefabricated plastic teeth used in dentures.(13) Wind Tr. 738, 764, 800-01. Dr. Wind analyzed the survey data in order to estimate what would be the impact of changes in distribution options to the market shares of particular brands. Wind Tr. 739, 801-04. The survey data was also used by Dr. Reitman in his econometric modeling of the market share and price effects resulting from Dentsply's exclusionary policies. Reitman Tr. 1469, 1530-32, 1539, 1692. 99. Dr. Wind concluded that the survey produced an appropriate representative sample, and provided a basis for establishing empirically the relative importance to dental labs of tooth brands, types of distribution, and price. Wind Tr. 801.
100. Dr. Wind concluded that the survey design he used here (conjoint analysis or tradeoff ) was the most appropriate to address the issue in this case -- where there is a clear tradeoff made by dental technicians between brand, distribution, and price when purchasing artificial teeth. Wind Tr. 739, 762, 764. Dr. Rossi agreed that conjoint analysis is appropriate for assessing dental labs' preferences for different brands of artificial teeth, and he would seriously consider using conjoint analysis if he were asked to conduct a survey in this case. Rossi Tr. 3134. 101. Generally, conjoint analysis is the best approach to assess consumer tradeoffs among any possible combination of variables, including those that do not actually exist in the marketplace, and it provides the ability to then decompose those preferences analytically to determine their relative importance. Wind Tr. 740, 746, 752, 766. Conjoint analysis is the best approach to estimate consumers' price sensitivity, and has been used to determine how consumers value different distribution outlets for other product offerings. Wind Tr. 753-54; see also Rossi Tr. 3118, 3132 (Dr. Rossi has used conjoint methodology in the only survey he actually executed himself, for the University of Chicago/ Sandoz, although that survey was vastly simpler than Dr. Wind's survey here(14) (Rossi Tr. 3120)). 102. Conjoint analysis is the most popular, widely-used and widely-accepted market research analytic methodology by both practitioners and academics, and it has been used in thousands of marketing studies. Wind Tr. 741, 755. Dr. Wind concluded that conjoint survey analysis is reliable as indicated by its widespread use in research, publications, and industry. Wind Tr. 765. Another indication of its reliability and validity is the fact that companies use conjoint analysis repeatedly, suggesting that the users of this methodology believe that it is reliable, valid, and provides very valuable insight and input into their business decisions. Wind Tr. 764-65, see also Wind Tr. 751-52. 103. Dr. Wind has conducted numerous surveys, including conjoint analysis surveys, in almost all product areas to address various marketing issues. Wind Tr. 748-758. Dr. Wind has used conjoint analysis surveys for many of his business consulting clients, including, for example, in the design of the Courtyard by Marriot hotel chain. Wind Tr. 748-51. Dr. Rossi does not dispute that hundreds of companies rely on surveys Dr. Wind has conducted to make significant business decisions, and that many of these companies are repeat clients. Rossi Tr. 3139-40. Dr. Wind has also used conjoint analysis, as well as other survey methodologies, when he has been retained as a marketing research expert in legal matters.(15) Wind Tr. 755-57. 104. Conjoint analysis is more reliable than simply asking survey respondents to rank the importance of their preferences because of the tendency to indicate that everything is important. Wind Tr. 739, 765, 787. Dr. Rossi agrees. Rossi Tr. 3132-33. Conjoint analysis is also superior to simply asking respondents which alternatives they prefer in a single question (Rossi Tr. 3133, Wind Tr. 790-91). In addition, conjoint analysis is preferable to analyzing historical data because it is not limited to actual marketplace situations. Wind Tr. 766. 105. The conjoint survey design uses constant sum allocation, where a respondent allocated 100 points among the brands on each of the scenario cards. Wind Tr. 785-86. Constant sum allocation is more appropriate for the task in this survey than other types of conjoint design. Wind Tr. 909. Dr. Wind has used a constant sum data collection procedure similar to the one here in most of the conjoint surveys he has conducted both in his consulting and research because it the best method among the various ones available for measuring preferences. Wind Tr. 787, 841. Constant sum allocation is widely used in marketing research and social science research. Wind Tr. 788. It is a very informative way of collecting data and is an easy and natural task for respondents to complete. Id. It is quite common for the hundred points to represent intended purchase shares, such as for example, in the survey conducted for Courtyard by Marriott and in pharmaceutical surveys. Wind Tr. 788-89. 106. Dr. Rossi agreed that it is sometimes appropriate to ask dental lab survey respondents about how they would allocate their purchases among brands in a conjoint survey. Dr. Rossi agreed that the selection and purchase of teeth by dental labs was an example of situations where customers split their choices among products, and that purchases of medical devices and supplies and business to business purchases are other examples. Rossi Tr. 3233-34. Dr. Rossi agreed that in these situations (and others) it may make sense to ask survey respondents to describe the allocation of their last ten or next ten purchases. Rossi Tr. 3235. 107. During the design stage, Dr. Reitman provided input into the objectives of the survey and played a role in formulating the questions that were asked, such as providing information about which brands were of interest and what the prices were for those brands. Reitman Tr. 1539. Dr. Wind typically relies on domain experts to provide information about the products and market in which he is conducting a survey. Wind Tr. 907. The prices used on the conjoint scenario base card were obtained from marketplace sources, such as the Zahn catalogue; the prices on the remaining scenario cards were determined using Dr. Wind's master experimental design. Reitman Tr. 1691; Wind Tr. 780; GX 140 at p. 8 & Ex. 2. 108. The brand variable is a summary measure that incorporates everything the respondent perceives about the specific product, such as tooth quality. Wind Tr. 782-83. Similarly, terms such as "local dealer" were deliberately left undefined so that the specific interpretation was left to the respondent. Wind Tr. 786. A dealer in one location can be local to one lab and mail order for another under this survey design. Wind Tr. 786-87; Reitman Tr. 1693-94. 109. Professor Wind disagreed with the assertion of Dentsply's counsel in her opening statement that the survey asked labs- - "would you like to buy from a dealer or manufacturer? But it kept prices the same" -- because the survey scenario cards varied prices. Wind Tr. 792; see Defendant's Opening Statement Tr. 56. As Dr. Wind explained, "[t]he whole logic of the design is to provide various combinations of brands and prices and distribution options." Wind Tr. 792. He identified several of the scenario cards given to the survey respondents in which the price for Ivoclar teeth was lower when it was available only direct from the manufacturer than it was when available from a dealer. Wind Tr. 793-94.
110. The survey was implemented according to generally accepted principles. Wind Tr. 763. Dr. Wind concluded that the specific data collection methodology used here (telephone-mail-telephone or "TMT") was the most appropriate one for the task here. Wind Tr. 762-63, 797-98. Dr. Rossi testified he would be "happy" to use a TMT approach in this case. Rossi Tr. 3135. Dr. Wind typically uses the TMT approach in the conjoint analysis surveys, as well as other studies where he is dealing with a difficult-to-reach population dispersed around the country. Wind. Tr. 798. It is a very commonly used approach by other researchers, too. Id. 111. The survey was conducted in 1998, during the investigation that preceded the decision by the United States to file suit in this matter. Wind Tr. 738; Reitman Tr. 1463-64. Once Dr. Wind decided to use conjoint analysis for this survey, he asked two of his colleagues at Wharton with whom he had worked on many previous surveys, Dr. Paul Green and Dr. Abba Krieger, to work with him on this project as a research team. Wind Tr. 767. Professional, trained interviewers with experience in marketing research conducted the interviews. Wind Tr. 799-800. The interviewers were trained specifically for this survey, including a dry-run of the questionnaire, and were monitored during both the screening and conjoint analysis phase. Wind Tr. 800. 112. Dr. Wind uses a sampling procedure to screen potential survey respondents to establish membership in the universe of interest, which is the general population to which the survey results are generalized or projected. Wind Tr. 767-69; GX 140 at pp. 4-5 & App. D, G. The universe of interest in this survey was defined as the dental lab technicians or other laboratory personnel who are responsible for the selection of plastic artificial teeth for use in making dentures. Wind Tr. 767-68; GX 140 at p.4. The screening process here was similar to those in other surveys Dr. Wind has conducted. Wind Tr. 775. 113. Once the respondents were identified and agreed to participate in the survey, they were mailed a two part survey -- Part A, which the respondents completed on their own, included a questionnaire regarding lab demographics, preferences for dealer and brand attributes, and brand familiarity, and Part B, which was the conjoint analysis task. Wind Tr. 770, 777-779; GX 140 at p. 11 & App. E; Reitman Tr. 1540. The respondents were then telephoned again several days later and walked through the conjoint task, which required them to respond to eight specific scenario cards selected from 140 different combinations of brand, price and distribution. Wind Tr. 779-82, 789-90; GX 140 at pp. 6-11 & App. C. The instructions for completing Part B were read to the respondent - they were not provided in writing. Wind Tr. 789-90 & App. E.(16) 114. The survey was a double-blind study, which means that neither the interviewer nor the respondent knew the purpose or the sponsor of the survey. Wind Tr. 796. 115. Out of 2,520 calls made, 667 labs satisfied the screening requirement and qualified for the survey. Wind Tr. 771-73 & GX 140 App. G. The target number of respondents was approximately 200, which is within the range of typical sample sizes in conjoint surveys. Wind Tr. 773-74. Here, 274 of the 667 qualifying labs provided completed questionnaires, for a response rate of slightly less than 40%, which is "spectacular . . . [i]n today's environment." Wind Tr. 772-73, 776. Dr. Wind performed a statistical test of the survey data, and concluded there was no evidence suggesting nonresponse bias here. Wind Tr. 913-14. With the proliferation of telemarketing, public opinion surveys and marketing research, respondents are increasingly reluctant to participate and it is becoming more difficult to obtain representative samples. Wind Tr. 777. Dr. Rossi agreed that response rates obtained in marketing research have been declining. Rossi Tr. 3138. The response rate in most commercial marketing research surveys is between 10 and 20 percent. Wind Tr. 776. Dr. Rossi conceded he has no basis for disagreeing with Dr. Wind that low response rates are typical in marketing research. Rossi Tr. 3141-42. 116. The result of this process resulted in a final random probability sample that included those people who qualified as members of the universe, who agreed to participate, and who completed both Parts A and B. Wind Tr. 770. Because this is a random probability sample, this sample of 274 laboratories can be projected to the larger population of thousands of laboratories. Wind. Tr. 774-75; Rossi Tr. 3184 (because there are approximately 10,000 laboratories in the United States, making it impossible to know the preferences of all those labs, "[t]hat's why a survey has to be done"). 117. Dr. Wind concluded that the survey was short, easy, and not complicated compared to others that he has conducted. Wind Tr. 794. He also testified that the survey was not too demanding compared to, for example, the survey he conducted for Courtyard by Marriott. Id. Dr. Rossi agreed that the Marriott survey was more complicated than Dr. Wind's survey in this case. Rossi Tr. 3121. The survey he conducted in this case had only three factors (brand, price and distribution), which is fewer than other surveys Dr. Wind has conducted. Wind Tr. 794-95. Also, the survey respondents here were professional buyers -- the people responsible for selecting the brands of artificial teeth the laboratory uses -- not consumers, and the task of allocating 100 points is similar to what they do in real life. Wind Tr. 794-95. 118. Professor Wind did not pretest the survey for two primary reasons. Wind Tr. 795. First, he has used this methodology in "many, many" other surveys, so there was no uncertainty regarding the straightforward instructions and scenario cards that would indicate a need to pretest. Wind Tr. 795-96. In addition, Dr. Wind and his colleagues checked frequently with the research firm conducting the survey fieldwork to check for any problems with the data collection, and they received no complaints that respondents were having difficulty completing the task. Wind Tr. 797. 119. Typically, Dr. Wind does not pretest surveys such as this one, where the research approach and methodology have been used before and, in effect, already tested: "Basically, it's a methodology which I used and I feel confident that consumers responding in this case would have no difficulty completing the task." Wind Tr. 796-97. The fact that the topic or subject matter changes from one survey to another does not affect the reliability and validity of the survey design and analysis. Wind Tr. 907-08. The same survey method can be applied to any number of subject matters, and Dr. Wind does not typically pretest the format each time the subject matter changes. Wind Tr. 908. Significantly, the fact that many companies use conjoint analysis on a repeated basis suggests that the methodology is reliable, valid, and an accurate input into their business decisions. Wind Tr. 765; see also Wind Tr. 751-52. 120. Second, Dr. Wind did not conduct a pretest because of the real concern that Dentsply, given its dominant position in the market and salespeople calling on labs, would discover the existence of the survey during the pretest and potentially bias the results of the survey itself. Wind Tr. 796. The risk of Dentsply biasing the survey results outweighed any benefit from conducting a pretest. Id.
121. Dentsply retained Dr. Peter Rossi to criticize Dr. Wind's survey and both Dr. Reitman's and Dr. Wind's analysis of the survey results. Rossi Tr. 2998, 3114. Dr. Rossi formed his opinion in this case after only two or three months of work on this matter. Rossi Tr. 3115. He was content simply to criticize the work of Dr. Wind and Dr. Reitman, rather than conducting a survey of his own in this case. Rossi Tr. 3151. The most appropriate method to critique and test a survey, however, is to conduct another survey to demonstrate empirically any problems, as Dr. Wind did when he evaluated the Pepsi Challenge survey. Wind Tr. 758-62. In addition to the numerous surveys he has conducted, Dr. Wind has also been retained on occasion to critique surveys conducted by others. Wind Tr. 758. 122. Dr. Rossi, however, has very limited experience designing and executing (conducting) surveys.(17) Rossi Tr. 3117-3132. In addition, although he has published articles on statistical analysis, he has authored or co-authored only four publications addressing any aspect of survey methodology, and even some of those pertained only to the analysis of survey data rather than designing or executing surveys. Rossi Tr. 3115-17. Accordingly, Dr. Rossi lacks the expertise necessary to evaluate and critique the design and execution of Dr. Wind's survey, two of the "three major areas" of Dr. Rossi's opinion here. Rossi Tr. 3130-32. (a) Dr. Rossi has had primary responsibility for conducting or executing only one survey. Rossi Tr. 3130-31. That survey was jointly sponsored by the University of Chicago, where Dr. Rossi teaches, and Sandoz Pharmaceuticals as part of the school's New Product Lab to help educate students. Rossi Tr. 3118-20. The students, rather than a survey research firm, did the actual interviewing as part of their course work. Id. at 3119. Dr. Rossi did not know whether Sandoz relied on or used in any way the survey results from a report his students prepared and submitted. Id. at 3121-22. Similarly, Miller Beer occasionally allows Dr. Rossi and a colleague to add questions to Miller's own surveys, which he is involved in "from an academic point of view" rather than to provide marketing research consulting advice. Rossi Tr. 3124-25. (b) Dr. Rossi has had a primary role in actually designing one survey — the University of Chicago/Sandoz survey. Rossi Tr. 3131. In the only other three surveys where he contributed at all to the design, he was either "uncertain" whether he had any role in the survey design (credit cards survey), had a "purely advisory" role in the design (Miller Beer survey), or participated "only to some extent" in the design (Body Time, Social Time article and survey). Rossi Tr. 3123, 3125, 3131-32. (c) The remainder of his involvement in the survey area is limited to analyzing data from a handful of surveys, providing advice to students and colleagues at the University of Chicago, or "being exposed" to people with whom he discusses surveys. Rossi Tr. 3132; see also Rossi Tr. 3126-30. (d) In the only legal matter, other than this case, where Dr. Rossi has testified in connection with a survey (Free v. Peters, 12 F.3d 700 (7th Cir. 1993), he was one of three experts who testified regarding the reliability of the data generated by a survey of jury instructions that yet a fourth expert had conducted, much like Dr. Rossi's role in this case. Rossi Tr. 3127. Dr. Rossi did not design or conduct the survey at issue in Free. Id. at 3127. Although Dr. Rossi testified in this case that he found the survey in Free "extremely reliable" with a "100 percent response rate," on appeal Judge Posner disagreed, stating that the survey was "so deficient" it "would not support the conclusion" that the jury instructions were confusing. Id. at 3128 (quoting Free, 12 F.3d at 705). (e) Although Dr. Rossi stated that the 40% response rate obtained in Dr. Wind's survey "raises concern" of response bias, he acknowledged that he could not state with certainty that there actually is any response bias in the survey data used by Drs. Wind and Reitman, and he could not measure the extent of it since he had not conducted any sort of statistical analysis to address the issue, as Dr. Wind had done. Rossi Tr. 3148-51. He testified that there is no way to prove or disprove that the sample used by Dr. Wind is representative, short of doing another survey in this case, which Dr. Rossi failed to do. Rossi Tr. 3150-51. 123. Dr. Rossi is not an expert in antitrust economics, or more specifically the economics of exclusive dealing or free-riding. Rossi Tr. 3162. He has never been involved in an antitrust case other than this one. Rossi Tr. 3163. Dr. Rossi has little or no knowledge regarding the artificial tooth market or the evidentiary record in this case. Rossi Tr. 3164-68, 3171-77, 3179-80, 3182-84. In forming his opinion, he reviewed no business documents and participated in no interviews of any lab, dealer, or artificial tooth supplier, including even Dentsply, and looked at only portions of a few dealer depositions without reviewing them in detail. Rossi Tr. 3164-66. Dr. Rossi also never reviewed any completed questionnaires of survey respondents prior to forming his opinion, and knows few details about the dental laboratory business. Rossi Tr. 3168, 3180, 3182-84. (a) Because of his near total lack of knowledge regarding the artificial tooth market, and particularly the dental laboratory business, Dr. Rossi is unable to prove that the survey respondents were confused or unwilling to provide informative answers, as he claims they were. Rossi Tr. 3162, 3169-70, 3179. Belying his assertion that the survey respondents were confused because the term "local dealer" was deliberately undefined in the survey, Dr. Rossi does not know what participants at all levels of the tooth market -- labs, dealers, rival suppliers and even Dentsply -- understand the term to mean. Rossi Tr. 3171-78. Dr. Rossi was unaware that the term local dealer is commonly understood in the dental laboratory supply business; for example, Robert Brennan testified that the term local dealer was "pretty common in the industry." Brennan Tr. 1712-13; Rossi Tr. 3171-72. Dr. Rossi also conceded that Zahn's use of the term "local service for teeth from coast to coast" in its catalogue was not intended to confuse its customers, and that Zahn expects its customers to understand the meaning of "local service dealer for teeth." GX 160; Rossi Tr. 3172-73. (b) Dr. Rossi's scant knowledge of the artificial tooth market also undercuts another one of his criticisms -- that Dr. Reitman's scenarios are unrelated to the market without Dealer Criterion 6. Rossi Tr. 3178. He admitted he could not identify what a "realistic" scenario would be. Rossi Tr. 3179. Dr. Reitman explained that his model analyzed the effect of Vita and Ivoclar being made available through local dealers, which only can happen if Dealer Criterion 6 is removed and those brands gain access to the dental laboratory dealer network. Reitman Tr. 3903-04. Dr. Reitman modeled one of several effects from removing Dealer Criterion 6 that he found based on the evidence in the record. Id. (c) Similarly, Dr. Rossi's criticism that some respondents were unwilling to take the survey seriously, based on the two indicia that some respondents did not vary their preferences across survey cards and that some have high shares, is nothing more than speculation. He acknowledged he did not know what respondents, even those who did not vary their preferences, were thinking in completing the survey, and knows little about dental laboratories in general. Rossi Tr. 3179-80, 3182-84. As a result, Dr. Rossi does not know whether these two indicia in fact show the laboratories' strong brand preference, as Dr. Wind believes, rather than confusion or unwillingness. Rossi Tr. 3184-85. In any event, Dr. Rossi agrees with Dr. Wind that these respondents should be included in the data set. Rossi Tr. 3185-86. (d) Dr. Rossi was unaware of the existence of surveys Dentsply commissioned or conducted in-house, on which it relied to make business decisions, that had response rates lower than the rate in Dr. Wind's survey. Rossi Tr. 3143-47 (1992 Division name survey: 25% and 26% response rates; 1991 Company image survey: 7% and 17% rates; 1994 Vita lab shade survey: 31% rate). Dentsply relied on each of these surveys in the course of its business: to change the name of the York Division to Trubyte (Rossi Tr. 3144-45); by Christopher Clark when he became General Manager to understand the brand equity of Trubyte products (Rossi Tr. 3145; Clark Tr. 2494); and again by Mr. Clark to determine that Dentsply needed Vita-shaded teeth (Rossi Tr. 3146-47; Clark Tr. 2497-2500).
124. Dr. Wind's analysis of the survey results demonstrated that when Vita and Ivoclar teeth were available to dental laboratories from either a local dealer, or from both a local dealer and mail order dealer, rather than being available only directly from the manufacturer, their relative market shares would increase by as much as 24 to 35 percent for Ivoclar and 10 to 32 percent for Vita. Wind Tr. 763, 803-04; GX 140 at 14-15. Dr. Wind concluded that a 10 to 35 percent share increase for a brand is "incredibly significant." Wind Tr. 804. 125. Dr. Wind used a PRIDEM/PRIDEL model to analyze the survey data. Wind Tr. 801. The model has been subject to peer review, and has been widely used by a number of companies and professional research houses. Wind Tr. 802-03. Dr. Wind testified that confidence intervals typically are not provided for any conjoint analysis applications, including both the most popular software, Sawtooth, and PRIDEM, the model Dr. Wind used here. Wind Tr. 904-05. It is impossible to calculate confidence intervals for the PRIDEM model. Wind Tr. 904. Dr. Rossi agrees it would require a very complicated fitting strategy and is unlikely that standard errors can be developed for PRIDEM. Rossi 3215-16. Instead, in Dr. Wind's and all other conjoint analysis studies, the survey results are presented to the client company itself to assess if the results are managerially significant. Wind Tr. 905-06. He concluded that the model is reliable based on its wide and repeated use by those companies to make key business decisions worth millions of dollars, its use in teaching marketing research, and the fact that it is widely publicized in the professional literature. Wind Tr. 915-16.
126. Dr. Reitman used a multinomial Logit model to evaluate the survey data. Reitman Tr. 1541. The multinomial Logit model is appropriate because it was designed for situations, such as the one here, where customers are selecting among a discrete set of choices. Reitman Tr. 1541, 3888, 3890-92, 3894. The logit model has been discussed frequently in refereed journals. Reitman Tr. 1542. Dr. Reitman himself has authored an article using a logit model similar to the analysis he did in this case, has used logit models in other cases in which he has been involved, and has reviewed the work of outside economists who used logit models. Reitman Tr. 1450, 1461. The logit model also has been applied frequently in disciplines and contexts other than antitrust investigations. Reitman Tr. 1461-62. 127. Although he criticizes the use of a Logit model here, Dr. Rossi states in one of his publications that Logit models perform well with aggregate market share data, and can perform exceptionally well when fit to the aggregated choices of many different consumers. Rossi Tr. 3212-14. 128. Logit models are used for analyzing conjoint survey data in several leading statistical analysis software packages. Reitman Tr. 3893. For example, Sawtooth software, the leading commercial software package for analyzing conjoint data (Rossi Tr. 3231; Wind Tr. 803, 904-05), has an application for using Logit models to analyze conjoint data using exactly the same procedure Dr. Reitman used here. Reitman Tr. 3893. 129. In addition, a paper recommending Dr. Reitman's Logit model over the one Dr. Rossi used, titled "Modeling Constant Sum with Logit: A Comparison of Four Methods," was presented at the 2001 Sawtooth software conference. Reitman Tr. 3893; Rossi Tr. 3231, 3238. Dr. Rossi is familiar with the paper itself and with both authors, one of whom provided data for a recent paper of his. Rossi Tr. 3231-32. The article compared fou |