Full opinion text
TABLE OF CONTENTS FINDINGS OF FACT...........................................................325 I. Relevant Market Definition..................................................325 A. Industry Overview......................................................325 B. Relevant Geographic Market.............................................326 C. Relevant Product Market................................................326 1. Consumer Dynamics ...............................................326 2. Retail Customer Perspectives........................................330 3. Competition From Products in Other Product Categories................331 4. Supply-Side Considerations.........................................332 5. Conclusion: the Relevant Product Market Is All RTE Cereals...........333 II. Market Share Measurement.................................................335 A. Actual Shares of Market in 1992...........................................335 B. Measurement of Nabisco Share ..........................................336 1. Nabisco was in decline before its acquisition by KKR...................336 2. Shift in Consumer Preferences.......................................336 a. Oat bran craze.................................................337 b. Taste enhanced cereals..........................................337 3. Lack of effective advertising and distribution..........................337 4. Lack of full line of products.........................................337 5. Harvest strategy...................................................338 a. “Harvest” limited to marginal brands.............................338 b. Nabisco continued to support Nabisco Shredded Wheat..............338 i. Defense against Kellogg ..................................338 Ü. Trade and consumer promotion spending....................339 iii. Reformulation of Spoon Size...............................339 iv. Transfer of RTE cereal to the Biscuit Division...............339 6. Plaintiff’s growth projection for Nabisco is unrealistic...................339 a. Dissatisfaction of lapsed users....................................339 b. Plaintiffs erroneous assumptions.................................340 c. Market share increase due to introduction of new cereals............341 7. Conclusion: Nabisco’s 1992 market share accurately reflects its competitive significance...........................................342 III. Likelihood that the Acquisition Will Produce Anticompetitive Coordinated Effects..................................................................342 A. Dimensions of Competition ..............................................342 1. There is No Evidence of Close Coordination in Price....................342 2. Rubinfeld Pricing Analysis..........................................343 3. Retailers’ Buying and Pricing Behavior...............................343 4. Consumer Promotions..............................................345 5. Advertising .......................................................346 6. New product introductions..........................................346 7. Quality improvements..............................................347 8. Recent changes in firms’ marketing strategies.........................347 9. Competition from private label cereals................................347 10. Shelving..........................................................349 B. Likelihood of Tacit Collusion with respect to Promotion Activities......... 349 C. There is No Evidence that RTE Cereal Manufacturers Can, or Do, Punish Competitors.........................................................350 D. Industry Profitability...................................................351 E. Retailers Who Testified Support the Acquisition............................351 F. Conclusion of Court-Appointed Expert....................................351 IV. Likelihood that the Acquisition Will Product Anticompetitive Unilateral Effects.....352 Grape-Nuts and Nabisco Shredded Wheat Are Not Each Other’s Closest A. Competitors.........................................................352 1. Grape-Nuts and Nabisco Shredded Wheat Are Physically Dissimilar.....352 2. Grape-Nuts and Nabisco Shredded Wheat Are Associated with Different Attributes..............................................352 3. Grape-Nuts and Nabisco Shredded Wheat Each Compete With a Broad Array of Products..........................................352 4. Grape-Nuts and Nabisco Shredded Wheat Each Have Direct Form Competitors.....................................................353 5. Grape-Nuts and Nabisco Shredded Wheat Appeal to Different Consumers .........................................................353 B. Grape-Nuts and Shredded Wheat Are Priced and Promoted Independently ..................................................................354 C. Post-Acquisition Pricing and Promotion of Grape-Nuts and Nabisco Shredded Wheat Does Not Support an Inference of Anticompetitive Unilateral Effects....................................................356 D. Expert Opinions........................................................356 1. Defendant’s Expert’s Conclusions....................................356 2. Plaintiffs Expert’s Conclusions......................................357 CONCLUSIONS OF LAW.......................................................358 I. The Relevant Market.......................................................359 II. Market Shares and Industry Concentration....................................361 III. Likelihood of Anticompetitive Effects.........................................363 A. Facilitation of Anticompetitive Coordinated Conduct........................363 B. Promotion of Anticompetitive Unilateral Effects............................365 CONCLUSION.................................................................366 OPINION & ORDER KIMBA M. WOOD, District Judge. Kraft General Foods, Inc. (“Kraft”), which owns Post cereals, entered into an agreement to buy the ready-to-eat (“RTE”) cereal assets of Nabisco on November 12, 1992 (the “Acquisition”). The Federal Trade Commission reviewed the Acquisition pursuant to the requirements of the Hart-Scott-Rodino Antitrust Improvements Act of 1976, 15 U.S.C. § 18a. That Act’s waiting period expired, without a challenge to the Acquisition, on December 24, 1992, and the Acquisition was consummated on January 4, 1993. Soon thereafter, the Nabisco assets were fully integrated into Kraft’s Post Cereals Division. On February 10, 1993, more than five weeks after the Acquisition was consummated, plaintiff, the State of New York’s Attorney General (the “State”), initiated this suit seeking divestiture or recission pursuant to Section 7 of the Clayton Act, 15 U.S.C. § 18, Section 1 of the Sherman Act, 15 U.S.C. § 1, and the Donnelly Act, N.Y.Gen.Bus.Law §§ 340-344. Much of the procedural background of this case is set forth in the court’s Opinion and Order of June 14,1993, denying the State’s first motion for a preliminary injunction. Familiarity with that opinion is assumed. The State maintains that the Acquisition may substantially lessen competition in what the State terms the adult RTE cereal market, or, in the alternative, in the entire RTE cereal market. The State seeks either (1) to rescind the transaction between Kraft and Nabisco, making it possible for Nabisco to reenter the RTE cereal business immediately, or (2) to divest Kraft of Nabisco’s assets to another firm (“Newco”) that could function in Nabiseo’s place as the sixth major competitor in the RTE market. By stipulation between the State and Nabisco, the action has been stayed as against Nabisco pending this court’s determination of liability. The State has twice moved for a preliminary injunction enjoining Kraft from implementing a proposed transition between its Post cereal brand and the Nabisco brand. The court denied both of the State’s preliminary injunction motions, finding an insufficient threat of irreparable harm. After a three-week liability trial during which the court heard the testimony of Kraft business people; retail grocery chain executives; a Nabisco business person; expert economists for both parties; and the court-appointed, independent expert, Dr. Alfred Kahn; and having evaluated the witnesses’ credibility, the exhibits received in evidence, and the parties’ legal contentions, the court finds that defendants are entitled to judgment in their favor. The court’s findings of facts and conclusions of law supporting that verdict are set forth below. FINDINGS OF FACT The court has jurisdiction over this action pursuant to 28 U.S.C. § 1331 and § 1337(a), and Section 16 of the Clayton Act, 15 U.S.C. § 26. Venue is proper in this district under 15 U.S.C. § 22 and 28 U.S.C. § 1391(c). I. Relevant Marked Definition A. Industry Overview There are over 200 RTE cereal products available for sale to consumers. Kraft currently produces and sells 28 RTE cereal products, 21 under the “Post” name and 7 under the “Nabisco” name. Kraft is permitted to sell the former Nabisco RTE cereal products under the “Nabisco” trademark pursuant to a license that expires in January 1997. Kellogg and General Mills are the two largest manufacturers in the RTE cereal industry, and jointly account for about 60% of all RTE cereal products sold in the United States. Kraft ranks a distant third behind Kellogg and General Mills in the RTE cereal industry. Smith Deel. ¶ 13. In 1992, Kraft accounted for 11.70% of the sales of RTE cereal in the United States. Carroll Deck ¶ 15; Rubinfeld Deck ¶ 9(ii)(a), Figure 1 and Table 1. Other branded RTE cereal manufacturers include Quaker, with a 1992 market share of 6.82%; Ralston, with a 1992 market share of 4.58%; and Malt-O-Meal, with a 1995 share of 1.95% (Malt-O-Meal’s market share for the first three quarters of 1994 is 2.3%). Rubinfeld Deck ¶ 9(ii), Figure 1 and Table 1. Carroll Deck ¶ 15; Smith Deck ¶ 14; DX 251A, Lettmann Dep. at 14-15; DX 251; Leekie Deel. ¶ 14. There are also several smaller, specialized manufacturers of branded RTE cereal products, including Health Valley and Sunshine Biscuit Co. Boehm Deel. ¶ 13. Nabisco, with a 1992 market share of 2.82%, was the sixth largest RTE cereal firm and a fairly weak competitor in the RTE cereal market prior to the Acquisition. Boehm Deck ¶ 14; Smith Deck ¶ 15; Thomas Deck ¶ 10; Carroll Deck ¶ 15; Rubinfeld Deck ¶ 9(ii)(a), Figure 1 and Table 1. Nabisco’s main strength was its shredded wheat line of cereals. Thomas Deck ¶ 11; Boehm Deck ¶ 14; Smith Deck ¶ 15. Although shredded wheat cereals are popular, Nabisco’s sales had been declining since 1988. Thomas Deck ¶ 10; Smith Deck ¶ 15. Private label (“store brand”) cereals are a significant and growing presence in the RTE cereal market. See Carroll Deck ¶ 15. Private label cereal sales have consistently exceeded the growth of the cereal industry as a whole in recent years. Boehm Deck ¶ 16; Smith Deck ¶24. Gilster-Mary Lee Corp. and Malt-O-Meal Co. produce both branded and private label RTE cereal products, and these companies have grown rapidly in the past few years. Rubinfeld Deck ¶ 9(ii)(c); DX 250; DX 251; DX 250A, Welge Dep. at 9-11,15,18-19; DX 251A, Lettmann Dep. at 14-15. Actual market shares of the manufacturers in the industry fluctuated significantly during the period 1988-1993. Carroll Deck ¶ 15. Between 1970 and July 30, 1994, there has been a decline in concentration in the RTE cereal industry of about 27 percent. Rubinfeld Deck ¶ 60(b);' Tr. at 251, 253 (Cotterill). RTE cereals differ from one another in important respects, including type of grain, degree of sweetness, product form (e.g., flake, nugget, shredded, etc.), texture, flavor, complexity, type of additional ingredients (e.g., nuts and fruit), and perceived health benefits. See Leekie Deck ¶4; Tr. at 822 (Rubinfeld). RTE cereal consumers demand a particularly high level of variety and new products, and firms in the industry compete with one another by, among other things, introducing, a high volume of new products to meet that consumer demand. Carroll Deck ¶¶ 17-18, 23. B. Relevant Geographic Market The parties agree that the relevant geographic market in which to analyze the Acquisition is the entire United States. Post products are manufactured in four facilities and are distributed throughout the United States. Leekie Deck ¶2. The former Nabisco RTE cereal products were manufactured in two facilities and were distributed throughout the United States. Thomas Deck ¶ 14. Post sells its RTE cereal products to retailers throughout the United States at a single wholesale price. Leekie Deck ¶ 50; Schena Deck ¶ 5; Rubinfeld Deel. ¶ 13(a). C. Relevant Product Market 1. Consumer Dynamics The parties agree that individual consumers have a preference for variety when they choose RTE cereals. See Tr. at 98 (Cotterill). This consumer preference for variety is reflected in the following buying and eating practices. The average household buys about 17 different RTE cereals in the course of a year. Leekie Deck ¶5; Tr. at 459 (Carroll); DX 42; DX 51 at 5; DX 56 at 19. Most households buy only 25 to 30 percent of their RTE cereal from any one segment of the market. Tr. at 478-79 (Carroll). Seventy-seven percent of RTE cereal purchasers switch products on consecutive purchase oceasions. Leekie Deel. ¶5; DX 56 at 20. Very few people limit their selection of cereals to one product or one type of product. Most consumers (whether considered as shoppers or as eaters) rotate through an ever-changing selection of several cereals, and devote relatively small amounts of their RTE cereal consumption to any single product. Carroll Deel. ¶¶ 22, 23; Rubinfeld Deel. ¶ 21; DX 47; DX 51. Most consumers express a willingness to try a wide range of cereals. Even the most popular RTE cereals account for only a small percentage of total RTE cereal sales. The best-selling RTE cereal, Kellogg’s Com Flakes, had only a 5.23% market share in 1993. Leekie Deel. ¶ 4; DX 51 at 3. Grape-Nuts, one of Post’s most successful products, had only a 2.20% market share in 1993, and had an annual household penetration of only approximately 13%. DX 144 at KGF 1166572; Parker Deel. ¶ 4. Only about one-third of all adults responding in a 1991 survey said they would not eat all types of cereals. Carroll Deel. ¶¶ 23, 45; DX 47 at KGF 0254434. Most households have a principal shopper who decides which RTE cereals to buy for the entire household. Carroll Decl. ¶¶ 24-25. Purchase decisions are determined by a number of factors, including taste, nutrition, type of grain, sugar content, habit, price and the availability of coupons. Carroll Decl. ¶ 24; Tr. at 570 (Carroll); DX 49 at KGF 0261922. In one survey conducted for Post, consumers ranked price as only the seventh most important factor in choosing cereal, behind taste, “brand you usually buy,” “nutritious,” “family member requested or likes it,” “sugar content,” “had a coupon,” and “type of grain.” Carroll Decl. ¶24. Consumers are not as loyal to particular RTE cereals as they are to particular brands of other dry grocery products. Leekie Decl. ¶¶ 4—5, 18; Carroll Decl. ¶ 22; Parker Deck ¶ 4; Thomas Deck ¶ 63; DX 51 at 4. If one looks at consumption from the standpoint of what percentage of a household’s consumption is devoted to a particular cereal, one finds that even the most popular RTE cereals account for only about 10% of the cereal consumption within the households that purchase them—a far lower percentage than is common in other grocery categories. Leekie Deck ¶ 5; DX 51 at 4. Grape-Nuts accounts for only about 10% of the cereal consumption by its buyers; the remainder of their cereal consumption is widely dispersed among many brands. Leekie Deck ¶ 18; DX 27 at KGF 1166242-246; see Parker Deck ¶4. Even among heavy users of particular products, those products represent a small fraction of their RTE cereal purchases during the year, which accounts for the small “share of requirements” that any one brand fills among its own buyers. DX 47 at KGF 0254426; Carroll Deck ¶ 23. Approximately 75% of the buyers of any particular RTE cereal are either “light” or “medium” buyers of that product, meaning that they purchase it less than four times a year. Tr. at 440 (Carroll). On average, 30% of the volume of each RTE cereal is bought by light to medium purchasers. Tr. at 499 (Carroll). Post focuses its marketing efforts mainly on light and medium buyers, in addition to potential new buyers. Tr. at 500 (Carroll). The ability to increase volume sales depends on getting hght buyers to purchase more and attracting new buyers. Tr. at 625 (Leekie). Light users disperse their cereal purchases broadly across the range of RTE cereals. Tr. at 626-27 (Leekie); see Tr. at 337-339 (Cotterill). Forty to fifty percent of the buyers of a cereal in one year will not purchase that cereal the next year. Tr. at 503-04 (Carroll). When buyers are lost, they are usually the fight and medium buyers. Tr. at 500 (Carroll). Light users and nonusers are more likely to respond to a price change than are “heavy” users. Tr. at 964 (Rubinfeld); see also Tr. at 630 (Leekie). Individual RTE cereals cannot easily be categorized in terms of the “benefits” they deliver. Many “adult” cereals (e.g., those promoted as “healthy” and high in fiber) actually have as much or more fat and sugar, and fewer vitamins, than many “kid” cereals. Carroll Deck ¶ 46; DX 65-104; Tr. 67-70 (Cotterill); Tr. 2068-73 (Leekie). Plaintiffs expert concedes that you cannot distinguish one RTE cereal from another on the basis of a single attribute. Tr. at 71 (Cotterill). Cereal selection is motivated primarily by taste. Tr. at 2075-76 (Leckie); DX 49 at KGF 0261922. Although consumers responding to surveys report a desire for wholesomeness and nutrition, that purported desire is not consistently acted upon if taste is not delivered. Tr. at 563-64 (Carroll). For comparison purposes, manufacturers of RTE cereal group RTE cereals into subgroups. Different manufacturers have chosen to categorize cereals differently. Rubinfeld Deel. ¶ 54(a). For example, Nabisco used three RTE cereal segments—“adult,” “all family,” and “kid”—while Post used five segments—“simple health nutrition,” “taste enhanced wholesome,” “all family basic,” “family acceptable Md,” and “traditional Md.” DX 310 at KGF 0069534; DX 54 at KGF 0261814. Post classified both Grape-Nuts and Nabisco Shredded Wheat as “simple health nutrition” products, while Nabisco classified Grape-Nuts as an “all family” product and Nabisco Shredded Wheat as an “adult” product. Parker Deel. ¶ 11; Rubinfeld Deel. ¶ 54(b); Thomas Deel. ¶ 93; DX 54; DX 310. Nabisco classified many cereals as “all family” that Post classified as “family acceptable Md” or “traditional Md,” and that plaintiff, relying on Post’s classifications alone, excludes from its “adult” market. Nabisco classified Nabisco Frosted Wheat Squares, Kellogg’s Frosted Mini Wheats, General Mills Cheerios (Flavored), General Mills Golden Grahams, General Mills Wheaties Honey Gold, and Quaker Life as “all family,” together with, for example, Grape-Nuts, Cheerios and Kellogg’s Com Flakes. DX 310 at KGF 0069534. Plaintiff excludes these products from the “adult” market. The market segments used by Post to assess the competitive environment of Post products (and the particular products included in a segment) change over time in response to changes in market conditions, consumer preferences, and Post’s then-current understanding of RTE cereal competition. LecMe Deck ¶ 22; Carroll Deck ¶ 61-62; Tr. at 110 (Cotterill). The former Post market segments did not clearly distinguish between any “Md” and “adult” segments in terms of consumption. Post considered cereals in the “middle”—“family acceptable Md” and “all family basic”—to be consumed by everyone in the household. DX 54 at KGF 0261814; Carroll Deck ¶ 64. Even the segments at either “end” of the spectrum—“traditional Md” and “simple health nutrition”—were viewed by Post as “primarily,” but not exclusively, eaten by children and adults, respectively. DX 54 at KGF 0261814; Carroll Deck ¶ 64. The former Post market structure under-emphasized the fact that most consumers buy and consume products from all of the categories. Carroll Deck ¶ 68. For example, Post placed Grape-Nuts in the “simple health nutrition” segment in 1990, but products in the “all family” category accounted for a far larger percentage of Grape-Nuts users’ cereal requirements. LecMe Deck ¶23. The five Post market segments on wMch plaintiff bases its proposed “adult” market definition are no longer used by Post, because, among other things, Post concluded that there is far broader competition among brands than Post’s 1990 model recognized. Rubinfeld Deck ¶ 55; Carroll Deck ¶ 68. The evidence does not support plaintiffs claim that there is one group of consumers who eat primarily “adult” cereals and a separate group that eats primarily “Md” cereals. People of all ages generally eat all types of cereal, in varying volumes. Carroll Deck ¶ 21; Rubinfeld Deck ¶25; DX 35; DX 38. Even plaintiffs expert admitted at trial that adults eat cereals classified by Post as “traditional Md” cereals, and that children eat “simple health nutrition” cereals. Tr. at 37-38, 92-93 (Cotterill). A 1993 study conducted for Post by National Eating Trends (“NET”), which looked at individual eating patterns by age over a nine-year period, showed that approximately 70% of children eat “all family cereals,” a NET-defined category that includes most of the products that the plaintiff categorized as “adult.” Carroll Decl. ¶37; DX 40. The same study also showed that approximately 20% of all adults eat “pre-sweet” cereals, which plaintiff generally categorizes as the “kid” market. Carroll Deck ¶ 37; DX 40. NET has considered consumption in terms of “eating occasions”; their data show that adults account for about 40% of all eating occasions of pre-sweet cereals. Adults account for over 60% of the eating occasions of certain “kid” cereals, such as Frosted Flakes and Golden Crisp. DX 37 at 11. Adults aceount for 58% of the Honey Nut Cheerios eating occasions, 48% of the Quaker Life eating occasions, and 30% of the Kix eating occasions. Carroll Deck ¶ 41; DX 37 at 12-13. Looked at from the standpoint of how many adult eating occasions are eatings of pre-sweet cereals, 1991-1994 NET data show that 14.6% of all adult cereal eating occasions were of pre-sweet cereals—18-34 year-olds ate pre-sweet cereals 17.4% of the time; even those over the age of 44 ate pre-sweet cereals about 9% of the time. Carroll Deck ¶ 38; DX 36. “Age of Eater” information compiled for Post by HTI (an organization Post used often for home use product testing) in May 1993 from a survey of 50,000 households is consistent with the NET data cited above. Carroll Deck ¶ 39. It showed that a high proportion of the eaters of “kid” cereals are adults: adults constituted 29% of the reported eaters of Fruity Pebbles; 31% of the eaters of Lucky Charms; 44% of the eaters of Cap’n Crunch; 51% of the eaters of Com Pops; 62% of the eaters of Golden Crisp; and 67% of the eaters of Frosted Flakes. DX 38. It also showed that large proportions of children ate so-called “adult” cereals: 39% for Cheerios; 34% for Rice Krispies; 23% for Crispix; 15% for Wheaties; 15% for Com Flakes; 13% for Post Raisin Bran; 7% for Special K; and even 7% for Grape-Nuts and Shredded Wheat. Carroll Deck ¶ 39; DX 38. Adults over 48 who eat pre-sweet cereals eat such cereals nearly as many times during the week as children do. Carroll Deck ¶ 40; DX 40. “Kid” cereals account for 25% of the cereal eaten by those who also ate “all family basic” cereals, 23% of the cereal eaten by those who also ate “taste enhanced wholesome” cereals, and 17% of the cereal eaten by those who also ate “simple/health nutrition” cereals. DX 48 at KGF 0397101. Twenty percent of the time that eaters of “adult” cereals eat cereal, they eat a “kid” cereal. Carroll Deck ¶ 36. Indeed, these eaters of “adult” eat “traditional kid” cereals twice as often as they eat “family acceptable kid” cereals. Eaters of “all family basic” and “taste enhanced wholesome” cereals eat “traditional kid” cereals as often as they eat “simple health nutrition” cereals. Carroll Deck ¶ 36. Conversely, those who ate “kid” cereals nearly half the time (45%) also ate “adult” cereals. Id. Based on its 1993 panel data, NET projects that one in four adult RTE cereal eaters eats both “adult” and “kid” cereals. DX 35. The proportion is even higher for children; 50% eat both “Md” and “adult” cereals. Id. The other 50% of children fall almost evenly into two categories: those who eat only “adult” cereals and those who eat only “Md” cereals. Carroll Deck ¶43; DX 35. Sixty percent of households with no children purchase both “Md” and “adult” RTE cereals. Carroll Deck ¶¶28, 31; DX 373. Four percent of households without children buy only “Md” RTE cereals. The remaining households with no children (36%) purchase only “adult” products. Carroll Deck P 31; DX 373. Households without children buy a significant amount, 24%, of the volume of “traditional Md” cereals. Carroll Deck ¶ 29; Rubinfeld Deck ¶ 23; DX 39. Households without children buy 24% of the volume of Post Cocoa Pebbles, 23% of Kellogg’s Froot Loops, 25% of General Mills Lucky Charms, 35% of Post Golden Crisp, and 25% of Kellogg’s Corn Pops—all of which are categorized as “Md” cereals in plaintiffs proposed market definition. Carroll Deck ¶ 29; DX 39. Similarly, households without children buy 40% of the volume of “family acceptable kid” cereals. Tr. at 116 (Cotterill). They bought 39% of Kellogg’s Frosted Flakes, 39% of Honey Nut Cheerios, 46% of Apple Cinnamon Cheerios, and 53% of Honey Gold Wheaties. Carroll Decl. ¶ 29; DX 39. Households without children bought 53% of Nabisco Frosted Wheat Squares (now Frosted Wheat Bites) and 46% of Kellogg’s Frosted Mini-Wheats. Carroll Decl. ¶ 29; DX 39. Conversely, households with children bought 35% of Grape-Nuts Nuggets, 24% of Nabisco Shredded Wheat (Big Biscuit and Spoon Size), 24% of Kellogg’s Complete Bran Flakes, 21% of Quaker Puffed Rice and Wheat, 23% of Product 19, and 31% of Special K—all of which were in Post’s “simple/health nutrition” segment. Carroll Decl. ¶ 29; DX 39. Although cereal manufacturers direct some of their advertising for certain sweetened/flavored cereals at children, a cereal’s advertising target group is not a reliable test of who eats the cereal. For example, Post classified Golden Crisp as a “traditional kid” cereal, and directed all Golden Crisp advertising to children. However, the HTI Age of Eater Information shows that 62.3% of all eaters of Golden Crisp are 18 and over. Leckie Decl. ¶ 19; DX 38; Tr. at 506 (Carroll). Popular pre-sweetened products, such as Kellogg’s Frosted Flakes, Frosted Mini Wheats, and Corn Pops, and General Mills Apple Cinnamon Cheerios, direct separate television advertising campaigns to children and to adults; the time at which they air varies depending on the target group. Leckie Decl. ¶ 19; DX 245; Tr. at 1137-38 (Rubinfeld). Similarly, popular cereals that the State classifies as “adult,” such as Kellogg’s Corn Flakes and Raisin Bran and General Mills Cheerios, have advertising campaigns that are directed at children, which are separate from the advertising campaigns that they direct at adults. Leckie Decl. ¶ 19. Consumption of RTE cereals has increased steadily for many years. DX 46; DX 290 at KGF 0130467. Per capita cereal consumption has also increased, whether measured in terms of pounds eaten or in terms of eatings per person. Carroll Decl. ¶ 16; DX 40 at 10; DX 290 at 0130468. Over ninety percent of all households purchase RTE cereals. Tr. at 425-26 (Carroll); DX 51. Retailers report that RTE cereals are one of their fastest growing sales categories. Smith Decl. ¶ 9; see also Boehm Decl. ¶ 7. During the past decade, consumer demand has shifted substantially, from established, relatively plain cereals, including Nabisco Shredded Wheat and Grape-Nuts, to cereals with complex tastes and textures (e.g., cereals composed of a mix of different ingredients, flavors, and forms) as these new products have been introduced by manufacturers. Carroll Decl. ¶ 17; Leckie Decl. ¶ 23; Thomas Decl. ¶24. During that time period, cereal manufacturers have introduced many new products in the complex and sweetened/flavored segments of the market. Carroll Decl. ¶ 17; DX 256 at 16, 18-21. The introduction of new products has increased substantially during the past 5-10 years, and is in large part responsible for the growth in consumption in the RTE cereal category. Carroll Decl. ¶ 18; Leckie Decl. ¶ 6; Rubinfeld Decl. ¶ 91; DX 256 at 5; DX 258 at P000557. 2. Retail Customer Perspectives RTE cereal manufacturers sell their cereals to retailers, not to consumers. Leckie Decl. ¶ 50; Tr. at 29 (Cotterill). Post does not control the prices that retailers charge consumers for its cereals. Leckie Decl. ¶ 50; Schena Decl. ¶ 6; Smith Decl. ¶ 49; Boehm Decl. ¶ 47; Rubinfeld Decl. ¶ 13(a). Retailers typically classify all RTE cereals in one category, along with hot cereals, breakfast bars, granola, instant breakfast products, and toaster pastries. Boehm Decl. ¶ 30; Tr. at 1012 (Boehm); Smith Decl. ¶ 31. Retailers usually assign only one category manager or category management team to manage all RTE cereals and the other breakfast products that they classify with RTE cereals. Boehm Decl. ¶ 30; Smith Decl. ¶ 31. Retailers buy, price, and shelve all RTE cereals together as a single product type. Boehm Decl. ¶ 31; Smith Decl. ¶ 32; Rubinfeld Decl. ¶ 26. Retailers typically place all RTE cereal products in a single aisle of their stores. Smith Decl. ¶ 31; Boehm Decl. ¶ 30; Rubinfeld Decl. ¶ 26. Retailers generally shelve RTE cereals by manufacturer, rather than by product type. Smith Deel. ¶ 43; Boehm Decl. ¶ 40; Rubinfeld Decl. ¶ 26; Tr. at 681-82 (Smith); Tr. at 29 (Cotterill). Retailers do not distinguish between “kid” and “adult” cereals in allocating shelf space. Boehm Decl. ¶ 40; Smith Deel. ¶ 43; Rubinfeld Decl. ¶ 26; Tr. at 1051-52 (Boehm). Retailers pay little, if any, attention to the marketing “segments” into which RTE cereals are categorized. Leckie Decl. ¶ 20; Boehm Decl. ¶ 31; DX 289. Retailers choose to carry the cereals that they believe will sell the best. See Smith Decl. «40-41; Boehm Decl. « 35-37; Tr. at 683 (Smith). All RTE cereals compete with all other RTE cereals, and often with all other breakfast products, for retail shelf space based on each cereal’s overall sales performance. Boehm Decl. ¶¶ 30, 37; Smith Decl. ¶¶ 30, 41; Tr. at 1013-1014, 1048-49, 1052 (Boehm). Tr. at 643-44 (Smith). Retailers have noted that, when a particular RTE cereal is promoted, that product’s sales will increase, while the sales of all other RTE cereals will tend to decrease. Tr. at 1048-49 (Boehm); see also Tr. at 2092-95 (Leckie); DX 389. All RTE cereals, regardless of type, compete with one another for scarce retailer end-aisle display space. In order to obtain retail feature-priced displays and other trade promotions, every Post or Nabisco RTE cereal, regardless of its classification, must be offered at the same or a lower retail feature price point as all other products competing for retail display promotions and other trade promotions. Leckie Decl. ¶ 30; Tr. 608-10,613-16 (Leckie). It is not uncommon to sell twelve times the normal weekly volume in one week when a cereal has a large end-aisle display and an attractive feature price. Tr. at 614 (Leckie); DX 385-389. Thirty percent of Post’s cereal sales volume comes through trade deals. For Post to be competitive, Post must convince retailers to accept its trade promotion offers instead of offers made by other manufacturers. In this way, all RTE cereals compete with one another. Tr. at 608-609 (Leckie). If Post is not competitive on its trade promotions and thus does not cause retailers to give it feature ads or store displays, Post loses a very large volume of sales. Leckie Deel. «12, 30; Tr. at 615, 2063 (Leckie). 3. Competition From, Products in Other Product Categories RTE cereal products experience some competition at the margin from other product categories. RTE cereals compete to some extent with all other breakfast products. Boehm Decl. ¶ 29; Smith Decl. ¶ 30; Rubinfeld Decl. ¶28; Tr. at 1025 (Boehm). Retailers have noted consumer shifts in purchasing among the various products offered on the breakfast aisle. Smith Decl. ¶31. Many consumers substitute a broad range of other breakfast food products for RTE cereal: Rubinfeld Decl. ¶ 28; Leckie Decl. « 32-33; Thomas Decl. ¶65. For example, the data suggest that New York area consumers substitute frozen bagels for RTE cereal. Leckie Deel. ¶ 33. Many consumers view hot cereal as a substitute for RTE cereal, particularly during the winter months. Leckie Decl. ¶ 32. RTE cereal sales decline overall during the first quarter of each year, and retailers have observed significant seasonal shifting between RTE cereal and hot cereal. Leckie Decl. ¶ 32; Smith Decl. ¶ 31. To take advantage of this, Nabisco promoted Nabisco Shredded Wheat as a hot cereal during winter months, and Post has promoted Grape-Nuts as a hot cereal during January since the mid-1980s. Leckie Decl. ¶ 32; Thomas Decl. ¶ 65; Rubinfeld Decl. ¶ 28(b). Both Spoon Size and Big Biscuit Shredded Wheat have lost substantial volume to hot cereals. One Nabisco document states that over 24% of Spoon Size losses and 50% of Big Biscuit losses went to hot cereals during one period in 1988-89, and that these products lost more volume to hot cereals than they did to Grape-Nuts and other RTE cereal products in what Nabisco characterized as an “adult health non-bran” segment. DX 301 at NBA 0067072-73; Rubinfeld Deel. ¶ 28(b). In addition to competing against other breakfast foods, RTE cereals also compete against snack foods. Smith Decl. ¶30; Tr. at 1065-66 (Boehm). 4. Supply-Side Considerations There is substantial supply-side substitutability among many RTE cereals, including many products inside and outside of plaintiffs proposed “adult” market. Leckie Decl. ¶¶ 25-28; Rubinfeld Decl. ¶ 16. Most firms in the industry make a wide variety of RTE cereals, using standard manufacturing processes such as flaking, puffing, extruding, granulation, and shredding. These processes are not dedicated to producing cereals in only one marketing segment; rather, the same process is used to make products in various segments. Leckie Decl. ¶ 25; Rubinfeld Decl. ¶ 16. If prices for products in one or more segments were to increase significantly, RTE cereal producers would have an incentive to switch some of their production capacity to produce more of the cereals in those segments. Leckie Decl. ¶¶ 27-28; Rubinfeld Decl. ¶ 16. More specifically, there are a number of different “families” of similar RTE cereals, consisting primarily of an established cereal and its line extensions, where the individual cereals within the “family” are produced by virtually identical manufacturing processes, but fall into different marketing categories. Examples of such “families” include (1) Kellogg’s Corn Flakes and Frosted Flakes, (2) Kellogg’s Rice Krispies and Cocoa Krispies, (3) Nabisco’s Spoon-Size Shredded Wheat and Frosted Wheat Bites, (4) General Mills Cheerios, Honey-Nut Cheerios, Apple Cinnamon Cheerios, and Multigrain Cheerios, and (5) General Mills Wheaties and Honey Gold Wheaties. Some complex products, such as Post’s Honey Bunches of Oats, Fruit ’N Fibre, and Blueberry Morning, contain flakes that could be used to make cereals that fall into other marketing categories. Leckie Decl. ¶ 26. Similar processing methods are used to make both products that plaintiff includes in the “adult” market and products that plaintiff excludes from that market. Rubinfeld Decl. ¶ 17; see also Leckie Decl. ¶¶ 25-28. If prices for cereals in the “adult” market were to increase significantly relative to the prices of other cereals, Kellogg, General Mills, Post and other manufacturers could increase production of “adult” cereals relatively quickly and easily by redirecting the use of some portion of the capacity currently used to produce “kid” cereals. Rubinfeld Decl. ¶ 17; see also Leckie Decl. ¶ 27. This capability is particularly obvious for “pairs” of “adult” and “kid” versions of essentially the same product. For example, Kellogg could switch from production of Cocoa Krispies to regular Rice Krispies; General Mills could switch from Honey-Nut Cheerios and Apple Cinnamon Cheerios to regular Cheerios; and Post could switch from Frosted Wheat Bites to Spoon Size Shredded Wheat in a few days, without incurring any costs, if there were a business reason to do so. Leckie Decl. ¶27. The switch could be completed in a matter of days, at little or no cost, for many cereals. More generally, manufacturers using a particular process—such as flaking—to produce “kid” cereals could readily use their existing capacity to produce more cereals in the “adult” segment. For example, Kellogg could immediately shift from producing Frosted Flakes to producing any of its “adult” flaked cereals, including Kellogg’s Corn Flakes, Product 19 and Special K. Rubinfeld Decl. ¶ 18. Post could quickly and inexpensively shift from producing its complex adult and all family cereals that contain sweetened flakes to producing the same cereals with unsweetened flakes. Leckie Decl. ¶27. Post could also, at no cost, quickly shift from making Fruity Pebbles and Cocoa Pebbles to making a basic crispy rice product. Id. Post could shift from producing Alpha-Bits or Marshmallow Alpha-Bits to producing an unsweetened, toasted oat type “adult” product in approximately two to three months, at a cost of about $50,000 for the equipment needed to extrude the new product in a different shape. Id. Conversely, Post also could convert assets now used to make “adult” cereals to the production of “kid” cereals. For example, if Post believed that it could increase profits by shifting from producing the Toasties line to producing a sweetened frosted flake product, it could do so almost immediately, at no cost, because the equipment used to produce Toasties already has coating and drying equipment associated with it. Leckie Deck ¶ 28. (Leckie). 5. Conclusion: the Relevant Product Market Is All RTE Cereals RTE cereals are so highly differentiated, and compete with one another along so many different dimensions, that there is no clear break in the chain of substitutes among cereals that would permit definition of a market smaller than all RTE cereals. Tr. at 828-9, 1400 (Rubinfeld); Tr. at 1394-97 (Falk). Consumers buy and eat a wide variety of RTE cereals. Any substantial price increase for any one type of RTE cereal would lead to significant demand-side substitution of many other RTE cereals. Rubinfeld Decl. ¶20. For example, cereals representing large shares of Grape-Nuts buyers’ requirements directly affect the pricing and marketing of Grape-Nuts, and vice versa. See Parker Decl. ¶¶ 8, 34. Consequently, the pricing and marketing strategies of Grape-Nuts are affected by a wide spectrum of products. Rubinfeld Decl. ¶ 21. In addition, an increase in the prices of products in one market segment would lead to the supply-side substitution by manufacturers referred to above. Although the calculation of cross-price elasticities of demand does not in and of itself provide a market definition, it does provide information relevant to assessing whether a proposed market definition is or is not reasonable. Rubinfeld Decl. ¶ 44. In this ease, cross-price elasticities confirm that there is demand-side substitution between cereals inside and cereals outside plaintiff’s proposed “adult” market. Rubinfeld Deck ¶ 24. There is a consistent and “robust” pattern of statistically significant positive cross-price elasticities among cereals in each of the five Post marketing segments. Rubinfeld Deck ¶ 31. The cross-price elasticities calculated by Professor Rubinfeld are significant for many cereals across a wide range of market segments. Rubinfeld Deck ¶ 32; Appendix 4 Regressions; Tr. at 954-5 (Rubinfeld). To take just one example, Grape-Nuts has a statistically significant cross-price elasticity with Kellogg’s Frosted Mini Wheats, and Frosted Mini Wheats has statistically significant cross-price elasticities with other “adult,” “family acceptable kid,” and “traditional kid” cereals. Rubinfeld Deck ¶21 n. 10 and Appendix 4, Regression 1. Cross-price elasticity is a more useful tool than own-price elasticity in defining a relevant antitrust market. Cross-price elasticity estimates tell one where the lost sales will go when the price is raised, while own-price elasticity estimates simply tell one that a price increase would cause a decline in volume. Tr. at 929-30 (Rubinfeld). Plaintiffs economist, Ronald W. Cotterill, relied on several types of Post data, as well as an econometric analysis, in forming his opinion that the relevant product market should include only “adult” RTE cereals. The “brand interaction indices” that he relied upon do not measure the degree of substitutability between products. Rubinfeld Deck ¶¶ 22, 29-32. Interaction indices show only the extent to which a given cereal contributes to the total volume of cereal purchased by consumers of another cereal relative to the first cereal’s share of market during a given period. Carroll Deck ¶ 9; see Tr. at 120-21 (Cotterill). Interaction indices do not show “switching” patterns over time; nor do they relate consumption patterns to price changes or other factors, such as promotions or advertising. Interaction indices measure only the propensity of households to purchase two products. Interaction indices are not equivalent to, or proxies for, cross-price elasticities, because they do not purport to measure changes in consumption as a function of changes in price. Carroll Deck ¶ 9; Tr. at 416-17 (Carroll); see Tr. at 119-20 (Cotterill); Tr. at 960 (Rubinfeld). With respect to RTE cereals, where there are hundreds of brands and no one brand has more than a tiny share of the market, interaction indices have limited significance. Carroll Deck ¶ 12. Post does not rely exclusively on interaction indices to make major business decisions. Interaction indices are simply one piece of information provided to Post once a year. Tr. at 409-410 (Carroll). To the extent that plaintiffs expert relies on interaction data, the data do not support his suggested market definition. In fact, interaction data show that buyers of “kid” cereals are nearly as likely to buy “adult” cereals as are buyers of RTE cereals generally. DX 31; Carroll Deel. ¶ 32. For example, Nabisco Frosted Wheat Squares, which plaintiff categorizes as an “adult” cereal, actually has higher interaction indices with the “traditional kid” (106) and “family acceptable kid” (144) segments, looked at as a whole, than with any of the “adult” segments. DX 30. The same is true for Kellogg’s Frosted Mini-Wheats, which plaintiff also includes in its “adult” market, which has a 163 interaction index with the “family acceptable kid” segment and a 142 interaction index with the “traditional kid” segment as a whole. DX 30. Plaintiffs reliance on interaction indices to define distinct “kid” and “adult” markets is misplaced. Interaction indices do not demonstrate any clear dividing line between “kid” and “adult” products. Carroll Deck ¶ 34. Professor Cotterill also relied on “shifting studies” performed for Post. See Cotterill Aff. ¶¶ 28-30. Shifting studies provide information about purchases from one period to the next, but they do not identify the factors that led to those changes. Carroll Dec. ¶ 49. They do not relate changes in consumption to changes in price, to changes in promotional activity, to changes in advertising, or to other factors that may affect purchasing. Carroll Dec. ¶ 49. In fact, shifting data indicate that a high proportion of the volume shifting that occurred in the RTE cereal market between 1992 and 1993 was among market segments, especially between so-called “adult” -and “kid” segments. A Nielsen analysis found, for example, that 71% of shifting gains and losses experienced in the “traditional kid” segment occurred with the “adult” segments of the market, compared to 26% with the “family acceptable kid” segment. Conversely, each of the “adult” segments experienced substantial exchanges of volume with the “kid” segments: 48% with the “all family basic” segment, 45% with the “taste enhanced wholesome” segment, and 35% with the “simple health nutrition” segment. It is not appropriate to use, as plaintiffs expert has done, the marketing segments adopted from time to time by any particular RTE cereal producer to define a relevant economic product market. Rubinfeld Deck ¶ 53. The market segments that Post managers use in analyzing the performance of particular products do not define the scope of competition for those products. Leckie Deck ¶ 21; Tr. at 605 (Leckie); Carroll Deck ¶ 64; Tr. at 412-13 (Carroll). In a differentiated product market such as the RTE cereal market, the decision whether to include a product in the market is inevitably somewhat arbitrary, because not all products in a relevant market compete equally with all other products. In such a market, any market definition is likely to exclude some products that are reasonable substitutes for products that are included in the market. Whether such errors of inclusion or exclusion are significant depends in part on the relevant cross-price elasticities of demand between the individual products. Professor Cotterill’s suggestion that cross-price elasticities are not relevant (Cotterill Aff., Exhibit C, p. 11) is incorrect. Professor Cotterill failed to provide a consistent and persuasive methodological basis for deciding which RTE cereal products to include in his “adult” market. Professor Cotterill’s model employed a three-stage decision tree for RTE cereal purchase choices: (1) whether or not to buy a RTE cereal; (2) whether to buy an “adult” cereal or a “kid” cereal; (3) whether to buy a particular cereal within a segment. Professor Cotterill’s assumptions about purchase and decision-making behavior are unreliable, because they are biased. Tr. at 1394-1407 (Falk); Tr. at 2349 (Kahn). For example, Professor Cotterill’s model merely assumed that an increase in the price of Grape-Nuts or Nabisco Shredded Wheat (or any other “simple health nutrition” cereal) will have no direct effect on products in other marketing segments. Rubinfeld Deck, Appendix 6. His model was biased in favor of finding very low cross-price elasticities between “adult” and “kid” cereals. Tr. at 973 (Rubinfeld); Tr. at 1394-1407 (Falk). Professor Cotterill’s conclusion that the demand for “adult” cereals is relatively inelastic is at odds with empirical evidence concerning cross-price elasticities and other credible evidence in this case. See, e.g., Parker Deck ¶¶ 11-12, 14; Carroll Deck ¶¶ 21, 56. Moreover, survey data do not support the notion that consumers consistently follow the decision tree structure used by Professor Cotterill. Tr. at 987 (Rubinfeld). Professor Cotterill’s estimated price elasticity of demand analysis was flawed also because it did not consider the likelihood of a supply response, that is, it does not consider the likelihood that a cereal manufacturer will shift production from “kid” to “adult” cereals (or vice versa). A price increase that would be profitable absent a supply response may well be unprofitable if potential supply responses are taken into account. Tr. at 1194-1196 (Rubinfeld). The independent economic expert appointed to assist the court, Alfred E. Kahn, concluded on the basis of all the evidence presented at trial that the relevant product market in which to assess the competitive significance of the Acquisition is all RTE cereals. Tr. at 2359 (Kahn). Professor Kahn based his conclusion in part on his views that (1) there is no clear break in the chain of substitutes among RTE cereals, and (2) the chain of substitutes for RTE cereals is not along straight, unidirectional lines radiating out in a uniform and orderly manner. Tr. at 2340 (Kahn). Professor Kahn also expressed the view that any market definition and any measure of market concentration that ignores the dynamic aspects of changing demands for RTE cereal would produce misleading results. Tr. at 2352-54, 2359 (Kahn). Professor Kahn concluded that “the proper market has to be the whole RTE cereal [market] because that is the locus of the competition of the particular kinds that seem to be so important in this industry.” Tr. at 2359 (Khan). The court agrees with his conclusions. II. Market Share Measurement A. Actual Shares of Market in 1992 In 1992, the market shares of the major RTE cereal manufacturers, as reported in Nielsen Scantrack data, were: Kellogg 37.03% General Mills 25.12% Post 11.70% Quaker 6.82% Ralston Purina 4.58% Nabisco 2.82% Private label/Generic 8.52% Other 3.40% Carroll Decl. ¶ 15; Rubinfeld Decl. ¶ 9, Figure 1. These market shares provide an accurate measure of concentration in the industry. Rubinfeld Decl. ¶ 66; see also Rubinfeld Decl. ¶¶ 61-65. The 1994 partial year market shares relied on by plaintiff are not an accurate measure of the competitive significance of the firms in the industry. Leckie Decl. ¶ 11 n. 2. General Mills suffered a contamination problem earlier in 1994 that had a dramatic, but probably temporary, depressing effect on General Mills sales during the second quarter of 1994. The 1994 year to date market shares data used by plaintiff therefore understate the significance of General Mills. Leckie Decl. ¶ 11 n. 2. Market share data that exclude the fourth quarter of a year are distorted because they are likely to understate Ralston’s market share; Ralston’s Chex products sell a disproportionate amount of their volume during the fourth quarter each year, when they are promoted as a holiday snack food. Leckie Decl. ¶ 11 n. 2; Rubinfeld Decl. ¶ 60 n. 27; see also Tr. at 1066 (Boehm). The pre-acquisition Herfindahl-Hirschman Index (“HHI”) for the RTE cereal market was 2215; the increase in the HHI resulting from the Acquisition is approximately 66 points. Rubinfeld Decl. ¶ 60 and ¶ 9 n. 2 Table 1. Concentration in the RTE cereal industry declined between 1970 and 1992. In 1970, the HHI for the RTE cereal industry was about 2955. By 1986, it had declined about 17%, to 2458. By 1992, it had declined approximately 10% more, to 2215. Rubinfeld Decl. ¶ 60(b) and Figure 2; Tr. at 250-253 (Cotterill); DX 336 at NBA 0021354-57 (volume data for 1980 through 1990). Since the Acquisition, concentration in the RTE cereal industry has continued to decline. Although the Acquisition increased the HHI by 66 points, the HHI declined in 1993 by 58 points. Accordingly, the net increase since the Acquisition has been only 8 points (to 2223 from 2215). Rubinfeld Decl. ¶ 60(b) and Figure 2. B. Measurement of Nabisco Share Nabisco’s 1992 actual market share, 2.82%, is the appropriate measure of Nabisco’s competitive significance at the time of the Acquisition. Rubinfeld Decl. ¶ 66. 1. Nabisco was in decline before its acquisition by KKR Between 1986 and 1989, the Nabisco Shredded Wheat line’s market share fell from 4.11% to 2.89%. DX 174; Tr. at 1429 (Cotterill); see also Thomas Decl. ¶¶ 17, 36; Tr. at 1877 (Thomas); Rubinfeld Decl. ¶ 66(b). Nabisco’s Shredded Wheat cereals began losing market share before Nabisco was acquired by Kohlberg, Kravis and Roberts (“KKR”) in 1989. Its loss of market share through 1992 is attributable both to a short term profit strategy (the so-called “harvest strategy”), stressed by plaintiff, and to a number of factors plaintiff downplays, including: (1) reduced consumer interest in Nabisco Shredded Wheat, and consumer movement to new, tastier, and more complex products; (2) the oatbran “craze”; (3) lack of effective advertising; (4) lack of successful product innovations and lack of a full line of products; (5) line extensions that cannibalized the core Nabisco Shredded Wheat cereals; (6) quality problems; (7) Kellogg’s introduction of an unsweetened shredded cereal; and (8) a change in Nabisco’s distribution. Thomas Decl. ¶¶ 21-33; Hinkes Decl. ¶ 23; Rubinfeld Decl. ¶ 66; Tr. at 1877-78 (Thomas); DX 176 at KGF 0118036-039; Leckie Decl. ¶ 65. The core Nabisco Shredded Wheat cereals—Spoon Size and Big Biscuit—had started their decline in market share even earlier. Between 1985 and 1989, Spoon Size’s market share declined about 43 percent. During the same time period, Big Biscuit’s market share declined about 35 percent. Thomas Decl. ¶ 18; DX 174; DX 317 at NBA 1043333. The decline in market share for the Nabisco Shredded Wheat line in the three years preceding the so-called “harvest” period from 1989 to 1992 was almost twice as big as the decline in market share during the “harvest” period. Tr. at 1430 (Cotterill). Nabisco’s temporary increase in its total market share in 1987 and 1988 was due to a very expensive launch of new products, which products were ultimately unsuccessful. 2. Shift in Consumer Preferences Beginning in the 1980s, several new RTE cereal products came on the market that were promoted as having health benefits, such as being high in fiber or fortified with vitamins and minerals. Although Nabisco had stressed traditionally that Nabisco Shredded Wheat was a natural product, it did not have the same health benefits of these new brands and lost market share to them. Thomas Decl. ¶ 22; Rubinfeld Decl. ¶ 66(c). a. Oat bran craze Nabisco Shredded Wheat’s market share was strongly affected by the oat bran “craze” that resulted from reports that a diet high in oat bran reduced blood levels of cholesterol. DX 330 at KGF 0068696; Rubinfeld Deel. ¶ 66(c). These reports resulted in a dramatic surge in the development and sale of oat bran RTE cereals, including Kellogg’s Common Sense Oat Bran and Cracklin’ Oat Bran, and line extensions of General Mills Cheerios. DX 319 at KGF 0067129; Thomas Deck ¶ 23; Tr. at 1573 (Cotterill). Between 1988 and 1989, a significant number of Nabisco Shredded Wheat consumers switched to oat bran products, which caused 38 percent of Nabisco Shredded Wheat’s volume loss. DX 319 at KGF 0067130; DX 321 at KGF 0089663; Thomas Deck ¶ 23; Tr. at 1859 (Thomas); Tr. at 1436 (Cotterill). When the oat bran craze ended in 1990, Tr. at 1859 (Thomas), many consumers did not return to Nabisco Shredded Wheat. Rubinfeld Deck ¶ 66(b) n. 38; DX 176. An October 1991 report prepared for Nabisco by Information Resources, Inc. (“IRI”) concluded that Nabisco cereals had contributed 9.2 percent of oat bran’s total volume gain during its period of growth, but received only 1.1 percent of oat bran’s loss when the craze ended. Thomas Deck ¶ 23; DX 343 at KGF 0070040a72. b. Taste enhanced cereals Between 1983 and 1992, the established, basic RTE cereals of all firms lost significant market share to new “complex” products, such as products that were enhanced with the addition of dried fruits, nuts, sweeteners, or spices. Leekie Deck ¶ 64. For example, between 1983 and 1993, General Mills established products declined 31% (or 5.6 share points), and Kellogg’s established products declined 18% (or 6.2 share points). DX 256 at 5; Leekie Deck ¶ 64. Nabisco Shredded Wheat is an unsweetened, basic cereal. Thomas Deck ¶ 22. Many consumers stopped buying Nabisco Shredded Wheat, in part, because they thought that it was not as tasty as other cereals perceived to be “healthy,” and Nabisco Shredded Wheat lost share to those new cereals. Rubinfeld Deck ¶ 66(b) n. 38; DX 299; Thomas Deck ¶ 24; Carroll Deck ¶ 17. 3. Lack of effective advertising and distribution In the late 1980s and early 1990s, Nabisco failed to develop an advertising campaign for Nabisco Shredded Wheat that was appealing to consumers. Thomas Deck ¶25; DX 333 at KGF 0064587; Tr. at 1860 (Thomas). When Nabisco was unable to develop an effective advertising campaign for Nabisco Shredded Wheat, Nabisco reduced its advertising expenditures. Thomas Deck ¶ 25; Tr. at 1860, 1906 (Thomas); DX 306. Nabisco arrested the decline of Nabisco Shredded Wheat only when it developed an effective advertising campaign in April 1992. Thomas Deck ¶ 39. At that point, Nabisco increased expenditures on advertising by $3.4 million over the 1992 budget, and, as a result, began to stabilize Nabisco Shredded Wheat’s market share in the last half of 1992. DX 276; PX 50; Thomas Deck ¶26; Tr. at 1878 (Thomas). Another factor contributing to Nabisco’s decline was that Nabisco changed from a direct sales force to a less effective broker sales force in 1990. Hinkes Deck ¶23; Thomas Deck ¶ 15. 4. Lack of full line of products Because consumers seek variety in RTE cereals, successful new product introductions are important to increasing market share. Thomas Deck 127; Leekie Deck ¶¶ 6, 63; Rubinfeld Deck ¶ 91; DX 256. Nabisco’s ability to develop successful new products was very limited compared with larger companies, such as Kellogg and General Mills, which had a variety of manufacturing processes, research and development resources, and the marketing ability to launch successful new products. Nabisco lacked diversified product lines that would permit a wide variety of line extensions and cross-promotion opportunities. Thomas Deck ¶ 27; Tr. at 1885 (Thomas). Nabisco essentially was limited to introducing new shredded products. Tr. at 1885 (Thomas). This limited ability to launch new products restricted Nabiseo’s ability to expand its market share. Thomas Deck ¶ 28. Although Nabisco launched several new products in the mid 1980s, Thomas Decl. ¶ 29, each of these new products met with only limited success. DX 804. Consumers tried them initially, but the new products’ market shares rapidly fell below 0.5 percent. DX 174; Thomas Decl. ¶ 29. These products also contributed to the decline in market share of Nabisco’s core products—Spoon Size Shredded Wheat and Big Biscuit—by cannibalizing them. Thomas Decl. ¶ 29. Other Nabisco new product introductions were also unsuccessful. In 1989, Nabisco introduced two new oat bran products, Shredded Wheat with Oat Bran and 100% Bran With Oat Bran, but did so too late to benefit from the oat bran “craze.” Those products never achieved a significant market share, and were phased out from 1990 to 1991. Thomas Decl. ¶30. To the extent that they attracted buyers, they did so at the expense of the core Nabisco Shredded Wheat products. For example, some Nabisco Shredded Wheat users stopped buying Spoon Size Shredded Wheat and tried Shredded Wheat With Oat Bran. DX 299 at NBA 1020451. However, when consumers abandoned Shredded Wheat With Oat Bran, they tended not to revert to Spoon Size Shredded Wheat. Thomas Decl. ¶ 31. Nabisco introduced Teddy Graham Breakfast Bears in 1989, hoping to capitalize on the success of Nabisco’s Teddy Graham cookies. However, Breakfast Bears failed because they disintegrated into mush when milk was added. Thomas Decl. ¶ 32; Rubinfeld Decl. ¶ 68(d). Nabisco spent over $20 million advertising and promoting Breakfast Bears, and it lost more than $11 million (plus its research and development costs) on the product. Thomas Decl. ¶ 32; DX 304 at KGF 1161925; Rubinfeld Decl. ¶ 68(d). Nabisco’s failure caused it to be chary of introducing new products; Nabisco did not introduce any new RTE cereal on a national basis after 1989. Thomas Decl. ¶ 33. 5. Harvest strategy Plaintiff contends that the decline in Nabisco’s market share from 1989 to 1992 was due primarily to an intentional effort by Nabisco to “harvest” its RTE cereal business for short-term profits at the expense of its market share. The evidence does not support this contention. The decline in