Cocoa-Cola Bottling Company of the Southwest

Text (OCR of the scan at left; may contain errors)

IN THE MA TTER OF THE COCA-COLA BOTTLING COMPANY OF THE SOUTHWEST FINAL ORDER, OPINION, ETe. , IN REGARD TO ALLEGED VIOLATION OF SEe. 7 OF THE CLA YTON ACT AND SEe. 5 OF THE FEDERAL TRADE COMMISSION ACT Docket 9215, Complaint July 1988--Fina/ Order, lIug. , 1994 This final order requires Coca-Cola Bottling Company of the Southwest to divest within 12 months, the Dr Pepper franchise it acquired from San Antonio Dr Pepper Bottling. If the divestiture is not completed within that period, the Commission may appoint a trustee to complete it. In addition, the order requires the respondent to obtain Commission approval before acquiring any branded carbonated soft drink interests in any area in which it already makes distributes or sells branded concentrate or syrup, or branded carbonated soft drinks.

Appearances For the Commission: James E. Elliott, Thomas B. Carter and Mary Lou Steptoe.

For the respondent: Gregory Huffman, Thompson Knight Dallas, TX.

INITIAL DECISION BY JAMES P. TIMONY. ADMINISTRATIVE LAW JUDGE JUNE 14, 1991 BACKGROUND Companies and Persons 1. Respondent Coca-Cola Bottling Company of the Southwest CCSW") is a privately-held corporation with headquarters in San Antonio, Texas. (CX 980- U; RX 549-A.) Its sales in 1988 were $145,496 000. (CX 3806- * Complaint previously published 112 FTC 588 (1989). THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 453 452 Initial Decision 2. In 1983 the Biedenharn family consolidated their holdings in Temple, Uvalde and San Antonio Coca-Cola Bottling Companies into CCSW, and established The Biedenharn Corporation to hold the stock of CCSW. (RX 232- C.) In December 1986, The Biedenhar Corporation merged with CCBG Merger Corp., a subsidiary of Texas Bottling Group, Inc. ("TBG"), resulting in the sale of the Biedenharn family s interest in CCSW. (CX 3052; RX 549- , B; R. Hoffman Tr. 5588.) The Biedenharn family of Vicksburg, Mississippi was the first bottler of Coca-Cola. (Howell, Tr. 4005; RX 232- 3. TBG is the sole shareholder of CCSW. (CX I 372-H; CX 1373- 23; RX 572-1.) Affiliates of Prudential Insurance Company of America hold 51 % of the stock of TBG and 49% is held by The Coca-Cola Bottling Group (Southwest), Inc. ("CCBG-Texas ), a Texas corporation, which is a wholly-owned subsidiary of The Coca- Cola Bottling Group (Southwest), Inc. ("CCBG-Delaware ), a Delaware corporation. (Hoffman, Tr. 5603; CX 1372-G, H. ) All of the voting stock of CCBG-Delaware is held by Edmund M. Hoffman and Robert K. Hoffman (the "Hoffmans ). (RX 572-H; RX 2805- , Z- I5.

4. Edmund M. Hoffman is the majority shareholder of CCBG- Delaware. He is also the Chairman and a member of the Board of Directors of each corporation controlled by CCBG-Delaware including CCSW, and is the father of Robert K. Hoffman. (RX 2805- 15; CX 1372- 37.

5. Robert K. Hoffman is the second largest shareholder of CCBG-Delaware, and the only other voting shareholder. (RX 2805- 15; CX 1372- 37.) Robert Hoffman is the President ofCCBG- Delaware and of all of its subsidiaries except CCSW, of which he is Vice-Chainnan; he is a Director of all entities in the corporate group. (CX 1373- 89.

6. Southwest Coca-Cola Bottling, Inc. ("SWCC"), a whollyowned subsidiary of CCBG- Texas, is the Coca-Cola bottler in West Texas, Eastern New Mexico, Western Oklahoma and parts of Colorado and Kansas. (CX 4; CX 2805- , Z- ) SWCC is a franchisee of The Coca-Cola Company. (RX 2805- 5, Z- 7. Snappy Snack is an operating division of CCSW which provides full-line vending and food service in the San Antonio area. (CX 3211.) Bev-Tex until 1986 was a division ofCCSW selling fountain syrup and service, and selling and leasing fountain Initial Decision 118 FT. refrigeration and institutional kitchen equipment in the San Antonio area. (CX 28-L; RX 232; CX 2068-A.) 8. E. T. ("Toby ) Summers II is President and Chief Operating Officer of CCSW. (Summers, Tr. 6360.) Norborne Cole was President of CCSW from 1982 until January 8, 1988. (RX 2805- 15.

9. The Dr Pepper Company was a publicly-held corporation with headquarters in Dallas, Texas until 1984, when Forstmann-Little & Co. acquired it in a leveraged buyout. (CX 614-B: RX 1447-D: RX 990- , N. ) After selling the headquarters building, bottling operations, and other assets, except the Dr Pepper franchise contracts and the syrup manufacturing facilities, Forstmann-Little sold Dr Pepper Company in 1986 to a group of investors led by Hicks & Haas Holdings, Inc. (RX 990- 10. In 1986, a group which included some Dr Pepper Company shareholders and bondholders bought Seven-Up Company and combined the administration for the two companies in Dallas, Texas and the manufacturing for the two companies in St. Louis, Missouri. (Knowles, Tr. 2640. ) In 1988, the Dr Pepper Company and the Seven-Up Company were combined into Dr Pepper/Seven-up Companies, Inc., the current franchiser of the Dr Pepper and Seven- Up bottling operations in the United States. (RX 1989, pp. 3- ) Dr Pepper/Seven-up Companies, Inc. is the owner of the trademark and manufacturer of concentrates for Dr Pepper and Seven-Up brand products. (Clarke, Tr. 4297-99: Knowles, Tr. 2638-41.) The term DPCSA" is used here to mean Dr Pepper Company and its successor Dr Pepper/Seven-up Companies, Inc.

11. Until 1984, DPUSA owned bottling operations in Dallas/Fort Worth, Waco, Houston, San Antonio, and Corpus Christi, Texas. (RX 1648- 29- 31; Turner, Tr. 916; Antle, Tr. 3041 , 3079. 12. San Antonio Dr Pepper Bottling Company ("DP-SA") was a wholly-owned subsidiary of DPUSA. (RX 1648- 29; Turner, Tr. 917-918; Antle, Tr. 3041.) DP-SA sold its bottling plant to Grant- Lydick, Inc. on October 31 , 1984. (RX 2409. 13. From 1982 until the company-owned bottling plants were sold, DP-SA and the other company-owned plants were overseen by Jim Turner, as executive officer in the DPUSA offices in Dallas Texas. (Turner, Tr. 914- , 1035-37; Antle, Tr. 3083-85. 14. Grant-Lydick Beverage Company ("Grant-Lydick") does business in San Antonio, Austin, Corpus Christi, Victoria and South THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 455 452 Initial Decision Texas; in San Antonio, Grant-Lydick uses the trade name Big Red Bottling Company. (Lydick, Tr. 2992-3008. ) Grant-Lydick was formed by Bud Grant and Lee Lydick in April 1984 to get into the soft drink bottling business by purchasing some of the assets of DP- SA. (RX 1648- ) Emery Bodnar is Executive Vice President general manager and part owner of Grant-Lydick. (Bodnar, Tr. 1225. 15. Pepsico, Inc., with headquarters in Purchase, New York, is in the snack, restaurant and soft drink businesses. (RX 2864-D; RX 1218, pp. PC027073-74; Davis, Tr. 4619-4624. ) Its sales in 1988 exceeded $13 bilion. (RX 1218, p. 116. ) Pepsico, Inc. receives one-third of its revenue from soft drinks, the rest coming from its snack and restaurant businesses. (Summers, Tr. 6767-68. 16. Pepsi-Cola Company ("Pepsi lTSA") is a division of Pepsico, Inc. (RX 2864- 34. ) Pepsico, Inc. owns the United States trademark, and produces concentrate for Pepsi-Cola and other brands of soft drinks. (Davis, Tr. 4463, 4638.

17. Pepsi USA owns bottling operations in various parts of the United States, including San Antonio, Houston, Danas/Fort Worth, and Austin, Texas. (Amrosowicz, Tr. 791-793 , 837- 838.) These company-owned bottling operations are responsible for 37% of Pepsi USA bottle and can sales. (RX 1218; p. PC027073. 18. Pepsi USA' s operations were known as the Pepsi Bottling Group. (RX 1213; RX 1216.) In 1987 the name was changed to Pepsi COBO (Company-Owned Bottling Operations). (Amrosowicz Tr. 787. ) The term "Pepsi COBO" is used here to refer to Pepsi company-owned bottling entitles, before and after 1987. 19. The Seven-Up Company ("7-Up USA") is currently part of DPUSA, with headquarters in Dallas, Texas. (Knowles, Tr. 2639. Philip Morris, Incorporated bought 7-Up USA in the mid-70' s to enter the soft drink business, but sold it on November 12, 1986 to an investor group hcaded by Hicks & Haas Holdings, Inc. (RX 1990, p. 3; Knowles, Tr. 2685.

20. 7-Up USA owned 7-Up bottling operations in various parts of the United States. (CX 3941 , pp. 263-64; CX 997.) From 1982 to January 1986, 7-UP USA owned the Seven-Up Bottling Company of San Antonio ("SA 7- ), which held the 7-Up franchise in the San Antonio area. (RX 2002; Lydick, Tr. 2996-97. ) Texas Bottlers, Inc. held the 7-Up franchise from January 1986 until May 1987 . when Grant-Lydick purchased the assets of Texas Bottlers, Inc. , for 800 000. (Bodnar, Tr. 1334.

Initial Decision 118 FTC. 21. RC Cola Company is a subsidiary of DWG, Inc., a conglomerate. (Coyne, Tr. 3495-96; RX 2836-39; RX 2841 , p. 3. ) RC Cola Company owns the trademark and produces concentrate for RC Cola products. (RX 2841 , pp. 9- 10.

22. Texas Beverage Packers ("Texas Beverage ) is a familyowned bottling company with headquarters in San Antonio. Texas Beverage contract packs soft drinks and sells its own "Texas" brand private label soft drinks to retailers throughout Texas. (Hixon, Tr. 7269- , 7271- , 7332-43.) Steven Hixon is General Manager of Texas Beverage. (Hixon, Tr. 7270.

23. Shasta Beverages ("Shasta ), with headquarters in Hayward California, manufactures concentrate and carbonated soft drinks. (RX IOOJ- , B; RX 1532. ) Shasta operates bottling plants throughout the United States, including Houston, Texas. (Skinner Test. , RX 3011 , p. 3166.) Shasta makes Shasta soft drinks which it distributes nationwide. (RX 1532. ) Shasta also contract packs other soft drinks such as IBC Root Beer. (Knowles, Tr. 2689, 2810. 24. Kroger Company owns and operates a chain of grocery stores in various parts of the United States. (Morath, Tr. 7654-7655. Garland Beverage Company, a soft drink production plant owned by Kroger in Garland, Texas (near Dallas), produces Kroger s own "Big K" private label line of soft drinks for sale in Kroger stores. (Kaiser Tr. 3254. ) Garland Beverage Co. also contract packs for other brands. (RX 1726.

25. Kroger has a large regional warehouse and administrative office in Houston, Texas which supervises its operations in most of CCSW' s territory. (Kaiser, Tr. 3155-57. ) Kroger is several times larger than HEB, but has fewer stores than HEB in CCSW' s territory. (Summers, Tr. 6617, 6627- , 6767.

26. Winn-Dixie, a large grocery chain, operates a bottling plant in Ft Worth, Texas which produces "Chek" brand private label soft drinks for sale in Winn-Dixie stores. (Hixon, Tr. 7278-79. 27. Beverage Packers Inc. is a privately-held company which owns and operates a bottling plant in Fort Worth, Texas. (Hixon, Tr. 7274. ) Beverage Packers Inc. produces a number of soft drinks including its own line of warehouse brand soft drinks. (RX 1819. 28. Philip Espinoza was an employee and part owner of the Royal Crown Bottling Company of San Antonio. (Espinoza, Tr. 4163-65. Since retiring in 1986, he has worked for a series of companies (the Espinoza companies ) selling soft drinks in and around San Antonio THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 457 452 Initia! Decision and the Rio Grande Valley. The Espinoza companies include La Hacienda, Premier Distributing, Apollo Distributing, and Star Distributing. The Espinoza companies have distributed Nehi soft drinks and other brands, in the San Antonio area since 1986. (Limon, Tr. 4956-57; Espinoza Tr. 4166, 4169-87; Coyne, Tr. 3431.) 29. The Coca-Cola Company has headquarters in Atlanta, Ga. Coca-Cola USA ("CCUSA") is the division of The Coca-Cola Company that manages domestic soft drink operations. (Howell, Tr. 4004.) CCUSA produces the concentrates for Coca-Cola soft drinks. I (Atchison, Tr. 5237-38.

30. Coca-Cola Enterprises ("CCE"), a publicly-held company with headquarters in Atlanta, Georgia, owns Coca-Cola bottling operations in various pans of the United States, including Dal1asfFon Worth, Houston, and Austin, Texas. (Howell, Tr. 4002-07.) The Coca-Cola Company owns 49% of the stock ofCCE. (RX 3131- 31. From 1939 to July 1982 the Big Red Bottling Company of San Antonio was an independent bottler which owned and operated a bottling plant in San Antonio selling Big Red and other brands of soft drinks. In July 1982, DP-SA acquired Big Red Bottling Company of San Antonio for stock, and a non-compete agreement valued at $6,000 000. (RX 1648-E: CX 3315- 32. From December 1982 to November 1984, DP-SA held the Royal Crown Cola franchise. (RX 3065-A; Bodnar, Tr. 1251-52; Turner, Tr. 1037.) On November 9, 1984, Grant-Lydick became the Royal Crown franchisee. (RX 3105- 33. The Huntress family owned a bottling plant which held Pepsi- Cola franchise in San Antonio until 1982, when they sold the operation to Pepsi COBO. (Lauterjung, Tr. 4844. 34. Oneta Company ("Oneta ) owns and operates the Pepsi-Cola bottling plant and franchise in Corpus Christi and Victoria, Texas and surrounding areas. Karl Koch is President and Chairman of the Board. (Koch, Tr. 1801.) 35. Better Beverages, Inc., a closely-held corporation with headquarters in Hallettsville, Texas, owns and operates Dr Pepper, Pepsi Seven-Up, A&W, Canada Dry, Country Time, Nesbitt s and Hawaiian Punch franchises in southeast Texas between San Antonio and Houston. (Antle, Tr. 3047-48; Campbell, Tr. 1922-23. ) Dale 1 Concentrate companies arc "syrup companies" or "parent companies, " (Knowles. Tr. 2699- 2700.

Initial Decision 118 FTC Campbell, his mother and his two brothers own Better Beverages. (Campbell, Tr. 1935-36.

36. The Dr Pepper Bottling Company of Texas ("Turner DP" owns and operates the former DPUSA company-owned bottling operations in Dallas/Ft. Worth, Waco, and Houston, Texas, with plants in Houston and Irving (near Dallas). (Turner, Tr. 915. ) Jim Turner is President and CEO of Turner DP, and owns a minority interest in the company. Turner DP holds franchises for DPUSA, RC Cola, 7-Up USA, Big Red, Canada Dry, A&W, Original New York Seltzer, Sunkist, and other flavor companies in various parts of its sales terrtory. (Turner, Tr. 926-28.

37. AbTex holds Pepsi-Cola and Dr Pepper franchises for West and Southwest Texas and operates a bottling operation in Abilene Texas. (Cole, RX 3008, pp. 90-91.) 38. H. E. Butt Grocery Company ("HEB") is a privately-owned regional grocery chain with headquarters in San Antonio, Texas. (Gonzala, Tr. 2024: Summers, Tr. 6767, 6589-93.) HEB is the largest volume grocery chain in CCSW' s territory. There are 153 regular HEB stores in Texas, with 86 located in CCSW franchise territory. There are 23 smaller "Pantry Stores" operated by HEB in areas outside the CCSW franchise territory. Robert Chapman is Vice President of procurement at HEB and Tim Brinkley is Manager of Information Services. (Summers, Tr. 6593. 39. Albertson is a national grocery chain which operates retail stores in parts of Texas. Albertson is several times the size of HEB although it has fewer stores in CCSW' s area. (Summers, Tr. 6767. 40. Other supermarket chains which operate stores in Texas include Handy Andy and Super S. (Howell, Tr. 4058: Sendelbach Tr. 7686- 89.) Convenience store chains which operate stores in Texas include: National Convenience Stores, which operates the Stop- Go stores, the largest volume convenicnce stores in South Texas (Summers, Tr. 6630-6631; Howell, Tr. 4063; Davis, Tr. 4604- 05), with 195 stores in San Antonio (Hiller, Tr. 5531-32); Circle K (Summers, Tr. 6631); and Maverick Markets. (E. Hoffman, Tr. 575. 41. Concentrate companies and "fountain wholesalers" sell postmix fountain syrup in this market including: CCUSA (RX 861); DPUSA; Martin-Brower, which supplies McDonald' s restaurants (Summer, Tr. 6515, 7060: Knowles, Tr. 2813- 17); Burger King Distribution Systems, formerly Distron, which supplies Burger King THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 459 452 Initial Decision restaurants; Sysco; Sugar Foods; White Swan; and McLane. (RX 861; Summers, Tr. 6503; Short, Tr. 7740-45. 42. Full-line vending companies operating in CCSW' s territory include: Servomation (Little, Tr. 657) and ARA (Summers Tr. 6655), L. C. Vending (a family-owned business, headed by Ladd Little) (Little, Tr. 632-33) and A&W Leasing. (Summers, Tr. 6655. This Proceeding 43. The original complaint was filed on July 29, 1988, naming CCSW and DPUSA as respondents. The complaint asked that CCSW be required to divest the Dr Pepper and Canada Dry licenses and assets acquired from DP-SA in 1984. On August 4 1989, complaint counsel and DPUSA entered into a settlement agreement and DPUSA was dismissed from the case. On November 18, 1988, an amended complaint was filed.

44. Trial in this matter commenced on July 10, 1990 and concluded on October 3 , 1990.

History of Challenged Acquisition 45. In 1984 Forstmann-Little began selling the Canada Dry business (Turner, Tr. 920-21) and DPUSA' s company-owned bottling plants. (CX 3817. ) Jim Turner (DPUSA President of companyowned Bottling Operations) and Don Antle (DPUSA Vice President Franchise Department) were appointed to handle the sale of the plants. (Turner, Tr. 1411- 12.

46. Bud Grant, a geologist and oilman, and Lee Lydick, owner of Triple XXX Root Beer, wanted to buy DP-SA but their offer of $ 1 6- 17 million was refused by DPUSA. (Lydick, Tr. 3023. ) They made a later offer, but were unable to obtain financing for the purchase. (Turner, Tr. J 097- , 1150) 47. CCSW wanted the franchises for Dr Pepper and Canada Dry. CCSW had no need for DP-SA' s main production facility, the fonner Big Red Bottling Company of San Antonio plant. CCSW indicated its interest but DPUSA wanted to sell the operation as a whole and initially rejected CCSW' s response. (Antle, Tr. 3059. 48. In 1984, DPUSA preferred granting Dr Pepper franchises to independent bottling companies not owned by competing concentrate companies. The Pepsi bottler in San Antonio was wholly-owned by Initial Decision lis FTC Pepsi USA. Further, Pepsi USA officials told DPUSA that the amount requested by DPUSA for the DP-SA bottling operation was too high. (Antle, Tr. 3059-60; Turner, Tr. 1095. 49. DPUSA sold the operation in two parts. (Turner, Tr. 1152. CCSW bid on the Dr Pepper and Canada Dry franchises. CCSW initially offered 55 million, later increased to $14.5 million. (CX 3; RX 2092-F; Turner, Tr. 1158.

50. On August 28, 1984, CCSW purchased from DP-SA assets for $14.5 milion (RX 1292, p. 1; CX 1662; CX 253): a warehouse adjacent to the CCSW bottling plant (Bodnar, Tr. 1276; 1518-20); 2150 DP-identified used vending machines wjth an average age of five to six years (Little, Tr. 653); 40% of the delivery and over-theroad trucks owned by DP- , with an average age of seven to ten years (Bodnar, Tr. 1689; CX 254); and DP-SA' s rights in contracts relating to the Dr Pepper and Canada Dry franchises were reissued to CCSW. (CX 3, p. 7: CX 247-C; CX 270.

51. In the same transaction, DPUSA agreed to issue Dr Pepper liccnse agreements to CCSW. (CX 3 , pp. 17- 18.) DPUSA and Canada Dry issued new franchise agreements for the Dr Pepper and Canada Dry brands to CCSW in 1984. (CX 266-67. 52. CCSW and DPUSA also entered into a sales agency agreement requiring CCSW to act as DPUSA' s agent in the sale of Dr Pepper products produced in DPUSA company-owned plants to customers in CCSW' s Dr Pepper territory until a specified number of cases had been sold. (CX 3, p. 276: CX 275; CX 276; CX 1838- Schwerdtfeger, Tr. 2571- , 2622.

53. After the sale to CCSW, DP-SA still owned the DPbottling plant, the bottling equipment, non- Dr Pepper-identified vending machines, the remaining 60% of the vehicles, and the franchises for Big Red, RC, Crush, and Hires. (Bodnar, Tr. 1668; CX 237.

54. DP-SA continued to operate its business as Big Red Bottling Company of San Antonio, until DPUSA' s assets were sold to Grant- Lydick. (CX 2052; CX 2484; CX 3254-A; CX 237-C.) 55. In October 1984, Grant-Lydick acquired the remaining assets ofDP- , including the bottling plant (RX 1663), 60% of the tmcks and some vending machines for 56.5 million. (RX 2408: RX 2409; Lydick, Tr. 2981-82; RX 1648. ) Grant-Lydick put up $100,000. (Lydick, Tr. 2977 , 2984.) The remaining $6.4 million was lent by THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 461 452 Initial Decision General Electric Credit Corporation, which received a 44% share of the business. (Lydick, Tr. 2983- 84; RX 2410; RX 2411. 56. Grant-Lydick hired Emery Bodnar, the manager of DP- , to run the business. (Bodnar, Tr. 1223. ) Grant-Lydick also hired half of the former employees of DP-SA. (Bodnar, Tr. 1294. 57. Grant-Lydick obtained licenses to produce and sell Big Red , Crush, Hires, and DP-SA' s other remaining brands (CX 3495 CX 3504, CX 3505), about 58% of DP-SA' s 1983 sales volume. (Knowles, Tr. 2874. ) Grant-Lydick operates its soft drink business in San Antonio as Big Red Bottling Company of San Antonio. (Bodnar, Tr. 1581.) 58. On December 3, 1986, TBG acquired the Biedenham ownership in CCSW (R. Hoffman, Tr. 5588, CX 3052; RX 2805for $211 million, consisting of $145 million in cash and the assumption of $65.4 million in existing debt. (CX 29; CX 28; CX 3123. ) Prudcntial Insurance Company ("Prudential") providcd financing in exchange for 57% of the stock of TBG. Prudential provided $20 million in cash and $40 million as Scnior Debt and $80. 5 million as Subordinated Debt. Additional financing was provided by a revolving loan of $95 million from Texas Commerce Bank. (R. Hoffman, Tr. 5601; RX 2874-75; Admit.) 59. DPUSA and Canada Dry Corporation then issued new franchise agreements to CCSW. (R. Hoffman, Tr. 5618-20; CX 1391-A; CX 1938- l and Z- IO- 13; CX 3113; RX 2902. ) The new Canada Dry franchise was for 34 counties in South Texas. (CX 2852; CX 3065-B; RX 2932.

60. In April 1987, CCSW acquired thc assets of the American Bottling Company, a Dunnam family partnership, for $54 million. (CX 2805.) The American Bottling Company held the franchises for Coca-Cola, Dr Pepper and several other brands around Corpus Christi, Texas. CCSW closed the Corpus Christi production facility and supplied the Corpus Christi sales center from San Antonio and Cuero. (Summers, Tr. 6365; E. Hoffman, Tr. 230-31.) 61. In March 1989, CCSW acquired the remaining interest held by CCE in Crossroads Canning Company, a canning co-operative located in Cuero, Texas, for $3 million. (Summers, Tr. 6397-98. 62. CCSW acquired Coca-Cola Bottling Company, Cuero, Texas from the Summers family in 1985 (CX 3261; CX 22) and the Del Rio and Mason/Menard Coca-Cola bottling operations in 1986. (CX 28- 29.

Initial Decision 118 FTC 63. Grant-Lydick has acquired additional soft drink brands and new geographic territories. (Bodnar, Tr. 1334-36; RX 2970. ) In 1987 , Grant-Lydick acquired Texas Bottlers Inc. (the Sevennon producing bottler in San Antonio and Austin, Texas) for $7. million (Bodnar, Tr. 1334) and the Seven-Up bottler in Corpus Christi from the Nielsen family in August 1987 for $1.2 million. (Lydick, Tr. 2999-3000.

64. Grant-Lydick purchased the assets of Big Red Bottling Company of Austin in December 1988 for $1.3 million. (Lydick, Tr. 3002-03.

65. In April 1990 Grant-Lydick purchased Timberline Corporation, an RC Cola distributor in LaGrange, Texas, for $134 000. (Lydick, Tr. 3005-06.

66. Pepsi COBO in the early 1980' s acquired the Pepsi bottlers in Dallas, San Antonio, Houston, Austin, and Harlingen. (Davis, Tr. 4451-54; CX 3971) 67. In September 1984, the Texas Attorney General' s Office fied suit to challenge the transactions whereby CCSW acquired the Dr Pepper and Canada Dry brands, charging that the transactions violated Texas antitrust law. (CX 2- 68. On July 1 , 1986, CCSW, DPUSA, and the Texas Attorney General entered into a Settlement Agreement. (CX 2- ) CCSW was enjoined until July 1 , 1993, from the following: selling to its vending subsidiary on terms different from those offered to third party vendors; placing vending equipment on an "exclusive" basis; seeking or accepting more than 65% of the shelf space "regularly allocated for the sale of soft drinks" in any store; seeking or accepting exclusive end-of-aisle display space" for "more than 65% of the weeks in any given calendar year; or "seeking or consenting to participate in, on the average, more than 65% of' promotional ads during any calendar year.

69. CCSW was required to offer to sell the vending machines acquired from DP-SA "to the owner of the site at which such vending machine(s) was currently located" or to any of CCSW' s third party vending customers at book value. For any vending machine not sold CCSW is required to make available at no charge two slots in each vending machine for the sale of products of CCSW' s competitors. (CX 2-G, Sec IV; Summers, Tr. 6665.

70. Texas Attorney General is entitled to seek an extension of the order for a period of up to three years. (CX 2- , Section VII). THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 463 452 Initial Decision 7 J. CCSW sent a letter to vending companies offering to sell the vending machines which CCSW acquired from DP-SA at book value. None of the machines was purchased. (Little, Tr. 73 I 32. COMPETITON Soft Drinks 72. CCSW' s primary business is bottling, distributing, and selling carbonated soft drinks' at wholesale. (F 236- 39. 73. Soft drinks are sold in cans, glass, and plastic (PET) containers. The term "bottles" sometimes refers to soft drinks sold in any container ready to drink. Soft drinks are also sold in five galion tanks to fountain outlets ready to drink ("pre-mix ) or as syrup which must be mixed with carbonated water ("post-mix ). (Turner Tr. 1085-86; Knowles, Tr. 268 I -82.

74. Soft drinks are produced by combining "concentrate sweetener, and carbonated or still water. "Concentrate" includes the flavors, extracts, and essences used to produce soft drinks. "Syrup is concentrate mixed with sweetener and some water. (Turner, Tr. 1046.

75. In 1987, national sales of carbonated soft drinks totaled $38 billion. (CX 833-X; CX 784- 76. The 1988 per capita consumption of carbonated soft drinks was 45.9 gallons. Carbonated soft drinks lead all bcverages in per capita consumption, including water. (RX 990- 77. Texas is the "heartland" of both Coca- Cola and Dr Pepper. (Hoffman, E., Tr. 227-28; Turner, Tr. 982. ) Texas is very weak for Pepsi and represents 90% of Pepsi' s national share gap with Coca- Cola. (Amrosowicz, Tr. 889; Limon, Tr. 4977. 78. The national carbonated soft drink industry s main flavors are cola, lemon-lime, pepper, orange, and root beer. (CX 2956- C; CX 2527-D; RX 990-S, Z- 19. ) These five flavors are 95% of all soft drink sales. (CX 3956- C; RX 990 Z- 19; CX 3982-E.) 79. Colas are about 65% of carbonated soft drink sales. (Bodnar Tr. 1253, 1263; RX 990- , Z- 19-21.) The cola category is dominated by Coca-Cola and Pepsico. Royal Crown is a weak third. (CX 41- Soft" drinks contain no alcohol.

Initial Decision 118 FTC V; RX 990- ) Most consumers of soft drinks regularly drink colas and look for other flavors as a change of pace. (CX 858- C, E. 80. In 1984, the national market shares for the other soft drink flavors were (CX 864 at p. 14; RX 990- 19): lemon-lime, 12.7%; pepper, 6.9%; orange, 7.0%; and root beer, 4.9%. 81. In 1984, the national market sales by brand were (RX 990- 18): Coca-Cola, 21.6%; Pepsi, 17. 1 %; Diet Coke, 5.5%; 7-Up, 5.0%; and Dr Pepper, 5.4%.

82. Market shares of soft drink brands in San Antonio food stores in October, November 1989 were (RX 34-D): Coca-Cola (Classic and New Coke), 25.7%; Pepsi, 9.5%; Dr Pepper, 7.4%; Diet Coke, 7.3%; Big Red, 6.9%; Sprite, 5.2%; 7-Up, 2.5%; Royal Crown, 2. 1 %; and control brand (private label), 11.6%.

83. In 1984, national sales of non-diet soft drinks by channels included (RX 990-U; CX 3218- K): grocery chain, 50. 8%; fountain 14.0%; vending, 10.2%; small grocery store, 5.7%; convenience store, 4.7%: discount store, 1.4%; and drug store, 0. 8%. 84. In 1985 , the number of independent bottlers of soft drinks in the United States by brand were (RX 990- 29): Coca-Cola - 206; Pepsi-Cola - 167; 7-Up - 24; Dr Pepper - 10; Royal Crown - 45; and Canada Dry - 2.

85. San Antonio is Big Red' s largest market, and Grant-Lydick Beverage Company is the largest Big Red bottler. (Turner, Tr. 953. CCSW introduced Cima Red to compete against Big Red. (Hoffman , Tr. 346.

86. Carbonated soft drink package sizes include 6. , 10 , 12, 16 20 and 32 ounce glass or PET bottles, I , 2 and 3 liter PET bottles and 12 oz. cans. (CX 53- , Y- ) Private label carbonated soft drinks are sold in 12 ounce cans and 2 and 3 liter PET bottles. (CX 3158- ) H. B.'s Plaza is only in loose cans and 2 liter bottles. (Chapman, Tr. 7165; CX 4022.

87. The sales of soft drinks are seasonal. (CX 3816. ) The peak selling months are from May to September. Soft drink sales are strong at the holidays: July 4, Memorial Day, and Labor Day. After a lull at Thanksgiving, sales increase during the Christmas/New Year holiday period. Sales are slowest in February. (Summers, Tr. 6609- 10.

THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 465 452 Initial Decision Fountain and 88. Concentrate firms, including CCUSA, Dr Pepper, Pepsico have exclusive geographic territories for their pre-mix fountain syrup. (Admit.) 89. Pepsico and RC Cola have exclusive geographic franchise territories for post-mix fountain syrup. (Knowles, Tr. 2681-82. CCUSA and Dr Pepper do not have exclusive franchise territories for post-mix franchise syrup.

90. CCUSA and DPUSA sell post-mix directly to some customers. (Howell, Tr. 4005; Turner, Tr. 1010- 11; Koch, Tr. 1804. ) Dr Pepper post-mix syrup manufactured by CCSW is sold by CCSW and resold by Pepsi COBO, and Grant-Lydick. (RX 2783; Summers Tr. 6509. ) Coca-Cola and Dr Pepper fountain products are available from many fountain wholesalers in the San Antonio area. (Short, Tr. 7741-42; RX 861; Turner, Tr. 1172-74; CX 33- 18. 91. Dr Pepper fountain is delivered directly to the customer, or to a bottler, commissary or food broker who services the customers. (RX 1919.) HEB , Kroger, Albertson, Skaggs and FUIT s are all national fountain accounts for DPUSA. (Knowles, Tr. 2831. 92. Larger fountain accounts qualify for "national account pricing" from both CCUSA and DPUSA. (Short, Tr. 7736; Cassagne Tr. 7585; Knowles, Tr. 2820-2823.

93. About 65-70% of CCSW' s sales of post-mix are made at the national account price. (Knowles, Tr. 2820; CX 4073. ) Coca-Cola fountain syrup is also distributed by food distributors McLane Sugar Foods, Frostex and Distron, the Burger King commissary (RX 3108; Summers, Tr. 6505-06, 6515- 16; CX 387- 103; CX 4039), and Martin-Brower, which supplies McDonald' s. (Short, Tr. 7759- 60; Turner, Tr. 1177).

94. Most ofCCUSA' s fountain business is through commissaries and distributors, with the rest through Coca-Cola bottlers like CCSW. (CX 387- 103; RX 636- 95. McDonald' s and other restaurant chains sell private label fountain products. The largest selling orange fountain soft drink is McDonald' s private brand. (Cassagne, Tr. 7759-60. Initial Decision 118 FTC Franchises 96. Franchises for bottled soft drinks are territorially exclusive. (CX 1666. ) The franchisor grants to the franchisee the exclusive right to make and sell soft drinks in bottles and cans bearing the franchisor s trademark and using the franchisor s formula, in a specified geographic territory. (RX 2848. 97. Concentrate companies historically required the bottler to own a facility to produce the product sold in the franchise territory. (RX 2848- , E (CCUSA); RX 2909-A (DPUSA); RX 2932- (Canada Dry); RX 2930-B (A&W).) Some concentrate companies now waive the production requirement and allow a bottler to become a "non-producing bottler" who may acquire product from elsewhere. (RX 602; RX 2925; RX 912- 98. Coca-Cola (RX 2848-E) and Dr Pepper (RX 2908franchises are perpetual Franchises for allied products of The Coca- Cola Company are granted for ten-year renewable terms. Both types can be tcrminated for cause. (Admit.) 99. CCSW has a Jicense to market Hi-C products to schools; all other marketing for Hi-C is conducted by Coca-Cola Foods division of The Coca-Cola Company. (Admit.) 100. CCSW sells New York Seltzer under a distributorship agrecment providing for termination on thirty days notice. (Admit.) 101. In many franchise agreements (but not including certain franchises issued by The Coca-Cola Company), a transfer of the franchise, including a change of ownership of the corporation which holds the franchise, constitutes a breach of the franchise agreement unless the franchisor has given prior written consent. The Coca-Cola s Bottling Contract Company First Line Bottling Contract and Bottler each restricts direct franchise transfers, but both are silent as to changes in control of corporate franchisees. (E. Hoffman, Tr. 220; R. Hoffman, Tr. 5618-20.

102. CCSW and SWCC are licensed under the First Line Bottling Contract for Coca-Cola (RX 2848) as amended by adding geographic territory. (RX 2849; RX 2851; RX 2852; RX 2856; Summers, Tr. 6734-38.

franchise to be transferred 103. DPUSA does not allow any without its consent. The sale of a bottling operation allows DPUSA to choose a different franchisee. (Knowles, Tr. 2802-03, 2877. , p.

THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 467 452 Initial Decision 104. Concentrate companies use "transfer restrictions" to control bottler performance. (Knowles, Tr. 2802; Treibelcock, Tr. 5839. They may refuse to grant a new license to the prospective purchaser. (E. Hoffman, Tr. 491-92.) Or they may revoke the existing license if the bottler is sold (or even refinanced) without their prior approval. (Knowles, Tr. 2872; RX 1390.

105. Bottling franchises prevent the bottler from selling more than one brand in a "flavor segment." (CX 1668; RX 2938-C.) These provisions are known as "imitative products provisions." (CX 1912. 106. Concentrate companies may waive imitative products provisions, allowing the bottler to sell more than one brand of a soft drink flavor. CCSW sells two orange flavors, Minute Maid and Sunkist. (RX 2936-A; RX 2937; RX 2136.) CCSW also sells two seltzers, Canada Dry and Original New York Seltzer. (RX 2877; Summers Tr. 6751; CX 3182.

107. Franchise agreements establish the standards for bottlers performance, including sales volume, logos, and vending. (R. Hoffman, Tr. 5625-26; Summers, Tr. 6747-49; RX 2933-34. 108. Canada Dry requited CCSW to agree to performance requirements to obtain the Canada Dry franchise following the change of control of CCSW, from the Biedenharns to TBG , in December 1986. (RX 2932-33.

109. CCUSA includes "right of first refusal" clauses in newlyissued franchises. (RX 914- ) By September 1988, 76.7% of Coca-Cola volume was subject to such restrictions. (RX 769. 110. The performance standard in CCSW' s Coca-Cola franchise requires that CCSW "vigorously push " and "use reasonable efforts to sell" Coca-Cola products. So does the DPUSA franchises. (RX 2848- , 0; CX 1861 (Coca-Cola franchise): RX 2850-D (1983 Amendment); Summers, Tr. 6486.

111. Concentrate companies enforce territorial-exclusivity of the bottling franchises by prohibiting a bottler from "transshipping, selling in another bottler s territory. (CX 1667; Davis, Tr. 4473-74; RX 2850-B; RX 2908-B; RX 2932-A.) J 12. Many bottlers are licensed by several concentrate companies to sell their brands of soft drinks. (Shanks Test., CX 3989 35.) CCSW sells Coca-Cola owned by CCUSA. Dr Pepper owned by DPUSA, Sunkist owned by Cadbury-Schweppes, and Original New York Seltzer owned by ONYS, among others. (RX 2931; E. Hoff- Initia! Decision 118FTC man, Tr. 507-09, 549; CX 2196- 37; CX 3716- 19. ) This practice is sometimes called "piggybacking." (Knowles, Tr. 2764-67. 1l3. Piggybacking facilitates entry of new brands. (E. Hoffman Tr. 507-09; Knowles, Tr. 2764- , 2770-74; R. Hoffman, Tr. 5627; CX 3646 (Quickick); CX 321; CX 3650 (ONYS Iced Coffee); CX 3782 (Promotion); CX 3726 (Topo Chico).

114. DPUSA built its business by franchising Coca-Cola and Pepsi-Cola bottlers, picking the most etfective distributor. (Knowles Tr. 2856, 2667-68: R. Hoffman, Tr. 5620-21; Turner, Tr. 1134- 1154-55; Clarke, Tr. 4374-76; Antle, Tr. 3078. 115. Dr Pepper uses mostly Coca-Cola bottlers (40-45% of Dr Pepper volume) and Pepsi bottlers (40% of Dr Pepper volume). (Knowles, Tr. 2765.) Only one or two bottlers remain who bottle just Dr Pepper products. (Knowles, Tr. 2769.

116. Other concentrate companies also have a similar policy of licensing the most effective bottler. (Coyne, Tr. 3597 (Re); CX 857 (Crush).

Production 117. A "case" of soft drinks includes: 24 twelve-ounce aluminum cans; 24 bottles of 6. 5-ounce, 10-ounce, 16-ounce or 20-ounce bottles; 6 two-liter PET bottles; 6 three-liter PET bottles; or 12 oneliter bottles. (Summers, Tr. 6491.) 118. Sixteen-ounce returnable is usually sold in 8-packs; sixteen ounce nonreturnable is usually sold in six-packs or singles. Twentyounce PET is always sold in singles, while 12 ounce cans may be packaged in six packs, 12 packs, 15 packs or 20 packs. Two and three-liter PET bottles are sold individually. (Summers, Tr. 6492. 119. Soft drinks are bottled and canned on automated production lines." (Cole Depo., RX 3008, p. 43.) A bottling plant usually includes a can line and one or more bottle lines. (Morath, Tr. 7662-64. 120. Equipment for a can line consists of a filler, a can seamer a pro portioner, high-side refrigeration equipment, a can warmer, a date coder, a can rinser, a tray former/case packer, a depalletizer, a Hi-Cone machine, a multi-pack machine, and a conveyor belt. (Summers, Tr. 6447-60.

121. A bottle line must also have a labeling machine. Returnable bottles also require bottle sorting capability and a bottle washer. (Summers, Tr,. 6373.

THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 469 452 Initial Decision 122. There are economies of scale in bottling and canning. (Turner, Tr. 1026-27.) Economies of scale are more significant in canning than in bottling. Most economies of scaje are achieved at a soft drink plant of three to five million cases per year of cans and two to four million cases per year of bottles. (CX 3218-P (Figure 16), Z- 14; Amrosowicz, Tr. 826; CX 570- 123. Small companies may achieve economies of scale by hiring others to produce the product ("contract" or "copacking (Campbell, Tr. 1926; Summers, Tr. 6465-66; Turner, Tr. 1119-22. 124. The contract packer spreads fixed overhead over a larger number of cases. (Turner, Tr. 1119-20. ) The customer does not have to invest in equipment, and can purchase the product for less than it would cost to produce it. (Turner, Tr. 1121.) 125. The 1983 Amendment to the Coca-Cola Bottler s Contract permits the Coca-Cola bottler to provide contract packing services even for another cola product. (Howell, Tr. 3998.) CCSW provides contract packing for other bottlers. (Cole, RX 3008 , pA5 , (1.5 million cases in 1986).

126. Bottlers who contract-pack in Texas include Turner DP (Turner, Tr. 929- , 1117- 18), Better Beverages (Campbell, Tr. 1925-26; Turner, Tr. 1120-21), the Pepsi COBO plants in Conroc and Dallas (Amrosowicz, Tr. 866), Temple Dr Pepper Bottling Company (Espinoza, Tr. 4193; Turner, Tr. 1120-21), Grant-Lydick (Bodnar, Tr. 1534- , 1656; RX 1607; RX 2015), AbTex (Turner, Tr. 1120), Garland Beverages (Morath, Tr. 7667, 7670; RX 2440; RX 171 J), Texas Beverage (Hixon, Tr. 7271), Beverage Packers, Inc. (Hixon Tr. 7274; Morath, Tr. 7670), the Shasta plant in Houston (Hixon, Tr. 7283; Morath, Tr. 7670; Skinner Test. , RX 3011 , pp. 3167-68), and the Winn-Dixie plant in Ft. Worth (Hixon, Tr. 727879). 127. Contract packers, price is slightly higher than the marginal cost of production. (Bodnar, Tr. 1657-68. 128. Some bottlers. including Grant-Lydick, have no can line and purchase all of their cans from contract packers. (Bodnar, Tr. 1256-57.

129. New brands have been introduced by contract packing, including Soho (Collier Test. , RX 3015 , pp. 4082-84), Original New York Seltzer (Miller Test. , RX 3013 , pp. 3441- , 3448), and Aga. (Limon, Tr. 4956.

130. Bottlers can join a cooperative canning or bottling plant. (Howell, Tr. 4011- 12; Turner, Tr. 1121-22; CX 3218-Q, R; Summers Initial Decision 118 F. Tr. 6405-06. ) Co-ops help bottlers lower their cost of goods and become more efficient. (Howell, Tr. 4012, Summers, Tr. 6405- CX 3218-Q, R.

131. Crossroads Canning Company was a production cooperative formed by Coca-Cola Bottling Company--Cuero, San Marcos Coca- Cola Bottling Company and Coca-Cola Bottling Company of McAllen. In 1989, CCSW acquired it. (Admit.) 132. CCSW and SWCC own Western Container, a cooperative which manufacturers PET bottles for its bottler members at facilities located in Houston and Big Spring, Texas. (Summers, Tr. 6404. Excess Capacity 133. There is excess capacity in bottling and canning in Texas. (RX 2939; Summers, Tr. 6465-66; Campbell, Tr. 1983-84; Morath Tr. 7662- , 7681-82 (Kroger); Turner, Tr. 1122-25; RX 2983. 134. During the busiest time of the year Grant-Lydick operates with 20-40% unused capacity. (Bodnar, Tr. 1651-53. 135. CCE has 23 milJon cases per year of unused capacity. (CX 167.

136. In Texas, Pepsi COBO has 42 million cases (65% of total capacity) of excess capacity for cans (CX 2380-J), 13.3 mmion cases (57%) of excess capacity for 2 liter bottles (CX 2380- K), and 7. million cases (53%) excess capacity for nonreturnable bottles. (CX 2380-J, L; Amrosowicz, Tr. 856-57, 892; RX 2986. 137. Better Beverages, Inc. has excess production capacity on the can line of six million cases annually, which could expand to ten milion cases with the addition of a second shift working six days. The capacity of the bottle line is one million cases, and 600 000 cases are produced annually. (Campbell, Tr. 1983-84. 138. The Turner DP production in Irving is 27 million cases with the capacity of 35 million, and in Houston production is 12 million cases with 20 million cases capacity. (Turner, Tr. 1122-25. Texas Beverage (CX 2710-E; Hixon, Tr. 7294) and Kroger (Morath, Tr. 7662-64) also have excess capacity.

139. In 1986 Procter and Gamble planned to manufacture its Hires/Crush lines through contract bottlers, based on "over capacity in the industry. " (CX 858- THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 471 452 Initial Decision Distribution 140. Soft drink bottlers distribute finished goods to retail outlets that sell soft drinks to consumers. For bottles and cans, the tasks include (Clarke, Tr. 4272-75): (a) warehousing (RX 329); (b) taking orders (Turner, Tr. 955); (c) delivering to the retailer s premises (Summers, Tr. 6468; E. Hoffman, Tr. 327); (d) placing on the shelves fronting," and pricing the product (E. Hoffman, Tr. 327-28; Howell Tr. 4032; Knowles, Tr. 2662); (e) removing old merchandise (E. Hoffman, Tr. 203, 327-28; Turner, Tr. 956-57); (f) ensuring "point of sale " signs are displayed (Summers, Tr. 6474; CX 2161- , E); and (g) changing space allocation. (Summers, Tr. 6960-61.) 141. Soft drinks are distributed to retail outlets by "direct-storedoor delivery" ("DSD") and warehouse delivery ("warehouse (Knowles, Tr. 2662-63.) In DSD the bottler s employees do (a) to (g). In warehouse, the bottler s employees do (a) to (c) and the retailer s employees do the rest. (Knowles, Tr. 2663-64.) Low quality merchandising can reduce sales volume. (Coyne, Tr. 3338- 39, 3341; E. Hoffman, Tr. 327- , 335-37. 142. In a DSD the driver drives to the store, carries the soft drinks inside, and merchandises the shelves. (Turner, Tr. 955-56. 143. "Bulk delivery" DSD is used with larger retailers. (Turner Tr. 1530-31.) Delivery is by a 45 foot tractor-trailer; unloading by a forklift. (Summers, Tr. 6414- 15.) A salesperson stocks the shelves. 144. Some bottlers telephone the customer to take the order for cold drink" the day before delivery is scheduled. This system is called "Tel-Sell." (Summers, Tr. 6640-41.) 145. CCSW (CX 2503- 5) and Pepsi COBO (Davis, Tr. 4471- 72), use all three types ofDSD. (Summers, Tr. 6414- 16. 146. Some bottlers rely on independent distributors. Half of Oneta s sales are handled by independent distributors. (Koch, Tr. 1901.) CCSW has used independent distributors to sell in the Rio Grande Valley. (E. Hoffman, Tr. 621.) DP-SA also used independent distributors. (Bodnar, Tr. 1235-36.

147. In addition to DSD and warehouse there are food brokers and beer distributors. Food brokers in Texas include Sweeny & Co. Gordon/Southtex, Fleming, Nelson Beverage, Bill Lyons, and Marketing Specialists. (CX 1999- W.) IBC Root Beer (Knowles, Tr. 2685), Canfield (RX 1823). Shasta (RX 1957), BPI (RX 2043; RX ) Initial Decision 118 FTC. 1827), Rocky Top (Morath, Tr. 7667), and Parade (RX 1829-B) have been sold by food brokers. (Knowles, Tr. 2809. 148. Beer distributors sell beer by DSD. They also sell soft drinks, including: Original New York Seltzer (CX 2725; RX 3013 pp. 3443 , 3449; Turner, Tr. 1016), Hawaiian Punch (Anderson, Tr. 3886-87), Jolt Cola (RX 1810), Soho (Collier Test. , RX 3015), DPUSA (Bodnar, Tr. 1235-36), RC Cola (Coyne, Tr. 3436-37), and Crush/Hires (CX 2609.

149. IBC Root Beer, a premium priced soft drink produced by DPUSA through contract packers, is distributed in brown nonreturnable bottles to the home market by food brokers. (Hiller, Tr. 5340; Kaiser, Tr. 3158.) It is better suited to warehouse delivery because it is a premium priced product in a long-necked glass bottle that does not permit high-speed manufacturing or high volume delivery. (Knowles, Tr. 2664-65. ) Crush and Hires have been delivered by DSD and warehouse delivery. (Turner, Tr. 954-55. 150. The "home" market includes soft drinks consumed at home. Cold drink" is immediately consumed. Cold drink includes vcnding and fountain sales, and sales from cold vaults in convenicnce stores. The home market is 83.5% of bottle and can sales, and cold drink is 14. 5%. (CX 883- 151. The A.C. Nielsen Company ("Nielsen ) tracks sales in the home market. (RX 875. Nielsen Audits" show total sales and market share by brand and package for bimonthly periods. (CX 109-A.) 152. The Nielsen Audit for San Antonio includcs Bexar County. (CX 3557-F.) 153. Nielsen collects "scanning" data from stores with electronic scanners at the checkout counters. In Texas, Scantrack data is available for Austin/San Antonio. (CX 752; CX 1165; CX 753; RX 780; Bodnar, Tr. 1573-74.

154. CCSW soft drink sales are 66% bottling and 34% fountain. Pre-mix is 15- 18% of fountain sales, three to five percent of CCSW' sales. (RX 405-E; Summers, Tr. 6497.

155. CCSW delivers Coca-Cola and Dr Pepper fountain syrup to national accounts for a fixed delivery fee per gallon; CCSW also sells Coca-Cola and Dr Pepper fountain syrup to smaller accounts on Bottles and cans sold to convenience stores may be sold to the consumer hot" or "cold. " Some of the products sold at wholesLlle in the home market arc purchased by third-party vending companies and placed in vending machines. (R, Hoffman. Tr 5520. THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 473 452 Initial Decision tenns negotiated between CCSW and the local account. (E. Hoffman Tr. 449-50, 548.

156. Convenience stores most often buy fountain soft drinks through their wholesale grocery supplier. (Summers, Tr. 6525. 157. Concentrate for Coca-Cola fountain syrup is supplied to CCSW and SWCC from the Coca-Cola syrup plant in Dallas, Texas. Dr Pepper fountain syrup and concentrate are supplied from the Dr Pepper syrup facility in St. Louis, Missouri. (E. Hoffman, Tr. 546- 47.) CCSW manufacturers Dr Pepper and Coca-Cola fountain syrup from concentrate. (Summers, Tr. 6508-09.

158. Vending companies in the San Antonio area include: CCSW' s vending division, Snappy Snack, ARA, MaITiott, Canteen Service America, Drappala, D&J, Tom s Peanuts, A&W Leasing and L.c. Vending. (Summers, Tr. 6655.

159. Vending customers ofCCSW also purchase soft drinks for their vending machines at Sam s Wholesale Club or other wholesale outlets, or at supermarkets when prices are discounted. (R. Hoffman Tr. 5713 , 5520; Jackson, Tr. 3375.

160. In 1988, CCSW' s vending sales were 12.6% of total sales. (Snappy Snack 2%, other vending firms 3.4%, and 7% through its own machines, CX 3418-F; Summers, Tr. 6668-73. Prices 161. Few soft drink wholesale sales are made at list price. The price is reduced by a discount or allowance. (RX 327. ) In 1990 at least 90% of CCSW' s sales were made at less than list price. (R. Hoffman, Tr. 5555, 5645.) only 2% of Pepsi COBO sales are at full list price. (Davis, Tr. 4684- 85.

162. Bottlers change promotional offers often. (Campbell, Tr. 1954; R. Hoffman, Tr. 5551-52; Summers, Tr. 6613 (monthly). ) In January 1986 CCSW issued 199 different promotional offers. (CX 2179. ) Wholesale prices vary by brand, package and geographic area. (CX 1979; CX 2180; Turner, Tr. 1474; Bodnar, Tr. 1648-49: Davis, Tr. 4702-03; Kaiser, Tr. 3224.

163. Promotional allowances reduce the price to the retailer and facilitate lower prices to the consumer. (Turner, Tr. 960.) When soft drinks are on sale, consumers consume faster and purchase more soft drinks. (Knowles, Tr. 2838-40. ) Soft drink promotions encourage volume purchases. (Coyne, Tr. 3474.

Iniri,tlDecision 118 F. 164. Promotional allowances involve a feature ad, an instore display, or a reduced retail price. (CX 1039- , C; CX 1041-H; CX 2373- , I.) 165. Soft drink bottling is a "volume-oriented" business. (Knowles, Tr. 2838-39; Bodnar, Tr. ) 271; Turner, Tr. 1395; CX 836- ) Bottlers seek additional volume to spread overhead over additional sales. (CX 3407-C; Knowles, Tr. 2846, 2899. ) Concentrate companies require volume increases from bottlers to increase the concentrate companies' sales of concentrate. (Howell, Tr. 4072-73; R. Hoffman, Tr. 5625-26.) The most effective means of increasing Bales unit volume is to reduce price. (Knowles, Tr. 2838- , 2845; Howell, Tr. 4020; Coyne, Tr. 3563-64.

Promotions 166. CCSW' s Coca-Cola franchise provides that Coca-Cola USA pays 100% of the national advertising for Coca-Cola Classic and 50% of the national advertising for all other brands, sharing all local media costs equally. (Howell, Tr. 3930-31; E. Hoffman, Tr. 406-07. DPUSA and Seven-Up Company also fund national and local media advertising and other promotions. (E. Hoffman, Tr. 40607. 167. In retail stores, soft drinks are in a beverage aisle of the store. Retailers also display soft drinks at the end of the aisle. (Summers, Tr. 6602.) Soft drinks are usually purchased on impulse. (CX 2008- , Q.

168. Retailers award display space to suppliers who offer the most attractive promotional deals. (Summers, Tr. 6602-03.) Bottlers offer discount pricing to retailers for displays and lower consumer prices. (Coyne, Tr. 3486, 3488; Summers, Tr. 6613 , 6621-22. 169. Retailers include soft drinks in their weekly newspaper advertising. (Turner, Tr. 1130-31.

170. )n order to obtain a feature ad, a bottler must offer greater discounts than those required to obtain an in-store display. (Gonzaba Tr. 2057; Davis, Tr. 4616.

171. Sales volume for products promoted in a feature ad may increase 500 or 600%. When products are promoted on display without a feature ad, sales may increase 250%. (Coyne, Tr. 345152. 172. Recently, the cost of ad payments has increased. (CX 203; CX 205-06; CX 212- 13; CX 3020: CX 1620 ($3.7 million to HEB); THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 475 452 Initial Decision CX 2464-N (DPUSA and CCE); Bodnar, Tr. 1481; CX 4018-G; Cole Depo. , CX 3843 , pp. 258-60.

173. In 1986 Pepsi COBO paid Kroger $275 000 for 22 feature ads in South Texas and in 1987 the payment increased to $1. 1 milion for 20 feature ads. (RX I 130- 174. A calendar marketing agreement CCMA") is an ad payment by a bottler to the retailer for displays, feature ads in supermarkets, or in-store advertising, such as window banners. (Kaiser, Tr. 3229- 31.) Bottlers 175. The number of bottling plants in the United States has been steadily declining since 1950. (CX 1671; CX 836-E; CX 3218- The number of bottlers decreased by almost 50% from 1980 to 1988. (CX 858- 176. CCUSA and Pepsico have acquired over half of the volume of their own bottling systems. (CX 858-E; RX 579. 177. Economies of scale led to production in larger, modern plants. (CX 3218- , N; RX 912-0; Bodnar, Tr. 1237-38; E. Hoffman Tr. 189-90, 277. ) Consolidation and the non-producer agreements allow production through more efficient bottlers. (Howell, Tr. 4007- 08, 4011- 12; Coyne, Tr. 3435.

178. The geographic consolidation of bottlers increased the efficiency of the bottlers, achicving economies of scale in distribution and administration. (Bodnar, Tr. 1232; Schwerdtfeger, Tr. 2290; Howell, Tr. 3935 , 4006; E. Hoffman, Tr. 190-92, 513; Lydick, Tr. 3008-09.

RELEV ANT PRODUCT MARKET 179. Complaint counsel contend that the relevant product market consists of "the manufacture, distribution, and sale of finished carbonated soft drinks (or syrups) produced from the concentrates of widely advertised branded, carbonated soft drinks, merchandised and distributed by direct-store-door delivery, in al1 channels of distribution" which includes: "branded soft drinks" carried by the Pepsi, Big Red, and Coca-Cola bottlers, and Mr. Espinoza s companies, including fountain soft drinks, mixers and club soda. (Hilke, Tr. 6153- 6176-77.) I find that relevant product market must be expanded to y, InitiaJDecision 118 FTC include: private and warehouse brand soft drinks, seltzers and other flavored waters, and non-carbonated soft drinks produced and sold by CCSW and competing bottlers.

Competing Brands 180. The Dr Pepper Company sells: Dr Pepper, Diet Dr Pepper Caffeine-Free Dr Pepper, Caffeine-Free Diet Dr Pepper, IBC Root Beer, IBC Cream Soda, Diet IBC Root Beer, Diet IBC Cream Soda Welch' s Grape, Welch' s Strawberry, Welch' s Orange, Welch' Pineapple, and Welch' s Punch. (Knowles, Tr. 2642) 181. The Canada Dry Company sells: Ginger Ale, diet Ginger Ale, Club Soda, Tonic, diet Tonic, Seltzer regular, Seltzer Lemon- Lime, and Collins Mixer. (RX 2932-34.

182. CCSW sells: Coca-Cola Classic, diet Coke, Caffeine free diet Coke, Caffeine-Free Coca-Cola Classic, Coca-Cola (New Coke), Caffeine-Free Coca Cola, Cherry Coke, diet Cherry Coke, TAB Sprite, diet Sprite, Minute Maid Orange, diet Minute Maid Orange Mello Yello, diet Mello Yello, Sun kist, diet Sunkist, Fresca, Mr. PIBB , A&W Root Beer, diet A&W Root Beer, A&W Creme Soda diet A&W Creme Soda, Welch' s Strawberry, Welch' s Grape, Lipton Tea, diet Lipton Tea, Delaware Punch, Dr Pepper, diet Dr Pepper Pepper Free, diet Pepper Free, Original New York Seltzer, Raspberry, diet Raspberry, Root Beer, diet Root Beer, Cream Soda, diet Cream Soda, Peach, dict Peach, Lemon Lime, diet Lemon Lime Cima Red, Canada Dry Ginger Ale, diet Ginger Ale, Club Soda, Tonic, diet Tonic, Tom Collins, diet Tom Collins, Spike Orange, Red punch and Lemon Lime, Hawaiian Punch (in Corpus Christi), and red cream, root beer, orange, strawberry, mixers and tonic Fanta in fountain. (Summers, Tr. 6581-82; Teague Depo. , RX 3007 , pp. 33- 34.) These brands are in cans (6-pack and 12-pack), l-liter, 2-liter and 3-liter PET bottles, la-ounce, l6-ounce and 20-ounce nonreturnable bottles, BIB and figals as post-mix and pre-mix fountain syrup, 6 ounce and 16-ounce returnable bottles. CCSW sells 145 different items. (Summers, Tr. 6582.

183. Pepsi COBO sells: Pepsi, Diet Pepsi, Pepsi Free, Caffeine Free Pepsi, Diet Caffeine Free Pepsi, Mountain Dew, Diet Mountain Dew, Orange Slice, Diet Orange Slice, Lemon-Lime Slice, Diet Lemon-Lime Slice, Wild Cherry Pepsi, Diet Wild Cherry Pepsi, and Apple Slice. (Davis, Tr. 4464, 4639.

THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 477 452 Initial Decision 184. Grant-Lydick sells: Big Red, 7-Up, Royal Crown, Crush Hires, Squirt, Diet Squirt, Country Time, Hawaiian Punch, Dr Pepper, Yoo Hoo, Upper 10, Schweppes, Canfields, and Diet Rite. (RX 1665; RX 1614- 15.

185. DP-SA sold: Dr Pepper, Frostie Root Beer, Country Time Lemonade, Hawaiian Punch, Salute Flavors, Canada Dry, Crush, Big Red, Royal Crown, Hires, and Barq s. (Turner, Tr. 1035-37; CX 3825; Bodnar, Tr. 1234.

186. Star Distributing, Mr. Espinoza s company, sells: Nehi flavors, Koala Springs Mineral Waters, and Mason Root Beer. (Espinoza, Tr. 4182-83.

187. Texas Beverage Packers produces: Canfield' s, Plaza flavors, and Texas Brand. (Hixon, Tr. 7275-83. 188. HEB sells Plaza brand in 2-Jiter PET bottles and cans in the same flavors as national brands, including colas. (Chapman, Tr. 7162-68.

189. Kroger produces and sells Big K brand in 2-liter PET bottles and cans. (RX 2444; RX 1685.

190. Shasta (RX 1957; RX 958- J) and Faygo (RX 1953; RX 958-J) sell flavors in 2-jiter PET bottles and cans. (CX 1084; RX 958 , pp. 810- 13; RX 1001; Skinner, RX 3011 , pp. 3161-62. 191 Yoo-Hoo, Artesia, and Ozarka are sold in the San Antonio area. (RX 3112; RX 2951.) 192. Independent soft drink warehouse brands include (CX 814- 8): Shasta, Faygo, Sunkist, Hires/CrushiSundrop, A&W, Dad' Bubble-Up, Welch' , Nesbitt s, No-Cal, Frostie, Nugrape, Sun Crest, Moxie, Mason, and Dr. Wells.

193. Royal Crown brands include (Coyne, Tr. 3828): Royal Crown, Nehi and Diet Rile.

194. National brand and private label' and other carbonated soft drinks are produced on the same equipment. (Summers, Tr. 6445-66; RX 2939.

4 "National brand" - brand of soft drinks distributed in most of the United Slates, generally by direct-store-door delivery 5 "Private label" (also private brand or control label) - brand of soft drinks owned by a grocery chain or other retailer.

Initial Decision 118 FT. 195. National brand and private/warehouse brands' are produced in the same plant. (Hixon, Tr. 7275-83.

196. Non-carbonated soft drinks (such as Lipton s Iced Tea, Hi- , Hawaiian Punch, and isotonic drinks like Spike) are bottled and canned on the same equipment and in the same containers used for carbonated soft drinks, except that nitrogen is used instead of carbon dioxide. (Summers, Tr. 6426-28.

197. The same tasks are required for distributing and merchandising private/warehouse brands and national brands and noncarbonated soft drinks. (Summers, Tr. 6469. 198. Consumers seldom are aware of what type of delivery method was used for soft drinks. (Kaiser, Tr. 3159; Gonzaba, Tr. 2125-26; Brinkley, Tr. 2249-50.

199. In retail stores, including HEB (Gonzaba, Tr. 2123-24; Chapman, Tr. 7156), Kroger (Morath, Tr. 7682; Kaiser, Tr. 3239), and Super S (Sendelbach, Tr. 7691- 92), private/warehouse, noncarbonated, and national brands are sold next to each other in the soft drink aisle. (Summers, Tr. 6595; Howell, Tr. 4024. 200. Private label soft drinks in stores in CCSW' s territory include: HEB ("Plaza ) (CX 4022). Kroger ("Big K"), Winn-Dixie ("Chek Cola ), Stop N' Go (" Parade ). (Hiller, Tr. 5337-38; Howell Tr. 4024-25; Kaiser, Tr. 3158, 3160; Turner, Tr. 1208; Bodnar, Tr. 1311.) 201. Grocery wholesalers and bottlers provide "warehouse brand" soft drinks to independent grocers. Examples include Shasta (RX 1531; RX 1957; Howell, Tr. 4031), Paygo (RX 1953; Summers, Tr. 6551), IBC Root Beer (CX 1294), Rainbow, Rocky Top, and Parade. (Hiller, Tr. 5337-38; R. Hoffman, Tr. 5534-35. 202. Some bottlers produce their own brand name products including the "Texas" brand of Texas Beverage Packers (Hixon, Tr. 7277-78) sold in Super S (Sendelbach, Tr. 7691). Rocky Top brand sold in Kroger (Morath, Tr. 7667 , 7668-69), and "BPI" brand of Beverage Packers, Inc. (RX 1819; CX 202; RX 2245. 203. Private label, non-carbonated soft drinks, and warehouse brands are delivered to the retailer s warehouse. The retailer delivers the product to the retail stores, stocking the shelves and displays, and 6 "Warehouse brand" - bfind of soft drinks distributed to retailers by delivery to their ware. houses. The brand may be owned by grocery wholesaler. contracl packer. bottler, or concentrate company.

THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 479 452 IllitiaJ Dccision merchandising the product. (Summers, Tr. 6468; Turner, Tr. 955; Hoffman, Tr. 327.

204. National brands are delivered by the "direct store door ("DSD") method of delivery, where the bottler or distributor delivers the product to the retailer s store and stocks and merchandises the product on the store s shelves and displays. Some brands like Shasta and Faygo are sold nationally but delivered by warehouse delivery. (RX 1001.) 205. The United States Department of Commerce s "Standard Industrial Classification" code for soft drinks, SIC No. 2086, includes private Jabel, non-carbonated and warehouse brands as well as national brands. (CX 4080, Hilke, Tr. 8540; CX 4160. 206. The National Soft Drink Association, the primary industry trade association, considers private label soft drinks, warehouse soft drinks, and non-carbonated soft drinks produced by soft drink bottlers (Lipton Tea, Delaware Punch, Hawaiian Punch) to be "soft drinks. (RX 3128; Strickland, Tr. 7956-57.

207. Companies which track the sales of pri vate label and warehouse brand soft drinks include Nielsen Audits (E. Hoffman, Tr. 7289-9 I: CX 27- V), Nielsen Scantracks (Summers, Tr. 6549-50; CX I 165- , W, Z-30-37), and Infoffation Resources, Inc. (CX 2392-A). 208. The Share of Intake Panel ("SIP"), prepared by NFO Research, tracks an beverages including private/warehouse brand soft drinks. (RX 2197, pp. 6707- 12; RX 2204.

209. Witnesses from the marketplace perceive private label warehouse brand, national brand and regional brand soft drinks to be generally competitive products. (Howell, Tr. 4028-29; Campbell, Tr. 1995; Knowles, Tr. 2806-07; Koch, Tr. 1875-76; Trebilcock, Tr. 5873- , Turner, Tr. 988.

210. Documents and testimony from soft drink bottlers and concentrate companies refer to competition from private label and warehouse brands. Concentrate firms includc Procter & Gamble (CX 774- , C; CX 858-A); CCCSA (CX 3436, RX 687- , M, RX 958- , CX 1084, CX 199 I - , CX 3436, pp. 870-71; CX 2230- , CX 169- , Howell, Tr. 4029, 4023-25); Pepsico (CX 4122-E): DPUSA (RX 1405-E); RC Cola (Coyne, Tr. 3602-03, RC Annual Report, RX 2837 , p. 10, RX 2838, RX 2841 , p. 10): 7-Up (RX 1990, p. 415); Schweppes (CX 287 J -B); Canada Dry (RX 2245); and Welch' s (RX 1937 , pp. J, L-M).

Initia! Decision 118 FTC 211. The bottlers include CCSW (RX 2060 at C- 11965, RX 226- , K, RX 480- , CX 3158- , CX 3784, CX 2974- , RX 398); Pepsi COBO (RX 2503- , D, RX 1259- , RX 1287- , CX 4122); and CCE (RX 1479-J).

212. The retailers include: HEB (Gonzaba, Tr. 2122-23); Super S (Sendelbach, Tr. 7691); Stop- Go. (RX 1506. 213 In 1984, a CCSW market report stated that (RX 2059 p. 11757): "We continue to watch price brands such as Shasta and private label store brands increase their space, share of market and even ad take.

214. Fanta, the Coca-Cola flavor line, competes directly with private label soft drinks. It is delivered direct-store-door. (RX 687; CX 8134- X; RX 958- Prices 215. When Jim Turner, the Dr Pepper bottler in Houston, sets his prices on the pepper and lemon-lime soft drinks he looks at branded competitors, Coca-Cola and Pepsi. But he watches the prices for private labels because they could affect his sales of Sunkist Nugrape, Squirt, Big Red and A&W. (Turner, Tr. 988. 216. Robert Chapman, of H. E. Butt, explained the price gap between private and national brands (Tr. 7190): Q. Does H B try to maintain Plaza as the cheapest brand? A. Yes, we do.

Q. Can you tell us why"

A. Yes. To be competitive with other private Jabels from other companies, other private label brands such as companies like Kroger or somebody else might have.

Also, we pay less for it, and the consumer can only buy it at H B. If the consumer is reaIJy a Plaza Jiker, then the consumer can only get it at our stores. So we want to keep them coming back there and keep them happy, so we try and price it below the other brands.

Q. Does H- B make any effort to try to maintain at least an everyday margin between national brands, DSD brands and its private label? A. We have set our markups based on cost, generally, and because the costs are different, there is a spread. We don t say, well, we are going to be 15 cents a six-pack or whatever difference, but we base it off of costs and the costs natmaIly do that.

Q. 1f DSD prices decreased, what impact would that have on your private labels, or would it necessarily have an impact? A. I believe the sales would decrease on private labels. THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 481 452 Initial Decision 217. In February 1989, Texas Bottling Group in San Antonio raised wholesale price six percent resulting in a three to four percent net price increase after discounts. Big Red matched the price increase in mid February. Pepsi matched the increase on March 1. The Nielsen Ratings for the February/March period indicated that private label market share increased up to 20%. (CX 3806- 56. 218. An RC bottler from Iowa testified that he priced off national brands but watched the price gap (20% in his market) between national brands and private labels. (Trebilcock, Tr. 5873. 219. Texas Beverage, a contract packer, has given up major holidays to national brands because their prices are so low. (Hixon Tr. 7303.

220. Shasta seeks a mid-point position between the prices of private label and national brands. (RX 3011 , p. 3197. 221. In 1983, a Coca-Cola offcial estimated that "private/control labels peg their net prices to those of the national brands (an average of 29% lower)." He estimated warehouse brands, like Shasta and Faygo, at 20% lower in price. (CX 814.

222. David Davis, Vice President of Pepsi USA, testified about the affect in San Antonio of price competition between national and private brands (Tr. 4528-29):

Q. With regard to San Antonio, did private labels come back or increase in their market share? A. Yes, they did.

Q. Was that a result of the branded price increase? A. It' s my opinion it is, yes.

Q. How A. We felt Jike when you re getting national brands down so low -- 99 cents you re taking market share out of private label then. When the prices are higher, then you still have the price shopper that s going to pick up the private label. Therefore, you re losing share back to the private label Q. Web, since 1988 have you seen any interaction between private labels and your Pepsi brands? A. You mean in the same ad? Q. No. With regard to either losing market share or losing volume. A. Yeah. We took a volume hit when prices came up. Share -- We saw private label pick up some share also.

Q. How significant? A. I don t recall. It just seems like it was out of both of us. Initial Decision 118 FTC Q. "Both of us" meaning A. Pepsi and Coke.

223. In 1989 Pepsi Cola report on Nielsen performance stated that (RX 2503-A):

Private label was the key beneficiary of 1988 Corp. Pepsi (- 9) and Corp. Coke (0.7) share losses in Pepsi-Cola South with a 1.3% share growth v. 1987. The Pepsi report stated that in San Antonio, private label increased market share by 2.4% in 1988 and Coca-Cola lost 2. 8% while Pepsi stayed the same. (RX 2503- 224. When setting the retail price for Coca-Cola, the HE Butt grocery chain does not consider private label or Pepsi prices, but uses cost-based pricing. (Gonzaba, Tr. 2106-07. 225. Toby Summers testified about the market share changes caused by price competition between national brands and private labels (Summers, Tr. 6556 6726-27):

Q. What is your opinion as to why control brands fell that particular bimonthly? A. It' s influenced by the ad feature activity. The summer ad feature activity. the summer of ' 89 was heavily inf1ucnced by national soft drinks and, therefore I think what you saw would be n What you should see is that the national soft drinks, when they go on ad, spike down or get down, whatever you want to call , and suck up and siphon off private label volume. And the inverse happens when the private Jabcls are on ad. They spike up into the national brand share and siphon off share. So you see a trade-out that s heavily influenced by the ad feature frequency. Q. Have you and I discussed that earlier in your testimony concerning the situation in 1989 on the FM and AM Nielsens, bimonthly Nielsens? A. Yes, we did.

Q. Can you tell us again what that relationship was? A. It was the same relationship. When private labels hit one of their two strongest months, which was FM at 18 share, I believe, Pepsi hit one of their lowest months.

The following month private labels were at 14-something, which was another strong month, and Pepsi continued to be somewhat depressed. Later on in the SUITJer months, Pepsi went up and the private label share went down to about seven -- or control brands went down to about seven, I believe. 226. In 1982, private label and control brand soft drinks had 7% of the San Antonio market. This is one of the lowest such market shares in the country. The United States average was about 10. , and in some markets the share is over 20%. (CX 1084- THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 483 452 Initial Decision 227. "Control brand" in Nielsen means private label controlled by retailer (excluding Shasta and Faygo). (Summer, Tr. 6551.) Controlled brands bimonthly share of food stores in San Antonio was (RX 2806-X pp. 17):

1988 1989 February 18. April 14. June August 11.% October 10. 132% December The 18% shares in February 1988 and February 1989 coincided with Plaza ads by HEB. The drop in share in December 1988 and June 1989 coincided with ad feature activity by the national brands. (Summers, Tr. 6553-57.

228. Nonnal1y, private/warehouse prices average between 20% and 30% below the prices of national DSD brands. (Adams Depo., CX 3814, p. 39: CX 814- 229. When national brands are promoted, the retail price of national brands drops near or below the price of private/warehouse brands. (Trebilcock, Tr. 5873-74; Bodnar, Tr. 1555-56: Summers, Tr. 6549. ) Retailers use reduced prices on national brand soft drinks to demonstrate to consumers that their prices in general to consumers are low. (Howell, Tr. 3951-52. ) Retail price reduction of national brand soft drinks reduces sales of private label brands. (CX 3031; RX 538- 99 ('The primary victims of lower DSD prices were the warehouse and private label brands, which experienced marked share loss and volume decline ); Hixon, Tr. 7303, 7360; Lydick, Tr. 2973; Chapman, Tr. 7190; Turner, Tr. 988; Campbell, Tr. 1999; Skinner Test. , RX 3011 , pp. 3171-78, 3197-98.

230. When the price difference between private/warehouse brands and national brands increases (Davis, Tr. 4528-29), or when retailers promote their private brands heavily, the market share of private label increases. (Kaiser, Tr. 3252; Bodnar, Tr. 1359-60; Sendelbach, Tr. 7692-93; Hixon, Tr. 7303; Howell, Tr. 4118.) Private brands in San Antonio had l8.3% share in February/March 1990. (Summers, Tr. 6554; CX 3708; CX 3784- , D.

Initial Decision 118 FTC 231. Dr. Hilke, complaint counsels economist, ran "price sign tests" comparing the movement of prices of national brand to that of private label and warehouse soft drinks. (Hilke, Tr. 5948-56; CX 1678) Prices moved in the same direction eight out of ten times. (CX 1678- 232. Respondent s economist, Dr. Strickland, calculated the probability that private label and national brand soft drinks would randomly move in the same direction eight out of ten times was less than six percent. (RX 3088; Strickland, Tr. 7979. Consumers 233. The quality of merchandising for DSD and warehouse brands can vary. Some bottlers' employees do a good job of merchandising their DSD products; others do a poor job. (CX 2627 - Y to lO; Hixon, Tr. 7362.) HEB does a better job of merchandising its Plaza private brand than Pepsi does of merchandising its DSDdelivercd brands. (Summers, Tr. 6472.

234. One market research report perceived that the use of private brand soft drinks is "significantly higher" among Hispanic consumers than it is among other consumers. (CX 2662- 66.) About 55% of San Antonio s population is Mexican-American. (Bodnar, Tr. 1224. 235. The three liter PET bottle is a much better seller than the two liter PET bottle in San Antonio. The opposite is true for the rest of the state. (Kaiser, Tr. 3189, 3249.

Retailers 236. Private label soft drinks are more profitable for the retailer than national brands. (Sendelbach, Tr. 7692. 237. Private label soft drinks have more space relative to sales than national brand soli drinks (Kaiser, Tr. 3267-68; Smith Test., Rx 3005, p. 3721) because the retailer controls the allocation of space. (Davis, Tr. 4761-62; RX 256; CX 3270; CX 3384- Similar Products 238. Canada Dry mixers and seltzers are in the relevant product market. (Hilke, Tr. 6177. 343) Similar products like Original New THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 485 452 Initial Decision York Seltzer, Perrier and Artesia should also be included. (Strickland, Tr. 8005 , 8012- 13.

239. CCSW sells Canada Dry mixers and Ginger Ale in bottles cans and fountain syrup. Canada Dry mixers include Tom Collins mix, club soda, sparkling water, and diet versions of these products. (Summers, Tr. 6529-30.

240. Bottled carbonated water and Canada Dry products are usually in the beverage section of a supermarket. (Summers, Tr. 7860.

241. Flavored seltzer is a premium-priced drink which is clear in color, premium priced, in flavors such as lemon, raspberry, peach and root beer. (Summers, Tr. 6532; CX 2916 (CD Sparklers).) The seltzer segment has grown recently. (CX 2914-Q, R; CX 2390; Espinoza, Tr. 4196-97.) New products have been introduced by both existing concentrate companies and new entrants. (RX 2235. 242. CCSW has developed a new product called "Spike " which is an isotonic soft drink similar to Gatorade. (CX 308; CX 3685. Other isotonic products sold in CCSW territory are QuicKick Promotion, and 1O-K. (Summers, Tr. 6534; Antle, Tr. 3111- 12. 243. CCSW produces and packages non-carbonated soft drinks, including Lipton Iced Tea (RX 345), Delaware Punch, and Hawaiian Punch. (Summers, Tr. 6426-27. ) Grant-Lydick sells Country Time Lemonade, a non-carbonated soft drink, in 12-ounce cans in food stores and vending machines. (Bodnar, Tr. J 547. ) These "still" drinks must be packaged with nitrogen to provide pressure to strengthen aluminum cans. (Turner, Tr. 1405-06.) CCSW packages these products, using the same production equipment, in the same sizes and types of containers that it packages carbonated drinks. (Summers, Tr. 6427-28.) CCSW generally prices these still products at the same prices charged for carbonated soft drink brands in food stores and vending machines. (Summers, Tr. 6538. 244. Pepsi USA is test-marketing H,Dh' , a bottled water Mountain Dew Sport (an isotonic beverage), and Tea Breeze (a canned tea). (Davis, Tr. 4639-43; Christian Depo. , CX 3912, pp. 79- 83; CX 387; CX 1934; CX 2903- , G ("Schweppes ); CX 2916- The differences between carbonated soft drinks and non-carbonated drinks have blurred as products with characteristics of both have been introduced. (RX 2200; CX 2330-0; CX 2903- , G; CX 2916-Q; RX 2255; RX 2267; RX 2963; Koch, Tr. 1876.

Initial Decision lis F.T.c. RELEV ANT GEOGRAPHIC MARKET 245. The geographic area of the Dr Pepper franchise acquired by CCSW in 1984 consisted of seven counties in Texas (Atascosa Bandera, Bexar, Frio, Kendall, Medina, and Wilson) and portions of three other counties (Blanco, Comal, and Karnes). This region will be referred to as "the ten-county area." (Amended Complaint, p. 3 Section 9; Hilke, Tr. 5988. ) 1 find that the relevant geographic market exceeds the ten-county area.

.. () ! .., THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 487 452 Initi,!l Decision 246. Here is a map of the ten-county area compared to the Pepsi Cola franchise area (RX 2973-A):

a: w c: c:

:2 c:

c w w ::

.. Z c: c:

() a:

u. 0 :: 0 u. z W en c:

::;

Initial Decision 118FTC 246a. Here is a map with cities and distances (RX 2964): EVE I?.

THE COCA COLA BOTTLI;\G COMPANY OF THE SOUTHWEST 489 452 Initial Decision Shipments 247. Much of the soft drinks sold within the ten-county area are produced outside that area. Grant-Lydick has no canning line, and has purchased 98-99% of its canned soft drinks from the Turner DP canning plant in Irving (Turner, Tr. 1117; Bodnar, Tr. 1526-27) and the Better Beverage, Inc. plant in Hallettsville (Campbell, Tr. 1926 1987), both of which are outside the ten-county area. The Sevenand RC Cola products sold by Grant-Lydick and its predecessors in the San Antonio area since 1982 have been produced in the Houston bottling facility presently operated by Turner DP. (Turner, Tr. 929; Espinoza, Tr. 4248-49; Bodnar, Tr. 1557.

248. Until June 1990, Pepsi COBO imported all canned soft drinks sold within the ten-county area from its plant in Houston (Davis, Tr. 4461- , 4464, 4630-32) which is outside the ten-county area. Fifty percent of Pepsi COBO' s sales are in cans. (Davis, Tr. 4630-31.) 249. Pepsi COBO also obtained 22% of its bottled soft drink products from outside the ten-county area. (Davis, Tr. 4632. 250. In 1990 Pepsi COBO moved a can line from Conroe to its bottling plant in San Antonio, at a cost of from $1.0 to 1.3 millon. (Amrosowicz, Tr. 808, 822-23.

251. Kroger produces its Big K soft drinks for Texas in its plant near Dallas. (Knowles, Tr. 2837; Morath, Tr. 7665-66; Kaiser, Tr 3254-56. ) Shasta s plant in Houston, Texas, produces all of Shasta soft drinks for Texas. (Knowles, Tr. 2689. 252. Beverage Packers, Inc. supplies all of Texas, including San Antonio, from its Fort Worth plant. (Hixon, Tr. 7274; Morath, Tr. 7670.

253. Star Distributing purchases Nehi finished products from Temple Dr Pepper Bottling Company in Temple, Texas (outside the ten-county area) for distribution in San Antonio and the Rio Grande Valley. (Espinoza, Tr. 4193; Coyne, Tr. 3433. 254. CCSW produces 12-pack cans in its Cuero facility (outside the ten-county area) for distribution throughout its franchise territory. (Summers, Tr. 6403-04.

255. USA supplies Coca-Cola concentrate and much of the Coca- Cola fountain syrup sold in the ten-county area and throughout Texas from its syrup plant in Dallas. (Short, Tr. 7734-35; Howell, Tr. 3984.

( Initial Decision 118 FTC. 256. Much of the soft drinks produced within the ten-county area is shipped and sold outside that area. CCSW ships soft drinks from San Antonio and Cuero throughout its terrtory to the Corpus Christi Victoria, Temple, Uva1de and Del Rio warehouses. (Summers, Tr. 6410; E. Hoffman, Tr. 130 201.) CCSW produced soft drinks for Fredericksburg Coca-Cola Bottling Company. (Schwerdtfeger, Tr. 2463.) About 45% ofCCSW' s sales are outside the ten-county area. (Summers, Tr. 6423-25.

257. Texas Beverage supplies soft drinks throughout Texas from its San Antonio plant. (Hixon, Tr. 7272- , 7278. ) About 50% of Texas Beverage s production is sold outside San Antonio. (Hixon Tr. 7290.

258. Grant-Lydick has one bottling plant, located in San Antonio. (RX 2939- ) Grant-Lydick supplies its sales centers in Austin Corpus Christi, Victoria, Rio Grande and La Grange with bottled products produced in the San Antonio plant. (Bodnar, Tr. 1338-40. 259. Pepsi COBO' s San Antonio bottling plant packages soft drinks in two-liter and three-liter PET bottles. (Davis, Tr. 4461.) Three-liter Pepsi bottles for shipment throughout Texas are produced in San Antonio. (Davis, Tr. 4636; Amrosowicz, Tr. 827-28; CX 2360-A.) About 35% of the San Antonio three-liter bottle production is sold outside the ten-county area. (Amrosowicz, Tr. 827-28. 260. In 1983, 75% of the product produced by plants in San Antonio was sold in the ten-county area, and 78% of the product sold within the ten-county area was produced within that area. (Strickland Tr. 8040- , 8672; RX 3129.

261. In 1988, 57% of the product produced by plants in San Antonio was sold in the ten-county area and 77% of the product sold in the ten-county area was produced within that area. (Strickland, Tr. 8046-50; RX 3130.

262. Hilke performed Elzinga-Hogarty (" ) calculations One set involved an E- H calculation based on a 1990 extrapolation of 1988 production and sales estimates (CX 4089-E), adjustcd for the fact that Pepsi has moved a can line to San Antonio in June 1990. (Hilke, Tr. 8516, 8554.) Another set re-calculated Dr. Strickland' H figures, but excluded private label and warehouse brand sales and production. (CX 4089-A; CX 4089-C; Hilke, Tr. 8516, 8555. 7 The E-H test measures actual shipments of relrvant product into and outside of a proposed region. To qualify as a relevant geographic market. an area must satisfy a two- pronged test: 'Liule in From Outside" ("LIFO" ) and "Little Out from Inside LOFI" THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 491 452 Initial Decision 263. The 1983 calculation gave a LOFj percentage of 81 % and a LIFO percentage of 77% (CX 4089-Al B), thus failing the most recent (90%) version of the E-H test. (Hilke, Tr. 8551-52. 264. The 1988 calculation gave a LOFj percentage of 59% and a LIFO percentage of76% (CX 4089- , D), failing the "weak" (75%) LOFj test. (Hilke, Tr. 8553-54.

265. The 1990 calculations gave a LOFj percentage of 62% and a LIFO percentage of 85% (CX 4098- , F) thereby failing the weak" 75% LOFI test. (Hilke, Tr. 8554.

266. The freight cost to ship a truckload of soft drinks is between $0.75 and $1.0 per mile. (Hixon, Tr. 7286; Amrosowicz, Tr. 807 859-60; Summers, Tr. 6884, 6915.) Truckload capacity varies with the type of soft drink package; a truck can carry 2200 cases of cans. (Amrosowicz, Tr. 859-60.

267. Using a cost figure of $0.75 per mile, Toby Summers calculated that a 10% increase ($0. 59) in the wholesale price of canned soft drinks would enable canned soft drinks to be shipped a distance of 793 additional miles on a round-trip basis without back hauling. (Summers, Tr. 6437-38, 6885, 6915- 17.) Back haul would reduce the shipping cost. (Bodnar, Tr. 1528-29. 268. CCSW has sales centers in Del Rio, Uvalde, Kerrville Victoria, Corpus Christi, Temple and San Antonio. (E. Hoffman, Tr. , 201-03; Summers, Tr. 6407-08.) CCSW ships from its San Antonio plant up to 150 miles to supply its sales centers. Three of the sales centers are about 150 miles from San Antonio, two are about 100 miles away and one is 60 miles. (RX 353. 269. Turner DP purchases Original New York Seltzer from a contract producer in Des Moines, jowa, 900 miles away. (Turner, Tr. 1006; Trebilcock, Tr. 5811 , 5867 5869.

270. Pepsi COBO ships throughout Texas from its plants in Conroe, Houston, San Antonio and Mesquite. (RX 1238- , F; Amrosowicz, Tr. 847-48.) Pepsi COBO ships 260 miles from its Conroe can plant. (Amrosowicz, Tr. 847-49; CX 2380-C.) 271. Grant-Lydick purchases soft drinks in cans from Dallas and ships them to San Antonio (280 miles) and from there an additional 240 miles to HarJingen, Texas, for a total cost of 25 per case. (Bodnar, Tr. 1528-30.

272. Kroger supplies soft drinks to its warehouses in Louisiana Tennessee and throughout Texas from the Garland production facility. (Morath, Tr. 7665-66; Kaiser, Tr. 3254-58. 492 FEDERAL TRADE COMMISSION DECISIOC'S Initial Decision 118 FTC Territories 273. The ten-county geographic market consists of the Dr Pepper franchise area acquired by CCSW in September 1984. (F 274. ) It is smaller than the territory in which: CCSW operated before and after September 1984 (CX 1854-B); the geographic territory in which DP- SA operated before and after September 1984 (Bodnar, Tr. 1522-24): and the 35 county franchise area for the Canada Dry brands acquired by CCSW in September 1984. (RX 2972. ) The ten-county area does not include the eleven additional counties in the Dr Pepper franchise territory acquired by CCSW after September 1984. (Rx 352. 274. In 1984, the Dr Pepper franchise acquired by CCSW was for ten counties including San Antonio. (RX 2964. ) Later the franchise was expanded to 21 counties, through the acquisition of American Bottling Company of Corpus Christi. (RX 352; R. Hoffman, Tr. 5597-98; RX 6- ) In 1984, CCSW was franchised by Coca-Cola Company in 29 counties, including San Antonio. By 1989, the franchise had increased to 51 counties. (RX 2971; Strickland, Tr. 8085- 86. ) In 1985 , CCSW operated primarily in the ten-county area but about 30% of its sales were distributed outside of that area. (CX 418- ) In 1986, CCSW operated in 39 counties in Texas. (CX 1854- 275. CCSW' s current franchise territory includes San Antonio and 60 counties in southern, central and eastern Texas. (RX 352; RX 6; E. Hoffman, Tr. 201 , 496-98.

276. Grant-Lydick' s current franchise territory includes San Antonio and 60 counties in southern central and eastern Texas. (RX 3; RX 5 (RC Territory); Coyne, Tr. 3502-04; RX 2970. 277. Pepsi COBO' s franchise territory includes San Antonio and 105 contiguous counties in the eastern half of Texas. (RX 2973; RX 2; Howell, Tr. 4013- 14; Davis, Tr. 4451-54. ) Pepsi COBO also has other counties in West Texas and in the Rio Grande Valley. (RX 2; F 246.

278. There are no territorial restrictions in the sale of CCUSA or DPUSA fountain syrup to retail accounts. (Howell, Tr. 4005; Cassagne, Tr. 7619-20.

279. There are no territorial restrictions in the sale of private label or warehouse soft drinks. (Hixon, Tr. 7277-78. 280. HEB currently operates 165 stores in South-Central Texas. (RX 4; Gonzaba, Tr. 2111- 13. ) Forty of these are within Bexar THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 493 452 Initial Decision County. (Chapman, Tr. 7144; Summers, 7843.) There are 86 HEB stores in CCSW' s franchise territory. (Summers, Tr. 6593: RX-4. HEB distributes grocery products (including its Plaza soft drinks) to all its stores from the warehouse located in San Antonio. (Chapman Tr. 7141; Gonzaba, Tr. 2114.

), is 281. Kroger s Houston "KMA" ("Kroger Marketing Area from Eastern Louisiana to West Texas, and includes San Antonio and the ten-county area. (CX 3966- 12; CX 2037-C; Kaiser, Tr. 3156. 282. Albertson s Texas Division marketing area includes 55 stores in North and South Texas and 12 stores in Louisiana. (Donald, Tr. 5287.

283. Eckerd' s Houston District marketing area includes Houston Beaumont, Corpus Christi, San Antonio and Austin, Texas. (CX 1144.

284. The media advertising measure for television and radio is The A.C. Nielsen Company s "Area of Dominant Influence ADI"). (Strickland, Tr. 8075.) The San Antonio AD! is 15 counties larger than the ten-county area. (RX 2967. 285. The advertising areas for the two major San Antonio papers the San Antonio Light and the San Antonio Express (Strickland, Tr. 8696-97), includes about 30 counties.

286. Arbitron sells warehouse shipment data for grocery items as a Selling Area Marketing, Inc. ("SAMI") report. (RX 1945; Strickland, Tr. 8077-78.) The SAMI region which includes San Antonio is about 50 counties. (RX 2696.

Transshipping 287. Transshipping is the movement of franchised soft drink products from the territory of one bottler for resale in the territory of another bottler. The franchise agreements issued by CCUSA DPUSA and Pepsi USA prohibit transshipping by bottlers. (F 111. Retailers are not parties to bottling franchises. (Ncslage, Tr. 8727; E. Hoffman, Tr. 391; Howell, Tr. 3977.

288. Almost a million cases of Coca-Cola products were transshipped into an area north of Houston in 1982. (RX 3122. 289. A Pepsi USA log of transshipment complaints against the Conroe can plant shows 230 complaints within a 62-month period mostly made by Oneta Company in Corpus Christ. (CX 2327; Davis Tr. , 4748: Koch, Tr. 8629-32.

Initial Decision 118 FTC 290. Quality Liquor Wholesalers, a beverage distributor in Amarillo, Texas, dealing primarily in liquor and beer, transships cases of soft drinks into SWCC territory. (R. Hoffman, Tr. 568990. 291. SWCC received over $200 000 in 1986 and 1987 for lost sales due to transshipping charged to CCE and other bottlers because of the activities of Quality Liquors. (R. Hoffman, Tr. 5691; CX 3623; CX 3645- 46 (38 000 cases in 2 months); CX 3624 (153 000 cases in 10 months).) Quality Liquors continues to transship. (CX 3636-A (20 000 cases in 1988); R. Hoffman, Tr. 5688, 5691. 292. In September 1988 CCUSA fined CCE $177 165 for 35,433 cases of transshipped product found in SWCC' s franchise territory. (CX 2409-C.) 293. In 1989, CCE paid more than a million dollars of transshipping fines to CCUSA. (RX 3131-R; Neslage, Tr. 8729-30. 294. Resellers of soft drinks in CCSW' s market sell to others who sell at retail either in or outside of that market. Such resellers include: Sam s Wholesale Club (RX 3121; CX 2199-1), Quality Liquors, and vending companies. (Jackson, Tr. 3365 , 3375. COMPETITIVE HISTORY Effect of Acquisition 295. The 1984 acquisition did not reduce the number of competing finns or soft drink plants in the market. (Turner, Tr.1158-59. DP-SA continued in operation until it was sold to Grant-Lydick. (F 54.

296. In 1982 DP-SA acquired the former Big Red Bottling Company of San Antonio. (F 31.) In 1987 Grant-Lydick acquired the San Antonio 7-Up bottler. (F 63.

297. Since 1982, Pepsico acquired the Huntress bottling company in San Antonio (F 33), the Pepsi-Cola bottler in Houston and in Dallas (F 17), thereby integrating vertically Pepsi operations throughout much of the eastern half of Texas. In September 1986 CCUSA acquired the JTL bottling operations in Dallas, Houston, and Austin, thereby integrating vertically Coca-Cola operations in much of the eastern half of Texas except for CCSW' s territory. (CX 1512- THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 495 452 Initial Decision 298. As part of the 1984 acquisition, CCSW purchased 40% of DP-SA' s fleet of used delivery and over-the-road trucks. (F 50. 299. CCSW could have acquired the trucks from many other sources, including lease companies that seJl trucks at the end of the lease period. (Summers, Tr. 6771.

300. DP-SA sold CCSW the warehouse that was located on property adjoining the CCSW property. (Summers, Tr. 6661.) DP- SA had its bottling operation in the building until it acquired the Big Red Bottling plant in 1982. (F 56; RX 1580- ) At the time of CCSW' s purchase, DP-SA had publicly listed the warehouse for sale but it had not been purchased. (Bodnar, Tr. 1519. 301. DP-SA sold CCSW 2 150 used vending machines (F 50), many of which were located in accounts where CCSW already had vending machines. (Summers, Tr. 6773.) Soft drink companies offer programs to finance new vendors, and used vendors are readily available from brokers. (Summers, Tr. 6671- , 6772 , 6957-58; Turner, Tr. 1194-95; F 60.

302. The average age of the machines was five to six years at the date of the 1984 transactions. (Little, Tr. 653. ) The average useful life of a vending machine is ten years. (Turner, Tr. 1194; Lauterjung, Tr. 4901; Little, Tr. 691.) 303. Most of the machines were in place at customer locations. (Schwerdtfeger, Tr. 2452.) Many of the locations already had a Coca- Cola or Pepsi-Cola vending machine in addition to the Dr Pepper machine. (Summers, Tr. 6773.

304. Dr Pepper products could be added to CCSW vending machines without reducing availability of other products. In locations where another vending machine could not be installed, CCSW replaced the second or third button allocated to Coca-Cola, thereby increasing the variety of products available to consumers without reducing competition among products. (E. Hoffman, Tr. 418.

305. Many of the DP-SA vending machines were located at military bases around San Antonio, pursuant to vending contracts between DP-SA and the Army-Air Force Exchange Service AAFES"). (CX 255- , D, Z- , Z-38, Z- , Z-67. Initial Decision J 18 F.TC. 306. After the acquisition of the Dr Pepper brand, CCSW dropped Mr. PIBB, which had a market share of 2. 1 % in 1983. (Hoffman, E., Tr. 342, 421; CX 122; CX 1681-C.) Finance 307. After adjustment for inflation, the retail and wholesale prices of soft drinks in the San Antonio area declined over the 1984- 1990 period. (Davis, Tr. 4697-99; Bodnar, Tr. 1569; Coyne, Tr. 3500; Campbell, Tr. 1999-2000; Atchison, Tr. 5242. 308. The costs to produce finished soft drinks increased over the 1984- 1990 period. (CX 3258; CX 1026.

309. CSW' s costs increased and financial support from CCUSA has been cut. The cost of an HEB ad buy was 45% higher in 1990 than in 1989, increasing CCSW' s marketing costs by $1.4 million. (R. Hoffman, Tr. 5635-36.

3 I O. Increasing costs and declining prices of soft drinks decreased profits ofCCSW (Schwerdtfeger, Tr. 2592-93; CX 154l-C), and Pepsi COBO. (Davis, Tr. 4695-96.

3 I I. Although the financial performance of Better Beverages, Inc. is improving in 1990, margins decreased fifty percent from the early 1980' s to 1989. (Campbell, Tr. 2001- ) The Victoria area, where Better Beverages, Inc. competes with CCSW, is one of the lowest priced soft drink markets in Better Beverages ' territory. (Campbell, Tr. 1950- 2000.

312. L.c. Vending Company s sales have increased since 1985 but profits have not. (Jackson, Tr. 3356. 3 I 3. Texas has the lowest soft drink prices in the United States due to high per capita consumption, the strength of Dr Pepper brands and promotional efforts of Pepsico to buy market share. (Turner, Tr. 979; Campbell, Tr. 1950-5 I; Trebilcock, Tr. 5874-75. 314. CCSW' s and SWCC's net prices per case were lower than the national average for Pepsi bottlers during 1988. (Strickland, Tr. 8433- 8444; CX 53; RX 2990.

315. Price competition has been intense in the San Antonio area since the 1984 transactions. (Lydick, Tr. 2974; Bodnar, Tr. J 480; CX 919-A; CX 1459.

THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 497 452 Initial Decision 316. Soft drink prices decreased in 1987 when Pepsi COBO increased its discounts on soft drinks in Texas and the San Antonio area. (RX 1126-M; RX 1129- , K.) CCSW matched these discounts follow d by further reductions by Pepsi COBO. (Davis, Tr. 4548-49, 4549-59.

317. The price of soft drinks in San Antonio in 1987 was below the price in 1977. (CX 1427- ) The average net effective price of Pepsi soft drinks in San Antonio during 1987 was $5.45 , 12.4% below the $6.22 average for 1986. (CX 2382- Y.) 318. Pepsi COBO forgoes profits at the bottler level to build sales volume and market share in Texas over the long run. (CX 422- C; CX 2389- , P; RX 2867; CX 2389- , P, Z-3; Howell, Tr. 4019: Davis 4559- 4653; Bodnar, Tr. 1482, 1568; CX 1427-F) Pepsi COBO has never made a profit at the bottler level in San Antonio. (Coyne, Tr. 3456; Davis, Tr. 4561.) 319. Pepsi COBO directed lower prices at CCSW, hoping to take advantage of CCSW' s financial burdens to generate sales, volume and market share. (Davis, Tr. 4605, 4614- , 4676-80; CX 3141- CX 2177- 320. Pepsi planned to offset $10.4 million in bottling losses in South Texas during 1989 with $9.5 million in concentrate profits. (CX 778- 25.

321. CCSW must meet its fixed costs and its interest expense while maintaining the cash flow ratios required by its loan agreements. (CX 1437; R. Hoffman, Tr. 5634.

322. In 1987, price competition and the inability to generate sufficient volume growth placed CCSW in financial difficulty. (R. Hoffman, Tr. 5643-44; E. Hoffman, Tr. 523-24; Howell, Tr. 3985- 86.

323. Based on the unsatisfactory financial perfonnance of CCSW in 1987, and the risk occasioned by violations of loan covenants TBG refinanced the acquisition loan. In addition, George Van Houten and David Green replaced Norb Cole and David Schwerdtfeger as President and Chief Financial Officer, respectively, on January 8, 1988. Toby Summers was promoted to Executive Vice President and Chief Operating Officer. The Vice President of Sales Larry Teague, had been terminated in September, 1987. (E. Hoffman, Tr. 428- , 525. ) In June 1988, Toby Summers replaced ) Initial Decision 118 FTC George Van Houten as President of CCSW. (E. Hoffman, Tr. 526; Summers, Tr. 6708.

324. CCSW' s bottling profitability has been below that of other Coca-Cola bottlers in recent years. (RX 759 (1987); RX 760 (1986); RX 598 (1985); RX 2049 (1983); RX 303- 325. The Coca-Cola Bottling Group (Southwest), Inc. (Texas) refinanced its debt in 1990 with a group of insurance companies at a fixed rate with no principal payments for seven years. (CX 891; E. Hoffman, Tr. 291-92.

Volume Share 326. "Brand loyal" consumers wil pay higher prices for their brand of soft drinks. (Turner, Tr. 1397-98. Brand loyalty" refers to the extent to which a consumer purchases only one flavor of soft drink. (CX 848- W; RX 642-E; RX 686- , 1.) 327. Brand loyalty for soft drinks is low and declining. (CX 1126- , K; RX 642-E; RX 2842; CX 972- 4; RX 1323-J; Koch, Tr. 1869; Davis, Tr. 4757-58; RX 2842- L; RX 1368-A: Coyne, Tr. 3574-75.

328. In this market, price competition, and the frequency oflow promotional" prices for soft drinks, induce consumers to buy on price. (Hixon, Tr. 7304-05; Knowles, Tr. 2837-38; RX 686- , I; CX 972- 4; RX 2843-A007006; RX 1533- , G; CX 2424-E; CX 2407- 329. At least one cola is always on sale. Cola drinkers are switchers who buy on price, especially in the sugar segment (nondiet). (Coyne, Tr. 3449-50; RX 686- , 1.) 330. The decline in brand loyalty is due in part to the proliferation of varieties of one trademark. (CX 1274- ) Before 1983 the trademarks only brand which carried the "Coca-Cola" and "Coke" was Coca-Cola. Since that time, diet Coke, Coca-Cola Classic and Caffeine-Free Caffeine-Free Coke, Caffeine-Free diet Coke, Coca-Cola Classic have been introduced. (F 332-36. 331. Sales of the new brands reduce the sales of existing brands. (Atchison, Tr. 5190-91; Stout, Tr. 5115.) The projected "cannibalization rate" of Cherry Coke was 49%. (Stout, Tr. 512627; CX 1140- ) T AS share declined 50% when diet Coke was introduced. (Sum- THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 499 452 Initial Dccision mers, Tr. 6730-31; CX 168-A.) New packers have the same effect. (RX 1365- 332. In May 1985, CCUSA substituted a refonnulated Coca-Cola ). (Stout, Tr.brand ("New Coke ) for the old fonnula ("Old Coke 5042.

333. Consumers, especially in Texas, rejected New Coke and Coca-Cola market share declined sharply. (RX 680- ) CCUSA reintroduced Old Coke as "Coca-Cola Classic" in September 1985. ) Sales of New Coke(Atchison, Tr. 5202, 5232; Stout, Tr. 5048. dropped dramatically. (Atchison, Tr. 5202, 5205.) New Coke had a 0.3% market share in the OctoberlNovember 1989 San Antonio Nielsen. (RX 2806- 334. During the May-September 1985 period, CCSW had 18.8% decrease in its sales/share. (CX 3557- ) Sales of Royal Crown Cola, Dr Pepper and Pepsi Cola increased. (Nicholson, Tr. 3718, 3804; Knowles, Tr. 2660.) Royal Crown s Nielsen share thejumped from 1.8% in the June/July 1985 to 5.2% in August/September 1985 Nielsens. (RX 2806- 335. In the summer of 1985, CCUSA introduced Cherry Coke. The sales share of Cherr Coke in the San Antonio Nielsens peaked at 3.3% during August/September 1985 , and has been declining ever since. (CX 3991-Q; CX 3558-P; RX 2806-R) Its 1989 annual share was 0.3%. (RX 2806- 336. Caffeine-Free Coca-Cola Classic was introduced in March 1990 and the national share is now 0.9%. (Atchison, Tr. 5219, 5222. 337. Caffeine-free soft drinks appeal to 15-20% of consumers. The caffeine-free category recently began with SevenUp and Sprite. (Coyne, Tr. 3475-76; Atchison, Tr. 5218.

338. Surveys of sales show large swings related to changes in retail pricing. (RX 452-M; Davis, Tr. 4563.) Texas consumers have low prices as a result of the "cola wars" and have become price sensitive, so that a change in price will produce a significant volume change. (Knowles, Tr. 2837 -38.

339. Pepsi sales surged 16% as a result of the price wars in 1987. (RX 2867; RX 2807. ) In 1988 Pepsi COBO increased net effective prices by an average of 6. 9%. (CX 4148.) During the first seven months of 1989, Pepsi shares were down 19% in bottle/can and 12% overall compared to 1988. (CX 4148- Initial Decision 118 FTC 340. COBO and Grant-Lydick have increased their Nielsen market share in CCSW' s territory since 1985. (Summers, Tr. 6766, 341. Private label sales volume wil increase if CCSW raises prices. When private label is featured on ad at a lower price, sales volume rises, eroding CCSW sales volume. (Summers, Tr. 6771.) 342. When private label brands, market share increases, Pepsi COBO' s share decreases, and when Pepsi COBO' s share increases, private label share decreases. (RX 2975; RX 43; Strickland, Tr. 7966-67; Summers, Tr. 6554-57.

343. Several retailers, led by HEB, promote private label soft drinks heavily. The market share of private label ("control brands has increased from 3.9% in 1982 (CX 3557-T), to 18% in 1989. (RX 2806; RX 2961; Summers, Tr. 6553; Howell, Tr. 4092. ) From 1984 to 1989, control brands grew 57%, sales of all other brands grew 20% , in Bexar County. (Summers, Tr. 6766. 344. The 1989 San Antonio private brand share for three flavor categories of soft drinks (7.3% of the market) was: grape 75%; root beer 44%; orange 25%. (CX 421-C; E. Hoffman, Tr. 624. 345. From 1984 to 1989 the sales share of Dr Pepper in Bexar County increased from five to eight percent, a 60% increase. (RX 34-A; Summers, Tr.6727.

346. During 1984 and 1985, market share for Dr Pepper, Pepsi (CX 27-Q) and RC Cola (CX 27-U) products increased, while Coca- Cola lost share. (E. Hoffman, Tr. 541-42. 347. Soft drink sales and market share in Texas are volatile. (CX 2392-H, N; RX 488- R; RX 666-E; CX 2533- 28; RX 1200; RX 1469- 6 (CCE).

Public Reaction 348. Retailer employees testified that there had been no adverse consequences from the 1984 acquisition. Chapman of HEB called it a non-event" (Chapman, Tr. 7249) and Sendelbach of Super S Foods said that the acquisition benefited Dr Pepper. (Sendelbach, Tr. 7690. 349. Ladd Little, president and owner ofl.C. Vending, and his sales manager, Terry Jackson, complained about the 1984 acquisition. (Little, Tr. 669-70, 705; Jackson, Tr. 3309- 10. ) L.c. Vending buys soft drinks from CCSW and sells them as a direct competitor of THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 501 452 Initial Decision CCSW' s vending operations. (Jackson, Tr. 3374; Little, Tr. 652 703-04.

350. L.c. Vending wants to raise the vending price on its soft drink machines, but competition with CCSW has undercut any higher price. (Little, Tr. 739-41.) 351. Emery Bodnar, former General Manager of DP-SA and currently General Manager of Grant-Lydick also complained about CCSW' s low prices. (Bodnar, Tr. 1571- , 1695. ) Mr. Bodnar was concerned about low pricing on CCSW' s Cima Red, which is similar to Grant-Lydick' s Big Red. (Bodnar, Tr. 1369. Innovation 352. New brands introduced in the San Antonio area since 1984 include Coca-Cola Classic, Caffeine-free diet Coke, Cherry Coke diet Cherry Coke, Minute Maid Orange and Lemon-Lime, Pepper Free, Original New York Seltzer natural fruit- flavored soda, and other seltzers, Lipton Tea, Cherry 7 Up and diet Cherry 7 Up 7 Up Gold, IBC Root Beer, and Slice. (CX 2038- , G; CX 2503- 3; CX 1673- , C; RX 803, p. CC36128633 (New brands introduced from 1978-87 rcached 20.2% share in 1987 Nielsen audit); RX 1183- (New Pepsi brands introduced since 1982 are 15% of business). Thirty-four new brands appeared from 1985 to October 1988. (CX 1673- , D; CX 3998.

353. New packages have been introduced or emphasized in CCSW' s territory since 1984, including I6-ounce PET, I-Liter PET 20-ounce PET, 3-Liter PET, Bag-in-Box, and multi-paks of 12 and 20 cans. (E. Hoffman, Tr. 563.

354. San Antonio was a test market for the 3-liter PET package introduced in 1984. (Atchison, Tr. 5226.

Efficiency 355. In January, 1987, CCSW had a "Reduction In Force reducing payroll by 20%. (CX 920- , CX 959; CX 241.) 356. CCSW' s acquisition of the Corpus Christi territory, from American Bottling Company, and the consolidation into San Antonio Initial Decision 118 FTC led to a cost savings. CCSW' s labor cost per case dropped 21.7% between 1986 and 1987. (CX 1399- 357. Consolidating production into one facility and using idle equipment reduced CCSW' s fixed overhead costs of manufacturing. (Summers, Tr. 6366.

358. CCSW delivers to the customer s warehouse in truck/trailer rigs rather than route trucks. Soft drinks are loaded on pal1ets. (Summers, Tr. 6411- 12, 359. Under ownership of The Coca-Cola Bottling Group (Southwest), Inc., CCSW has had cost savings in consolidation and volume discounts on raw materials. (E. Hoffman, Tr. 277- , 523. Dr Pepper USA 360. The Dr Pepper brand and DPUSA have been helped by the 1984 acquisition.

361. Sales volume and share of Dr Pepper brand soft drink in the San Antonio area increased since the 1984 acquisition. (CX 3946; RX 2823; Knowles, Tr. 2784-85.

362. Dr Pepper per capita sales in the San Antonio area increased 40% between 1984 and 1988. (CX 709-H.) The rate of Dr Pepper sales growth for the nation was about half that rate. (Knowles, Tr. 2848-49.) CCSW provided Dr Pepper products an excellent distribution system and worked to develop the brand. (Knowles, Tr. 2668, 2784- 2853- 2848; Coyne, Tr. 3598; E. Hoffman, Tr. 413.) Dr Pepper brands benefit when advertised with Coca- Cola. (Kaiser, Tr. 3232-33.

363. In 1984, per capita sales of Dr Pepper in CCSW territory were 74. 5 gallons, lower than the 85. 1 gallon per capita sales of surrounding bottlers, but by 1988, per capita sales of Dr Pepper in CCSW territory were 104.7, higher than surrounding bottlers. (RX 2826; RX 2828; Knowles, Tr. 2794-96; Clarke, Tr. 4380. 364. In San Antonio, Dr Pepper bottle/can sales decreased from 1982 to 1984 but began to increase in 1985 to 1988. (Knowles, Tr. 2878; RX 2823; RX 2980.

365. Military bids require that 80% of the can vending business be from Coca-Cola and Pepsi bottlers. Dr Pepper Company brands are in many vending machines as a result. (Summers, Tr. 6676-77. THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 503 452 Initial Decision Other Competitors 366. Grant-Lydick purchased the remaining assets of DPincluding the bottling plant and equipment and approximately 60% of the trucks (Lydick, Tr. 2978-79), for $6. 5 milion. (Antle, Tr. 3074, 3099; Turner, Tr. 1158; Lydick, Tr. 2981-82. 367. Grant-Lydick estimated that the assets and franchises they acquired were worth over $12 million. (RX 1648; Bodnar, Tr. 1645- 46; Lydick, Tr. 2982; Antle, Tr. 3074, 3099. 368. The brands which Grant-Lydick took over from DP-SA in 1984 accounted for 60% of DP-SA' s 1983 volume. (Lydick, Tr. 2978-79.

369. Respondent s accounting expert compared Grant-Lydick' profitability to the average profitability of 120 bottling companies. (RX 204; RX 205-K; Goode, Tr. 7427-33.) He concluded that Grant- Lydick was "doing very well in relation to the average for the industry." (Goode, Tr. 7439, 7444.

370. Grant-Lydick has been successful in obtaining feature grocery ads and in-store promotions for its brands. (CX 2954- B; CX 3248- E; RX 256- , C; RX 461; RX 1678. 371. Nielsen data show that Grant-Lydick receives a higher percentage of the total shelf space than its percentage share of sales. (Bodnar, Tr. 1613- 14.

372. Grant-Lydick increased profits from 1984 to 1988. (RX 2991.) Grant-Lydick has had geographic expansion in recent years. (RX 2970.

373. Grant-Lydick' s brands have had increased sales and share. (RX 201-A; CX 438- , C; Lydick, Tr. 3011- 12. ) Sales of the Big Red brand have grown. (Sharp, Tr. 7546-47. 374. Royal Crown s sales records (RX 2846; RX 1793), and Grant-Lydick' s reported sales ofRC brand products (RX 2784- , D) show growth of RC products sold by Grant-Lydick. (RX 2954- RX 2955- , W; RX 2956- , W; RX 2957-Z-4-7, 10; RX 2958- 12.

375. Emery Bodnar believes that he has caused Grant-Lydick to be a "tremendous success story. " (Bodnar, Tr. 1692. 376. The financial statemcnts of Texas Beverage Packers show a growth in profitability from 1981 to 1988. (Hixon, Tr. 7319-21; RX 2953; RX 1845-49.

Initial Decision 118 F.TC. 377. TBP's sales increased from 1981 to 1988. (RX 1850-56: RX 2952; Hixon, Tr. 7316- 18.

EASE OF ENTRY Distributors 378. Bottling plants in Texas are willing to facilitate new entry by producing new products (RX 2273), and new soft drink distributors have entered by having contract packers produce their product. (Limon, Tr. 4956 (AGA Beverages); Hixon, Tr. 2698; RX 2699. New entrants need not invest the capital required to build a new bottling plant. (Howell, Tr. 3999.

379. The physical requirements for distributing soft drinks consist of: a warehouse to store the product; trucks to deliver the product to retailers; and delivery and administrative employees. (Espinoza, Tr. 4237; Summers, Tr. 6478-79. 380. The cost of the equipment to enter into the business of distributing soft drinks is relatively low. The cost of developing a DSD distribution system to serve the San Antonio area is about $25 000. (Espinoza, Tr. 4237.) A 1988 Nehi business plan estimated that the start-up would cost $30 000 (Rx 2858-G), and take three and a half months (RX 2858- , F), and that profits during the first five months would recoup those costs. (Espinoza, Tr. 4231-33; RX 2858- 381. A soft drink brand must be accepted by retailers and be allotted shelf space, and have access to ad features or instore displays. (Espinoza, Tr. 4210 (with HEB); Donald, Tr. 5293. 382. CCSW and Pepsi COBO obtained ad features with HEB but lost money because the sales increase did not offset the cost of obtaining the ad. (Summers, Tr. 7829-33; Clarke, Tr. 4387- 88 (Dr Pepper); Davis, Tr. 4705; CX 2394- 67 (Pepsi). 383. Brands by newer, smaller distributors, such as IBC Root Beer (Nelson Brokerage) and Nehi (Espinoza), have acquired ad features and sales with San Antonio retailers. (CX 1295; CX 1299 (IBC in HEB ad); CX 88.

384. Retailers can feature their private label brands in ads or instore displays without incurring any direct costs. (Hilke, Tr. 6282- 83.

THE COCA COLA BOTTING COMPANY OF THE SOUTHWEST 505 452 Initia! Decision 385. Access to ad features does not guarantee the success of a soft drink product. (Knowles, Tr. 2656-57. 386. Grant-Lydick and Texas Beverage, which have obtained fewer feature ads than CCSW and Pepsi COBO in the San Antonio area in recent years (Bodnar, Tr. 1378- 80), have been profitable during this period. (F 372.

387. Mr. Espinoza has fonned distribution companies for Nehi flavors and other brands in the San Antonio area and the Rio Grande Valley. (Espinoza, Tr. 4163-67; RX 1777- , U- Bottlers 388. The cost to install a can line to produce five million cases of cans per shift per year with used equipment is $825 000. (Summers, Tr. 6460.

389. Used equipment is available because of recent consolidation in the soft drink industry. (Hixon, Tr. 7296; Bodnar, Tr. 1653-54. Such equipment costs less than half of the cost of new equipment. (Summers, Tr. 6447-60.

390. Other requirements for entering into the bottling and canning business are: a plant, a warehouse and trucks. (Summers, Tr. 6464 6467 6478-79.

391. Due to the depressed real estate market in South Texas, a prospective bottler could easily lease a suitable facility to install a bottling line (4 000 square feet). (Summers, Tr. 6465 , 6479. 392. Just-in-time inventory requires little warehouse space; space for seasonal1y higher inventory is readily available for lease. (Summers, Tr. 6463-64,) 393. Since 984 new firms have entered the bottling business in competition with CCSW. Entry has been quick and inexpensive. Kroger purchased the Safeway bottling and canning plant in Garland Texas in the fall of 1987 for $1. million. (CX 2827; CX 2828- Morath, Tr. 7661-62; RX 2304; RX 2441; RX 1740- , H N; RX 1741; RX 1744; RX 1745; RX 1750; RX 2441- A; RX 1711.) Kroger spent $600-700 000 to get the plant into production (Morath, Tr. 7661-62; RX 2441; RX 1760), which took four months. (Morath, Tr. 7662. ) The Garland plant produces five million cases per year including Kroger s private label brand and contract-packed brands. (Morath, Tr. 7662-64.

Initial Decision 118 FTC. 394. HEB previously produced soft drinks (Chapman, Tr. 7 I 55), but now uses Texas Beverages Packers ("TBP"), a contract packer to produce Plaza, its private label soft drink. (Chapman, Tr. 7147; Hixon, Tr. 7298; Summers, Tr. 6562. ) In 1987 HEB determined the costs of installing a bottling line in an HEB warehouse. (CX 201- C: RX 2040.) The project would cost $2.7 million and take a year. (Chapman, Tr. 7152-53; CX 201-B; RX 2040-A.) HEB projected that the annual contribution from running the bottling line would be $449 000. (Chapman, Tr. 7153; CX 201-B; RX 2040-A.) 395. HEB compared this cost with price offered by TBP, their current contract-packer. (CX 201- , E; RX 2040-B; RX 2041.) HEB decided to extend their current contract-packing with TBP for two years. (Chapman, Tr. 7150-51; Hixon, Tr. 7298-7301; CX 201- , M; RX 2040- , B.) HEB reserved the right to build their own bottling plant during the life of the contract. (CX 201-A; RX 2039. Recently, while remodeling an existing warehouse in San Antonio HEB installed water and sewage equipment to facilitate the installation of a bottling line. (Chapman, Tr. 7150. 396. Bottling of nationally branded soft drinks, to be deliveredstore-door in the San Antonio market, has comparatively high entry barriers. "(NJew entrants are scverely restricted and are relegated primarily to additional regional or other non-major brands with relatively insignificant market positions. " (CX 1406- 9; CX 102- Concentrate Manufacturing 397. "Flavor houses" inexpensively provide concentrates for new products. (CX 650 (Monarch); Antle, Tr. 3115- 16; Turner, Tr. 1427; Bonica Test. , RX 301 0, pp. 3373-74.) A new entrant like Soho (Collier Test. , RX 3015, pp. 4080-82) can rely on flavor houses to produce concentrates for their products. (Morath, Tr. 7668 (Kroger). CCSW makes and sells Cima Red and Spike. (CX 436; RX 541; Summers, Tr. 6687.) The flavor extracts for these two products are purchased by CCSW from Universal Flavors. Flavor extracts from a flavor house like Universal Flavors are less expensive than the bottling concentrate sold to CCSW by its soft drink franchisors. (RX 541-B; Summers, Tr. 6546-47.

398. A bottler could introduce a new product within a short time. (Bodnar, Tr. 1681-82; Clarke, Tr. 4372 ("four weeks ); Turner, Tr. THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 507 452 Initial Decision 1425-27 ("8- 12 weeks ); Coyne, Tr. 3584-91 ("3-6 weeks Amrosowicz, Tr. 869-70 ("60 days 399. CCSW introduced Cima Red, a red cream soda product similar to Big Red (Schwerdtfeger, Tr. 2366-69; E. Hoffman, Tr. 345-46; CX 436) within six months at a cost of$7 500 - $15 000 (CX 428-C; CX 436-0). Cima Red reached a 1. 1 % market share in the OctobcrlNovember 1988 Bexar County Nielsen audit. (RX 2806- CCSW' s new isotonic soft drink, Spike (RX 541; CX 3685; CX 308) reached full production within three months after the name was selected. (Summers, Tr. 6692.

400. Better Beverages introduced Promotion, an isotonic soft drink product, in May 1990, two months after signing the franchise agreement, with newspaper coupons, point of sale material, and retailer authorization in place. (Campbell, Tr. 1991-93. 401. Better Beverages introduced Red Red, a red soft drink product of the Monarch Company that is similar to Cima and Big Red, in May 1990 after two months of preparation. (Campbell, Tr. 1995.

402. Better Beverages introduced Nesbitt s Orange, Strawberry, and Nugrape immediately after obtaining the franchise. (Campbell Tr. 1994-95. ) Oneta Company developed and introduced Everest Seltzer in a two-month period. (Koch, Tr.1902. Piggybacking 403. New entry at the concentrate level has been facilitated by piggybacking." (F 112- 16.) A new concentrate can enter a market readily by distribution through a bottler already distributing competing products. (Espinoza, Tr. 4185 , 4189. ) Dr Pepper distribution through CCSW is piggybacking.

404. Piggybacking allows new entrants to take advantage of the distribution systems developed by established concentrate companies. (Knowles, Tr. 2765- , 2772-73.

405. Piggybacking allowed fast, low-cost new entry or gcographic expansion of Dr Pepper, W e1ch' s, A& W, Sunkist, and Canfield. (Lydick, Tr. 2975 , 2975-76; Knowles, Tr. 2767- 2772-73. 406. As a result of the decision to license cola bottlers, Dr Pepper s national market share grew from 2% in 1960 to 5% in 1978, a growth rate faster than the national average for the soft drink industry. (Knowles, Tr. 2767-69.

Initial Decision 118 FTC 407. CCSW quickly distributed new drinks like Lipton Tea Delaware Punch and Original New York Seltzer. Caffeine-free Classic Coca-Cola took less than five weeks to introduce in the San Antonio market. (Summers, Tr. 6687.

POTENTIAL EFFECTS Market Power 408. CCSW attempted to raise its prices in 1988, but was unable to do so. (Summers, Tr. 6763; R. Hoffman, Tr. 5546-47, 5550-51.) CCSW had to match Pepsi price reductions or lose market share. (F 316.

409. CCSW in 1989 increased its average list price from $9. to $10. , or 69 per case. (RX 2990.) CCSW was unable to increase its prices above the amount required by cost increases. (Strickland Tr. 8134, 8186.

410. In 1989 Pepsi COBO attempted a series of price increases averaging 6.9% in South Texas. This led to a 19% reduction in Pepsi COBO' s sales during the first seven months of 1989. (RX 2987; Strickland, Tr. 7987- , 8000-04.

411. Bottlers' profit margins on soft drinks have shrunk since the early 1980' s. (F 310.) This has forced bottlers to cut operating costs and pursue increased sales. (R. Hotfman, Tr. 5634-35; Turner, Tr. 1431.) 412. Pepsi COBO is aware of CCSW' s financial difficulties (Davis, Tr. 4605; Schwerdtfeger, Tr. 2375-76, 2601) and is unlikely to allow CCSW to increase prices. (Summers, Tr. 6763. Collusion 413. The 1984 acquisition did not reduce the number of competitors in the San Antonio area. (Turner, Tr. 1158- 59; F 295. 414. Soft drinks are sold by thirteen bottling companies in CCSW' s territory. This does not include sales of private label and warehouse brands, contract packers, or fountain wholesalers. (RX 3109; Strickland, Tr. 8142-44.

415. There are numerous fountain wholesalers selling Coca-Cola fountain syrup in Texas. (RX 1869.) DPUSA also has many fountain wholesalers in Texas. (Cassagne, Tr. 7598-99; RX 2799. THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 509 452 Initial Decision 416. Pepsi USA profits from the sale of concentrate in Pepsi COBO' s products. (Knowles, Tr. 2840-42, 2894; Howell, Tr. 4019; F 320.) Pepsi USA' s concentrate profits are used to offset Pepsi COBO' s operating losses at the bottling level in South Texas. (CX 778- 25; F 320.

417. Pepsi COBO has a large bottling and canning plant in Conroe, as well as smaller bottling plants in San Antonio and Houston. (RX 2939.) Because of the volume produced at Conroe Pepsi COBO has lower production cost for cans than CCSW. (Cole RX 3008, pp. 118- 19.

418. Pepsi COBO is low priced and buys its way into the ad cycle. (Turner, Tr. 989- , 1056.

419. CCUSA profits from the sale of concentrate and syrup to bottlers like CCSW and CCE. CCUSA wants bottlers to reduce prices of soft drinks to stimulate retail soft drink sales, which leads to higher concentrate sales and profits. (Howell, Tr. 4072-73. 420. CCUSA sells fountain syrup directly to fountain customers and to fountain wholesalers. (Howell, Tr. 4005; F 90, 93. 421. DPUSA' s cost of concentrate sold to bottlers like CCSW and CCE is less than 10% of the DPUSA' s price. (Knowles, Tr. 2665.

422. DPUSA negotiates the price of fountain syrup sold to most Dr Pepper fountain customers and to fountain wholesalers. (Cassagne, Tr. 7590; RX 1919-C.) 423. CCSW must increase unit sales volume. (Summers, Tr. 6636 6763-64. If CCSW increased prices, volume would be reduced and the loan covenants could be violated. CCSW has $220 million of debt and interest expense of $27 million per year. (R. Hoffman Tr. 5471 , 5481- , 5569, 5614, 5600, 5634, 5718, 5633; CX 1354- ) Cash flow, rather than profitability, is success for CCSW because TBG' s lenders look to cash flow as the source of debt repayment. (R. Hoffman, Tr. 5417, 5481- , 5612- , 5706. 424. Kroger discounts soft drinks to draw consumers to its stores to increase grocery sales. (Howell, Tr. 3951-52. 425. HEB also uses soft drinks as a loss leader to increase consumer traffic in its stores. (Gonzaba, Tr. 2032; Howell, Tr. 3951; Summers, Tr. 7004.

426. Convenience stores sell fountain soft drinks because the cost to the retailer is lower than finished soft drinks and the consumer serves himself. (Summers, Tr. 6935.

Initial Decision 118 F.T. 427. Fountain wholesalers like Martin-Brower, Sysco and Sugar Foods purchase fountain syrup from CCUSA and resell it to fast-food restaurants and other customers. (Short, Tr. 7740- , 7753; RX 1869.

428. None of the sequentially-operated Espinoza companies has owned a bottling plant; each purchased all of its finished soft drinks from contract-packers in Fort Worth and Temple, Texas and in Mexico. (Espinoza, Tr. 4193 , 4249-51; Limon, Tr. 4956. 429. Texas bottlers who contract pack for other soft drink distributors include: Texas Beverages in San Antonio, Beverage Packers in Ft. Worth; Temple Dr Pepper Bottling Company; Better Beverages in Halletsvile; CCE at various locations; and Dr Pepper Bottling Company of Texas in Dallas and Houston. (Summers, Tr. 6466; F 126.

430. There is excess capacity in the bottling and canning of soft drinks in Texas. (F 133-39.) 431. Soft drink price competition in Texas makes collusion difficult. (Knowles, Tr. 2899.

432. HEB is the leading retail grocery chain in San Antonio and Corpus Christi. (Knowles, Tr. 2836; Howell, Tr. 4041; Bodnar, Tr. 1743. ) HEB has 50% of the retail grocery business in the San Antonio area. (CX 3138-B; CX 2088- ) In 1990 25% ofCCSW' sales were to HEB. (Summers, Tr. 6589; CX 3806- 37. ) From 20- 25% of Pepsi sales in San Antonio were to HEB. (Davis, Tr. 4525. 433. HEB buys more than five million cases a year from CCSW. (CX 956-A) HEB has a larger market share in the San Antonio area than both Albertson and Kroger (Davis, Tr. 4525 (each 8-9% share)) but Kroger and Albertson are national grocery chains which are much larger than HEB. (Summers, Tr. 6767; Howell, Tr. 4130-31.) 434. Kroger is the second largest customer of CCSW, purchasing 12% ofCCSW' s total unit sales. (Summers, Tr. 6589. 435. Sam s Wholesale Clubs purchase 7-8% ofCCSW total unit sales. (Summers, Tr. 6638.

436. Sales to convenience stores are 15% of CCSW' s total case sales. (CX 53-I.) 437. The Stop- , operated by National Convenience Stores is a nationwide chain, and has 203 stores served by CCSW. (Summers, Tr. 6631.

438. Circle K, with 45 stores, was the second largest convenience store chain served by CCSW. (Summers, Tr. 6631.) THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 511 452 Inilial Decision 439. Super S is a major retailer in rural markets with 45 stores. (Summers, Tr. 6629-30.

440. The Army-Air Force Exchange Service (AAFESJ and the United States Navy operate stores on military bases in the San Antonio area. (Summers, Tr. 6675-76.

441. HEB requires that CCSW and other bottlers offer HEB the lowest net wholesale price available to retailers from each bottler. (Brinkley, Tr. 2234; Bodnar, Tr. 1660-6!; Chapman, Tr. 7245; Turner, Tr. 1200; Summers, Tr. 6646; CX 3700- ) Kroger and Albertson have similar policies. (Donald, Tr. 5320- , 532728; Kaiser, Tr. 3264.

442. HEB expects that bottlers will offer to other retailers the same prices offered to HEB. (Chapman, Tr. 7245; Howell, Tr. 4055. 443. HEB pressures CCSW to offer the same wholesale price as CCE on Coca-Cola products. (Summers, Tr. 6626. 444. Retailers specify the type of payments for promotions including flat payments for HEB, and flat payments plus per case rebates for Kroger. (Howell, Tr. 3943-44. ) Stop- Go requires payment in advance. (Summers, Tr. 6638; Howell, Tr. 3988-89, 4059- 4063; CX 1068.

445. Retailers can limit promotions and display activities of soft drink products. (Coyne, Tr. 3487.) Ads and in-store displays are important to soft drink companies. (Turner, Tr. 1130; Coyne, Tr. 3449-50; F 171.) 446. HEB sometimes promotes its private-label soft drinks rather than national brands. Other chains run 52 weeks of national brands. Kroger may run private label on top of national brands. (CX 2379- Hixon, Tr. 7303; Brinkley, Tr. 2199; Davis, Tr. 4526; Donald, Tr. 5324.

447. In Fall 1989, HEB promoted Pepsi products at the same time as Plaza private label products. (Knowles, Tr. 2753-55. 448. In 1986, Kroger did not buy outside bottlers, 3-liter product so that Big K, its private label soft drink line, could be the only 3-liter package available from its stores. (Howell, Tr. 4063. 449. In 1988 , HEB notified all vendors that it would not accept price increases for four months. CCSW complied rather than risk retribution for HEB. (Summers, Tr. 6769.

Initial Decision 118 F.T. 450. In 1986, Stop- Go refused to feature Coca-Cola products for six months in South Texas, because CCSW would not agree to S top- N - s terms for promotional programs. (Howell, Tr. 4061-63. 451. HEB required On eta, the Pepsi-Cola bottler in Corpus Christi, to remove its vending machines from all HEB stores because Oneta offered Sam s Wholesale Club a lower price than Oneta offered to HEB. (Davis, Tr. 4745-46.

452. HEB and Kroger have each canceled scheduled ads because the price was not competitive. (Summers, Tr. 6626-27 (HEB); Kaiser, Tr. 3218 (Kroger).

453. There are thirteen private brands of soft drinks in the CCSW market, usually with a retail price of six cans for $1.00. (Summers Tr. 6549.

454. CCUSA and DPUSA pressure CCSW to keep prices down to increase sales volume, criticizing its performance by comparison to sales records of other bottlers, and granting or withholding marketing support. (R. Hoffman, Tr. 5646-48.

455. DPUSA provides inducements to bottlers to assure that pricing for Dr Pepper products is low. (Knowles, Tr. 2698, 2846.) If a bottler experiences intense competition, DPUSA provides funds to assist the bottler s efforts to meet competition. (Knowles, Tr. 2747. 456. Concentrate companies pay part of the cost of promotions by their bottlers. (RX 498-C; RX 337; Coyne, Tr. 3417 - 18; Howell Tr. 3928- 29; Turner, Tr. 963-65; Knowles, Tr. 2698 , 2745-48; Bodnar, Tr. 1484-88.) In 1986 CCUSA' s promotional payments to CCSW totaled $3.37 million (CX 3205-A), and DPUSA' s funding for the San Antonio area totaled $644 851. (CX 3204- 457. Concentrate companies use "best efforts" requirements in franchise agreement to threaten to terminate the franchises of bottlers who have not increased sales. (RX 2835; CX 2676; Nicholson, Tr. 3775-76; Summers, Tr. 6759.

458. Low consumer prices increase volume and the purchase of concentrate which bottlers must buy from concentrate companies at a high-margin, fixed price. (Knowles, Tr. 2912, 2838-39. 459. Personal income is relatively low in San Antonio and consumers are very price sensitive, even more price sensitive (Davis Tr. 4811) than consumers in other Texas cities. (CX 1489; CX 108- G; CX 3778-A; CX 3162; CX 1054-P; Bodnar, Tr. 1545- , 1664; Davis, Tr. 4758; Kaiser, Tr. 3234-35 ("San Antonio more blue collar THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 513 452 Initial Decision 460. The 3-liter bottle provides consumers in San Antonio the lowest price per ounce nonreturnable soft drink package. (CX 1999- , D; Summers, Tr. 6770.

461. Recent demographic and economic trends in the San Antonio and South Texas (CX 3705- 28) areas have led to increasingly price-sensitive consumers. (Knowles, Tr. 2837, Summers, Tr. 6770.

462. The Texas Attorney General' s Office has authority and incentive to deter any collusive price increase by CCSW. (CX-2; F 68-70.) The provisions of the AG' s order impose constraints on CCSW' s use of marketing programs and practices in the San Antonio area. (CX 2; F 68.

DISCUSSION The complaint challenges CCSW' s acquisition in 1984 of the Dr 8 alleging a violation ofPepper and Canada Dry bottling franchises Section 5 of the Federal Trade Commission Act, 15 U. c. 45, and Section 7 of the Clayton Act, 15 U. c. 18. The acquisition allegedly lessened competition by weakening Grant-Lydick, reducing competition between Dr Pepper and Canada Dry brands and other brands, and by increasing the likelihood of collusion and the likelihood that respondent will unilaterally exercise market power. Amended Complaint paragraph 13.

L THE RELEV ANT PRODUCT MARKET Complaint counsel argue that the relevant product market is widely advertised, brand, finished carbonated soft drinks or syrup merchandised and distributed by direct-store-door delivery, in all channels of distribution. This definition includes the national brands of carbonated soft drinks sold by CCSW, Pepsi COBO, Grant-Lydick and the Espinoza companies. (F 179.

8 CCSW also acquired from San Antonio Dr Pepper Bottling Company trucks, a warehouse, and Dr Pepper vending machines. (F 50.

9 CCUSA and DPUSA also sell fountain soft drinks. The parties agree that those sales are also in the relevant market. Nationally. fountain sales are about one-third of all soft drink sales. (eX 34J8-F.) Initial Decision 118 FTC Respondent argues that the relevant product market includes private and warehouse brand soft drinks, and non-carbonated soft drinks, delivered by DSD or warehouse.

A. Law Product markets are defined by the "cross-elasticity of demand" or the "reasonable interchangeability of use" between the product in question and potential substitutes. Grand Union Co. 102 FTC 812 1041-42 (1983). When reliable evidence of cross-elasticity (the extent to which a change in price of the product will cause customers to switch to substitutes) is available, it can be "most important " in product market definition. Less direct evidence may also be considered such as, Olin Corp. 5 Trade Reg. Rep. 22 540 at p. 543 (1990): "perceptions of buyers that the products are or are not substitutes, certain differences in price movements that are not explained by parallel trends, similarities or differences in use, design physical composition and technical characteristics, and the perceptions of sellers that the products are substitutes. B. Private Label 1. Prices The issue on which this case turns is whether private label soft drinks are in the relevant product market. Private Jabel products sell at prices lower on average than national brand products, in this A lower pricemarket traditionally about 30% Jower. (F 221 228.)11 alone does not create a submarket. Brown Shoe Co. v. United States 370 U. S. 294, 326 (1962). Here, national brands on discount draw customers from private labels, and vice versa. (F 222, 225 , 227, 229- 30.) Although private label prices average below the prices of national brands, that difference diminishes during the almost constant promotions (F 229), and private label market share in San Antonio has increased to 18% when on promotion. (F 227. ) Similarly, in 10 Since private hrands and warehouse brands differ solely in ownership of the lahel, they will be treated together as "private bbcl."

1\ Despite lower retail prices. priv8te labels have been held to be in the same relevant market as national brands. Uniled SEales v. Jus. Schlir: Brewing Cn.. 253 F. Supp. 129. 133, 143 (;\. 0. Cal), afl'd perclIriall 385 U. S. 37 (J966); 111lernnliOlw! Tel. (lnd Tel Curp. 104 FfC 280. 410- 1 J (1984). THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 515 452 Initial Dccision Olin Corp. 5 Trade Reg. Rep. at 22 545 , the two swimming pool sanitizers were in the same relevant market because, after the traditional price spread between them had narrowed, a small price increase would cause consumers to switch. In Grand Union, 102 FTC 812, 1046 (1983), despite their lower prices smaller food retailers were held to be in the same relevant product market as the merging supermarkets. 102 FTC at 1046. And in Beatrice Foods Co. 101 FTC 773, 802-03 (1983), chilled orange juice glass containers and cartons were in the same product market, despite a wide price difference between the containers, because their prices were mutually responsive.

2. Characteristics Private label carbonated soft drinks on store shelves are in the same package sizes and flavors as national brand drinks. (F 188-89. Private label soft drinks have no peculiar characteristics different from national brand soft drinks, and are formulated, mixed, packaged and consumed in the same manner as national brand soft drinks. (F 194 197 199.) Much of the "image" ofa soft drink brand is created by advertising. (CX 858-C.) To a great extent, any perceived difference among soft drinks exists in the mind. Private label soft drinks and national brands are made in the same way. HEB , the largest grocery and private label seller in CCSW' territory, contracts with a local bottler to manufacture and package its Plaza line of soft drinks. (F 394. ) Kroger, another private label vendor, purchased its own plant in Garland, Texas (near Dallas) from which it supplies the state. (F 24 393.) These private label bottling plants are just like national brand bottling plants. (F 194-95. Most national brand carbonated soft drinks are delivered and stocked on store shelves by bottler employees ("direct-store-door delivery" or "DSD"). (F 204.) Some (like Shasta) (F 201) are delivered to the retailer s warehouse and then transported and stocked in the stores by the retailer s employees ("warehouse delivery ). (F 203-04.) Some national brands like Crush and Hires are sometimes sold by the DSD method and sometimes by the warehouse delivery. (F 149. ) Consumers are generally unaware of how different soft drinks are delivered. (F 198.

Initial Decision 118 FTC 3. Industry perception Most market analysts put private labels and national brand soft drinks in the same category. The National Soft Drink Association includes all carbonated soft drinks (bottled, canned, or fountain), along with carbonated mixers, seltzers and waters and non-carbonated waters. (F 206.) Government agencies and market reports put private labels with national brands. (F 205 207-08. CCSW focuses on its strongest competitor, Pepsi COBO. That does not mean, however, that other competitors are outside the product market. Grand Union Co. 102 FTC at 1045; Beatrice Foods Co. 101 FTC at 811. CCSW and Pepsi COBO watch private labels. (F209, 211.) Other firms in this market recognize that private labels compete with national brands. (F 209- 10. 4. Price changes Similarity in price movements indicates product substitutability. BAT. Industries, Ltd. J04 FTC 852, 909 n. 328 (1984). Here, price movements indicate that private labels are in the same market with national brand soft drinks. In one study, prices of national brand and private label soft drinks moved together eight out of ten times. (F 231.) The price movements were not random and were consistent with both being in the same product market. (F 232. Direct-Store-Door Distribution Most national brands are delivered to the retailer by "direct-storedoor." Employees of the bottler deliver to the retailers ' stores, and stock the store shelves and displays. (F 142.) Most private label soft drinks and some national brands are delivered to retailers warehouses and later distributed and stocked on store shelves by the retailers, employees. Complaint counsel would exclude these sales from the relevant market.

The consumer is unaware of which distribution method is used for the different brands (F 198); private label and national brand soft drinks are displayed in the same aisle of the store, often side-by-side. (F 199.) Concentrate companies, bottlers, and grocery chains believe that private label and warehouse brands compete with branded soft drinks. (F 209- 13.) Because of the prevailing industry recognition THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 517 452 Initial Decision that private label and warehouse brands compete, the argument that they do not compete because they tend to use different delivery methods is overstated and-not persuasive. Beatrice Foods Co., 101 FTC at 808, and n. 29.

D. Non- Carbonated Beverages Lipton Iced Tea, Country Time Lemonade, and Hawaiian Punch appear on the same shelves, fountain dispensers, and vending machines with carbonated soft drinks. (P 243. ) Minute Maid Orange Soda and Slice, containing 10% fruit juice, appear in J 2-ounce cans side-by-side with carbonated soft drinks like Hires Root Beer and 7- Up. (F 244; RX 2200, pp. 107, 116; CX 2330- ) Canned and bottled Lipton Iced Tea and isotonic drinks such as Spike and Gatorade are in the same market as carbonated soft drinks. (F 242- 44.

Consumers sometimes choose sparkling waters to replace carbonated soft drinks. (CX 31O- E; RX 752-C.) Mixers and seltzers belong in the product market. (F 238.

E. Conclusion on Product Market The relevant product market includes national brand, private label and warehouse brands of soft drinks, as well as mixers, seltzers, noncarbonated beverages such as Lipton Iced Tea, Country Time Lemonade, and Hawaiian Punch, and isotonic drinks. II. THE RELEV ANT GEOGRAPHIC MARKET A relevant geographic market must conform to the commercial realities of the industry and be economically significant. Brown Shoe Co. v. United States 370 U.S. 294, 336-37 (1962). The economically significant area is the area of effective competition. United States Philadelphia Natl Bank 374 U. S. 321 , 359 (1963). The area of effective competition is "the market area in which the seller operates and to which the purchaser can practically turn for supplies. Tampa Electric Co. v. Nashville Coal Co. 365 U. S. 320, 327 (1961); B. Goodrich Co., 110 FTC 207 , 289 (1988). Initial Decision 118 FTC. Complaint counsel argue that the effective area of competition is the ten-county area of San Antonio and suburbs . That was the area of the Dr Pepper franchise acquired by respondent. Respondent argues that the relevant geographic market is most of the eastern half of Texas. 13 It is complaint counsels burden to show the size of the market. Respondent is entitled to show that that market is erroneous without proving the size of the market it claims is proper. Topps Chewing Gum Inc. , Docket No. 8463, Interlocutory Order, Nov. 15 1962.

CCSW puts most of its competitive effort into the San Antonio ten-county area. (CX 1405-Z, Z- ) On the other hand, the area of effective competition "must be charted by careful selection of the market area in which the seller operates and to which buyers can practicably turn of supplies. Tampa Electric Co. v. Nashvile Coal, 365 u.s. 320, 327 (1961). CCSW' s Dr Pepper terrtory now includes 21 counties. (F 274.) CCSW' s sells its other brands in a 60 county territory. (F 275.) Grant-Lydick also has 60 counties. (F 276.) Pepsi also has more than 105 counties. (F 277.) While much of this area may be mostly jack rabbits and sagebrush and sparsely populated compared to the city and its suburbs, the issue of geographic relevant market must be looked at more deeply, beyond what appears to be the marketplace at first glance. Factors which may be considered include B.F. Goodrich Co. 110 FTC at 289: "persistent price differences; price change differences, similarities or differences in price movements; impediments to trade, such as transportation costs that are high relative to product value; shipment patterns and transshipment levels and industry perceptions. A. Shipment Patterns Shipping patterns are perhaps the best test in determining a geographic market. General Foods Corp. 103 FTC 204, 234- (1984) (Initial Decision by AU Parker).

12 The arbitrar nature of the alleged market is indicated by the fact that one of the three counties in the San Antonio Standard Metropolitan Statistical Area is not included in the ten-county area. (RX 2965-A; Strickland Tr. 8071-72.

13 Dr. Strickland identified a relevant geographic market of the eastern half of Texas, which excluded Harlingcn and the Rio Grande Valley but included San Antonio. Austin. Dallas a!1d Houston. (RX 2983. 3107: Strickland, Tr. 8094- , 8702. THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 519 452 Initial Decision 1. The Elzinga-Hogarty Test The Elzinga-Hogarty (" ) test evaluates whether a proposed geographic market is too small. Hospital Corp. of America 106 FTC 361 396 (1985). It measures shipments into and out of an area. An appropriate market area must satisfy LIFO (little in from outside) and LOFI (little out from inside). When Professors Elzinga and Hogarty first published the test in 1973, they proposed that 75% or more not shipped in or out shows a "weak" market and 90% or more not shipped in or out shows a "strong " market. Elzinga and Hogarty, The Problem of Geographic Market Delineation in Antimerger Suits Antitrust Bull. 45, 74-75 (1973). They now feel that the 90% test is more accurate. Elzinga and Hogarty, The Problem of Geographic Market Delineations Revisited: The Case of Coal 23 Antitrust Bull. 2 (1978). (Hilke, Tr. 8551.) a. LIFO Dr. Strickland analyzed shipment patterns in the ten-county area using shipment data for 1983 and 1988. Under the E-H test the relevant geographic market is larger than the ten-county area. Shipments into the ten-county area include the following soft drinks: -- Grant-Lydick canned soft drinks produced by the Turner DP plant in Dallas. (F 247.

-- Pepsi COBO canned soft drinks produced at a canning plant close to Houston until 1990 (most of Pepsi COBO' s cans are now produced in San Antonio). (F 248-50.

-- Shasta s soft drinks produced in Houston. (F 251.) -- 7-Up soft drinks produced in Houston. (F 247. -- Kroger s Big K soft drinks produced in Dallas. (F 251.) -- Original New York Seltzer produced outside the ten-county area. (F 269.

-- CCUSA' s fountain syrup produced in Dallas. (F 255. Dr. Strickland testified that 78% of soft drinks sold in the tencounty area in 1983 was produced in that area. (F 260. ) In 1988 the amount was 77%. (F 261.

Initial Decision 118 FTC b. LOn Much of the soft drinks packaged in the ten-county area is shipped outside for sale:

-- CCSW' s San Antonio plant ships throughout its Texas territory. (F 256.

-- Texas Beverage ships HEB' s Plaza brand and other brands to all parts of the state from its plant in San Antonio. (F 257. -- Grant-Lydick supplies its sales centers in Austin, Corpus Christi, and Victoria from its plant in San Antonio. (F 258. -- Pepsi COBO' s three-liter PET bottles are produced in San Antonio and shipped throughout Texas. (F 259. Dr. Strickland testified that 75% of all soft drinks produced in San Antonio in 1983 were sold inside the ten-county area. In 1988 the amount was 57%. (F 260-61.

The ten-county area therefore fails the more accurate and newer version of the E-H Test.

2. Shipping costs Products with low shipping costs relative to price are more likely to be traded in a broader geographic market. General Foods Corp. 103 Frc 204, 232 (1984). Soft drinks are shipped from $.75 to $1.0 per mile, with about 2000 cases per truckload. (F 266.) A 5% increase price would increase the shipping radius by 390 miles. A 10% increase in price would increase it by 780 miles. (F 267.) A price increase in San Antonio could be undercut by shipment from DallasfFort Worth, Austin or Houston. All of these cities are outside of the ten-county area, yet within Pepsi COBO' s franchise territory, and thus are not subject to Pepsi transshipment prohibitions. (Strickland, Tr. 8088.

CCSW ships from its San Antonio plant to Corpus Christi and Temple, about 100 and 150 miles. (F 268. ) Pepsi COBO shipped cans from its Conroe plant to Harlingen, about 260 miles. (F 270. Grant-Lydick purchases cans from Dallas and ships them to San Antonio and from there to Harlingen, a total distance of 500 miles. THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 521 452 Initial Decision 14 (F 269, Warehouse and private brands also are shipped widely. 272) If prices were to increase in the ten-county area, low shipment costs would increase the supply of soft drinks from outside of that area. (Hilke, Tr. 8559.

B. Prices Another factor delineating a geographic market is similarity in prices. Grand Union Co. 102 FTC 812 1041 (1982). Soft drink prices are uniform in a trade area beyond the ten-county area. The HEB stores in CCSW' s and CCE' s terrtories have a leveling effect on prices because of HEB' s preference for the same price throughout its territory. (Chapman, Tr. 7246-47.) Pepsi COBO offers HEB unified pricing throughout its terrtory. CCSW, CCE and Grant-Lydick provide similar prices across HEB' s marketing area. (RX 2985.

Other Market Factors The marketing areas of wholesale purchasers show that the tencounty area is not a realistic geographic market. The largest retailer in CCSW' s territory is HEB. (F 433.) About half of HEB' s stores are in CCSW' s franchise territory. The others are in CCE' s territory adjoining CCSW' s territory. (F 280. ) Except for the area around Corpus Christi and Hal1etsvil1e, HEB' s territory is within the Pepsi COBO franchise area of more than 100 counties. (RX 2; RX 4; F 277.

Kroger s marketing area includes Eastern Louisiana to Western Texas and both San Antonio and Houston (F 281); Albertson marketing area includes 55 stores in North and South Texas, and 12 stores in Louisiana (F 282); Eckerd' s marketing area includes Houston, Beaumont, Corpus Christi, San Antonio and Austin, Texas. (F 283.

14 A company whJC measures trade areas of supermarkets. Sellng Area Marketing. Inc. CSAMI"), indicates that warehouse shipping patterns for supemlarkcls located in San Antonio includes an area of about 50 counties. (F 286.

Initial Decision 118 FTC D. Transshipment Prohibitions Concentrate companies, franchise agreements restrict bottlers from transshipping their national brand soft drinks outside of the franchise territory. (CX 102-G; CX 166- E; CX 418-F.) These market restrictions, while severe, are authorized by statute. They are not, however, completely effective. Transshipment prohibitions do not apply to private labels and to some fountain soft drinks, nor to customers who purchase soft drinks from the bottlers to resell. (F 287.

Despite transshipment prohibitions, soft drinks are shipped, to some extent, between bottlers' franchise territories. Unauthorized transshipments have occurred in the San Antonio market. (F 288-89, 291.) Concentrate companies do seek to restrict bottlers from transshipping. But defiant transshipment indicates that such barriers might be discounted in defining the geographic market. "(T)heoretical concepts must yield to the facts which have persisted in the industry through the years and reflect an industry pattern." United States v. Bethlehem Steel Corp., 168 F. Supp. 576, 599 (S. 1958).

E. Conclusion on Area of Competition The relevant geographic area of effective competition is larger than the ten-county area around San Antonio. The respondent sells in a larger area, and customers turn to a larger area for supplies of competing products.

II COMPETITVE HISTORY The alleged relevant market having failed for lack of proof, no accurate concentration analysis is possible. There is, however, a wealth of proof of competition in respondent s trade. A. Post-Acquisition Evidence Post-acquisition evidence is relevant in a Section 7 case when it is reliable and cannot be manipulated by the respondent. United States v. General Dynamics Corp. 415 U.S. 486, 506 (1974). When THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 523 452 Initial Decision so much time goes by between the acquisition and the trial, business records may be prepared with litigation in mind CX 3806- (the history of respondent' s 1989 attempt to raise prices). When such evidence is uncha1Jenged on cross-examination, or is corroborated however, it must stand regardless of its unnatural clarity and intent. B. Number of Competitors The acquisition did not reduce the number of competitors in the ls Theremarket. The asset acquisition left DP-SA as a viable bottler. was no transfer of any production plant or capacity. The physical assets which were transferred were used and of relatively sma1J 6 The important assets transferred were the Dr Pepper and thevalue. Canada Dry franchises.

In 1982, DP-SA had acquired Big Red bottling Company of San Antonio, an independent bottler. (F 31.) After CCSW obtained the Dr Pepper and Canada Dry franchises in 1984 , Grant-Lydick acquired the DP-SA bottling plant and the rest of DP-SA' s brands. (F 53-55. ) In 1987, Grant-Lydick purchased the Seven-UP distributor reducing the number of soft drink bottlers. (F 63. C. Prices Since 1984 Soft drink prices in San Antonio have declined since 1984. (F 307.) Soft drink prices in Texas are among the lowest in the United States. (F 313- 14.

Concentrate companies profit from increases in bottler sales s 17 pricesvolumes. Pepsi USA reduced its bottling subsidiary order to boost bottling sales volume and market share. (F 318, 320. 15 DP-SA continued as a bottler of a number of products including Big Red and Royal Crown unlil November 1984, when it sold its plant to Grant- Lydick. Grant-Lydick continued and expanded the bottling operations.

16 CCSW purchased approximately 400;0 of DP-SA' s used delivery trucks. Also purchased was a warehouse adjacent to CCSW is hottling facility which DPSA no longer used and which had been listed for sale with a real estate agent for several months. (F 50, 300.) CCSW also purchased 2150 used vending machines, the average age of which was three to five years at the time of the 1984 acquisition. (F 50, 301.) The useful life of the average vending machine is seven to ten years. (F 302. ) The acquisition of these assets had little competitive significance. (Hilke, Tr. 6321-24. The Pepsi bottler in San Antonio, Austin, Houston, Dallas, and much of the rest of the state is Pepsi COBO, which is a wholly-owned subsidiar of Pepsi USA. (F 17- 18. Initial Decision 118 F.T.C Pepsi USA also hoped to increase the sale of its own high profit concentrate. (F 320, 421.) Pepsi COBO is aware that CCSW has financial difficulty and directed its lower prices at CCSW. (F 319; CX 3141-C; RX 2465- Further, costs have been increasing at the bottling level. The costs of concentrate, sweetener, and containers have risen since 1984. (F 314.) The effect of increasing costs and declining prices pushed CCSW to the edge of default on its loan. (F 3 I 0, 321 , 323. Pepsi COBO can incur losses more easily than CCSW. Pepsi can afford low prices. (F 15.) This disparity of size must be considered in assessing competitive effects. "(T)he (Clayton Act) would not impede, for example, a merger between two small companies to enable the combination to compete more effectively with larger corporations dominating the relevant market Brown Shoe Co. v. United States, 370 U.S. 294, 319 (1962).

Another reason for the low prices of soft drinks in CCSW' s trade area is the competition from private labels. Private labels have increased market share in Bexar County (San Antonio) grocery stores from 3.2% in 1981 to 11.6% in 1989 and 18.3% in 1990. (CX 27- RX 2806-X; F 230.) This increase was at the expense of Pepsi and Coca-Cola hrands. (F 222-23.) Pepsi COBO was battling CCSW and private label sales were increasing at HEB and Kroger. (F 230. CCSW attempted to raise list prices in 1987 and in 1989, and was forced to discount prices back to the former levels due to lost sales. In 1989, CCSW raised its list price by $.69 per case, but over the year had a net profit increase of $. 01 per case. (F 419.) CCSW came close to default, and had to refinance. I' Pepsi also unsuccessfully attempted to raise prices in 1989. (F 410. ) Pepsi COBO lost 19% of its Nielsen share during the first seven months of 1989. (F 410.

D. Brand Loyalty Brand loyal" consumers attach a premium to a soft drink brand and are willing to pay more for it. (F 326.) Recently brand loyalty in Texas has eroded due to intense price competition which induces IS, Pepsi USA' s gross profit from the sale ofconccnlrate is approximately 90 to 95%. (Drewes Dep.. ex 3913. pp. 32- 33.

19 CCSW' s profitability has been below that of other Coca-Cola hottlers in recent years. (F 324, THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 525 452 Initial Decision consumers to shop for lower-priced soft drinks. (F 327- 31.) The trend is also due to "brand dilution" caused by the influx into the market of new brands. (F 330. ) The "New Coke" episode shows that consumers easily substitute other brands. (F 332. Consumers in CCSW' s territory are more price- sensitive than elsewhere. (F 458.) Consumers like the economical three-liter package (F 459), and buy private labels and national brands when put on sale.

E. Benefits to Dr Pepper and Grant-Lydick Since the acquisition, sales of Dr Pepper have increased in the San Antonio market, in volume and compared to the sales by neighboring bottlers. (F 361-62.) This contrasts with the decline in sales Dr Pepper experienced when piggybacked wjth Big Red from 1982- 1984. (F 364.

Grant-Lydick has operated profitably since 1984 and has acquired other bottlers. (F 63, 372.) Grant-Lydick increased sales through instore displays while avoiding costly CMA expenses .'o (F 369. Grant-Lydick has outperformed both CCSW and Pepsi COBO and is a "success story" of this marketplace 21 (F 375. F. Impact of the Acquisition A key factor to consider in analyzing whether an acquisition violates Section 7 is the impact of the transaction on customers. FTC v. Great Lakes Chem. Corp. 528 F. Supp. 84, 94-95 (N.D. Ill. 1981). The Commission in Weyerhaeuser Co. 106 FTC 172, 286 (1985), said:

In considering Lanecdotal1 testimony we do find it significant that complaint counsel did not offer any evidence of opposition to the acquisition, either from the integrated box producers without medium mills in the west, or from customers of the box companies. Although lack of customer complaints is not always a reliable indicator of the competitive effect of an acquisition, the fact that the representatives from groups likely to be harmed by any diminution of competition in the western 2D "CMA" is a lump dollar payment to a retail chain which agrees to promote the soft drink. typically over a holiday weekend. (F 172-74. Texas Beverage, the fourth bottler located in San Antonio has grown and also continues to grow. Its sales have expanded substantially over the last seven years. (F 377. Initial Decision 118 FTC market in fact have only testified in support of the acquisition suggests to us, in this case, that Weyerhaeuser s move into North Bend is unlikely to promote collusion. No retailer complained about the transaction. Some felt that CCSW' s acquisition of the Dr Pepper branches benefited competition. HEB felt that the 1984 licensing was a "non-event. " (F 348. A competing third-party vendor, L.c. Vending Co. , complained that its supplier/competitor CCSW kept the price of soft drinks in vending machines down to $. 50. (F 350.) Emery Bodnar of Grant- Lydick complained because ofCCSW' s low prices in competing with Grant-Lydick' s Big Red product. (F 351.) That an acquisition would allow the acquiring company to lower prices and capture market share states no antitrust injury si vigorous price competition is what antitrust laws were designed to S. 104,promote. Cargill, Inc. v. Monfort of Colorado, Inc. 479 U. 115- 16 (1986). The testimony of injury in this case is the wish of two competitors for higher prices.

. POTENTIAL EFFECTS Effective competition in the soft drink industry in this part of Texas rebuts the allegations that interbrand competition is deficient in the relevant market. There is no credible proof that the 1984 acquisition will allow CCSW "to collude, expressly or tacitly, and thereby force prices above or farther above the competitive level." United States v. Rockford Memorial Corp. 898 F.2d 1278, 1283 (7th Cir. 1990).

Even if the relevant market had been more narrowly drawn in this case, the most the evidence shows is high concentration. A high HHI alone "cannot guarantee litigation victories. United States v. Baker Hughes, Inc., 908 F.2d 981 , 992 (D. C. Cir. 1990). Market share alone is not conclusive proof of market power, but may be rebutted by other market considerations. United States v. General Dynamics, 415 U.S. 486, 498-504 (1974).

Competition rather than preservation of rivals is the "lodestar that shall guide the contemporary application of the antitrust laws, not Hospital Corp. of America v. FTC, 807excluding the Clayton Act. F.2d 1381 , 1386 (7th Cir. 1986), cert. denied, 481 U. S. 1038 (1987). Market share cannot supplant a careful analysis of the factors pertinent to predicting future competitive conditions in a market. United States v. Baker Hughes, Inc. 908 F.2d 981, 988 (D. C. Cir. 1990). y, THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 527 452 Initial Decision A. Entry Into This Market In the absence of barriers to entry, an acquisition cannot violate Section 7. F. Goodrich Co. 110 FTC 207, 296 (1988). This case involves an acquisition by a bottler of licenses to be used in soft drink bottling and distribution.

1. Entry as a distributor The cost of equipment Entry as a soft drink distributor is easy.22 and facilities necessary to warehouse and move finished soft drinks is low. (F 379-80.

The cost to lease the delivery trucks and warehouse is about $25, 000. The time to set up as a distributor is about 3 months. (F 380.) A distributor does not need to be a bottler; the excess capacity in Texas allows a distributor to purchase contract-packed bottled and canned soft drinks at low prices and without any capital expenditures for bottling equipment. (F 378.

Numerous non-bottling distributors exist in this market. There are many fountain distributors." (F 90, 93. ) Independent bottle and can distributors actively compete. Approximately 50% of the Pepsi distribution in the Victoria/Corpus Christi area is through independent distributors. (F 146.

Promotional payments paid to retailers can be expensive in sales to food chains.24 However, as Grant-Lydick has demonstrated, instore promotions are available, at no cost other than the discounts granted. The companies which engage in CMA programs spent mightily and have lost money, and the companies with the least promotional cost have been profitable.

Major competitors are able to advertise and promote soft drink products without the necessity of any payment program. Retailers like HEB and Kroger promote and advertise their private label If the prevalent product and geographic market had been found as alleged by complaint counsel entry barriers exist. (F 396.

23 Fountain accounts for 34'1 of al1 CCSW carbonated soft drink sales. (F 154. 24 Both CCSW and Pepsi COBO have spent millions of dollars on CMAs in the last seven years. IF Inn) 25 Grant-Lydick and Texas Beverage have been profitable during the same time period. (F386. Initial Decision 118 F. products without any promotional cost other than the 10w cost of a newspaper ad. These promotions have caused increases in volume and market share. CF 230.) Retailers face no barrier to entry as far as promotional costs are concerned. CHilke, Tr. 6282-83.) The retailer opens as much shelf space as it chooses for its private label products. 2. Entry as a bottler All of the bottlers in the relevant market are operating with excess capacity. CF 133-38.) Each may add a new product to its production line of products and ship it in weeks. CF 398402.) Entry does not depend on the construction of a bottling plant." Expansion of existing capacity to produce is just as effecti ve entry as the construction of new facilities. Weyerhaeuser Co., 106 FTC 172, 287-88 (1985); Grand Union Co., 102 FTC 812 1064 (1983). Since 1984, existing bottlers have added many new products: Coca-Cola Classic. Caffeine-free diet Coke, Cherr Coke, diet Cherry Coke, Minute Maid Orange and Lemon Lime, Pepper Free, Original New York Seltzer natural fruit flavored soda and other seltzers, Lipton Tea Cherry 7-Up and diet Cherry 7-Up, 7-Up Gold, IBC root beer, Pepsi Free and Slice. CF 352.

Used bottling equipment is cheaply available to facilitate entry. Kroger entered the market as a new bottler since 1984 and HEB stands poised to do so. CF 393, 395.) Entry as a bottler is easy, rapid )27 The recent trend in clos-and relatively inexpensive. CF. 393-94. ing bottling plants leaves physical facilities available which indicates Inc., 102 FTCbarriers to entry are not high. CF 175. Dairymen, 1151 1158 (1983).

Economies of scale can easily be achieved in the bottling industry. CF 122.) Kroger and Winn-Dixie have entered the Texas market with very little capital investment. HEB anticipates the expenditure of $2.7 million to erect a canning facility to serve its South Texas area which would rival Texas Beverage s existing plant in efficiency. CF 394.

26 The flavor exclusivity provisions in the hottlers' franchises do not prevent a new flavor from coming into the market. Contract packers such as Texas Beverage, Kroger, Beverage Packers, Better Beverages, and Tumer OP have excess capacity available. (F 137-38. Kroger spent $600,000 - $700,000 and four months to start up the old Safeway plant. (F 393. HEB estimated $2.7 million and 12 months would be required to start up a new production facility. (F390S.

THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 529 452 Initial Decision B. Unilateral Price Increase By CCSW CCSW tried in 1987 and 1989, to raise prices but was forced to lower prices within a short time. (F 408-09.) In 1988, Pepsi COBO' unilateral price increase failed. (F 410.) CCSW, Pepsi COBO Grant-Lydick, and other sellers in the market, have been forced to keep prices low despite rising costs.

Collusion Collusion is a primary concern underlying Section 7. United States v. Rockford Memorial Corp. 898 F.2d 1278 , 1282-83 (7th Cir. 1990). Collusion here is unlikely to occur because of the number of sellers (F 414), varied cost structures and profit incentives (F 416- 28), excess capacity (F 133-39), price competition (F 161), and strong buyers. (F 441 449-52.

1. Competitive conditions a. Competitors Collusion is easier as the number of competitors decreases. FTC v. Elders Grain, Inc. 868 F.2d 901 905 (7th Cir. 1989). The 1984 acquisition left unchanged the number of competitors. (F 295. ) Here there are a large number of competitors. In the ten-county market four DSD companies (CCSW, Pepsi COBO, Grant-Lydick, and the Espinoza companies), two concentrate companies (CCUSA and DPUSA), and dozens of fountain distributors compete. In a larger market which recognizes actual shipment patterns and product substitutability, 13 national brand bottlers (F 414), private label bottlers (F 424-25), and many distributors (F 415), also sell. Collusion in this market is unlikely.

b. Costs and profit incentives The concentrate companies (Pepsi USA, CCUSA, and DPUSA) profit on sales of concentrate (F 416, 419 421); their interest is in keeping bottler prices low to spur retail sales and sales of concentrate to the bottler. Pepsi COBO is a wholly-owned subsidiary of Pepsi USA. (F 17- 18.) Pepsi COBO' s prices sacrifice bottler profits to Initial Decision 118 F. increase sales volume (F 320), which increases the parent s sales of concentrate on which Pepsi USA makes a 95% gross profit. CCSW as an independent bottler, makes no profit from CCUSA' s concentrate sales. Any bottler collusion would be less likely because of Pepsi COBO' s and CCSW's different profit motivations. Grant-Lydick operates with a different cost structure. Unlike CCSW and Pepsi COBO, Grant-Lydick must purchase its cans of soft drinks from an independent packer in Dallas. (F 247.) Grant-Lydick has a greater incentive to keep can prices high relative to other packages which Grant-Lydick produces itself in San Antonio. " cost functions vary widely from one firm to another, each will prefer a different industry price level, and developing a collusive consensus price will consequently be more difficult. F. Goodrich Co., 110 FTC 207, 321 (1988).

In addition, HEB and Kroger, which sell soft drinks to increase store traffic, have little incentive to maintain higher prices on private label soft drinks. Higher-priced soft drinks would be less of a consumer draw, and HEB and Kroger would lose profits from their grocery sales if they were to raise their private label soft drink prices.

The variety of brands, packages, and flavors for soft drinks further complicates the market. (F 73 180-93.) With more variety of relevant products, price collusion is more diffcult. Cj, United States v. Container Corp. of America 393 U.S. 333 , 337 (1969); Hospital Corp. of America v. FTC 807 F.2d 1381 , 1390 (7th Cir. 1986). c. Price competition Prices in this market fluctuate. (F 347.) Only 10% of CCSW' soft drinks sell at list price and 90% is discounted, changing monthly and varying store-to-store. (F 161.) In order to increase volume, especially during holidays, discounts vary. (F 162. Collusion is more likely when prices are relatively steady and change gradually. "Greater stability and predictability make it easier to create and sustain a collusive arrangement. B.F. Goodrich Co., lid FTC 207, 326 (1988). In a volatile market, parties to the collusive agreement can cheat more easily without detection by the others thereby frustrating any collusion.

28 Supermarkets like HEB and Kroger have the incentive to keep prices of all soft drinks low as loss leaders. General Foods Co. 103 FTC 204 , 362 and n. 68 (1984). . . . , .

THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 531 452 Initial Decision d. Ad and display competition Colluders would also have to agree on advertising before any collusive agreement could succeed. Pepsi COBO and CCSW promote their products through CMAs (F 172-74); Grant-Lydick relies on in-store promotions. (F 370.) Private label competitors advertise and rely on in-store promotions and consistently lower price to boost sales. (F 222; Turner, Tr. 1208.) Colluders would have to agree on promotional programs so that volume changes would not disrupt each colluders' profit.

2. Buyers Large retailers have the power and incentive to thwart any collusive agreements made by bottlers. Grocery stores account for 40% of soft drink case sales in San Antonio, (CX 53-I.) HEB sells half of the soft drinks sold through supermarkets in San Antonio. (F 432. Convenience stores account for 15% of soft drink case sales. (CX 53-I.) Large sophisticated buyers deter collusion and price discretion by sellers. FTC v. Elders Grain, Inc. 868 F.2d 901 , 905 (7th Cir. 1989); Hospital Corp. of America v. FTC, 807 F.2d 1381 , 1391 (7th Cir. 1986); B. F. Goodrich Co. 110 FTC 207, 323-24 (1988). Here, HEB controls the most important channel in the soft drink business.29 It and other large retailers assert power over soft drink suppliers. HEB , Kroger and Stop- Go allocate and control bottlers, promotions in ads or point-of-sale displays within the store. (F 445. HEB and Kroger have their own private label soft drinks to supplant national brands on the shelf, in ads, or on displays. (F 188- 446. In the face of a price rise among national brand soft drinks, HEB Kroger, and other retailers who stock private labels could easily promote those brands in place of national brands (F 446.) HEB and other retailers can shatter any collusive agreements to raise soft drink prices. Soft drinks are a favorite loss leader in San Antonio to HEB alone buys more than five null10n cases of soft drinks a year from CCSW. (F 433.) That is 50% of al1 volume in Bexar County, (F 432.29 " HEB demands non-dlscnmmatory pncmg from soft drmk sellers. (F 442. ) This power hasan effect30in detening collusion. Private label soft drinks take volume from the national brands. FOI example, private labels attained an 18% Nielsen share in San Antonio when !feb advertised Plaza two weeks out of nine in a bi-monthly period. (F 230. Initial Decision 118 FTC generate store traffic for the purchase of all grocery items, Retailers watch the national brand prices closely and would quickly spot collusive agreements to raise prices.

3. Concentrate companies Concentrate companies such as CCUSA, Pepsi U. A. and DPUSA have the power and incentive to deter collusion at the bottler level. (F 454-56.) Bottlers lack power in the fountain segment of the market. (F 92. ) Most of the fountain sales of Coca-Cola and Dr Pepper are made on the account of the concentrate companies or by grocery wholesalers and distributors other than CCSW.3I (F 93-94, CCSW services the accounts sold directly by CCUSA and DPUSA but does not set the price or terms for the sale. (F 92.) Much of this market is not subject to control by CCSW. The concentrate companies also fund and arrange for advertising and promotions in selling national brands. (F 456. ) If CCSW were to collude with other bottlers to reduce promotional allowances, not only the retailers but also the concentrate manufacturers would know. Like the retailers, the concentrate companies can deter bottler misconduct by reduction of funding, and even the threat of litigation under the terms of the franchise agreements. (F 166, 457. Soft drink licenses contain best-efforts clauses requiring the bottler vigorously to promote and sell that line of products. (RX 2932- ) The bottler could face nonrenewal of the contract." (F 104 457.) If a bottler wants to sell its business, it must request the concentrate company to approve the purchaser as a new franchisee. (F 101-03.

4. Consumers If consumer demand drops in response to price increases, suppliers are constrained. Soft drink sales are particularly susceptible to price. (Strickland, Tr. 7982-85. ) The sensitivity of soft drinks to ;\ I Usually syrup and carbonated water are mixed after the sale ("post-mix ) at the customer place of business. but some fountain accounts prefer a single container of already mixed beverage ("premix ). (F 73.) The sale of pre-mix is usually governed hy an exclusive franchise. (r 88. ) Coca-Cola and Dr Pepper post- mix is not sold through an exclusive franchise. (F 89. . - Sprite. Tab. Fanta and hesca licenses afc for ten-year terms. (F 98.) CCSW' s Original New York Seltzer distributorship agreement is an at- will license. (F 100. THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 533 452 Initial Decision price, and the growth of national brand soft drinks is due to consumer demand by price promotion. (F 229.

Consumers in San Antonio are particularly price sensitive. (F 459, 461.) The economical three-liter PET bottle sells well (F 460), and private labels went from 3.2% to 18.2% in the Nielsen ratings from 1981 to 1989. (CX 27-W; RX 2806- 5. The Texas Attorney General CCSW signed a consent decree with the Texas Attorney General under which CCSW is constrained competitively. (F 68-69.) CCSW unlike the other competitors in the relevant market, is subject to this decree and to court supervision until 1993, or to 1996 if the decree is extended. Collusion therefore seems unlikely. CONCLUSION AND ORDER Respondent s acquired assets from a competitor in 1984, the most important of which were the franchises for Dr Pepper and Canada Dry for the San Antonio area. The record in this case shows a failure of proof that this transaction may substantially lessen competition. The relevant product and geographic markets are broader than alleged, including private label and other soft drinks in a market which extends well beyond the environs of San Antonio. Further, the market was competitive in 1984 and competition is healthy now, with over capacity and low prices being hallmarks. Respondent lacks market power and collusion appears unlikely. The complaint must, therefore, be dismissed. Opinion 118 FTC OPINION OF THE COMMISSION BY Y AD Commissioner:

L INTRODUCTION' In 1984, Coca-Cola Bottling Company of the Southwest CCSW") acquired the Dr Pepper and Canada Dry bottling franchises for certain areas around and including San Antonio, Texas. Previously, these franchises were held and operated by a so-called third bottler " San Antonio Dr Pepper Bottling Company ("DP- SA"), a wholly-owned subsidiary of Dr Pepper Company. 3 Certain other assets held by DP-SA -- including franchise rights for a regionally distributed branded soft drink, Big Red -- were subsequently acquired by Grant-Lydick Beverage Company ("Grant-Lydick" ), a successor "third bottler" in the market Complaint counsel alleges that this acquisition substantially lessened competition in violation of Section 5 of the Federal Trade Commission Act, 15 U. c. 45, and Section 7 of the Clayton Act, 15 c. 18. The administrative law judge ("AU") who tried the case found that a reduction of competition was unlikely and thus ordered dismissal of the complaint. Complaint counsel now appeals. Our review of this matter is de novo, and our assessment of the evidence differs from that of the AU. We rcverse the initial decision, find violations of the FTC and Clayton Acts resulting from 1 The following abbreviations arc used in this opinion Initial Decision (page no.

IDFF Initial Decision Findings of Fact (paragraph no. CCPFF Complaint Counsel' s Proposed Findings of Fact (paragraph no. RPFF Respondent s Proposed Findings of Fact (paragraph no. RRCCPFF Respondent s Reply to Complaint Counsel' s Proposed Findings of Fact (paragraph no.

CCAPB Complaint Counsel's Appeal Brief (page no. ABR- Answering Brief of Respondent-Appellee (page no. Complaint Counsel and Respondent Exhibits RCX Rejected Complaint Counsel Exhibit Name, Tr. Administrative Hearing Transcript - CX3.

3 Turner, Tr. 918 , 928. 1035. As we noted in Coca-Cola Co., Ok!. No. 9207 (June 28 , 1994), lmJost local markets for carbonated soft drinks have a Coca-Cola bottler, a Pepsi-Cola bottler, and a so-called ' third hottler,' which carries soft drinks other than Coca- Cola or Pepsi-Cola brands," Slip op. at 57.

4 Lydick, Tr. 2978-79.

, THE COCA COLA BOTTLING COMP ANY OF THE SOUTHWEST 535 452 Opinion CCSW' s acquisition of the Dr Pepper franchise, and now enter an franchise. For reasonsorder of divestiture as to the Dr Pepper differing from those of the AU, we find that CCSW' s acquisition of the Canada Dry franchise did not violate the FTC and Clayton Acts. II. THE BACKGROUND OF THE ACQUISITON Respondent CCSW is a privately held corporation with headquarters in San Antonio, Texas CCSW holds the Coca-Cola franchise (among others) for San Antonio and the surrounding area CCSW' s sole shareholder is Texas Bottling Group ("TBG")'; a sister corporation is Southwest Coca-Cola Bottling, Inc. ("SWCC"), which is the Coca-Cola bottler in West Texas, Eastern New Mexico, Western Oklahoma, and parts of Colorado and Kansas. CCSW' s primary business is bottling, distributing, and selling carbonated soft drinks pursuant to franchises from several concentrate companies. mFF paragraph 72. The franchisor grants the franchisee the exclusive right in a specified geographic territory to make and sell soft drinks in bottles and cans bearing the franchisor trademark and using the franchisor s fomula. CCSW sells Coca-Cola 'o Dr Pepper brands, II and Sunkist brands among others.brands 5 ex 980 R.U; RX 549 A.

6 RX 232 A.

7 ex !372 H; ex !373 2-23; RX 572 1. The stock of TBG in turn is held by affliates of Prudential Insurance Company of America, which hold 51 % of the stock, and a 49% stockholder, the Coca-Cola Bottling Group (Southwest), Inc. ("CCBG-Texas ), which in turn is a wholly-owned subsidiary of the Coca-Cola Botting Group. Inc. ("CCBG-De\aware ). Hoffman, Tr. 5603; ex 1372 G, H. All of the voting stock of CCBG-Delawarc is held by Edmund M. Hoffman and his son Raben K. Hoffman. RX 572 H; RX 2805 J, K, Z 15. At the time of the acquisition at issue in this case. CCSW was held by the Biedcnhom COf1oration. RX 232 A-C, which sold its interest in CCSW to a TEG subsidiary in 1986. CX 3052; RX 549 A. B.

8 SWCC is a wholly-owned subsidiary ofCCBG-Texas, which is controlled by CCBG-Delaware which is owned by the Hoffmans. ex 4; CX 2805 23. 24; RX 2805 25 , Z6. 9 RX 2848 10 Coca-Cola USA ("CCUSA") is the division of the Coca-Cola Company that manages domestic soft drink operations and produces the concentrate that CCSW purchases to make Coca-Cola soft drinks. Howe!1. Tr. 4004; Atchison, Tr. 5237-38. The Coca-Cola Company also owns 49% of the stock of Coca-Cola Enterprises ("CCE"), which owns Coca-Cola bottling operations in various parts of the United States, i!1c1uding Dal\asfFort Worth. Houston, and Austin, Texas. Howe!!, Tr. 4002-07; RX 3 \3) As noted above, the Dr Pepper franchises were previously held by San AntonlO Dr Pepper Tr. 918 Bottling11Company ("DP-SA"), a wholly-owned- subsidiary of Dr Pepper Company. Turner. 928, 1035. The Dr Pepper Company was a publicly held corporation until J984, when it was bought Opinion 118 FTC The practice of having a single bottler licensed by each of several concentrate companies to sell their brands of soft drinks is sometimes called "piggybacking.

Prior to CCSW' s acquisition of the Dr Pepper franchise, the franchise was held by DP-SA, a wholly-owned subsidiary of DPUSA, Until 1984, DPUSA owned bottling operations in San Antonio, as well as in Dal1asfFort Worth, Waco, Houston, and Corpus Christi 5 After DPUSA was bought in a leveraged buyout, its acquir-Texas. , Forstmann-Little, began selling off the DPUSA company-owned bottling plants and the Canada Dry business. " CCSW wanted the San Antonio area franchises for Dr Pepper and Canada Dry, but had no interest in DP-SA' s main production facility, the former Big Red 17 Although DPUSA initial-Bottling Company of San Antonio plant. 18 it eventuallyly wanted to sell the operation as a whole sold the operation in two partsl9 CCSW bid on both the Dr Pepper and Canada Dry franchises, initially offering $5 million, but subsequently On August 28increasing its offer to $14. 5 million.2() , 1984, CCSW bought the Dr Pepper and Canada Dry franchises, along with other assets, from DP-SA for $14.5 million.

in a leveraged buyout. ex 614 B; RX 1447 D; RX 990 E. N. After some assets were divested. the Dr Pepper franchise contracts and syrup manufacturing facilities were sold to an investment group. which subsequently bought the Seven- Cp Company. Knowles, Tr. 2640. The current franchiser of Dr Pepper bottling operations in the United States is Dr Pepper/Seven-Up Companies, Inc. ("DPUSA" ). The Dr Pepper and Seven-Up companies were combined in 1988. RX 1989 at 3-4. Dr Pepper/Seven-Up owns the trademarks and manufactures the concentrates for Dr Pepper and Sevcn-Cp brand products. Clarke Tr. 4297-99; Knowles, Tr. 2638-41. For convenience, we will use "DPUSA" to refer La both the Dr Pepper Company and its successor, Dr PeppcrlSeven-Cp Companies, Inc. 12 RX 293J; E. Hoffman, Tr. 507-09.

13 Knowles, Tr. 2764- 67. 14 RX 1648 Z29; Turner, Tr. 917- 18; Antle, Tr. 304l. 15 RX 1648 Z29-Z31; Turner, Tr. 916; Antle. Tr. 3041. 3079 16 CX 3817: Turner, Tr. 920- 17 Antle, Tr. 3059.

18 Antle, Tr. 3059.

19 Turner, Tr. 1152.

2() CX 3; RX 2092 F; Turner, Tr. 1158.

21 CX 3 at 7; CX 247 C; CX 270; RX 1292 at 1; CX 1662; CX 253. In the same transaction DPUSA agreed to issue Dr Pepper license agreements to CCSW ex 3 at 17- 18. DPUSA and Canada Dry issued new franchise agreements for the Dr Pepper and Canada Dry brands to CCSW in 1984. CX 266, CX 267. CCSW also received a warehouse, 2150 used vending rnachines, and 40% of DP- SA' used delivery and over-the-road trucks. Bodnar, Tr. 1276 , 1518- , 1689; Little, Tr. 653. THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 537 452 Opinion After the sale, DP-SA still owned the franchises for Big Red, RC, Crush, and Hires, and various equipment including the DP-SA bottling plant." DP-SA continued to operate its business as Big Red Bottling Company of San Antonio until DPUSA' s assets were sold to Grant-Lydick in October, 1984. Grant-Lydick obtained DP-SA' franchises to produce and sell Big Red, RC, Crush, Hires, and DP- SA' s other remaining brands, which accounted for about 58% of DP- SA' s 1983 sales volume. Grant-Lydick also hired DP-SA' s manag- , Emery Bodnar, to run its business, as well as about half of DP- SA' s other employees.

Grant-Lydick operates its soft drink business in San Antonio as the Big Red Bottling Company of San Antonio 26 and has subsequently acquired additional soft drink brands and new geographic territories. In 1987, Grant-Lydick acquired the Seven-Up bottler in San Antonio and Austin, as well as the Seven-Up bottler in Corpus Christi." In 1988, Grant-Lydick purchased the assets of Big Red Bottling Company of Austin, and, in 1990, an RC Cola distributorship in La Grange, Texas.

The other major branded carbonated soft drink ("CSD") bottler in San Antonio is the Pepsi COBO (Company-Owned Bottling Operation), owned by the Pepsi-Cola Company ("Pepsi USA" Pepsi USA also owns bottling operations in various parts of the United States, including San Antonio, Houston, DallaslFort Worth and Austin, Texas. These company-owned bottling operations ac- - ex 237; Bodnar, Tr. 1668.

ex 20S2;CX 2484;CX 3254A; ex 237 C: RX 1663; RX 2408; RX 2409; Lydick, Tr. 2981- 82; RX 1648.

ex 3495: ex 3504; ex 3505; Knowles, Tr. 2874 25 Bodnar, Tr. 1223, 1294.

26 Bodnar, Tr. 1581.

27 RX 2970: Bodnar, Tr. 1334-36: Lydick, Tr. 2999-3000. From 1982 to Januar, 1986, the 7franchise was held by the Seven-Up Bouling Company of San Antonio, which was owned by Seven- USA. RX 2002; Lydick. Tr. 2996-97. The franchise was then held by Texas Bottlers, Inc. ("TBI") until May 1987 , when G-L purchased TBL Bodnar, Tr. 1334 28 Lydick, Tr. 3002- , 3005-06.

29 Pepsi USA is a division of Pepsico, Inc., which owns the united States trademark and produces concentrate for Pepsi-Cola and other brands of soft drinks RX 2864 Z34; Davis, Tr. 4463 4638; Amrosowicz, Tr. 787.

538 FEDERAL TRADE COMMISSION DECISIO:-S Opinion 118 FTC count for about 37% of Pepsi USA bottle and can sales. 3D In addition there is a small, branded CSD distributor, Star Distributing, that has undergone three corporate restructurings in the last three years. II THE HISTORY OF THE PROCEEDING The Commssion s complaint in this matter was issued on August 1988 , and was amended on November 18 1988. Administrative hearings on the merits began before Administrative Law Judge James P. Timony on July 10, 1990. The hearings on the merits were concluded on October 3, 1990. IDFF paragraph 44. On June 14, 1991 , the AU issued his opinion, finding a failure of proof that CCSW' s acquisition of the Dr Pepper and Canada Dry franchises may substantially lessen competition. He found that "(tJhe relevant product and geographic markets are broader than alleged including private label and other soft drinks in a market that extends well beyond the environs of San Antonio." ID 78. He found further that "Ir Jespondent lacks market power and collusion appears unlikelyl,)" and that "the market was competitive in 1984 and competition is healthy now, with over capacity I sicj and low prices being hallmarks. ld.

For the reasons set forth below, we reverse the ALl's findings as to the relevant product and geographic markets and as to the likely competitive effects of CCSW' s acquisition of the Dr Pepper and Canada Dry franchises, and find that CCSW' s acquisition of the Dr Pepper franchise constitutes a violation of the FTC and Clayton Acts. Although we agree with the AU that CCSW' s acquisition of the Canada Dry franchise did not violate the FTC and Clayton Acts, we reach our conclusion based on factual findings and legal reasoning that differs from that of the AU.

IV. THE RELEVANT PRODUCT MARKET Bottlers may selllo retailers a variety of beverages, ranging from nationally known, branded CSDs to non-branded CSDs, non-carbon- 30 RX l::lS; Amrosowicl., Tr. 791- 93, 837-38. Espinoza, Tr. 4166-67: Bodnar, Tr. 1559- , 1713. THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 539 452 Opinion ated soft drinks, seltzers, juices, and even iced tea drinks. Here, the franchises that were transferred were those of branded CSDs: Dr 33 The issue is whether the relevant product Pepper and Canada Dry. market is confined to branded CSDs or conversely includes certain beverages in addition to branded CSDs. As we explain in detail below, we define "branded CSDs" as widely available carbonated soft drinks distributed by direct-store-door delivery and heavily promoted by concentrate companies, bottlers, and retailers. "Private label" carbonated soft drinks are less heavily promoted and are available in fewer channels of distribution since they are sold by retail chains that own the trademark. "Warehouse" carbonated soft drinks use warehouse delivery, are less heavily promoted, and are also available in fewer channels of distribution. See Section IV. C. infra. Complaint counsel has asserted that all branded CSDs comprise the relevant product market CCAPB at J 2. By contrast, CCSW has claimed that the relevant product market consists of all carbonated soft drinks (including private label and warehouse brands) and certain non-carbonated soft drinks packaged and sold in the same manner as CSDS. RPFF paragraph 348. The AU found that the rele.vant product market includes "national brand, private label and warehouse hrands of soft drinks, as well as mixers, seltzers, noncarbonated beverages such as Lipton Iced Tea, Country Time Lemonade, and Hawaiian Punch, and isotonic drinks." ID 62. For the reasons set forth below, we find that the evidence in this case supports a relevant product market consisting of branded CSDs. 32 For example, CCSW at various times has distributed in the San Amonio area the following heverages: Coke (and allied brands, such as Sprite, Frcsca, and \1r. PrEB). Sunkist, A & W (and allied brands). Welch' s Grape and Strawberr, Cima Red, Minute Maid CSDs, Old New York Seltzer, Spike. Hawaiian Punch, Delaware Punch, Lipton Iced Tea, and Country Time Lemonade. ex 3489 Z29. Z 1 0- !3; ex 3483 R Q: Summers. Tr. 6581.

As a result 0 acqumng the franchrses, CCSW added the following Dr Pepper and Canada Dry products to its list of brands for sale and distribution in the San Antonio area' Dr Peppcr products: Dr Pepper, Sugar Free Dr Pepper, Pepper Free, Sugar Free Pepper Free. CX3 at 396.

Canada Dry products: Ginger Ale, Sugar Free Ginger Ale, Club Soda, Tonic Water, Sugar Free Tonic Water, Seltzer, Collins Mixer. CXJ at 397. Complaint counsel presented evidence that the relevant product market IS "the manufacture,distribution,34and sale offinished carbonated soft drinks (or syrups) produccd from the concentrates of widely-advertised, branded, carbonated soft drinks, merchandised and distributed by direcHtore-door delivery, in all channels of distribution. " IDFF; see Hilke, Tr. 5944-86. , .

Opinion 118 FTC A. The Legal Standard for Defining the Relevant Product Market The purpose of defining a relevant market is to identify a market in which market power might be exercised and competition thereby diminished. HJ. Inc. v. l Tel. Tel. Corp. 867 F.2d 1531 1537 (8th Cir. 1989). Product markets may be defined either by "the reasonable interchangeability of use or the cross-elasticity of demand. Brown Shoe Co. v. United States 370 U.S. 294, 325 (1962). To assess whether market power might be exercised, the courts and the antitrust enforcement agencies have sought to define a market in which "sellers, if unified by a hypothetical cartel or merger, could raise prices significantly above the competitive level" HI. Inc., 867 2d at 1537. Under the Merger Guidelines 35 the federal antitrust agencies seek to identify a product market as a "product or group of products such that a hypothetical profit-maximizing firm that was the only present and future seller of those products ("monopolist ) likely would impose at least a ' small but significant and nontransitory' increase in price. " Merger Guidelines, Section J . 11. 36 This inquiry focuses on whether other products are sufficiently substitutable that customers would turn to them in the event of a "small but significant and non transitory" price increase by the hypothetical monopolist. At the point at which other products are not substitutable in that sense the contours of a relevant product market have been defined. Because a "small but significant and nontransitory" price increase is generally interpreted to be 5%, this test is known as the "5% test. Merger Guidelines, Section 1. 11.

In Beatrice Foods Co. 101 FTC 733, 801 (1983), the Commission stated that "cross-elasticity of demand (is) the most important factor in product market definition. Although the Commission considers all reliable evidence of interchangeability, Olin Corp. , 113 FTC 400, 594-95 (1990), the Commission has recognized the utility of evidence of cross-elasticity of demand such as the 5% test is designed to elicit, despite some of the difficulties in calculating such 35 United Slates Department oj Justice and Federal Trade Commission Hori;:rilfo/ .Iverger Glide/illes, reprinted il1 4 Trade Reg. Rep. (CCH) paragraph 13. 104 (Apr. 2, 1992) ("Merger Guidelines The version of the Merger GUIdelines that was generally used by both enforcement agencies at the time of the AU' s decision. Uniled Srmes Deparrmenl of ils/ice lWerger Guidelines, repril1ed in 4 Trade Reg. Rep. (CCH) paragraph 13 103 (June 14 , 1984) (" 1984 Guidelines ). uses essentially (he same methodology for product market definition as the 1992 Merger Guidelines. Coca-Cola Co., slip op. at 26 n.50, THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 541 452 Opinion elasticities. Coca Cola Co., slip op. at 27- 29; see also Merger Guidelines, Section 1. 11. The AU here, however, failed to discuss any of the testimony relating to the 5% test. The testimony in this case is undisputed that bottlers of branded CSDs in the San Antonio area could profitably raise prices more than 5%.37 Moreover, the . weight of the other evidence relevant to this issue -- including the opinions of market participants, historical evidence of price interactions, and industry business records -- also supports a product market limited to branded CSDs.

The ALl's narrow focus on certain selected pieces of evidence concerning industry perception, characteristics of the product, and price movements failed to give an accurate and complete picture of the relevant product market. E.g. 1D 60-62. We find that the AU erred in asking only whether certain beverages "competed" against each other in a broad sense, without focusing on which products were suffciently substitutable that they could constrain a small but significant, nontransitory price increase. For example, the ALJ implies that an inverse relationship between branded and non-branded CSD market shares shows that they are in the same product market. ID 60. That this alone is an insufficient basis on which to reach such a conclusion is easily ilustrated by considering the case of two different product markets that are arbitrarily lumped together to calculate shares, such as two unrelated products: branded CSDs and mouthwash. Assuming that mouthwash sales are stable throughout the year, an increase in branded CSD sales (because of feature activity with consumers stocking-up on favorite brands or seasonal swings in consumption) will produce a share increase for branded CSDs and a share decrease for mouthwash. However, this inverse relationship provides no reasonable basis for claiming that branded CSDs are in the same product market as mouthwash; rather, it is an artifact of arbitrarily treating the unrelated products as though they are in the same market.

Moreover, even if branded CSD price increases produced some consumer switching to non-branded CSDS, that would not establish that both products are in the same antitrust product market. The key to product market definition is not whether some consumers will switch to other products in the event of some price increase. Unless demand for a product is perfectly inelastic, some consumers will See Section IV, D. Infra.

Opinion 118 FTC switch in response to a minimal price increase. Rather, the question is which beverages are sufficiently substitutable that they could constrain i.e. make unprofitable, a price increase in the relevant market. The evidence here estabEshes that consumers wil not switch to other products in the event of a small but significant, non transitory price increase of branded CSDs in sufficient numbers to make such a price increase unprofitable.

B. The Concentrate and Carbonated Soft Drink Industry In order to assess the extent to which branded CSDs face competition from other beverages, it is necessary to understand some aspects of the soft drink industry. Soft drinks are produced combining concentrate, sweetener, and carbonated or still water. Concentrate" includes the flavors, extracts, and essences used to produce soft drinks. "Syrup" is concentrate mixed with sweetener and some water. mFF paragraph 74.

Bottlers purchase concentrate from concentrate companies, such as CCUSA, DPUSA, and Pepsico, Inc. ("Pepsi"). mFF paragraphs , 16 29. Bottlers generally sell soft drinks to retailers in cans, glass, and plastic (PET) containers; retailers in turn sell the finished soft drinks to consumers. IDFF paragraph 140. Concentrate companyies, bottlers, and wholesale grocery suppliers sell soft drinks to fountain outlets in ready to drink form ("pre-mix ) or as a syrup that must be mixed with carbonated water ("post-mix ). mFF paragraph 73. The record in this case estabEshes that soft drinks are differentiated products." One obvious difference among soft drinks involves flavors, such as colas, lemon/times, and oranges. However, in addition to flavor differences, soft drinks are also differentiated in other, less obvious ways. For instance, there are differences among soft drinks as to the image that their advertising projects to 38 The Commission also recently found this to be the case in Coca- Cola Co., slip op.at 30.

, , , .

THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 543 452 Opinion 39 and even whether the soft drink is advertised signifi-consumers cantly at all.

There are also differences among soft drinks as to their availability in either the "take home" distribution channel (cans and bottles to be consumed later) or the "cold drink" distribution channel (chilled soft drinks, usually sold for immediate consumption fountain ) or dispensed by vending machines ("vending ) through convenience stores and restaurants). Soft drinks that are available through fountain or vending outlets are typically branded CSDs that use "direct-store-door" or "DSD" delivery'" or are private label CSDs of the outlet itself (such as McDonald' s private soft drink brands).42 Warehouse and private label brands are generally not available in the cold drink channel In the "take-home" distribution channel, soft drinks also may be differentiated by the services that the bottler provides to the retailers, such as grocery and convenience stores. Typically, bottlers provide only delivery to the retailer s central warehouse for private label and warehouse brand soft drinks, whereas bottlers provide DSD delivery for branded soft drinks such as Coke and Pepsi. See Section IV.C.2 infra. The in-store merchandising" by the bottlers' own employees in DSD delivery provides advantages generally not available through 39 Mr. Carew, Vice President for Planning of CCE, which owns Coca-Cola bottling operations in various parts of the United states, testified that "soft drink service is called a necktie product. They are sold on image. If you have any success, you have built an image up. " ex 3967 at 205-06. Mr. Carew testified that brands that have the kind of consumer demand that allow them to " sit back and do nothing" for a long time while "selling off share" include Coca-Cob, Pepsico, Dr Pepper Company, Seven Up Company and Royal Crown. ex 3967 at 205. Most pnvate label brands arc not advertised on tclc:vision or radio, but may appear In the retailer s newspaper ads or circulars. Turner, Tr. 1208; Summers. Tr. 6546-47; Hmvell, Tr. 4025: Hixon Tr. 7344. Some warehouse brands, notably Shasta, had engaged in television and radio advertising at one time, although Shasta now markets itseJfmore as a private label brand. Chupman. Tr. 7171-72. By contrast, concentrate Finns allocate millions of dollars annually toward acquiring, improving, securing, protecting, and capitalizing on the value of trademark equity they develop for their trademark names and brandedCSD products. Summers, Tr. 6523, 6547-48. Branded CSD bottlers and their concentrate firms realize that it is important to manage and protect the equity of the brand. Knowles, Tr. lR02 CX 3915 at 29 (Clements); Amrosowicz. Tr. 89!; Summers, Tr. 6547-48, 6523 41 CX 3989 at65-66lShanksJ.

42 Summers. Tr. 6517: Short, Tr. 7759-60. 41See Section IV infra. 44 "Merchandising" the product includes the tasks of placing the product on the shelves or other displays fronting" the product to ensure the label is facing forward and, if necessary, individually pricing the product rotating" the product to remove older, out- of-date merchandise from the shelves and ensuring that the price and other merchandising signs (called "point of sale" or " POS" ) are adequately displayed. Coyne, Tr. 3439-41; ex 21610 , E. . .

Opinion 118 FTC warehouse delivery, such as: (a) ensuring the visual impact of trade- 45 (b) ensuring marked brands quality control of damaged or out-ofdate stock 46 (c) maintaining shelf space:' (d) facilitating responsiveness to competitive situations 48 (e) maintaining and promoting a full stock ofproduct 49 (f) maintaining a good relationship with the retail account.

A review of the evidence shows that soft drinks are divided into at least three distinct categories: major national and regional brands; warehouse" brands; and private label brands. Major national and regional brands are characterized by: wide availability in both the DSD delivery; take home and cold drink distribution channels;52 and heavy advertising to promote a particular image and trademark. For convenience, we wil refer to these as "branded CSDs. The remaining soft drinks consist of those that have brand names but use warehouse distribution ("warehouse brands ), such as Shasta and Faygo 55 and private label products, such as H. B.'s Plaza, that 45 ex 505 c; ex 194H; ex 2240 D-ZI3; ex 2243 0- 46 ex 505 D, G, H; ex 2240 F, 0 , z; ex 2243 1-217. 4' ex 505 E. Alt DUg the shelf space t at retatlcrs allocate to thelr own private labels may be considered "untouchable " (Summers, Tr. 6624; Davis. Tr. 4526, 4764; Bodnar, Tr. 1763; Howell, Tr. 4050: Sendelbach, Tr. 7718), bonlers still compete among themselves for shelf space not allocated to the retailer s own private label. Summers, Tr. 7119. 48 ex 505 E.

49 ex 50S . 1-1,0; ex 2240; ex 2240; ex 2627 Y-ZlO; Summers. Tr. 7119. 50 ex 505 E.

51 In Coca- Cola Co., this Commission reached the same conclusion. Slip op. at 30-32. 52 Donald, Tr. 5291; RX 990 E.

See a/so Section IV infra. There was also testimony that to have a fully effective merchandising uperation, carbonated soft drinks must be distributed in aji channels of distribution. Turner, Tr. 934; CX 3915 at 17- 18 lCJementsJ; CX 3988 at 530-531 (O' Donnelly; CX 1853 N; CX 1909.

53 See Section IV C.2 infra.

S4 For example, the trademark "Coca-Cola" is "the most widely known brand name in the world. CX 131 D. Concentrate firms typically make available marketing support to local branded CSD bonJers. CX 3989 at 78-79, 104lShanksJ; CX 3987 at 2085 lLowenkronJ: CX 3976 at 2129 (QuirkJ; Coyne, Tr. 3413- 17; Knowles, Tr. 2745-49; Trebilcock, Tr. 5812; Turner, Tr. 963-65; Howell, Tr. 3928-31; RX 990 RX IS31; RX 1957: Howell, Tr. 403!: Summers, Tr. 6551. Other examples include' mc Root Beer, ex 1294; Rainbow, Rocky Top and Parade. Hiller, Tr. 5337-38; Hoffman, R., Tr. 5534-35. 55 This category includes the proprietary brand name products produced by bottlers, such as the "Texas brand of Texas Beverage Packers. Hixon, Tr. 7277- . . ..

THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 545 452 Opinion are sold by the particular store chains that own the trademark. 56 Warehouse brands are available primarily in large retail chains; are generally not available in the cold drink channel; are less heavily advertised than major national and regional brands; and are less The private label products are also expensive than branded CSDs.60 not usually available in the cold-drink channel;" use little or no advertising;" and are even less expensive than warehouse brands. For convenience, this opinion wil refer to warehouse and private label brands collectively as "unbranded" or "non branded" products. C. The Distribution and Marketing of Branded Carbonated Soft Drinks I. Channels of Distribution Soft drinks are sold through various "channels" of distribution. One broad distinction is between the "home market" or "take home channel, which consists of sales for later consumption, and the "cold drink" channel, which consists of sales for immediate consumption 56 ex 4022. Private label soft drinks are usually proprietary brand names of retail chains. Hixon, Tr. 7278-79. See also Morath, Tr. 7674-75; Howell, Tr. 4023-24; Knowles, Tr. 2860-6J. The A.c. Nielsen Company ("Nielsen ) tracks sales in the home market segments of the bottling market, including sales to supermarkc!s and convenience stores. RX 875. Nielsen refers to private label brands as "control" brands. RX 2H06 X.

57 See Section IV C.2 infra.

See Section IV. infra.

Although one warehouse brand, Shasta, has engaged In television and radio advertlsmg, most do not. See a/so Section IV C.3 infra.(Chapman,59Tr. 7171-721. See Section IV infra.

See Section IV infra; ex 3989 at 65-66. Most pnvate label brands arc not advertised on television or radio, hut may appear In the retailer s newspaper ads or circulars. Turner, Tr. 1208; Summers, Tr. 6546-47; Howell, Tr. 4025. 63 See Section IVD. infra. In addition, private label soft drinks arc available in many fewer package sizes than branded eSDs. Branded eSDs come in a variety of package sizes, including 6.5, 10, , J 6 , 20 or 32 ounce glass or PET bOllles, 1. 2 and 3 liter PET bonles. and 12 oz cans. ex 53 G, Y- Z6. Typically, private label eSDs arc sold in J2 ounce cans and 2 and 3 liter PET bottles. ex 3J58 E.B.'s Plaza is available only in loose cans and 2 liter bollies. Chapman, Tr. 7165; CX 4022. Warehouse-delivered eSDs are also limited in their package availability. Hixon, Tr. 7279 , 7285- 7300. 7342.

CX 836 H, S.

65 Knowles, Tr. 2647-48; Turner. Tr. J 185- 86; ex 4 18 J, K. Opinion 118 F. The take-home channel is primarily served by chain supermarkets and independent grocery stores, mass merchandisers, and conven- The cold drink channel is served by stores and otherience stores.66 locations that offer (a) vending sales, (b) fountain sales, andjor (c) single drink sales Concentrate companies and bottlers recognize significant differences between the take-home and cold-drink channels. As described in a 1985 CCSW "Corporate information Memorandum Almost all Coca-Cola bottlers divide their business into two broad categories the home market and the cold drink market. The home market consists of all soft drinks which are sold for consumption at some place other than where they are purchased - hence for "home" consumption. The major types of outlets which comprise the home market are supermarket chain stores, mass merchandisers and discount stores, drug stores, independent supermarkets, and convenience stores. The coJd drink market segment is composed of those outlets where soft drinks are purchased for immediate consumption: vending machines, restaurants and bars athletic and other social events, and convenience stores. It is obvious that almost alj cold drink accounts require some fonn of special equipment since the product must he delivered cold, while home market accounts generally sell soft drinks off the shelf or possibly off of a special rack. Soft drinks are sold in different packages in different market channels. In the home market, soft drinks are sold in bottles and cans. In the cold drink market product is sold in bottles, cans, and cups. Approximately 76% of a11 soft drinks are sold in bottles and cans. The remaining 24% are sold in cups or similar containers. Cups are filled using either a post-mix or pre-mix system. Pre-mix, which is the same as the product in bottles and cans, and accounts for only 18% of cup sales today, is distributed in five gallon metal tanks. It is pumped out under pressure and is used primarily where no local water hook-up is available. Post-mix is also distributed in five gallon tanks, as well as one gallon jugs. It is very similar to bottling syrup and must be mixed with carbonated water at the point of serving. CX 418 J, K.

1n addition to these differences, there are other significant differences between the take-home and the cold-drink channels, especially the fountain portion of the cold-drink channel. For example, both CCUSA and DPUSA handle fountain sales differently than sales of take-home, branded CSDs in that CCUSA and DPUSA -- not bottlers - set the price at which a large proportion of Coca-Cola and Dr Pepper fountain sales are made. Large fountain accounts qualify for 66 ex 883 V; RX 990 U; ex 418 J . K.

67 ex 783 E; ex 3419 Z56; RX 990 C.

THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 547 452 Opinion national account pricing" from both CCUSA and DPUSA,68 About 65-70% ofCCSW' s sales of post-mix fountain syrup are made at the national account price.

In addition, CCUSA and DPUSA do not have exclusive franchise terrtories for post-mix fountain syrup, although DPUSA does restrict each bottler s sales of post-mix fountain syrup to its specified terri- This means that Coca- tory for bottle and can sales.'o Cola and Dr Pepper post-mix fountain syrup can be sold by a variety of entities such as wholesalers, in addition to concentrate companies and bottlers. As a result, Coca-Cola and Dr Pepper fountain products are available from many fountain wholesalers in the San Antonio area in addition to the two franchised bottlers." Indeed, Mr. Carew, Vice President for Planning of CCE, the owner of Coca-Cola s bottling operations, described the marketing of post-mix fountain syrup as " totally different from bottle/can marketing that efforts to merge the two are not in the best long term interest of either system. Finally, there are often significant price differences between the take-home and cold-drink channels. For example, an individual branded CSD can is typically $.50 in a vending machine in the San ByAntonio arean contrast, a six-pack of Pepsi take-home cans in San Antonio sells at an everyday price of $1.99 and may be sold at a promotional price of $1.49 or even $.99 on occasion. 2. Direct-Store-Door Delivery Nationally and regionally branded CSD manufacturers overwhelmingly use "direct-store-door" ("DSD") delivery for their prod- 68 Short, Tr. 7736; Cassagne, Tr. 7585; Knowles, Tr. 2820- 69 Knowles, Tr. 2820.

70 Knowles, Tr. 2681; Turner, Tr. 1086 DPL'SA docs allow post-mix fountain to be distributed by food wholesalers and brokers within a DottIer's exclusive territory. Turner, Tr. 1086. By contrast, Pepsico and RC Cola.QQ have exclusive geographic territories for post-mix fountain syrup. Knowles, Tr. 2681- 82.

71 CX33Z!8;RX861;Short Tr.774!-42;Turner Tr 1172-74. 72 ex 799 M.

73 Turner, Tr. 646.

74 ex 3973: Davis, Tr. 4526.

, Opinion 118 F. ucts as opposed to warehouse delivery." For DSD delivery, the bottler s own employees will: place the product on the shelf front it to make sure that the label is properly displayed, and price the signsproduct; remove old merchandise; ensure that "point of sale" For warehouseare properly displayed; and change space allocation77 delivery, the bottler relies on the retailer s employees to perform these tasks. In such circumstances, the private label and warehouse soft drinks are delivered to the retailer s warehouse Under DSD delivery, the DSD vendor bears the cost of distribution, stocking, and in-store checks on promotional efforts past the point of the warehouse; in the warehouse delivery sequence, this cost Distribution costs typically account foris borne by the retailer80 about 35% of a branded CSD bottler s overall costs. The DSD delivery system provides at least two strengths which justify its added expense to the bottlers. First, it allows bottler control. Second, given sufficient overall volume, the DSD delivery system allows the bottler to reach smaller outlets. 75 All of the major carbonated soft drink brands are distributed by DSD distribution, using soft drink bottkrs or soft drink distributors. ex 3967 at 18! . ex 3976 at 211 J; ex 3582 at 2238. Nicholson, Tr. 3713. In fact. major franchises prohibit warehouse distribution. Turner, Tr. 956: Koch Tr. 1814. The Coca-Cola Company s soft drink products arc distributed entirely by Coca-Cola franchised bottlers through DSD delivery. No warehouse delivery is used for retail channels. ex 3967 at 181 (CarewJ: ex 793 A. Pepsi built its business on the merchandising advantages of DSD distribution. David Davis. Vice President for Trade Development for Pepsi USA. testified that Pepsi had better control of where its products went and how to merchandise them and move business by keeping itself vertically integrated. Davis . Tr. 4471-72. Consequently. Pepsi has not explored warehouse and beer distributors as an alternative to DSD distribution. Davis. Tr. 4471-72. 76 Turner, Tr. 956-57; Nicholson, Tr. 3713- 14; CX 3582 at 2238 (Clements). The two largest systems of DSD delivery are the Coke bottler system and the Pepsi bouler system. CX 3976 at 2128 (QuirkJ; CX 3978 at 2066-67 fLowenkronJ; CX 3990 at 929 lKalilJ; CX 864 H , 1. 77 Knowles, Tr. 2662-63; Turner, Tr. 956-58; Nicholson, Tr. 3711; CX 3989 at 27 (Shanks); CX 3988 at 505 (O' Donnelly: CX 3921 at 355lCurrieJ; Hoffman, E. , Tr. 327-28. 78 Knowles, Tr. 2663-64. In some cases, hQttlers have relied on independent distributors to perform DSD distribution for them. Koch, Tr. 1901. It is significant that the bottlers have hired independent distributors to ensure that these tasks are performed, rather than relying on retailers employees for them.

79 Warehouse delivery is used for retailers' private labels (also known as " control brands Private labels are a retailer s proprietary brand of soft drink. Howell. Tr. 403!; E. Hoffman, Tr, 412- 13. There are a few national brands"- Shasta, Faygo, and mCRoot Beer -- that also use v.' warehouse delivery. Howell, Tr. 4031. mc Root Beer, produced by DPUSA, uses warehouse delivery among other reasons because of its unique bottle. IDFF paragraph J 49 80 Summers, Tr. 6469.

81 See RX 0867 . . .

THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 549 452 Opinion Bottler control means that the bottler has someone in the store This marketing push is extremely important pushing the brand,82 given the degree to which sales respond to advertising, promotions It also gives the bottler the displays, and price (see next section),83 ability to get the product merchandised, priced, rotated " and looking fresh. A bottler would lose this with the warehouse delivery system.85 Bottlers characterized the services performed by bottler employees in DSD as extremely important in producing volume sales of soft drinks. Toby Summers, President of CCSW, stated: "We are an impulse item. If you don t have a display to execute it, you can sell it off the shelf" Summers, Tr. 7117- 19, 86 In response to questioning by complaint counsel, Mr. Summers stated: "Apparent- 82 Davis, Tr. 4471-72. David Davis, Vice President of Trade Development for Pepsi USA, testified that he viewed a new 7Up brand as a competitive threat. but not a comparable warehouse delivered brand, in part because of the differences in the delivery system: " L'p is a DSD brand. You ve actually got people in the store pushing the brand versus a Jolt Cola that s warehouse. They have to kind of depend on the store personnel doing it themselves. So you ve got more selling involved with a DSD brand on the store level, which is where the product gets moved or not." Davis, Tr. 4569. Texas Beverage Packer ("TBP")'s jack ofvoJume for its private labcI and warehouse brands is blamed on its failure to gain proper distribution. Hixon, Tr. 7332. 83 Hoffman, E., Tr. 358, 362; CX 3814 at 22- 23; Koch, Tr. 1831; Turner, Tr. 974, When soft drinks are on sale, consumers purchase more soft drinks, Knowles, Tr, 2838-40. Soft dnnks--espeClally diet soft dnnks--delenorate 1n quality over time, so careful attention must be paid to stocking and rotation of these items. Carbonation, flavorings, and aspartame are all sensitive to heat. CX 85 I. The level of carbonation in plastic containers, the quality of flavorings in all containers and the sweetening effect of aspartame decline over time. CX 851; E. Hoffman, Tr. 330- 31. Regular soft drinks after 150 days, and diel products after 90 days have diminished quality suffcient eeswto adverselybelievesaffectthaIrepeatits CSOssaleswithand consumeraspartamepreferences.have an expeetedsheJfex 85 I; CX 4005lifealof 63approximatelyR, 90Hoffman).days. E. Hoffman, Tr. 328-29. Consumers are sensitive to aspartame breakdown . ex 2281, When aspartame breaks down it turns biller and the flavor and quality become substandard due to deterioration. CX 4005 at 63; E. Hoffman, Tr. 328-29. Although this substandard product can be consumed safely, bottlers run the risk that consumers might never buy that product again, resulting in loss of volume E. Hoffman, Tr. 32829; Turner, Tr. 956-57. See also eX85 1, ex 3186 B. 85 Turner. Tr. 956-57; Knowles, Tr. 2663; E. Hoffman, Tr. 327- 28; CX 505 K; CX 3145 Y. Mr. Clements, President and CEO of Or Pepper from 1974 through 1986, testified that his attempts to use warehouse delivery for Dr Pepper in Indianapolis and Los Angeles in the 1950' s had convinced him "that with a product like Dr Pepper, and if you want to develop a consumer Franchise and if you want to develop an equity in that market, that we could not do il anyway except the store door delivery." CX 3582 at 2238 rClementsj. He explained that retailers "didn t reorder because they were not accustomed to having soft drinks that way, they were accustomed to having store door delivery, and if they did reorder, they didn t reorder in suffcient quantities, and so we went out of stock and after about six months we determined that that test was a failure and voted off. " ex 3582 at 2236 (Clements) cesw disputed this point. citing testimony by CCSW President Summers that, in his opinion, the retailer H. B. merchandised its private label, Plaza, better than Pepsi merchandised ils OSD delivered brands. Summers. Tr. 6472. Summers, testimony, however, supports the importance of control over distribmion and merchandising by the entity that ultimately would benefit most from volume sales of the producl.

g6 See also CX 2008 P, Q.

Opinion 118 FTC. ly you don t understand what sells volume in the soft drink industry. So let me tell you, it is not just price. You can have thelrest price in the world. If you can t get the product deliverea, if you can t get the display, you can t keep the display properly priced and stocked then the price becomes insignificant." Summers, Tr. 7117- 19. Other evidence similarly confirmed the importance of the services involved in DSD delivery for increasing the volume of soft drink sales.

In addition, the DSD delivery system also makes deliveries to smaller outlets economically feasible.89 Such outlets, while having a relatively small direct volume effect, are important for image90 and permit sampling that can lead to later sales. 3. The Importance of Advertising Branded CSD bottlers and concentrate companies invest significantly in advertising and promotion of their products. Concentrate 87 Other CCSW documents indicate the same view. When CCSW considered developing a house" control brand to compete with private label, il analysis recommended use of DSD over a broker system for several reasons:

1. DSD gives us an opportunity to reach more channels convenience stores, mom & pops, mass merchandisers, ete. without increasing OUTcasts dramatically. It also establishes an image and consumer sampling point of difference versus other private labels 2. DSD allows us more flexibility to respond to changes in the marketplace (i. lack of Coke ad feature activity, high volume hurdles, packaging emphasis changes, competitive features, etc RX398 D.

See Section IV supra.

89 Use of a warehouse-delivered system of distrihut on limits a firm to the large retail chains. Turner, Tr. 941. As a result, warehouse-distributed products cannot gain access to retail outlets such as drug stores, convenience stores, and smaller retailers that do not have the capacity to store the product. ex 3921 at 355 (CurTie); Turner, Tr. 941; CX 3943 aT 15 (Rapp); CX 3944 at 3511- 12 (RappJ; Coyne Tr. 3438, 3445. Even CCSW admitted that "private/warehouse brands are less available in other market segments, including convenience stores, vending and fountain." RPFF paragraph 332 (citing Knowles, Tr. 2662, 2892).

90 CCSW's president testified that presence in the fountain segment is important to develop the consumer s image of a product. Summers, Tr. 6513- 91 Mr. Clements explained that Dr Pepper was not able to reach all of the types of outlets that they wanted to reach with warehouse delivery in Indianapolis and Los Angeles: "We were only able to get the people like the chains, and not all of them, and some of the independents like IGA that had a warehouse that could deliver. What we couldn t reach were the outlets we needed most, and that's the single drink sales - the moms and the pops and the cafes and beauty shops and places like that. We did not have enough availability 10 create any great sampling of the product in order to develop the brand. ex 35H2 at 2236 lClementsj. See a/sa RX 398 D. Sampling occurs largely through cold drink sale rather than take- home sales. Turner, Tr. 1028-29. THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 551 452 Opinion firms pay milions of dollars annually in total marketing funding. Huge amounts of monies, in the aggregate and as a percen.tae of total marketing, are spent by concentrate firms in support of 10caJ branded CSD bottler activities. For example, the largest component of Pepsi Cola total cost is allocated to markcting. With respect to advertising by retail stores, major retailers typically run two types of carbonated soft drink promotions: "ad features" and "in-store promotions." An ad feature is typically a newspaper advertisement featuring a branded CSD at an attractive reduced price, often at or below cost An in-store promotion typically involves a branded CSD in-store display also featuring a reduced price, though not usually as low as the ad feature price and without any accompanying newspaper advertisement. An ad feature may give a bottler 10 times the non-featured sales volume, while an in-store display may give just twice to 2 Y2 times the normal sales volume. " The volume lift is much lower on the instore display in part because the retail price lo the consumer is usually higher." Thus, bottlers are willing to pay thousands to hundreds of thousands of dollars to obtain ad features."" Bottlers also offer and pay large dollar amounts in order to have exclusive promo- IOOtion and advertisement for their branded CSDs. For example CCSW' s 1988 Calcndar Marketing Agrcement with Diamond Sham- Turner, Tr. 965.

93 ex 3913 at 38 fDrewesj.

94 ex 3806 Z50: Turner, Tr. 973-74; Davis, Tr. 4515 (at or below cost Lids are termed " hot arls"Gonzaba,95 Tr. 2032 Turner, Tr. 1 ) 26. 96 Bodnar, Tr. 1498; Davis, Tr. 4504; Koch, Tr. 1831-32. Consumers also tend (0 stock-up during ad features, depending on the attractiveness of the ad feature price- Bodnar, Tr. J 766. 97 Bodnar, Tr. !498.

98 Bodnar, Tr. 1498; Turner, Tr. 974; E. Hoffman, Tr. 362-63. Increased sales volume due to an ad promotion or reduced price end-aisle display is known in the industry as volume " lift." E. Hoffman Tr. 358, 362 Turner, Tr. 1129-30.

100 ex 1040 (Pepsi); ex 1041 A- K (Grant-Lydick); ex 1042 A-V (CCSW). Opinion 118 FTC. rock stated that "(nJo national brand soft drink may be co-featured during these promotional periods."lol - Ad features are run as part ofretailers' promotional " ad cycles which include bottlers, branded CSDs as part of the advertising. 102 Most major chain retailers advertise one branded CSD in each of their weekly ads during a 52-week cycle lOJ Major convenience stores usually offer a monthly ad cycle. 104 Bottlers believe that you cannot grow. your brands without being in the ad cycle. 105 In fact, some believe that if a bottler never gets an ad feature, the effect will be volume deterioration in the marketplace. I06 Nor can the lost volume necessarily be made up for through l07 in-store displays.

There are promotional periods that are more advantageous than others. Holiday periods are the most advantageous and create consid- 10' For thaterable volume lift. reason, a retailer s holiday ad features cost hundreds of thousands of dollars for bottlers and concentrate firms. 109 Additionally, to obtain such ads from a retailer, the bottler must provide a greater discount than normal on its product. I 10 Retail- 101 ex 1039-B. Pursuant to Calendar Marketing Agreements ("CMAs ), the bottler and retailer agree to a schedule of promotional activities and the payments to be made to the retailer. CMAs were originally developed to help the retailer offset the cost of advertisements for their chain stores. Davis Tr. 4506. CMAs usually involve a base payment by the bottler to a retailer for a set number of ads. There are additional incentive payments for incremental volume growth. The bottler and retailer agree to sales projections and various requirements. CMAs are also known as "soft drink agreements " or " buy" agreements, "ad a.sistance " or "volume incentives." Davis, Tr. 4509, 4706; Gonzaba, Tr. 2055; Hiller, Tr. 5355.

102 Turner, Tr. 970.

103 Turner, Tr. 970; Davis, Tr. 4526; Kaiser, Tr. 3177. 104 E. Hoffman, Tr. 362.

105 Turner, Tr. 974; ex 3941 at 288-89 (Schmidl 106 ex 3941 at 288- 89 lSchmidJ.

107 Turner, Tr. 974.

108 Turner, Tr. 971; Summers, Tr. 6919; Davis, Tr. 45 !4- 16. The July 4th ad is usually considered the most valuable, followed by other summer holidays, then the Thanksgiving. ehristma. and New Year s holiday periods. Turner, Tr. 971; E. Hoffman, Tr. 367-68; Turner, Tr. 4514. As to non-holiday ad periods, pay week periods are more valuable than non-pay week ads. Turner, Tr. 97 I; E. Hoffman Tr. 368; 109Davis,Summers, Tr.Tr.69! 9. For example, the holiday4514.ads of H. , a very large retailer in the San Antonio area (and other areas in Texas), run from a low of $175 000 for Easter to a high of $500,000 for summer holiday ads. Summers, Tr. 6919; Gonzaba, Tr. 2055. 110 Gonzaba, Tr. 2057. However, there is a safety net of $50 000 for holiday ads ifvolume falls short. Summers, Tr. 6918- 19.

THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 553 452 Opinion ers, such as H.E. , must meet volume requirements in order to receive the ad payments in full.

Bottlers also compete with each other for retail space in retaij outlets which sell branded CSDs. I I I Bottlers attempt to convince retailers that their branded CSD products wil generate sufficient traffic to warrant display space and end-aisle displays. I 12 D. Branded CSDs as a Relevant Product Market With this background information in place, we can now properly address the question that the parties have presented to us: whether beverages other than branded CSDs could constrain a price increase by branded CSDs in the relevant geographic market. For this inquiry, we examine all of the relevant evidence concerning price and nonprice competition that could affect the Jikelihood that non branded CSDs would constrain a small but significant, nontransitory price increase by branded CSDs. Such evidence includes the opinions of market participants concerning price and advertising differences among different categories of soft drinks, historical evidence of price interactions among different categories of soft drinks, and industry perceptions about the degree of competition between different categories of soft drinks.

As we will discuss, nonbranded CSDs are largely unavailable in 13the cold drink channel. I Therefore, we will focus in particular on the likely substitution responses if branded CSD bottlers in the relevant geographic market raised their prices to retailers in the takehome channel, who purchase branded CSDs for sale to the ultimate 14 The retailer typically receives a discount or allowanceconsumer. I off the wholesale list price in return for its promotion of the product. I 15 "Net price" charged to the retailer equals the list price minus III ex 4005 at 55 (R. Hoffmann.

112 ex 4005 at 55-56 lR. Hoffmann.

ee eetlon IV mfra.

1!4 The Merger Guidelines advise that "(iJn genera!, the price for which an increase will be postulated will be whatever is considered to be the price of the product at the stage of the industry being examined." Section 1. 11. The same sentence appears in the 1984 Guidelines. 115 Por example , in 1990, at least 95% of CCSW' s sales were made at less than list price. Summers, Tr. 6721. Only 2% of the sales of the Pepsi COBO in the San Antonio area were made at list price. Davis, Tr. 4684-85. See also RX 327. , .

Opinion 118 FT. discounts and allowances. II6 "Net/net" or "net/net/net" prices are list prices minus discounts, allowances, and ad payments. It is also relevant here to examine the likely substitution responses if retailers raised the prices of branded CSDs to consumers, since the demand for the bottlers' products derives from consumer demand 17 Some agreements between bottlers and for those finished products. I retailers regarding advertising funds attempt to influence the retailer to offer a certain price to consumers for t e finished product, a further indication of the interrelationship between prices to retailers and prices to consumers. I For the reasons set forth below, we find that the evidence demonstrates a relevant product market of branded CSDs. I. Overall Substitution Possibilities:

Views of Branded SO Bottlers Both of CCSW' s primary branded CSD competitors in the San Antonio area stated that if branded CSD bottlers in San Antonio raised their prices by 10%, and everything else remained constant "9 Bottlers of brandedthey could profitably raise their price by 10%. CSDs in other South and Central Texas areas gave similar responses. 120 This evidence was uncontroverted. 121 116 R. Hoffman, Tr. 5652-53; Summers, Tr. 6713- 14; ex 414 B. 117 The Merger Guidelines state that, among other evidence, the Commission may take into account "the influence of downstream competition faced by buyers in their output markets" in evaluating market definition. Section I, 118Promotional a lowanccs are usually related to a performance requirement. his most commonly takes the form of a feature ad, in-store display, or a reduced retail price. ex 1039 B , c; ex lO4J H ("lowest retail price on featured package ). Although bottlers do not usually suggest retail prices, they often set discounts at levels calculated to drive a desired retail price, based on the margin usually added by a particular retailer. Campbell, Tr. 1972-73. In addition, when soft drinks are in a feature ad, retailers often add little or no margin to the wholesale price, or use incremental funding from the bottler to further reduce the retail price. TUrner. Tr. 960, 973-74. As explained by Mr. E. Hoffman lwJhat we re really trying to have happen is for the retailer to pass the cost, the lower cost, on to the consumer so that the benefit of the consumer -- the lower price is to induce more consumption or purchases." E. Hoffman, Tr. 380.

119 Bodnar, Tr. 1492, 1496, 1762-63; Davis, Tr. 4610. 120 Koch, Tr. 1815- 16; Turner, Tr. 988-89; CX 3931 at l801-04 (WestermanJ. 121 Respondent CC argues that this testimony !s not probative because comp amt counse not specify a time frame for the hypothetical price increase. ABR- A at 8. An example orthe testimony elicited by complaint counsel is given by Mr. Davis. a Pepsi official: Q. If Coke SW and Big Red raised their prices tcn percent in San Antonio, would Pepsi find it profitable to raise its prices the same? THE COCA COLA BOTTING COMPANY OF THE SOUTHWEST 555 452 Opinion The internal documents of the three bottlers of branded CSDs in the San Antonio area confirm that they take into account o!Jthe prices of other branded CSD products in deciding on pricing for their own branded CSD products. CCSW' s own business records indicate that CCSW does not consider the price of private label or warehousedelivered soft drinks when it considers increasing the price of its 12 Rather, CCSW considers the prices of otherbranded CSDs. branded CSDs in determining the price of its. branded CSDs. l23 Moreover, CCSW' s business records characterize its major competition as limited to manufacturers, distributors, and sellers of branded CSDs. '24 CCSW markets its branded CSDs against other branded CSDs. 125 A. Yes, they would.

Davis, Tr. 4610.

As discussed above, the hypothetical "5%" price increase test is set forth in the Merger Guidelines which typically define a smail but significant and "nontransitory" price increase as a 5% price increase maintained for a year or more. Section 1. 11. Although we agree that complaint counsel could have clarified the precise implications of this question by specifying a time frame, we do not find that the absence of a specified time frame renders such testimony worthless. "Profit" is generally understood as the gain stil! left after expenditures; this is not a short-run concept, hut rather something that businesses typically calculate over a time frame of months or year, not days. Thus, we believe that the question implied a "nonuansitory" time frame. In any case, the witnesses' responses indicate that the answer may well have been the same whether a short or long time frame had been specified, since no witness asked "Do you mean in the short run orthe long runs' Finally, we note that this is just one piece of the evidence supporting a branded CSD product market definition. We interpret the responses to complaint cOllnsel' s questions in light of that surrounding evidence, thc weight of which also supports a branded eSD product market.

122 ex 2244; ex 198; ex 3101 C-H, J: ex 3102 B- , J, L. 123 ex 104 D, G. H, M-N; ex 198.

124 ex 418 Z2- J, Z9. Z12, Z16. Z20; ex 1406 Z9- 10; ex 1854 H- 1. K-L, T- , X. Z2-5. Z7ex 1866 K-L. For example, ecsw' s records reveal that it viewed Mr. PIBB as the closest substitute to and a direct competitor of Dr Pepper. ex 596 A-I. Indeed, that ecsw recognizes the difference between branded and nonhranded eSDs is well-evidenced by their consideration of a proposal to establish a house product flavor line in the take home segment that would fiu the gap between branded CSDs and private label. The proposal was to " liJntroduce aDSD house line of flavors to include a Cola Cherr Cola, Red, Rombcer, and Orange. The line should be positioned as an image product with a low price (slightly higher than the private labels). Image development can be achieved through quality graphics, package availability, broad channel distribution and a unique trademark (perhaps the Buck Brand label)." RX 398. This document is consistent with other eesw documents that express concern that eesw needed a flavor line to compete with an expanding private label market. See, e. RX 2059: RX 2060; RX 226 A, K; ex 2974 Q. R.

125 ex 3760 ("In summar, beat the hell out of Pepsi!"); ex 104; ex 108 H; ex 1854 R , 22 24; ex 2255 S, T; ex 3109 C. Messrs. R. Hoffman and Summers, Tr. 6853, testified that CCSW' branded eSDs compete in a broad sense with virtually all liquids (See. qr, R. Hoffman, Tr. 5524: cesw competes with water in the sense that ail beverages vie for the same shelf space), but ecsw documents do not evidence the same approach. g., Opinion 118 FTC. CCE bottlers in Texas, Coke-Austin and Coke-Houston, create periodic reports in which they monitor the activities of thcicompetitors. Such activities -- which include pricing, package availability, marketing activities, sales, market share and pricing strategies -- are generally limited to observing the activities of bottlers of branded 126 Similarly, the bottling operations of CCE use KeystoneCSDs. reports that provide information only with regard to branded CSDs. When Coke-Austin introduced diet Coke, its introductory plans included volume and share forecasts. These projections were limited to branded CSDs and did not include private label or warehouse soft drinks 128 When Coke-Austin did a competitive analysis entitled Competitive Corporate Brands " it discussed only branded CSDs. 129 Pepsi official Davis testified that at the bottler level, Coke products are the only products to which the Pepsi bottler in San "Antonio would react with regard to price."o .. . Coke (CCSW) is usually the leader in the market. They go up and then we usually Davis stated thatfollow, depending on our pricing structure."I3I Pepsi does not follow private label CSDs closely enough to know whether they had price increases. 132 Pepsi bottler-related testimony and documents evidence a similar distinction between branded CSDs and nonbranded CSDs. For example, Pepsi official Davis testified that Pepsi would not be worried about promoting its products in conjunction with private labels I3 When thebut would not want Pepsi jointly marketed with Coke. Pepsi COBO bottler serving the San Antonio area performs comparisons with its competitors, it looks in detail to bottlers of branded 126 ex 2689; ex 2690; ex 2691; ex 2693.

Some documents note an increasing private label market share (e. ex 2623 F, ex 2561 N, Q), but very few suggest a price response from branded CSDs 10 such brands, and this evidence is much weaker than that pointing in the opposite direction. In one exception, an SWCC employee apparently suggested that a response to private label brands was-necessary to forestall the "expense of regaining price leadership long term," RX 1479 J.

127 ex 2680; ex 2688 A-D; ex 2695; ex 2918. 12H ex 503 B- 129 ex 17J.

\30 Davis, Tr. 4532-33.

!J I Davis, Tr. 4532; ex 441 c; ex 445 H-I, K; ex 448; ex 449 R- 132 Davis, Tr.4531 4829.

133 Davis, Tr. 4824.

THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 557 452 Opinion CSDs for their pricing and other competitive activity, 134 as well as the ad feature" or in-store allowances and ad assistance that tlmyar.e offering. 135 Emery Bodnar, former General Manager of DPSA and current Executive Vice President, General Manager and part owner of Grant- Lydick, similarly testified that Grant-Lydick considers and reacts only to prices of other branded CSDs in setting Grant-Lydick' branded CSD prices, and does not consider tM prices of private label or warehouse-delivered soft drinks in setting branded CSD prices. 136 Other bottlers also consider and react only to prices of the products of branded CSD bottlers in their areas when setting the prices of 134 ex 455 G-L, Z-24; ex 456 B-C; ex 457 c; ex 458 B; ex 459 E; ex 460 1; ex 461 J, L; RX 1013 U-W; ex 380. At leas! one document notes that, in 1989, private !abel's market share had increa. ed at the expense of Pepsi. RX 1287 E. Me Davis of Pepsi COBO explained that this fe11awed the deep discounting of 1987-88, when branded CSD prices had become so low that they were "taking share out of private label" Davis, Tr. 4528-29. When the branded CSD prices went back up, however then you stilt have "the price shopper that s going to pick up private label " and so "you re losing share back to the private label." Davis. Tr. 4528-29. See a/. Section IV. infra 135 RX 1013 J- Y; ex 455 H- , K, Z-!, Z-3; ex 456 E, F; ex 457 e, F; ex 458 G, 136 Bodnar, Tr. IJ59 , 1364, 1490, 1492- , 1762-63. Emery Bodnar, as manager of the Big Red bottler in San Antonio, cxplalned why he would not lower Big Red' s price to retailers if a warehouse or private label lowered its price 10%: Let me ask you this question. If Texas Beverage Packers lowered its triple net price in the ten-Q. county area including and surrounding San Antonio ten percent arid aJl other thirlgs remain constant, again for a sustained period of time, would you find it profilable to lower your prices A. I don t know what Texas Beverage Packers' triple net price is. I wouldn t know if they lowered it or not. See, because that doesn t come through the same charlnel as we do. re a direct store and they re through warehouses and through, you know, private label. Let s assume you did know.

If I did know that they went down ten percent? Yes.

Would I do anything? No, I've got to -- Let me just, if I can, state why. Private label or control brands, at least from where J sit, are not direct competition, as I look at Coke and Pepsi in San Antonio and maybe whatever they're callng themselves today, Premiere. Okay? Those brands that are essentially the warehouse or private label, first of all, space is dictated by somebody at headquarters and we re not going to change that. Number two, the product is displayed by somebody in the store or has to be handled by somebody in the store.

Jf you rcally go out and look at beverage sections, most often than not if you look at a beverage section Ihat looks ragged, it is the section that is supposed to be controlled by store personnel. As far as display space, that is pretty much, again, dictatcd, not at store level but at some buyer level or higher.

So really, there s not much I can do to compete, if r really wanted to. I mean, it s there, just the same say that Kool-Aid is, as we talked about earlier. So jf he lowered his price 15 20 percent, r wouldn t do anything. Fifty percent. He doesn t have that kind of margin to do iI, but if he did. They just can t execute. I mean, they just don t have the force to execute such a thing. Bodnar, Tr. 1762-63.

Opinion 118 F.T.C. their branded CSDS. 137 Moreover, bottler collusion cases indicate that branded CSD bottlers in other geographic areas beli that it is possible to raise price successfully together without having to involve bottlers of nonbranded CSDs. 138 2. Substitution Possibilities: The Cold Drink Channel In the cold drink channel, which ncludes fountain, vending machine, and single drink sales 139 there is relatively little availability of nonbranded CSDs that is, warehouse-delivered and private label CSDS. Respondent admitted that warehouse delivered brands are generally not available in the cold drink channel 140 and stated that private/warehouse brands are less available in other market segments, including convenience stores, vending and fountain.,,141 The evidence confirms that warehouse distribution does not provide access to the vending and fountain channels. l42 In addition, the evidence shows that carbonated soft drinks sold in vending machines are almost entirely brands that are direct-store- 143door delivered, not warehouse-delivered or private label brands. Vending machines are stocked with nationally branded CSDs, with l44 virtually no private label brands available. Moreover, although private label brands may be marginally more available in the fountain channel, since a few restaurant chains sell certain flavors as their own private label brands, 145 the record does not establish that the occasion- 137 Trebilcock, Tr. 5844- , 5848-50; Davis, Tr. 4532-33; ex 3990 at 923 lKaliIJ. 138 As we discuss in Section VI.

C.3 infra we find the bottler collusion cases relevant to this case and we therefore find that the AU cited in excluding evidence relating to them. For the cases cited by complaint counsel (See Section VI.C.3. infra), warehouse-delivered and private label firms in areas where branded CSD bottlers have been convicted of fixing prices were not named as defendants. The price-fixing cases involved only branded CSD bottler:;. See Section IV supra.

RRCCPFF paragraph 876.

141 RPFF paragraph 332, citing Knowles, Tr. 2662, 2892. 142 RX 3005 at 3759lSmithJ; ex 3978 ar2063-64 fLowenkronJ; Tumer Tr. 941 1403; CX 3945 at 177 fRappJ; CX 3944 at 3511- 12 fRappJ; CX 3977 at 72 lCarcwJ; Coyne, Tr. 3438; CX 3942 at 1905 (WilsonJ.

143 CX 804 G; ex 3989 at 65-66 (Shanks); RX 3003 at 82-84lHucyJ. 144 Koch, Tf. t835; Clarke, Tf. 4284; Turner, Tf. 1007; RX 3003 al84 (HueyJ. 145 Summers, Tr. 6517; Short, Tf. 7759-60. For example, McDonald' s sells its own orange private label fountain product. Short, Tr. 7759. THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 559 452 Opinion al presence of nonbranded CSDs in the cold drink channel would provide a constraint on the pricing of branded CSDs. 3. Substitution Possibilities: The Take Home Channel The record shows that private label and warehouse brands are available in this channel. Therefore, we must examine in greater detail whether their presence would be sufficie!l to constrain a small but significant, non transitory price increase in branded CSDs. a. Views of Bottlers of Warehouse and Private Label Texas Beverage Packers, Inc. ("TBP") is a manufacturer of private label and warehouse-delivered CSDs on its own account and for some of the major supermarkets in San Antonio. '46 Steve Hixon its general manager, testified that his carbonated soft drinks do not compete with those of CCSW and San Antonio Pepsi 147 and that to "I4' Hedo so would render his company "dead meat. ees manufacturers and distributors of private label and warehouse-delivered CSDs 149 and 150 Withas his direct competitors, not CCSW or Pepsi. respect to pricing, he reported the following:

Q. Now, in your opinion, there has not been an impact on your business by Coke Southwest's purchasing of the San Antonio Dr Pepper Bottling Company: is that correct? A. Yes.

Q. Your basic opinion is we re dealing with apples and oranges in this case? A. We re dealing with apples and oranges other than if there s some kind of pricewar going on. If they get down to 99 cents, then they do impact me, but I don t feel the -- If Coke had bought Pepsi, yes. Q. And you feel that that s because -- You don t see a relationship between you and Coke Southwest because you basically selno different clientele on different bases? A. No. We re -- Well, we re sitting in a grocery store next to each other, but I don t -- For the people to take my product over Coca-Cola, there s got to be a substantial differential in price to make them select the private label. 146 Hixon, Tr. 7269.

147 Hixon, Tr. 7354- 148 Hixon, Tr. 7356.

149 Hixon, Tr. 7359.

J 50 Hixon, Tr. 7360.

Opinion 118 F. Q. In fact, when you were first interviewed by FfC staff, you told them this doesn t have anything to do with you and you wish we lea,, .Ju alone?A. Absolutely. Still feel that way. Hixon, Tr. 7354-55.

With respect to the "impact" when branded CSDs reduce their prices, Mr. Hixon explained:

Q. In your experience, have national brands gotten down to the level of private labels in their pricing? A. They haven t gonen quite that low but it s been kind of -- They ve gotten close enough to make it scary.

Q. Have they in fact begun to squeeze out private label with low prices? A. That s a tough question. Certainly, to a limjted extent, I think they have. When they get in their 99-cent a six-pack wars with cans, yeah, at that point they re driving out private label. It s so low. We virtually have given up the m jor holidays to the national brands. We no longer try to compete with them.

Hixon, Tr. 7303.

Hixon views CCSW and Pepsi as "just out there screwing up the market with (their) occasional low prices." Hixon, Tr. 7360. He sees these bottlers as not trying to get his business, nor as having an impact on his business. Hixon, Tr. 7360-61. Hixon described himself and his "fellow copackers" as competing with branded CSDs only on the fringe:

Weare J out there scrambling over the ten percent of the business that Coke and Pepsi don t realize really exists or have slipped through their fingers, or whatever that they choose to ignore. So yeah, if Coke or Pepsi drop their prices to 99 cents it impacts our ten percent that we re fighting over. It takes business away from us. Hixon, Tr. 7360.

The Kroger Company operates a CSD manufacturing plant in Garland, Texas, called Garland Beverage Company ("GBC" 151 GBC does not consider the prices of branded CSDs in determining the price of its private label and warehouse-delivered products. The 52 GBC monitors onlyrecord does not show such a comparison. other private label and warehouse-delivered soft drinks, such as 151 Morath, Tr. 7672-73.

152 RX 1716- 17; RX 1721-22; RX L 726; RX 1740-41; RX l744-45; RX 1750; RX 1754-57; RX 1760.

THE COCA COLA Bottling COMPANY OF THE SOUTHWEST 561 452 Opinion 153 GBC alsoRocky Top, Big K, Mega, Parade, and Cragmont. monitors TBP. 154 This evidence also supports the existence of a branded CSD product market. The weight of the testimony by and documentary evidence of bottlers of both branded CSDs and non branded CSDs indicates that branded and nonbranded CSDs generally do not compete in the sense that a branded CSD price increase could be IS5 The evidence does establish thatconstrained by nonbranded CSDs. branded CSDs occasionally may constrain pricing by private labels and warehouse-delivered soft drinks, but it does not provide any reason to believe that non branded CSDs could constrain price increases by branded CSDs. IS6 b. Consumer Conduct: The Typical Price Gap Between Branded and Nonbranded CSD Retail Prices to Consumers Prices of CSDs appear generally to fall into three separated groupings. Most expensive are the branded CSDs; less expensive are warehouse-delivered brands; and cheapest are the private label 57 The price gaps separating these groupings may indicateproductS. 158 Althoughthat these soft drinks are in different product markets. the Commission and the courts do not always divide premium and 153 RX 1760 (991440; 991475-79; 991482-85); ex 2827 D. E. 154 RX 1756; RX 1757.

155 For example, Mr. Campbell. warehouse manager for a Pepsi/Dr Pcpperl7-Up bottler in Hallctsville, Texas, was asked whether he competed with H. s Plaza brand with his Dr Pepper and Pepsi brands. Mr. CampbciJ responded: " Wel yes and no. r mean, notreally. I mean, I don t -- r don think about competing against those people. I mean, that's not who I go to look in the grocery store to see if they ve reduced their price by one cent a can and then I adjust my pricing and my promotional strategies based upon that. I base my competing more against other direct store delivery products. Campbell, Tr. 2007. Even Mr. Howe!! of CCUSA admitted that he had never seen the price of Coke drop in response to private label prices. Howe\!, Tr. 4123. And Mr. Summers explained that CCSW created a private label to compete with private label and warehouse brands, being careful not to cannibalize CCSW' s branded products. Summers, Tr. 696284. 156 Similarly, Pepsi' s research shows that it is very hard for a private label to steal from a national brand, but that a national brand can gain shan: from a private label temporarily if its price comes down low enough. ex 3912 at 65- , 97 fChristianiJ. 157 ex 814 E; ex 3989 at 92- 93 (ShanksJ.

158 See. e.!?, United Stales v. Archer-Daniels-Midland Co., 866 F.2d 246 (8th Cir. 1988), cert. denied 493 U.S. 809 (!989) (despite functional interchangeability of sugar and high fructose corn syrup, persistent price difference of !O'io to 30% resulting from price support system required treatment as separate product markets).

Opinion 118 FTC. lower-priced brands into separate markets 159 the existence of a price gap calls for some examination of its degree and posme .significance.

We note first that the wholesale prices available to retailers vary, because bottlers may change their promotional offers on a weekly to monthly basis. 160 At any time, there may be a variety of effective wholesale prices for any given brand and package within any given geographic area. 161 Retail prices to consumers also vary frequently, depending on the extent to which and whether particular brands are on "promotion." The promoted prices of branded CSDs may be substantial discounts off the everyday or list retail price to consumers. The differences between the promoted retail prices of branded CSDs and the nonpromoted prices of branded CSDs vary from 20% to over 100%. 162 Despite these variations in price differences, there are clear distinctions between the average prices of branded CSDs and nonbranded CSDs, at both the wholesale and retail price level. As respondent CCSW has explained, the wholesale prices paid by the retailer for most pri vate/warehouse brands generally are Jess than the price charged by the bottler for branded CSDs. l63 Much of this differential is attributable to the labor cost of stocking and merchandising the product, which is usually borne by the bottler using DSD delivery for branded CSDs, but by the retailer for private/warehouse brands. l64 An additional cost difference is that national concentrate companies often spend significant sums of money advertising and 159 Coca- Cola Co. Dkt. No. 9207, slip up. at 32 n. 62; see also Olin Corp. \ 13 FIC 400, 595- 600 (1990), a/rd 986 F.2d 1295 (9th Cir. (993) (finding two relevant product markets, one consisting only of premium-priced product and one consisting of the premiurn-priced product and its functional equivalent), cert. denied, 114 S. Ct. 1051 (1994). 160 Campbell, Tr. 1954; R. Hoffman, Tr. 5551-52; Summers, Tr. 6613-H. However. some retaiters set their promotional schedule for an entire year at the beginning of the year. Summers, Tr. 6618. 161 ex 1979; ex 2180; Turner, Tr. 1474; Bodnar, Tr. 1648-49; Davis, Tr. 4702-03; RX 1200; Kaiser. Tr. 3224 ex 3973 (20- 100%); ex 3926 A (30- 50%); ex 3832 (20%); ex 3835 (20%). 163 Howell, Tr. 4028-29; RX 2423.

164 ex 3700 J; Brinkley, Tr. 2191-92; Kaiser, Tr. 3159; Turner, Tr. J 40 1-02. See aisf! Section IV. C.2 supra.

THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 563 452 Opinion promoting their branded CSDs. 165 These costs are often reflected in a higher concentrate price to bottlers. 166 Similarly, the average retail prices of most private/warehouse brands are less than the average retail prices of branded CSDS. 167 Estimates of the price differential var, but a common estimate is that private/warehouse prices average between 20% to 30% below the prices of branded CSDs. 168 There was testimony that the retail price gap between branded CSDs and private label CSDs is normally two to three price points, per unit. 169 In 1984, CCUSA found that, on average, private and controllabels net retail prices were an average of 29% lower than those of the national brands, while warehousedelivered Shastalaygo net retail prices were about 20% below the 170 In 1988, an analysis of the average case pricenational brands. differences for several bottler groups was performed, comparing Fanta, Shasta, Faygo, Controlled label and Coke Classic in 34 geographic areas. 171 The average case price difference between Coke Classic and the highest priced control label products was $0.94. Branded flavor lines were priced above control labels at an average price difference of $0.77 a case. 17 This retail price gap shows that certain consumers are willing to pay more for branded CSDs than for private label or warehousedelivered brands. Many consumers perceive a quality difference 165 ex 3158 ZI !- Z21; ex 8!4 B. Most private label brands are not advertised on television or radio, but may appear in the retailer s newspaper ads or circulars. Some warehouse hrands, notably Shasta, have engaged in te\cvision and radio advertising in the past. See Section IV. C.3 supra. Bodnar, Tr. 1739.

167 Hixon, Tr. 7356-57; Trebilcock, Tr. 5841-42. \68 Trehilcock, Tr. 5841-42; Howell, Tr. 4082; ex 3814 at 39lAdams1; ex 814 at 874. At different times, the retail price gap between branded CSDs and private label/warehouse soft drinks may range from 10% to 130%, depending on whether special promotions arc offered. Trebilcock, Tr. 5841- 42 (20-30%); Hixon, Tr. 735657 (30-40%); CX 3989 at 89-90 (Shanks); Bodnar, 'fr. 1715- 16; CX 3835; CX 3832; CX 39268 (20-70%); Limon, Tr. 4981 (6-pack ("6-pk") cans: private label CSD is 99 cents; Pepsi is $1.49 - $1.69 (49-69%)); Sendelbach, Tr. 7703-06 (6-pk cans: private label CSD is $1.20; branded carbonated soft drink is $ 1.59 133%)); Brinkley, Tr. 2! 94-95 (6-pk cans: private !abel CSD is $1.20-$1.2; branded CSD is $2.50 l50%)); Chapman, Tr. 7208. 721 L (6-pk cans: private label CSD is $1.06-$1.26; branded eSD is $1.59 $2.00 (26-88%)); Davis, Tr. 4519- 169 ex 3967 at 186 fCarewJ. Each price point has significance for a bottlers' revenue; for example, for CCSW, a ten-cent increase in the net price of a six-pack can increase cash flow by an additional $8 million a year, holding all else constant. E. Hoffman, Tr. 284; ex 875 G. 170 CX814A.

171 ex 3436 S- 172 CX3436F.

Opinion 118 FTC. between branded CSDs and private label/warehouse delivered CSDs. I7 Because of that perception, branded CSDs nave greater consumer appeal than do private label/warehouse-delivered CSDs, 174 and brand switching by consumers is generally limited to branded products. I7 The perceived differences in quality apparently account for the 176 Thefact that branded CSDs have some degree of brand loyalty. extent of brand loyalty has decreased recently, and consumers more readily switch between branded CSDs if prices differ significantly; however, there is little evidence of switching from branded CSDs to private label/warehouse-delivered CSDS I7 at least until the price If the retail price of branded CSDsdifferences are very large.l78 drops near or below the price of private/warehouse brands, then private/warehouse brands may lose sales to the branded CSDs. 179 Again, such evidence indicates that branded CSDs may constrain non branded CSD pricing on occasion, but not the converse. 180 173 Morath, Tr. 7676.

174 ex 3912 at 65- , 97 fChristianiJ. 175 ex 3942 at 1911- 12 (Wilson).

The AU concluded that branded and private label CSDs have similar functional characteristics. implying that they are in the same product market, although he acknowledged that "(t)o a great extent, any perceived differences among soft drinks exists in the mind." lD at 61. In evaluating the likelihood of customer switching in the event of a smail but significant, nontransitory price increase, such perceptions in the mind arc more relevant than a chemica! test of whether the ingredients are basically the same.

ex 3967 at 205 lCarewJ; Morath, Tr. 7676. 177 CX 3942 at 1911- , 1940-41. 178 CX 3921 at 386-87.

Mr. Koch, President of Oneta Company, the Pepsi-Cola bottler in Corpus Christi, testified that as to supermarkets only, "jwJc know that private labels have about a 10 percent residual share of the market that s based on existing price structures and existing price differences." Koch, Tr. !876. He stated that "private labels do a lot to keep us honest in the sense that "(gJo maybe 20 percent higher with a national brand than you can with private label, and then you start to lose volume." Koch, Tr. 1875. Other than the responses of branded CSD bottlers to the 5% question see Section IV supra this is virtually the only piece of testimony that directly addresses whether private label could constrain upward pricing of branded CSDS. We do not find this testimony suffcient to outweigh the weight of the evidence related to this point, especia!Jy since Mr. Koch only states his belief that branded CSDs would lose volume to private label, but does not state that such losses would make it unprofitable for branded CSDs to raise price.

179 Bodnar, Tr. 1555-56; Summers, Tr. 6549 (branded CSD discount to $.99 will pick up some private label share, hut "usual!" discount of $ 1.49 does not); Hixon, Tr. 7303- , 7360; Chapman, Tr. 7190; Turner, Tr. 988; Campbell, Tr. 1999; RX 30ll at 317 !- , 3197-98lSkinnerJ. ISO Olin Corp. 113 FTC at 598-600 (lo\'u- priced swimming pool chemical could not constrain upward price movement of premium-priced swimming pool chemical). See also ya-Cola Co. , slip op. at 36.

THE COCA COLA BOTTING COMPANY OF THE SOUTHWEST 565 452 Opinion Indeed, the preponderance of the evidence concerning brand loyalty suggests that consumers may be reluctant to switch to nonbnmed CSDs in the event of branded CSD price increases. For purposes of product market definition, the relevant question is whether, if a wholesale branded CSD price increase were passed on as a retail price increase, consumers would switch to non branded CSDs and thereby force a rol1back of the wholesale branded CSD price increase? The record contains a study designed specifically to address the issue of what magnitude of branded CSD price would cause consumers to switch to non branded CSDs, albeit in a different geographic market. When Procter & Gamble owned Coke-Mideast Bottling Company, it did an elasticity analysis, comparing warehouse-delivered CSDs and Coca-Cola products. It found that an acceptable spread between Coke products and Big K' s private label products was between 80 and 100%. If the prices of Coke products were above this level, consumers' normal preferences for branded CSDs began to diminish.

On the other hand, if the Cokc was for sale for 99 cents and Big K was for sale for 95 cents, Big K didn t sell almost at all because the spread was so small, consumers would virtually all opt for Coca-Cola.

CX 3921 at 386 (Curre).

This study supports the conclusion that sales of non branded CSDs would not constrain a retail price increase to consumers of branded CSDs when the initial price gap is the average size that we observe -that is, branded CSD prices averaging 20-30% above the prices for private/warehouse soft drinks. l8I Since the study indicates that consumers' preferences for Coke products would not diminish until the prices for Coke products were more than 80- 100% above the prices for private label products, the study indicates that retailers most likely could pass along to consumers any 5% or other small but significant, non transitory price increase by branded CSD bottlers. This ability would likely diminish the incentives of retailers to fight The ALl cited testimony of Robert Chapman of H. , a retailer, that his belief was that, if prices of branded CSDs decreased, sales of private labels would decrease. IDFF paragraph 2 J 6. That such substitution might occasionally occur, however, does not establish that jf prices of branded CSDs. generally increased, then sales of private !abel would increase suffciently to make the price increase unprofitable for the branded CSDS.

See note 168 supra.

g., Opinion 118 FTC. such a price increase, since they would not have to absorb the price increase themselves.

c. Views of Concentrate Companies Evidence from concentrate firms also is consistent with a product market confined to branded CSDs. CCUSA analyzes the market with regard to branded CSDS. I82 CCUSA performs business reviews of retail accounts in which it evaluates the performance of Coke products. In its 1988 H.E.B. business review, it listed the top ten brands of CSDs in both San Antonio and Austin; not one private label or warehouse brand was listed. l83 The 1990 Marketing Program presented to H. B. by CCUSA and Coca-Cola bottlers discusses and makes comparisons among only branded CSDs. '84 CCUSA generally has compared its prices of cherry Coke and Mr. PIBB only to Dr Pepper and not to any non branded CSD. I85 Pepsico as a concentrate company looks at the retail prices of only branded CSDs. 186 Pepsi performs periodic -competitive analyses comparing Pepsi brands to CSDs of its competitors. 187 These studies generally do not involve private label CSDs. l88 Arthur Christiani 182 ex 2547; ex 801; ex 803; ex 1892 L.

In the instances where CCUSA compared its products with nonbranded CSDS, the comparison involved a CCUSA product that diverges from the profile of a branded CSD. For example, CCUSA analyzes Fanta, which is not nationally advertised, in comparison to warehouse-delivered Shasta and private labe! brands. See, e. ex 3436 B, C; RX 687 D, M; RX 958 8-D; ex 1084; ex 1991-231; Howell, Tr. 4029- , 4023-25. In assessing whether to create a Fanta line offIavors, CCUSA believed that "(a) Fanta line would be unlikely 10 incur competition from Pepsi Cola CSA," and that "Fanta cola would compete with Coke and Coca-Cola classic only on the fringe. and thus not have any significant negative effect on these two brands." CX 799 F For CCUSA' s branded CSDs, the documents reveal only infrequent references to concern about competition from private label and warehouse brands. E.g., CX l69 C (concern that some Coke sales had been lost to private label or warehouse brands). 183 CX 506 T , Q, lJ.

184 ex 2263.

185 ex 790 E; ex 791 M.

186 ex 381 L Concentrate companies subscribe to Nielsen s retail sales report service, which provides onc collective entry for most private label and warehouse-delivered CSDs (except Shasta and Faygo). RX 694 at 13, 16; RX 2806. Concentrate r"inns do not subscribe to SAMI, which provides detailed analyses of the sales of warehousc-delivered brands. including CSDs. Clarke, Tr. 4279. . 187 CX38!.

188 ex 3912 at 121 (Chrislianil.

Two documents have compared the prof"itability of DSD versus warehouse delivery for retailers and noted that, to compete with warehousc on price, it would be necessar for Pepsi to Jowercosts, since050 costs more than warehouse delivery. ex 385 X to Z-53; ex 1922. THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 567 452 Opinion manager of business analysis for Pepsi, explained the process by which he performs retail elasticity studies for Pepsi. 1H9 Chrini. concluded when a retailer promotes its private label CSDs it does not supplant purchases of Coke or Pepsi products. 190 He also concludes that, generally, "(iJt is hard for a private label brand to gain share from a national brand because of the types of consumer dynamics. ex 3912 at 65- 97 (ChristianiJ. 191 When Dr Pepper performs reviews of retait accounts, it does not compare the performance of Dr Pepper brands with private label or warehouse soft drinks. 192 Dr Pepper looks to Coke, Pepsi, 7-Up and RC products when comparng sales volume movements, not to those of private label and warehouse soft drinks. 193 During the period of time that Dr Pepper Company owned and operated production facilities, its review of these operations involved analysis of Dr Pepper products' performance with branded CSDs. 194 The testimony of other market participants similarly confirmed that the pricing and marketing of branded CSDs is separate from that for non-branded CSDs. For example, A&W does not market its A&W brand products against private label products, nor does it develop marketing strategies with respect to private label products. 195 Michael Skinner of Shasta testified that increasing the price difference between his warehouse-delivered CSDs and DSD-delivered CSDs was not profitable 196 that he saw little response by Pepsi or Coca-Cola to Shasta s prices 197 and that he experienced price pressure from private label brands only in limited areas of the 189 ex 3912 at 139-40 (Christiani).

190 ex 3912 at 70 (Christiani). 191 Mr. Christiani testified that "a national brand can gain share from a private labe! brand temporarly if its price came down lowenough." ex 3912 at 65-66 (emphasis added). This statement does not establish the converse, of course. Mr. Christiani stated that private labels on sale would typically cannibalize Pepsi only if private label were included in the top three brands in the market which was not the case in the six markets he examined. ex 3912 at 24-27 (ChristianiJ. 192 ex 504: ex 206: ex 212 K- M: ex 214 H.

193 ex 600: ex 836 J-Q: ex 2524: ex 2526: ex 834 194 ex 834; ex 2526.

195 ex 3978 at 2096-97 (LowenkronJ.

196 RX3011 at 3198-320! !Skinner). 197 RX 301 I at 3201 (Skinner).

Opinion 118 FTC. country, not including Texas. 198 The testimony indicated that socalled ' boutique ' firms such as lolt Cola Co., Qriginatew York Seltzer, Soho, Sun dance and Snapple have no effect on the prices of branded CSDS. 199 The president of Double Cola stated that private 2OO label CSDs compete primarily with warehouse-delivered CSDs. The ALl relied on evidence that the National Soft Drink Association includes a large variety of CSD and non-CSD beverages in its reports, and that government agencies put private labels with national brands in certain reports, as evidence of a product market broader than branded CSDs. TO at 61; TOFF paragraphs 205 , 206. But neither the government's "SIC" categories nor the NSDA' categories track whether all of the items within each category could constrain price increases of the other items. We find the business records and testimony of market participants to be more probative of the relevant competitive issues and the weight of that evidence supports a product market confined to branded CSDs. d. Pricing History and Price Patterns (i) 1987- I 990 Branded CSD Pricing in San Antonio and Other Areas The history of price changes by branded CSD bottlers during the 1987- 1990 time period also provides some insight into whether branded CSDs are a relevant product market. In contrast to the ALl 2IJI we find that this evidence supports the existence of a product market of branded CSDs.

In 1987, the Pepsi COBO significantly increased its discounts in its territories overall, starting in San Antonio in particular 02 Pepsi official Davis explained that the Pepsi COBO became concerned because they had only a few ads scheduled for the year and feared a 198 RX30I1 at 3202 (Skinner).

199 ex 3989 at 99- 100. 169 (Shanks (Double Cola)); ex 3941 at 320 (Schmid (7-L'p)J; ex 3921 at 408 (Currie (Pructer & Gamble)J; ex 3990 at 928 (Kalil (Kalil Bottling)); RX 3014 at 3557lGreenberg (Unadulterated Food Products)). 200 ex 3989 at 93 (Shanks). 201The ALl tnte'1reted the evl ence SUfTOun mg t ese pncc c angcs as supportive 0 a pro uct market including non-branded CSDS. See, e. IDFF paragraph 222; ID 60. As we will discuss, we believe that the ALl' s conclusion resulted from a misinterpretation of the evidence. 202 Davis, Tr. 4527 4548-59.

. . ,, THE COCA COLA BOTTING COMPANY OF THE SOUTHWEST 569 452 Opinion loss of volume if something was not done to compensate for the lack of ads.20J They decided that, in San Antonio, they wanted to oe" least one, if not two, (price) levels below Coke " and that "( w)herever ,,204Coke went, we d go a little lower.

Davis stated that the Pepsi COBO serving San Antonio "led" pricing down in 1987 205 According to Davis, during this period usually Pepsi would move its prices down first, and Coke would then 06 There was some variability in thematch Pepsi' s lower prices. 988,direction of prices during this period. For example, in Januar J Pepsi and CCSW raised prices.207 Similarly, in March, 1988, Coke moved its pricing for cans back up, and Pepsi followed. However during the summer of 1988, Pepsi led pricing back down 0H Finally, in the fall of 1988, the price war abated when Pepsi led pricing back .'09 This conduct was in contrast to Coke s usual position as price leader, which Mr. Davis described: "Coke is usually the leader in the market. They go up and then we usually follow, depending on our pricing structure.,,2ID Davis described Pepsi' s 1987-88 pricing as Pepsi' s attempt to He stated thatgain market share at the expense of losing money2II the Pepsi COBO probably gained about four share points as a result but reported that the Pepsi COBO has of the deep discounting,2I by bringing upfound that it can "drive (its) business a lot easier" prices and giving ad payments to retailers than by "trying to drive it just with price."m Since the Pepsi COBO wanted to become more 203 aals testl Ie t at epsi gains a sign! leant Increase In va ume - - from 4 to !O times the usual! volume -- when an ad feature for Pepsi is on. Davis, Tr. 4504. 204 Davis, Tr. 4548.

205 Davis, Tr. 4527.

206 Davis, Tr. 4550-59.

207 Davis. Tr. 4557-58; Hilke, Tr. 5959.

208 Davis, Tr. 4558-59.

209 Davis. Tr. 4559. Overall, CCSW' s San Antonio wholesale netlnet/nc! prices (that is, net of discounts, allowances, and ad payments) increased 2. /r (from $6. ! 1 to $6. 28) between! 987 and 1988. ex 4114; RX 3085.

210 Davis, Tr. 4532.

211 Davis, Tr. 4560.

212 Davis, Tr. 4564.

213 Davis, Tr. 4528.

, .

Opinion J J8 F. profitable, it has now adopted a strategy of working on getting ads and not dropping prices so Both the AU and CCSW10w.'14rely heavily on two Pepsi documents,'I5 stating that private label gained share at the expense of both Pepsi and Coke because of the 1988 price increases to show that branded and nonbranded CSDs are in the same product market. However, this reliance misses the point.

As Mr. Davis explained, private labels had lost share when branded CSD prices became so low during the deep discounting in 1987- 88.'16 When branded CSD prices rose again, branded CSDs lost share back to the private labels because "you still have the price shopper that s going to pick up the private label."2I Mr. Davis agreed that both Pepsi and Coke had taken "a volume hit" when branded CSD 'I8 but pointcd outprices rose again that, since Pepsi' s Nielsen data include Big Red in the category of private label re not really sure how much of it is Big Red and how much of it is private label."'I9 A resolution of whether Pepsi actually lost volume to private label instead of to Big Red is not necessary for disposition of this issue, however. The question is not simply whether a branded CSD price increase caused branded CSDs to lose share to private labels. Rather, the question is whether any loss of share made the price increases so unprofitable that Pepsi or Coke rescinded them. If no rollback of the price increases occurred, then one can assume that Pepsi and Coke found them profitable despite any loss of volume to private label, and that therefore their pricing was unconstrained by pri vate label.

The testimony shows that whatever losses of volume to private label might, have occurred were insufficient to constrain price increases of branded CSDs in the San Antonio area in 1987-88 or in other, more recent times. Mr. Davis testified that Pepsi had no concern about possible volume losses to private label and that Pepsi 214 Davis, Tr. 4528, Davis testified that, although the Pepsi CaBO has always lost money in the San Antonio area, it lost a lot more as a result of the deep discounting in 1987-88. Davis, Tr. 4561. E.g.. RX 2503 A. 216215Davis, Tr. 4528- 29. 2!7 Davis, Tr. 4529.

218 Davis, Tr. 4529, 219 Davis, Tr. 4829-30.

. . . g., . . THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 571 452 Opinion had not rolled back any wholesale prices or increased any discounts to retailers in response to increased private label sales resultingffom He noted that Pepsi had rolled back some B. private label ads."o wholesale price increases in 1990 due to competition with Coke, not because of private label price competition. '2I Grant-Lydick' s Big Red followed CCSW' s pricing increase in 1989 as confirmed in an internal1ender report which relates a conversation that the author of the report had 'with Toby Summers CCSW' s president, and David Green, CCSW' s chief financial officer: Toby (Summers) and David (GreenJ analyzed month-by-month perfonnanee beginning with January, 1989.

In February, TBG implemented a 6% price increase at the wholesale level resulting in a 3% to 4% net price increase after discounts. Big Red matched the price increase immediately in mid-February. Pepsi matched the price increase on March 1.

CX 3806 Z56.

In addition, Mr. Bodnar of Grant-Lydick testified that if CCSW raises its prices, the convenience stores and independent stores will raise Big Red's retail prices to match CCSW' s prices even if Grant- Lydick does not raise Big Red' s wholesale prices. Thus, Mr. Bodnar s practice is to raise his prices when CCSW' s increases its prices.'" For example, in early 1990, CCSW raised the wholesale prices of certain of its package sizes, and Grant-Lydick maintained its same price levels. A month later Mr. Bodnar surveyed 100 accounts and found that:

OUf retails went up to match Coke s. So we had no choice but to raise our levels. I mean, the retailer was taking the long margin on us. Bodnar, Tr. 1493.

Although this evidence alone is not dispositive, it is overall supportive of the existence of a branded CSD product market. 220 Davis. Tr. 4530- , 4760. 221 Davis, Tr. 453!- 32. 222 Bodnar, Tr. 1492-96.

22 orne relallers testl led that they would not raise the pnce 0 one branded CSD based on a price risc for another Gonzaba. Tr. 2 !06-07, but Mr. Bodnar s review of 100 accounts indicates that it can happen.

Opinion 118 FTC. (ii) Price Relationship Studies When the prices for two products move in different directions over time, it indicates that the products are in different antitrust productmarkets.22 In this case, complaint counsel presented evidence attempting to show that prices for branded and non-branded CSDs have moved in different directions over time. We find, however, that complaint counsel' s study is inconclusive and cannot be afforded much weight.

Complaint counsel's economic expert, Dr. Hilke, compared the relative net price movements of branded CSDs with those of private Forlabel and warehouse brands for 1987 through 1989.22 his analysis, he used a "sign test " which simply tests whether the prices moved in the same direction, but does not provide any information on the relative differences in magnitude of any price movements. 226 Using comparisons of quarterly data, Dr. Hilke found different direction price movements 2 times out of 10; using monthly data, he found different direction price movements 11 times out of 32. Complaint counsel argues that these data show that branded and nonbranded CSDs are in different product markets, especially since the sign test does not take into account possible large differences in same direction price movements. By contrast, respondent s economic expert, Dr. Strickland, asserted, and the AU agreed, that the data show highly parallel price movements that are not random .'28 Overall, we find that the price movement data is not particularly useful in resolving the product market question. Especially for branded CSDs, which are frequently sold on low-priced ad features prices may change on a week-to-week basis; the unusually large swings in price attest to this.229 Comparisons of monthly data, in our view, are overly sensitive to this problem. Moreover, even assuming that this problem can be overcome by comparisons of quarterly (instead of monthly) data -- which is not 224 See United States v. Aluminum CO. OJ America 377 U. S. 271 , 276-77 (1964); see generally Antitrust Law Developments (3d) at 285 (1992). 225 eX! 678 A- 226 Hilke, Tr. 5954.

227 ex 1678 A- 228 Strickland, Tr. 8060; lDFF paragraph 232. 229 See Section IV C.3 supra.

THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 573 452 Opinion obvious to us -- and that the quarterly data show similar di!ection price movements, such nonrandomness can be caused for reader1s other than that branded and non branded CSDs are in the same product market. For example, there is ample testimony that soft drink prices in general (as is true for many retail prices) are seasonal.230 The prices of products in two different antitrust product markets could well exhibit some movements in the safe direction simply because of seasonal price changes or holiday-oriented discounting. Such may well be the case here.

In sum, we do not rely on the price movement data because we believe they are unreliable and should be given little weight. In any case, the results are inconclusive and therefore do not add to the substantive analysis.

e. Views of Retailers Evidence from retailers is consistent with a product market confined to branded CSDS. Trish Adams, the senior DSD buyer for all Target Corporation stores2JI testified that Target department stores have a limited amount of shelf space to dedicate to CSDs. Consequently, Target meets consumer demand head-on by offering only branded CSDs. This demand includes Big Red in San Antonio. Even a 20% increase in branded CSD prices would not motivate Target to include private label CSDs in its beverage aisle.23 When Target carried private label CSDs, branded CSDs were not affected by placing private label CSDs on sale'J4 Circle K and 7-Eleven convenience stores had private label CSDs Mass merchandisers in Sanat one time, but discontinued them.23 Antonio also do not car the private label CSD, Texas Cola, because they only want branded CSDs. 236 230See Section IV.C.3 supra.

231 ex 3814 at 5 (AdamsJ. 232 ex 38 14 at 9- 11 (AdamsJ, San Antonio is Big Red' s largest market, and Grant-Lydick is the largest Big Red battIer. IDFF paragraph 85; Turner, Tr. 953. ex 3814- 54 (AdamsJ. See also ex 3821-48 (Imper). 234 ex 3814- 36, 39-41. 235 Howell, Tr. 4000', Knowles, Tr. 2892. 236 Hixon, Tr. 7358-59.

Opinion 118 FTC. 4. Summary We find that the weight of the evidence establishes the existence of a relevant product market limited to branded CSDs. With respect to the cross-elasticity of wholesale prices, there is consistent San Antonio bottler testimony that they could profitably raise branded CSD prices by 10%. Similarly, the documents and testimony of bottlers, concentrate companies, and retailers overall indicated that branded CSDs are priced in comparison to other branded CSDs, not private label or warehouse brands. With respect to retail pricing of finished products, the weight of the evidence demonstrated a persistent price gap between branded and non-branded CSDs, reflecting a premium that consumers are willing to pay for branded CSDs. There was no testimony or other evidence that retailers would be unable to pass along any cost increases for branded CSDs, thus possibly putting pressure on bottlers to refrain from price increases. With respect to industry perceptions, the documents and testimony consistently supported significant distinctions between branded and non-branded CSDs in terms of prices, level of brand name recognition and advertising support, method of distribution, and availability in different channels of distribution. Thus, we conclude that the weight of the evidence shows a relevant product market of branded CSDs.

V. THE RELEVANT GEOGRAPHIC MARKET Having determined the product market to be branded CSDs, we turn now to defining the geographic market, the second "necessary predicate" for analyzing an acquisition s effect on competition. See United States v. Marine Bancorporation 418 U.S. 602, 618 (1974); Brown Shoe Co. v. United States, 370 U.S. 294 , 324 (1962); United States v. E.I Dupont de Nemours Co. 353 U.S. 586, 593 (1957). Such an inquiry is, of course, a prerequisite to determining whether the acquisition may result in a substantial lessening of competition in branded CSDs "in any section of the country" (Clayton Act Section 15 U. c. 18). See United States v. Marine Bancorporation, 418 S. at 618.

Complaint counsel alleges that the geographic market within which to assess this acquisition consists of a ten-county area centered THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 575 452 Opinion 2J These tenaround San Antonio, Texas (the "San Antonio market counties comprised the original territory granted through the 1984 sale of the Dr Pepper franchise to CCSW. Respondent contends in response that the San Antonio market is inappropriately narrow, and suggests instead a far larger market that includes the major cities of , and Houston. See RX 2983.San Antonio, Austin, Waco, Dallas Although the AU ultimately failed to delinea\e a specific geographic market, he rejected the San Antonio market, finding that the relevant geographic market was larger than the ten-county area around San Antonio. See lOFF paragraph 245; 10 67. For the reasons discussed below, we reject the AU' s findings and conclude, based upon our own review of the record, that complaint counsel carried its burden of proving that the relevant geographic market is the San Antonio market. In reaching this conclusion, we note that the ALl's geographic market evaluation was erroneous in several important respects. First and foremost, the ALl's assessment must be disregarded because it was premised on an incorrect and unreasonably broad view of the product market as encompassing not only branded CSDs, but also private label and warehouse (non- , non-carbonated beverages (e.branded) CSDs, mixers, seltzers Lipton Iced Tea), and isotonic drinks (e. Gatorade). Second, the AU failed to apply the proper standard for defining a geographic market, as set forth in the Merger Guidelines, at Section 1.21 , 4 Trade See AdventistReg. Rep. (CCH) paragraph 13 104 at 20,573 238 Health System/West, FTC Dkt. 9234 (Apr. I , 1994), 5 Trade Reg. Rep. (CCH) paragraph 23,591 at 23,258. In addition, the AU gave undue weight to, and otherwise misapplied, the Elzinga-Hogarty test concerning shipping patterns.

Under Section 1.21 of the Merger Guidelines, the relevant geographic market is defined as the smallest region within which a hypothetical monopolist could "profitably impose at least a ' small but significant and nontransitory' increase in price, holding constant the terms of sale for all products produced elsewhere." The "profitably impose" language implicitly recognizes that, in the face of a price 237 IS propose mar ct \!1C u es seven counties (Atascosa, Bandera, Bexar, Fno, Kenda!! Medina, Wilson) and portions of thr c others (Blanco, Comal, and Karnes). IDFFparagraph 245. 238 The approach to geographic market in the Merger Guidelines is essentially identical to that taken in Section 2.3 of the 1984 Merger Guidelines, 4 Trade Reg. Rep. (CCH) paragraph 13, !03 at 558, Opinion 118 F.T. increase, some sales wil inevitably be diverted elsewhere, as would be expected. Consequently, a geographic market wil e ist notwithstanding some diversion of trade, so long as the additional profit from the price increase over the remaining customers exceeds the profit lost from the trade that was diverted. .

In defining the geographic market using the methodology described in the Merger Guidelines, the Commission begins with the location of the merging firms and asks what would happen if a hypothetical monopolist imposed at least a "small but significant and nontransitory" price increase, typically 5% over a one-year period. , in response to the price increase, the reduction in sales would be sufficiently large to render the price increase unprofitable, then the agency adds the next best substitute location to the proposed market and the test is repeated. See Adventist Health-System/est, 5 Trade Reg. Rep. (CCH) paragraph 23 591 at 23 258. The record contains direct evidence establishing that a hypothetical monopolist selling branded CSDs in the San Antonio market could profitably raise prices by more than 5% for a nontransitory period. Most significantly, branded CSD bottlers in the San Antonio market provided undisputed testimony to the effect that they could profitably -- and without fear of outside competition -- raise their prices by as much as 10% if other branded CSD bottlers in this market did the same.23 Consistent with the foregoing evidence, bottlers of branded CSDs outside the San Antonio market testified that they would not ship into the San Antonio market, even if the price of branded CSDs in that market increased by 10%. 240 Another consideration that directly bears upon the likely response to a price increase is the fact that competition in the local soft drink industry is characterized by the use of exclusive territories.24 Concentrate firms grant bottlers exclusive rights (franchises) to manufac- 239 Bodnar, Tr. 1492, 1496, 1762; Davis, Tr. 4610; Koch, Tr. 1815- 16; Turner, Tr. 995-96. Respondent argues that this testimony should be disregarded because the hypothetical question posed by complaint counsel failed explicitly to incorporate a one-year time frame. We disagree. Because the question was framed in terms of profitability (see, e. Davis. Tr. 4610). we believe that this question was correctly understood by the witnesses as referring to a nontransitory price increase, a price increase that would be maintained for more than an insignificant period of time. Consequently, we accept the responses as constituting probative evidence of the ex.istence of a San Antonio market.

240 Turner, Tr. 8598-99; Campbe\!, Tr. 1946-47; Van Houtcn, Tr. 8470-76; Koch, Tr. 8625-26; Davis, Tr. 4476-78; Bodnar, Tr. 1712- , 1372; Nes!age, Tr. 8720-23. 241 ex 1666.

g., THE COCA COLA BOTTING COMPANY OF THE SOUTHWEST 577 452 Opinion ture and sell, in a specified geographic territory, soft drinks in bottles and cans bearing the concentrate company s trademark and using its formula.24 This feature is universal for packaged branded CSDs and for pre-mix fountain syrup, and partial for post-mix fountain branded CSDs.243 We wil first address the operation of exclusive territories for sales of packaged branded CSDs and pre-mix fountain syrup. Under the exclusive franchise agreements, concentrate firms prohibit their franchised bottlers from transshipping, that is, from shipping packaged CSD products, pre-mix fountain syrup, and concentrate outside of the exclusive franchise territory for which they are licensed into the franchise territory of another bottler.244 The restrictions against transshipping are vigorously enforced?45 Indeed they have become stricter over time, with increased penalties and tighter monitoring. 246 Because terrtories of the branded CSD bottlers are exclusive branded CSD bottlers outside of the San Antonio market would be contractually prohibited from selling packaged CSDs to a San Antonio customer that looked for an alternative seller outside the San Antonio market in order to a small but significant, nontransitory price increase by a San Antonio branded CSD bottler. Moreover, the impact of territorial exclusivity within the San Antonio market is highlighted by the fact that the major branded CSD bottlers in San Antonio also possess exclusive rights in various other portions of the immediately surrounding area beyond the ten-county San Antonio market. See, e. TOFF paragraph 275-77 (showing CCSW, Pepsi and Grant-Lydick' s exclusive franchises in Texas)247 Consequently, 242 See ex 102 G; ex 1666 A-E; ex 418 F; ex 1853 Zl; RX 2850 A; Howe!!, Tr. 4004; Dr Pepper Company, RX 2908 A-D; E. Hoffman, Tr. 381-82; Bodnar. Tr. nn; Nes!age, Tr. 8720; ex 574 A; ex 896 Z6; ex 891 K.

243 All major concentrate firms provide exclusive geographic tenitories for their pre-mix fountain syrup. Summers, Tr. 6894 Davis, Tr. 4470; Knowles, Tr. 267 ( 2681; ex 379 Z71-Z84; ex J 667 B- Strickland, Tr. 868\; Turner, Tr. 1086; ex 1406 Z5. While Pepsico provides exclusive geographic franchise terrtories for its post-mix fountain syrup (Davis, Tr. 4470; Knowles, Tr. 2670; Strickland, Tr. 8681; Turner, Tr. 1086), CCUSA and Dr Pepper do not (Knowles, Tr. 2681; Summers, Tr. 6895; Strickland, Tr. 8681; Turner, Tr. 1086; Cassagne, Tr. 7619; CX 418 K). 244 ex !667 A-D; CX 185321; Davis, Tr. 4473-74; Knowles, Tr. 2742; Turner, Tr. 1055; Schwerdtfeger, Tr. 2414- 15.

245 ex 1667 A-D; ex 300 A-B; ex 3432 B; RX 2850 B: ex 3414 A-C, RX 2908 B; Knowles, Tr. 2743-44; ex 3976 at 2111- 12 (QuirkJ; Summers, Tr. 690 I , 6920; ex 2203 A; ex 30 II; Schwerdtfeger, Tr. 2414- 15.

246 Davis, Tr. 4473-74; Little, Tr. 674- , 679. 247 See also RX 352 (eeSW); RX 2973 (Pepsi). Opinion 118 F. a collective price increase by the branded CSD bottlers in the San Antonio market would be far more difficult to defeat thartm-Qnventional markets.

San Antonio retailers uniformly testified that they would not purchase their branded CSD requirement" from an outside bottler Thiseven if the outside bottler offered substantially lower prices.248 is because retail accounts will abide by bottlers' geographic terrtory limitations, and therefore will not purchase outside of those terrtories or transship into their own territories, even if CSD prices were to go up significantly.'49 In addition, retailer transshipment is unlikely due Retailers are presumably alsoto the high cost of DSD delivery.25o reluctant to purchase transshipped products because the retailers would have to compensate for the loss of DSD marketing assistance,'SI a factor as important as price in the sale of CSDs '52 We therefore turn next to the possibility of unauthorized transshipments by branded CSD bottlers.

Importantly, the record shows very few, if any, significant instances of transshipment of branded CSDs into the San Antonio 248 ex 3963 at28-29lThurmondJ; E. Hoffman, Tr. 388-90; Davis, Tr. 4476;Chapman, Tr. 7213; Hiller, Tr. 5367; ex 3815 at 28-29 (JoynerJ; ex 3814 at 35 lAdamsJ; ex 3985 at 89 (Daub); ex 1853 ZI; Little, Tr. 659-60, 674- , 679.

249 Id.

Another possible explanation for retailers ' llnwiJ!ingn ss to purchase outside the San Antonio market if faced with a price increase may be that retailers do nO! absorb price increases charged by bottlers, but rather typically respond by passing the price increase along to the ultimate consumers. Kaiser, Tr. 3196; Chapman, Tr. 7212, 7255-56; Brinkley, Tr. 2235; Anderson, Tr. 3904; Davis, Tr. 4533; Donald, Tr. 5300 0!; Turner, Tr. 991. Because retailers would therefore likely pass on a 10% price increase (Anderson, Tr. 3904 Turner, Tr. 99!; Donald, Tr. 5300-01), profitability would be relatively unaffected and retailers would have little motivalion to undermine the increase. 250 unpac mg, an DS ehvery, whlc generally mvolves delivery to the actual retad outlet, reshclving (Turner, Tr. 955-56. 1530-31, 6414- 15), obviously entails high costs. See RX 0867 supra (study by CCUSA indicated that distribution accounts for about 35% of a bottler s costs). This would be even more true for retailer transshipment, where the retailer would also have to gather previouslydelivered bottles and cans and rcpack the trucks. The record evidence indicating that shipping costs are relatively low (Amrosowicz, Tr. 807, 859-60) refers only to the freight costs incurred in shipping in bulk quantities from warehouse to warehouse.

251 Bottlers pay for DSD, but retailers must pay for warehouse delivery. Summers, Tr. 6469. '52 Summers, Tr. 71 !9.

. . . . .

THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 579 452 Opinion Since it therefore appears that onlymarket by franchised bottlers.25 relatively small quantities of branded CSDs have been transshIpp into San Antonio, the issue is whether a price increase would induce transshipments in quantities sufficient to undermine that increase. We do not believe that this is a likely occurrence, for a variety of reasons.

First of all, as previously noted, exclusive territories are defined and enforced by concentrate companies, whicli strictly prohibit transshipping. Substantial penalties are imposed against an offender creating monetary disincentives against transshipping.254 Also, the offended bottler is compensated for the loss of sales 25 thus creating an additional incentive to monitor and enforce prohibitions against transhipping. In fact, CCSW has complained against a bottler for as little as ten cases of transshipped product. Transshipping is also relatively easy to detect. DSD delivery provides bottlers with day-to-day contact with retail stores, and bottlers can often identify products by means of date _codes and proprietary labels.'57 End-use consumers wil not undermine a price increase because it would not be worthwhile for most consumers to drive the substantial distance required to exit the ten-county San Antonio market 253In the rnitlal...decision,' the ALl listed the major Instances at transhIpping shown In the record. See IDFF paragraph 288- . Notably, however, these examples all involved transshipment between areas outside of San Antonio (ID ' paragraph 288-92; but see IDFF paragraph 293 (transshipping fines paid for undefined transgressions)) or by retailers (IDFF paragraph 294; RX 3121). None of them involved significant shipments of branded CSDs into San Antonio by franchised hottlers. See ex 3645 Z46 (69 cases of Coca-Cola and 20 cases of Dr. Pepper estimated to have been shipped into San Antonio over a two-month period in 1988). Moreover, the evidence of transshipment outside of San Antonio demonstrates that, in general, transshipment is minima! relative to total sales volume. Thus, for example the fact that CCE paid more than $1 milion in transshipment penalties in !989 (IDFF paragraph 293) pales in comparison to the company s sales of almost $4 billion (Standard and Poor s Register of Corporations 624 (J 990)); this penalty represents transshipments of far less than 1% of sales. Similarly, although SWCC received over $200,000 in transshipping penalties in 1986 and 1987 (IDFF paragraph 291), this is less than .2% of SWCC's 1989 net sales See CX 891 Z3; CX 1357 Z3. The 230 complaints against Pepsi COBO, mostly by the Oneta Company (IDFF paragraph 289), are likewise de minimis comparison to Pepsi CODO' s annual sales of around I J million ca. es (RX 1238) and Oneta s annual sales volume of around $) 0 millon, representing about 1.5 million cases per year (Koch, Tr. 1906-07; CX 4114; RX 3085).

254 ex 3432 B; ex 531; ex 538; ex 2927; ex 534; ex 539; ex 1667. An additional disincentive bottlers face is that parent concentrate companies may regulate the amount of concentrate that bottlers obtain in order to prevent them from transshipping. CX 1853 ZI. 255 Davis, Tr. 4822; E. Hoffman, Tr. 383; CX 379 Z41; ex 2327 A- 256 Summers, Tr. 6903-04; ex 2296 B.

257 E. Hoffman, Tr. 393-94; CX 3667 E.

Opinion J 18 F. simply to purchase CSDs at a slightly lower price. This commonsense conclusion is supported by ample record evidence, - Thus, with respect to the overwhelming majority of branded CSD sales for which exclusive geographic territories exist, we conclude that there is no competitive force that would effectively defeat a small but significant, nontransitory price increase on branded CSDs in the San Antonio market. We next examine whether the existence of nonexclusive geographic territories as to certain post-mix fountain sales requires any different conclusion.

Dr Pepper franchises assign exclusive territories for bottlers' sales of post-mix fountain syrup, but allow food wholesaler and broker sales within a bottler s exclusive territory.'59 Coca-Cola franchises do not grant any exclusive geographic territories for post-mix foun- In tain sales."'o addition, as noted earlier, both CCUSA and DPUSA control national account pricing for branded CSDs. 26I Thus, for postmix fountain sales of Coke and Dr Pepper products from outside a San Antonio market, customers may look to Coke bottlers outside of San Antonio, food wholesalers and brokers, and the parent concentrate companies. Other branded concentrate companies assign exclusive territories for post-mix fountain syrup sales. 26' These somewhat different facts do not lead us to any different geographic market definition, however. The Coke bottlers around San Antonio are CCSW itself and its sister corporation, SWCC; SWCC would be unlikely to constrain a price increase by CCSW. 263 Food wholesalers and brokers must obtain fountain syrup from some 258 The San Antonio market is a compact population center surrounded by large . sparsely populated areas. See ex 1684 C; see also ex 4131 , ex 4149. Indeed, a single county in this market, Bexar contains approximately 86% of the total tcn-county population. See ex 4131 A. This population distribution suggests limited alternatives for San Antonio consumers beyond the immediate market. Also, national and regional retailers view San Antonio as a separate retail market (see ex 3963 at 10- J J (ThunnondJ; Hiller, Tr. 5332, 5347; ex 3985 at 8-9 (Daub)), thus demonstrating that consumers tend to shop in this area, and not beyond.

259 Turner, Tr. 1086-87; Short, Tr. 7597 , 7619-20. 260 Knowles, Tr. 2681; Summers, Tr. 6895. See note 68 supra.

262 This includes sales of post-mix fountain of Pepsi, RC , and 7-Up, all of which grant exclusive geographic territories as to post-mix fountain syrup. Davis, Tr. 4470; Knowles, Tr. 2670, 2681-82; Turner, Tr. ! 086.

263 Both corporations are controlled by TBG. See Section II Stpra. CCSW' s current franchise territory includes San Antonio and sixty counties in southern, central, and eastern Texas. IDFF paragraph 275. SWCC is the Coca-Cola bonIer in west Texas. See Section II supra. . .

THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 581 452 Opinion source; in the face of a hypothetical collusive price increase by input suppliers in the San Antonio market, food wholesalers and broke as a practical matter, would have to rely on the same sources we have outlined above: bottlers with exclusive rights in other portions of the immediately surrounding area and national concentrate companies that also sold inputs in the immediately surrounding area. 264 We therefore conclude that, for the entire branded CSD product market, there is no competitive force that would effectively defeat a small, but significant and nontransitory price increase in the San Antonio market. Thus, we disagree with the ALl's assessment of the relevant geographic market issue.

Rather than attempting to ascertain whether branded CSD bottlers in the San Antonio market could collusively impose a small but nontransitory price increase, the AU instead relied primarily, and almost exclusively, on the Elzinga-Hogarty test of shipping patterns. ,"5 See 10 64-65. However, the Commission has previously found no basis for "definitive reliance" on the Elzinga-Hogarty test to establish a geographic market under the Clayton Act. Adventist Health System, 5 Trade Reg. Rep. (CCH) paragraph 23 591 at 259. Consequently, the Commission "does not. . . endorse either the ' strong' or the ' weak' test as the basis for establishing a relevant market. Id. at 23 260.

Shipping patterns, whether analyzed using the Elzinga-Hogarty methodology or in some other fashion, clearly constitute one source of information in analyzing the possible exercise of market power. But other evidence is equally relevant. Adventist Health System, 5 Trade Reg. Rep. (CCH) paragraph 23,591 at 23,259. In other words shipping patterns are only one of many surrogates for assessing market power (see, e. g., B. F. Goodrich Co. 110 FTC at 289), and 264 Although in theory food wholesalers and brokers ight purchase fountain syrup from far outside the San Antonio area and ship it into San Antonio, the record is silent on whether it would be cost effective fof wholesalers or brokers to do so in the face of a 5% or similar price increase on fountain syrup.

We arc not implying that national concentrate companies would necessarily participate in any collusive branded CSD price increase; that is an issue we wil address later. See Section VI. infra. Rather, we are simply assessing the a!tematives available to customers in the face of a hypothetical collusive price increase in post-mix fountain syrup. If the national concentrate companies participatcd in such a price increase, then it would be highly unlikely that they would undermine their own price increase in San Antonio by permitting food wholesalers and brokers to obtain fountain syrup at a lower price outside San Antonio.

See Kenneth Elz!fga & Thomas Hogarty, The Problem of Geographic Market Del1neatlOn Revisited: the Case of Coal, 23 Antitrust Bu!!. I (!978 ; Kenneth Elzinga & Thomas Hogarty, The Problem of Geographic Market Delineation in Antitrust Suits, \8 Antitrust Bul 45 (\973). g., Opinion 118 FTC. therefore should not be overemphasized, as the ALl erroneously did here in describing them as "perhaps the best test in determining a geographic market. " ID 64.

Moreover, the Elzinga-Hogarty test is less relevant to settings like this one, where territorial exclusivity imposes legal and contractual impediments to transshipping by competitors, 266 By virtue of exclusive territories, legal bottler shipments from outside the geographic area (e. contract packing) are controlled by franchised bottlers within the area; 26? other shipments are in violation of contract. Shipments from outside the San Antonio market that are under the control of a hypothetical collusive group would obviously not be used to defeat a price increase. Because the Elzinga-Hogarty test nonetheless takes such shipments into account, the test is an especially imperfect measure of market power in this case. Finally, to the extent that the Elzinga-Hogarty test has some limited value in the present context the ALl completely undermined the test by using the wrong product market, thus skewing the analysis. 268 Relative prices and movements of those relative prices are additional surrogates for the ability to exercise market power and, as such, can be useful considerations to assist in defining a geographic market. See B. F. Goodrich Co. 110 Ftc at 289. If prices of branded CSDs in the San Antonio area moved together with one another, and independently from prices in other areas, this would support the conclusion that there is a San Antonio market. 266 Hilke, Tr. 6240-41.

267 CCSW produces cans in its Cuero facility outside the San Antonio market for distribution in San Antonio and elsewhere throughout its franchise tCITitory. Summers. Tr. 6403-04. Grant-Lydick purchases its products from contracl packers outside the San Antonio market. Turner. Tr. 929, (! 17; Bodnar, Tf. 152627, 1557; Campbell, Tr. !926, 1987; Espinoza, Tr. 4248-49. Pepsi COBO also imports both bottles and cans from outside the San Antonio market. Davis. Tr. 4461- , 4464, 4630- 32. 11 il shipments a ran ed SDs mto and out 0 the San Antonio market! at were not controlled by the franchised San Antonio bottlers were taken into account, the LIFO calculations would ral1ge from 78% to 85% , indicating a market. See CX 4089. (If al1alyzed in this manner, we would cO!1sider LOB calculations iITe!evant to the question of whether a price increase could be constrained in the present case.) Moreover, if an ElzingaHogarty analysis were conducted to measure shipments of take-home branded CSDs into and out of the San Antonio market that were not controlled by the franchised San Antonio bottlers both the UFO ("little in from outside ) and LOFI ("little out from inside ) figures would approach 100%, thus satisfying even the most stringent Elzinga-Hogary test. RX 3062 A.

In reaching this conclusion, we reject respondent s assertion that this is a "tauto!ogical" approach that "assum(esj away the data of any supplier whom (the Commission) might choose to include in the market." ABR-A 43. We simply believe it is inappropriate to reduce the LIFO and LOFI numbers by including shipments into or out of San Antonio that are controlled by the franchised San Antonio bottlers, because such shipments clearly would not be used to defeat a price increase. !:.

THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 583 452 Opinion While the only systematic price data in the record show differences in percentage price changes between the areas of San Antonio '69 there is also at least some weak evi- Waco, and Corpus Christi dence suggesting price uniformity throughout at least some bottling territories.no On balance, we tind the evidence on relative prices and price movements to be weakly supportive of complaint counsel, but We do not relatively unreliable, and therefore we do not rely on it27 view the essentially inconclusive nature of this evidence as significant, however, because the more probative evidence strongly points to the existence of a San Antonio market, and also because the confounding effects of ad features and in-store displays would, in any event, limit our ability to discern true price variations'72 Finally, it is clear from the record that recognition of a San Antonio market comports with both economic and geographic realities. From an economic perspective, a number of trade and marketing factors support this market definition. For example, national and regional retailers view San Antonio as a separateIetail market. These retailers run localized advertising and marketing campaigns that treat the San Antonio market as a separate and distinct marketing area.'74 Retail prices and sales of CSDs in the San Antonio market are compiled separately, and compared to prices and sales in other The behavior of retailers thus constitutesgeographic markets.'75 strong confirmation of the existence of a San Antonio market. Viewed from a geographic and demographic perspective, a San Antonio market is eminently sensible. The San Antonio area is a See. e. ex 3999 A, E.

270 See, e. Summers, Tr. 6711 , 6719 RX 2985. 271 In reaching this conclusion, we reject the All' s essentially unsupported finding that CSD prices "arc uniform in a trade area beyond the tcn-county area. !D 66. Although the AU found that l-I.E.B. preferred a uniform price throughollt its territory. the cited testimony demonstrates that local competitive conditions generally prevcnted this (Chapman, Tr.7247. 7200-0J). The exhibit (RX 2985) cited for the proposition that numerous bottlers (Pepsi COBO, CCSW, CCE and Grant-Lydick) offered H.E.B. unifom1 pricing applies only to Pepsi eOBO, not to the other companies, and fails to reflect the fact that those prices were often not accepted. See ex 41 11; Hilke, Tr. 8507-08. 272 Ano( er problem WIt pnce movement ala IS that prlCCS change seasonally. so that pnces of different brands and products wi!! reflect this seasonality whether they arc in the same market or not. This means any simple "statistical analysis" of sign changes can be misleading, and is yet another reason why we believe that comparsons of price movements arc of Itttje value here. Hiller, Tr. 5333 , 5347; ex 580; ex 3963 at 10- 1! !ThurmondJ. 274 Chapman, Tr. 7198-200; Kaiser, Tr. 3187-89; Hiller. Tr. 5347; ex 1054 P- 275 ex 2263 F- , U-Z6; ex 580; Kaiser, Tr. 3188-89; ex 1014 A- Opinion 118 F.TC. compact population center, with 86% of its population in a single county, that is surrounded by large, sparsely populated ars.'76 By virtue of this population distribution, consumers in the San Antonio market would appear to have only limited realistic alternatives beyond the immediate market.

In sum, we conclude that the ten-county San Antonio market is the relevant geographic market within which to assess the challenged acquisition.

VI. THE LIKELY COMPETITVE EFFECTS OF THE ACQUISITON The purpose of Section 7 of the Clayton Act is to prevent mergers or acquisitions whose effect "may be substantially to lessen competi- To fulfill tion, or to tend to create a monopoly."m this purpose, we seek to discern whether a particular transaction is likely to create or enhance market power or to facilitate its exercise. Market power" is "the ability profitably to maintain prices above competitive levels for a significant period of time " or to "lessen competition on dimensions other than price, such as product quality, service, or innovation. ,,279 In certain circumstances, firms may exercise market power jointly through collusive conduct. Thus, one prong of our inquiry focuses on whether the transaction under scrutiny here may enable the acquiring firm to cooperate (or cooperate better) with other leading competitors in raising price or reducing output or colluding on other aspects of competition.280 In other circumstances, a firm may exercise market power unilaterally by raising price and reducing output.281 Thus, the other prong of our inquiry focuses on whether the acquisition at issue here may facilitate the exercise of unilateral market power. ' The AU found that, since complaint counsel had failed to establish a relevant product market, an accurate measure of concentration 276 See note 258 supra.

277 15 U. c. 8. Mergers arc subject to Section 5 of the Federal Trade Commission Act if they constitute an "unfair method of competition, 278 Merger Guidelines , Section O. J.

279 Merger Guidelines , Section 0. 1 & n. 6; Owens-Illinois, slip op. at 4-5 (quoting 1984 Guidelines).

280 See HCA v. FTC 807 F.2d at 1386; B.F. Goodrich J 10 FTC at 294. 281 Merger Guidelines, Section 2.

, THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 585 452 Opinion levels was not possible, and that, in any case, there was a "wealth of proof of competition" in CCSW' s trade. ID 67. As set forth-abve we find that the ALl erred in his assessment of the relevant product and geographic market. Using the correct relevant market -- branded CSDs in San Antonio and the immediately surrounding counties -we find that there is ample evidence of the likelihood of competitive harm from the acquisition at issue here, both in terms of likely coordinated interaction and unilateral effects. A. Market Concentration "2 In United States v. Philadelphia National Bank the Supreme Court noted that a crucial initial question in merger cases is whether the transaction at issue "produces a firm controlling an undue percentage share of the relevant market, and results in a significant increase in the concentration of firms in that market, (such that) it is..... inherently likely to lessen competition substantially...." 374 S. at 363; accord, B.F.Goodrich 110 FTC at 303-304. The transaction at issue in this case raised concentration levels significantly in an already highly concentrated market, as measured by the Herfindahl-Hirschmann Index ("HHI")2RJ The following are the pre- and post-acquisition HHIs in the relevant market: Pre-acquisition HHI 2807 Post-acquisition 3421 HHI Increase 614284 Under the Merger Guidelines (wJhere the post-merger HHI exceeds 1800, it will be presumed that mergers producing an increase in the 282 372 U.S. 321 (1963).

283 The HHI is calculated by summing the squares of the market shares of the market participants. The HHI ranges from 10,000 in a pure monopoly to near zero in a purely atomistic market. MergerGuidelines, Section 1.5 & n. 17.

284 ex 4146A. H. These data were provided by complaint counsel. but they are based on data used by respondent s expert, Dr. Strickland. in RX 3057 and RX 3058, with adjustments made to equate fountain units with package units. We relyon these data, rather than complaint counsel' s proposed HI-I calculations because these data include CCUSA, DPL'SA, and fountain wholesalers in the market as sellers of post-mix fountain syrup. We agree with respondent that those post-mix fountain sales must be attributed to the entity that sets the price for the sales, not to CCSW, although CCSW does deliver many of these sales for a delivery fee from the parent concentrate company. See Summers, Tr. 6500-6501.6507. , _ Opinion 118 FTC. HHI of more than 100 points are likely to create or enhance market power or facilitate its exercise." Merger Guidelines, Seon 1.51. These figures show that the relevant market was highly concentrated before the acquisition and became significantly more so as a result of the acquisition.

The resulting post-acquisition HHls are in the same range as or higher than those in most of the cases in which the Commission has successfully litigated a challenge to a merger or acquisition in the last ten years. '" For example, they significantly exceed the HHIs in the VCM market in P. Goodrich, which were found to justify a "relatively strong presumption of anticompetitive effects."'86 These HHIs which are much larger, create a strong presumption of possible anticompetitive effects; thus, relatively strong evidence from other factors will be necessary to rebut that presumption'87 B. The Significance of Increased Concentration The ALJ and respondent assert that these HHIs do not carry the same significance as other HHIs because, although they show a large increase in concentration, the number of market participants has remained the same. This argument ignores certain aspects of the information conveyed by HHIs, information that is particularly crucial to an accurate understanding of competition and the likeli- :!85 See Coca-Cola, slip op. at 44 (HHI increase of 443 to post-merger HHI of 3572); Occidental Petroleum Corp., Dkt. No. 9205 (Dec. 22 1992), slip op. at 27 (post-acquisition HHI in one market of 1305 with increase of 158 points); Owens- Ilinois. slip op. at 27 (using production figures, postacquisition HHI of 2478 with increase of 852 points); Olin Corp.. 113 FTC 400. 6! 0- 11 (J 990), (iffd, 986 F.2d 1295 (9th Cir. 1993), cert. denied J 14 5-. Ct. 105! (1994) (based on production, pastacquisition HHI of4122. with increase of 1186); Hosipital Corp. of America 106 FTC 361 . 488 (1985) (post-acquisition HHI of 2416 with increase of 395 points). 286 In B.F. Goodrich, the paries presented various measures of the HHIs in the VCM market (e. nameplate capacity, practical production capacity, and actual production). The highest HHI figures were those for actual production, which showed an HHI increase of 304 to produce a post-acquisition HHI of 1663. P. Goodrich I ! 0 FTC at 3 I 3. The Commission found that the data were " well above those that created a presumption of illegality in United States v. Geneml Dynamics and Weverhauser and that the data supPoJ1ed a "relatively strong presumption of anticompetitive effects." 110 FTC at 314. 287 See PPG, 798 F.2d at 1502-03 (acquisition resulting in 1352 point increase in HI-i! to postacquisition HHI of 3295 put merger "well within the range where abst;nt really extraordinary circumstances, the Department and the Commission will proceed against an acquisition under section 7 of the Clayton Act on the theory that the increased concentration raises a likelihood of ' interdependent anticompetitive conduct '" rcitations omitted)); F. Goodrich. 110 FTC at 314. THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 587 452 Opinion hood of collusion among and/or unilateral anticompetitive conduct by branded CSD The HHI conveys informationbottlers.288about - both the number of market participants and the size disparity of the market shares among market participants. As explained by then-Judge Bork writing for the Court of Appeals for the District of Columbia Circuit in FTC v. PPG Industries, Inc.:

Market power or the lack of it is often measured by the HHi. The FTC and the Department of Justice, as well as most economists, consider the measure superior to such cruder measures as the four- or eight-firm concentration ratios which merely sum up the market shares of the four or eight largest finns. The HHI, by contrast is calculated by squaring the individual market shares of all firms in the market and adding up the squares. This method, unlike the four- and eight-firm concentration ratios, shows higher market power as the disparity in size between firms increases and as the number of firms outside the first four or eight decreases. PPG, 798 F.2d at 1503 (emphasis added). Market share, of course, is an initial proxy for market power, since we typically have no direct means by which to measure market power. One premise underlying antitrust jurisprudence is that, absent other factors, a firm s market power is likely to increase as its market share increases, and that its market power relative to other market participants increases as its share becomes disproportionately larger than the shares of other market participants .'89 In this case, the three main soft drink bottlers in the relevant market stayed the same -- CCSW, Pepsi COBO, and Big Red Bottling (now owned by Grant-Lydick) -- and the other sellers of postmix fountain syrup (CCUSA, DPUSA, and fountain wholesalers) also remained the same. However, the acquisition increased CCSW' s pre- 290acquisition market share from 44.7% to 54.5%. We conclude, based on the record, that CCSW' s acquisition of the Canada Dry franchise, which accounted for only about 1% of this market share increase, had no anticompetitive effect. If only the 288 As we discuss in Section VI.c.! infra we do not rind that tacit collusion among the branded CSD bottlers in the relevant market would likely be prevented or disrupted by the other market participants that sell post-mix fountain syrup that is, the parent concentrate companies and food wholesalers and brokers.

289 See Warner lambert Co,. 87 FTC 8 J 2 , 870 (1976); see (dso Heublein, Inc. 96 FTC 385 , 577 n. 10 (1980); W, Shepherd, Market Power and Economic Welfare 40 (1970). 290 CX4146H.

Opinion 118 FTC. Canada Dry franchise had been transferred to CCSW, the pre- and post-acquisition HHIs would be as follows: Pre-acquisition HHI 2807 Post-acquisition HHI 2862 HHI Increase See CX 4146 H; CX 4079. Under the. Merger Guidelines, such a change would be viewed as "potentially rais(ings significant competitive concerns. . .." Merger Guidelines Section 1.51. In terms of the competitive issues we discuss next, however, we find that virtually no evidence exists to demonstrate that a one-percent increase in CCSW' s market share due to an acquisition of the Canada Dry franchise would provide CCSW with significantly greater market power than it already had and thus would substantially lessen competition. We note that a one percent -- or even less -- market share increase might have competitive significance in circumstances where the one percent was being combined with several other low-percentage shares. In this transaction, however, it is clear that the 8.6% market share increase from the Dr Pepper franchise acquisition is the true source of the likely anticompetitive effects that we describe in the following sections.

The acquisition of the Dr Pepper franchise increased CCSW' market share by about 8.6%.291 This acquisition changed the number of product offerings that each firm had available and thus changed CCSW' s and Big Red Bottling s relative costs of and advantages with respect to producing and marketing their branded CSDs. As we explain in more detail below, such changes can significantly affect the ability and incentive of smaller bottlers such as Big Red Bottling to compete. In this ca, the evidence confirms that CCSW' s acquisition of the Dr Pepperfranchise provided CCSW with increased market power and left Big Red Bottling facing significant disadvantages?92 As we discuss further below, this situation increases both the likelihood that CCSW and the Pepsi COBO could tacitly and successfully collude with Big Red Bottling -- since Big Red Bottling would have little or no ability or incentive to do any- 291 SeeCX4146H;CX4079B;CXI681C.

292 See Section VLC.2 infra.

THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 589 452 Opinion thing other than follow -- and the potential for an exercise of eral market power by CCSW. unilat- C. The Likelihood Of Successful Collusion 1. The Market Participants Required for Successful Tacit Collusion The first issue to be addressed is whether the presence of CCUSA, DPUSA, and fountain wholesalers as sellers of post-mix fountain syrup could prevent or disrupt any tacit or explicit collusive arangement among the bottlers of branded CSDs in the San Antonio market. We find that fountain wholesalers would be unlikely to disrupt a hypothetical collusive arrangement that included bottlers and the parent concentrate companies, because fountain wholesalers must obtain fountain syrup either from bottlers or from the parent concentrate companies. Thus, the key issue is whether CCUSA and DPUSA might be likely to participate in a collusive arrangement with botters as to San Antonio sales of branded CSDs, including post-mix fountain syrup.

The evidence indicates that it is highly unlikely that either CCUSA or DPUSA would have the incentive to become par of a collusive arangement in the San Antonio area. Both CCUSA and DPUSA use "national account" pricing for their post-mix fountain sales -- that is, pricing that is uniform across the geographic areas in which the chains and other purchasers of post-mix fountain syrup operate.'93 CCUSA representative Short testified that this is one of the advantages perceived by the chain customers.294 Thus, there does not appear to be any incentive for either CCUSA or DPUSA to deviate from their national account pricing solely in the San Antonio area. Moreover, such a deviation would be highly noticeable and presumably hard to justify.

This does not mean that CCUSA and DPUSA post-mix fountain syrup sales would necessarily be sufficient to constrain an overall collusive branded CSD price increase instituted by the bottlers however. First, as to post-mix fountain sales, both the CCUSA and the DPUSA representatives testified that they did not compete with 293 Short, Tr. 7739, 7797; Knowles, Tr. 2820. 294 Short, Tr. 7797.

, Opinion 118 FTC. the bottlers for the accounts to which the bottlers sold post-mix fountain syrup, because the bottlers typically served smal!eccounts Ththan those served by CCUSA or DPUSA.'95 s, there is - some reason to doubt that CCUSA or DPUSA would respond aggressively to a collusive branded CSD price increase by bottlers. Second because post-mix fountain products (the channel in which CCUSA and DPUSA are present) are differentiated from other branded CSD products, it does not appear that expanded sales of post-mix fountain syrup by CCUSA and DPUSA would be sufficient to constrain a collusive overall branded CSD price increase by bottlers. Therefore we turn next to whether CCSW's acquisition of the Dr Pepper franchise increased the likelihood of tacit collusion by branded CSD bottlers in San Antonio.

2. The Increased Likelihood of Tacit Collusion in the Take-Home Channel: Ad Features The record suggests that CCSW' s acquisition of the Dr Pepper franchise may have had an effect on Big Red Bottling s ability to obtain ad features, a significant element of competition. As we explained earlier, the most significant discounting and volume generation for take-home sales of branded CSDs occurs through ad features296 The loss of significant franchises could reduce the ability of the smaller bottler to obtain ad features in retail chains, a key component in effective competition among branded CSDs. Without ad features, price decreases have much less effect on attracting volume.'9? Moreover, even if the smaller bottler were still able to 295 CCUSA will provide national account pricing to any qualified entity with five outlets; DPUSA will provide national account pricing to any qualified entity with three outlets. Short, Tr. 7735- 36; Knowles, Tr. 2821. Mr. Short of CCUSA testifi Q. SO then are you in competition with Coke Southwest for fountain accounts? A. Not for fountain accounts, no. Not to pick up a fountain account. I have a segment of the business that I -- We manage the whole business. We allow them to go manage the local side of the business, and that s the part they manage. But we don t compete fof that business. Does Coke Southwest compete for your national accounts?Q.A. No. Short, Tr. 7800-01. Mr. Knowles of DPUSA testified that, for fountain sales not covered by DPUSA' s national account price (tJhat s basically the bottler selling up and down the street to the buy downstairs, and he doesn t have a contract. So, I mean, who knows what he s paying for his syrup." Knowles, Tr. 2820. Indeed, Mr. Knowles testified that the prices at which bottlers sold post-mix fountain syrup probably already ran higher than DPUSA' s national account price, but that he didn t know because "(wJe just don t get into it." Knowles; Tr.- 2824. 296 See Section IV.C.3 supra. 297 See id THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 591 452 Opinion obtain some ad features, the loss of significant franchises might mean that those ad features would cost the smaller bottler significatly more than was previously the case -- another deterrent to effective competition. In addition, smaller retail outlets with limited shelf space are more likely to car the high volume brands, other things being equal. 298 Conversely, the addition of significant franchises to the holdings of a larger bottler such as CCSW could increase its advantages in terms of ad features. As a result, the smaller bottler may become less wiling to challenge the market strength of the larger bottler through vigorous price competition and more willing to become a follower of noncompetitive market activity. We next examine whether the evidence demonstrates an increased advantage for CCSW and a decreased ability by Big Red Bottling to obtain ad features. a. Big Red Bottling s Loss of "Critical Mass Ad features involve significant retailer advertising of specially discounted products. Retailers use ad features offering sharply discounted branded CSD prices as a means to "pull" customers into their stores. For example, Kaiser representative Mr. Kroger testified that: We consider soft drinks the best customer count produced of any feature we run. It is the best item in grocery to run as a feature. ,,209 Thus, branded CSDs are used by retailers as volume generators and 00 They are often sold at cost orto increase consumer foot traffc 01 Aseven as loss leaders in order to generate retail store volume. with retailers nationally, San Antonio retailers recognize CSDs as one of the largest, if not the largest, retail food item and promote themaccordingly.302 298 See id.

299 Kaiser. Tr. 3231-33.

300 Davis, Tr. 4709; Turner, Tr. 1206; ex 3815 at 153 (Joyner); Anderson" Tr. 3840- , 3896; Chapman. Tr. 7256; Clarke, Tr. 4280; Brinkley, Tr. 2188; Knowles, Tr. 2840; ex 382J at 48 (ImperJ; ex 3814 at 54 (Adamsl.

3m Kaiser, Tr. 3185-86; Coyne, Tr. 3485-86; Chapman, Tr. 7256; TUrner. Tr. 973- 1206-07; Anderson, Tr. 3840-41; Clarke, Tr. 4280; Donald. Tr. 5289. 5297-98; Gonzaba, Tr. 2085; Brinkley, Tr. 2188; Sendelbach, Tr. 7696; Bodnar, Tr. 1570. These low prices are often known as "hot prices. Howell, Tr. 3952. See also lDFF paragraph 425 ("H. B. uses soft drinks as a loss leader 302 Sendelbach, Tr. 7695; Donald, Tr. 5288; Brinkley, Tr. 2187. San Antonio retailers advertise and promote branded CSDS, often at prices which are at or below cost. Sendelbach, Tr. 7698;. Anderson, Tr. 3841: Turner, Tr. 973-74.

Opinion 118 F.T. For branded CSD bottlers, the first priority for promotions is to get into the ad cycle. The second priority would be to ha in-store displays.'OJ A bottler cannot grow its brands without attaining the volume lift benefit associated with an ad feature in the ad cycle; instore specials alone are not enough to obtain the necessary volume increases..'o4 As discussed in Section IV supra bottlers are aware that ad features give much more volume "lift" than do in-store displays. 305 Without "critical mass" or market share, however, a bottler ability to get into the ad cycle is reduced because the retailer will not give a week of its ad cycle to a product that will not attract significant numbers of customers .'06 "Critical mass" as related to advertising means that a bottler has a significant enough market share and consumer appeal that retailers believe it draws customers into the store..'07 The more products or flavors a bottler has in its stable products, the greater the overall ability it has to sell product. .'o8 If a bottler only has a single brand, it requires significant brand equity or In order to put a: CSD in an ad cyclerecognition to sell products..'o9 the retailer must be convinced that the CSD would be a good The bottler must have the market share or "customer draw.'lo pullthrough" necessary to obtain the critical mass necessary to get into 303 E. Hoffman, Tr. 366.

304 Turner, Tr. 974 305 Bodnar, Tr. 1498; Turner, Tr. 974; E. Hoffman, Tr. 362. An ad feature may give a bottler!O times the non-featured sales volume, Bodnar, Tr. 1498; Davis, Tr. 4504; Koch, Tr. J 831 while an instore display givesjusl twice to 21/2 times the normal sales volume, Bodnar, Tr. 1498. A month long display at an attractive price produces close to the same volume as a one week ad, Bodnar, Tr. 149R. The volume !ift is much lower on the instore display because the retail price to the consumer is usually higher. Turner, Tr. 974; E. Hoffman, Tr. 362-63. e also Section IV.C.3 .'t/pra. 306 Turner, Tr. 1040-44; ex 3941 at 287 (Schmidj. 307 Turner, Tr. 1040 308 CX 3989 at 37 rShanksJ.

309 CX 3989 at 37 rShanksj.

For example, Mr. Kaiser of Kroger testified that he would probably not run Dr Pepper on its own as an ad feature because "(ilt would be too weak on its own to offset a Pepsi and/or a Coke feature, Kaiser, Tr. 3232-33. Although Dr Pepper has received a few exclusive ads from some of Kroger competitors, RX 438, even Dr Pcpper s own study advised that it should be advertised with Coke to build sales. RX 2825 C.

310 Gonzaba, Tr. 2053. Feature support is very expensive unless there is enough volume to justify it. Coyne, Tr. 3480; CX 97!.

THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 593 452 Opinion the promotional rotation. '" A bottler is "locked-out" when it receives no promotional period during a particular calendar span. The testimony of Mr. Kaiser of Kroger is illustrative: Q. Now in selecting a particular brand to be in an ad-buy program, how important is the customer draw of the product that s being put into the ad? A. The most important consideration we have is how strong the brand is and how many cases we can sell of it.

The evidence in this case shows that, after Big Red Bottling lost the Dr Pepper franchise, the Big Red bottler became significantly less able to obtain ad features at major supermarkets than it was before the acquisition. Mr. Bodnar, former General Manager of DPSA and currently Executive Vice President, General Manager, and owner of Grant-Lydick, explained that, pre-acquisition, the Big Red bottler had just acquired "critical mass A. You know, you have to have the necessary market share or pull-through of the products you represent to have the critical mass, if you will, to get into a promotional rotation, to get the shelf space that you need or the promotional efforts behind the brands that you represent. Q. At what point. . . did you acquire critical mass. . . A. I'm going to say in 1983 we stared to acquire it with the addition of RC Cola and the fact that we had a cola to offer... . Come mid- 1983 we started to get feature ads from the major chain supermarkets on a regular basis, again; because of the lineup of brands. . . . would say we more than tripled our ad rate.

Bodnar, Tr. 1254- 55. See also Turner, Tr. 1043. This situation changed drastically after the Dr Pepper franchise went to CCSW however.

Q. On average how many ads did you get during the course of the year pre (sick? A. With the major chains -- and by that I mean H. , Kroger, Handy Andy, Albcrtsons, and Warehouse Grocery -- we were averaging a minimum of one chain per month.

Q. And then after? A. Never got a Kroger or an Albertsons ad. . . . We did have an ad with Kroger in part of the stores. . . . So half of an ad. . . . Handy Andy, the tirst year I 311 Bodnar, Tr. 1254. It takes Jess critical mass to obtain in-store displays. Turner, Tr. 1043. 312 Davis, Tr. 4740.

313 Kaiser, Tr. 3231-32.

Opinion 118 FT. would say we continued to get one a month. Warehouse Grocery, one a month. And H.E.B., maybe three ads for the year. Bodnar, Tr. 1308-09. CCSW' s own document confirms Mr. Bodnar s recollections.314 Indeed, CCSW has conceded that "(iJn recent years, CCSW and Pepsi COBO have used their CMA programs to obtain the majority of ad features offered by San Antonio area retailers." RPFF paragraph 595. Mr. Bodnar testified that Big Red Bottling tried to interest San Antonio retailers in CMAs but was unsuccessful, because Big Red Bottling lacked the necessary volume throughput.3I That Big Red Bottling has been able to obtain some ad features does not negate the fact that the large majority of ad features have gone to Coke or Pepsi in the San Antonio area. 3I6 b. Big Red Bottling s Increased Costs Dr. Hilke testified that there are economies of scale associated with several aspects of the branded CSD bottling industry and that 314 ex 2954 H lists the number of feature ads in various San Antonio retailers for 1984 and 1985 for five branded CSDs: Coke, Pepsi, Dr Pepper, 7Up, and Big Red. As Mr. Bodnar recalled, it indicates that Big Red continued to receive about one ad per month from Handy Andy, but that the number of ads from other major chains sllch as AIbertsons and H. B. had sharply declined from 1984 to 1985. The precise amount of the decline is difficult to discern, because it is not possible to know how many of the Dr Pepper feature ads took place while the franchise was sti!1 held by the same entity as Big Red (DPSA), prior to the August 1984 acquisition of the Dr Pepper franchise by CCSW. However, one can compare the 1984 and 1985 totals with Dr Pepper as pari of CCSW and, hypothetically, if it had remained as one of the franchises held by the Big Red bottling operation. According to CX 2954 H, Big Red and Dr Pepper combined throughout all of 1984 would have had 57 feature ads; by contrast, Big Red alone in 1984 would have had only 6 feature ads. For 1985, Dr Pepper and Big Red combined actually would have had 80 feature ads; Big Red alone actually had only 43 feature ads. CCSWalone had 211 feature ads in each of 1984 and 1985; Pepsi alone had 62 feature ads in 1984 and 123 in J985. Even these comparisons do not show the full extent of the decline for Big Red, but a look at the stores at which Big Red continued to obtain feature ads shows that they are the smaller retailers, not the larger ones like HEB.

315 Bodnar, Tr. 1383- 316 CX 2954 B indicates that, in 1985, for chain supermarkets, Big Red obtained promotions accounting forabout5.3% of a!! commodities volume, as compared to Coke s46.9% and Pepsi' s 20.9%. For the number of store weeks of ads in major independents, Big Red' s share was somewhat higher-- 9%, as compared to Coke s 54.6% and Pepsi' s 18.6%. However, for share of convenience store ad months, Big Red was practically shut out as was Pepsi -- 5.7% (Big Red) and 5.9% (Pepsi) as compared to Coke s 85.6%. For drugstores and ma. s merchandisers, Big Red had a 9.6% share of store weeks compared with Coke s 49.4% and Pepsi' s 41. 1 %. CX 3248 A-E shows that Big Red was able to obtain about 15-20% of promotional activity in the summer of !985. RX 1678 lists some small independents that gave ad features to Big Red in 1988. A 1989 CCUSA survey found that, for total supermarket displays (not just ad features) in San Antonio. Big Red accounted for 13.6% of the total displays, compared to Coke s 55.4% and Pepsi' s 26.6%; the document shows Big Red as totally shut out of convenience store displays. RX 256 B, C. Mr. Bodnar testified that Stop- Go ran only Coke features in 1987 and 1988, as did other convenience stores. Bodnar, Tr. 1381. THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 595 452 Opinion these economies of scale seem to have increased over time. 31 These economies of scale have been quantified in a study sponsored by the National Soft Drink Association ("NSDA") and executed by the The NSDA report indicates Boston Consulting Group ("BCG")318 that economies of scale are present in at least three major aspects of bottler manufacturing operations: direct labor, equipment, and According to the NSDA stu materials costs.3I9 y, decreasing bottling output from 4 million cases to 2 millon cases, for example, would on average, entail an increase in the total of these costs from roughly $2.40 to $2.60 per case. 0 This represents an increase of about 8%.-21 In addition, as respondent admitted32, the per case distribution costs for a botter generally decrease as the volume or market share of the bottler increases.

The acquisition in this case resulted in the transfer of approximately 42% of DP-SA volume to CCSW prior to the sale of the remaining franchises and assets to Grant-Lydick.32 Mr. Bodnar noted that it was not until 1989 that Grant-Lydick Beverage Company matched DP-SA' s pre-acquisition sales volume level. This did not occur until after Grant-Lydick acquired a number of additional brands, including 7-Up, Dad' s Root Beer, Squirt, and Yoo Hoo Chocolate and five more sales locations in 40 additional counties. In a March, 1987 letter to the Federal Trade Commission, Grant- Lydick Beverage Company supplied pre- and post-acquisition revenue and costs estimates on a per case basis 32r' Grant- Lydick estimated that its average total cost per case increased from $6.37 per case in 1984 to $6. 90 in 1985. Grant-Lydick further estimated that 317 Hilke. Tr. 6054- , 6042-43; ex 1671; ex 1696.

318 Hilke, Tr. 6102-05; ex 1697.

319 ex 16971- 320 ex 1697 K.

321 ex 1697 RRCCPFF paragraph 1465.

323 ex 394! at 288 (Schmid). See a150 RX 0867 (CCUSA study indicated that distribution typically accounts for about 35% of a bottler s overall costs). 324 ex 4079.

325 Bodnar, Tr. 1347; ex 3830.

326 ex 1697 E- 327 CX1697F.

g., Opinion IISP. its production labor cost per case increased from $. 12 in 1984 to $.21 in 1985 and that its production overhead cost per case increased from 37 in 1984 to $.45 in 1985. ' CCSW disputed these figures claiming that Grant-Lydick did not sustain any substantial increase ih total cost as a result of losing the Dr Pepper and Canada Dry franchises. 329 However, even CCSW conceded that Grant-Lydick' s operating costs increased 7.9% from 1983 to f986, .Iargely as a result of "JJO Although respondent char-increased "promotional variable cost. acterizes the increased costs of promotion as resulting from increased bottler competition producing higher rebates to retailers, we find it more likely that this substantial increase in promotional costs occurred because, without the Dr Pepper franchise, Big Red was being Retailers expect better offersrequired to pay more for promotions.33I on ad features from bottlers whose products do not sell as much volume as those of other bottlers; as Mr. Kaiser of Kroger testified: Generally Pepsi wil offer more than Coke (on ad feature payments Thus, inper casel because they don t sell as much product."m addition to making it more difficult for Grant-Lydick to obtain ad features at all, the loss of the Dr Pepper franchise increased the cost to Grant-Lydick of competing against Coke and Pepsi in obtaining ad features -- the most significant means by which to obtain increased sales. J3 c. The Likely Competitive Effects The evidence demonstrates that CCSW' s acquisition of the Dr Pepper franchise significantly impaired the ability of Big Red Bottling to compete with Coke and Pepsi for ad features, the form of competition that generates by far- the largest volume of sales for branded CSD bottlers and retailers. This diminished ability to compete in such an important arena of branded CSD competition would 328 ex 1697 329 Goode, Tr. 7424; RX 200; RX 20!.

330 RRCCPFF paragraph 2047 (citing RX 200); ex 4056. 331 The evidence that shows increased promotional costs for branded CSD bottlers relates to the !986 time period, not to the 1983- 86 time frame. See. e. !OFF paragraph 172 . 173. 309. 332 Kaiser, Tr. 3210. See also ex 129, ex 3814 at 28-29 (Adams). 333 See Section IV C.3 slIpra.

!!. , THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 597 452 Opinion likely reduce Big Red Bottling s incentives and ability to contest any anticompetitive branded CSD price increases. Therefore, we-fucus. next on whether the other evidence similarly indicates a likelihood of anticompetitive effects, or whether it provides sufficient grounds for rebutting the presumption of anticompetitive effects that has been created by the degree of increased concentration in the relevant market.

3. The Increased Likelihood of Tacit Collusion in All Branded CSD Channels As we have previously noted (tJhe effective coordination of price and output strategies requires developing a consensus concerning price and output levels, and a means of enforcing its terms. B.F. Goodrich 110 FTC at 294. However, collusion may occur without .'34 firms reaching complex terms concerning price and output levels Instead, the terms of coordination may be imperfect and incomplete -- inasmuch as they omit some market participants, omit some dimensions of competition, omit some customers, yield elevated prices short of monopoly levels, or lapse into episodic price wars -- and still result in significant competitive harm." Merger Guidelines, Section 11.

Factors relevant to an evaluation of the likelihood of collusion include: the extent to which market information is available to market participants; whether there is a history of collusion in such markets; the number of market participants; the pricing and marketing practices used by market participants; the characteristics of sellers and buyers; and the heterogeneity (or lack thereof) of products and We begin with an examination of the avail-market participants.JJ ability of market information to branded GSD bottlers. a. Availability of Pricing Information The evidence in this record indicates that branded CSD bottlers have access to key information about their competitors ' prices and 334 Merger Guidelines, Section 2. 11. For example, coordinating firms may "follow simple terms such as a common price, fixed price differentials, stable market shares, or customer or territorial restrictions." Merger Guidelines. Section 2. 335 Merger Guidelines, Section 2.

Opinion 118 F.T.c. promotions, and that retailers provide such information. The availability of this information could facilitate collusion. For example, bottlers are aware of their competitors ' wholesale prices because they can obtain pricing information "from a retailer or from some other source." Clarke, Tr. 4424. CCSW obtained a copy of Pepsi' s 1987 Cooperative Marketing Program for Independent CCSW was shown a copyGrocery chains in the San Antonio area.J36 of Pepsi' s proposed 1988 Ad Buy Program to 1's Convenience Store chain by 1's personnel.JJ CCSW obtained a copy of a Pepsi- Cola eight-week summer 1988 promotion with the Payless convenience CCSW routinely collects andstore chain.JJ aggregates information regarding Pepsi' s ad and instore retail prices. One incident is particularly telling. In January, 1989, CCSW personnel obtained a copy of the carbonated soft drink promotional materials for National Convenience Stores, Inc. ("NCS"), which owns Stop- Go. CX 465; Hiler, Tr. 5367. Included in the materials was Pepsi' s wholesale price to NCS in the San Antonio area. CX 465 B. When NCS confronted CCSW about the materials, James Doege of CCSW was quoted as stating that "all of my sales people bring (such) information in all of the time." CX 465 A. Indeed, the day-to-day interactions with retailers that are necessitated by use of the DSD delivery system mean that branded CSD bottlers have the opportunity for almost immediate market information about their competitors, marketing, promotions, and pricing. The easy availability of such information suggests that any deviations from a collusive agreement could be quickly detected, thus enabling quick retaliatory action.

b. Branded CSD Pricing in San Antonio Respondents contend that soft drink price competition in San Antonio has been fierce since the acquisition 339 and that we should 336 ex 87. CCSW admitted this, but denied any interference that it was obtained from Pepsi. RRCCPFF paragraph J 874 337 ex 20070.

338 ex 87. CCSW admitted this, but denied any inference that it was obtained from Pepsi. RRCCPFF paragraph 1872.

ABR-A at 74. Respondent cites evidence that: after adjustment for inflation. soft dnnk pnces 339 .in San Antonio have declined "significantly" since 1984 (IDFF paragraph 307j: that this real decline in soft drink prices has occurred while production and promotion costs were increasing (IDFF paragraph THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 599 452 Opinion Wetake such evidence as confirming that collusion is unlikely.'40 agree that the record does not contain any evidence of expres-s-cl- Ius ion among branded CSD bottlers in the San Antonio market, and that the record shows a period of particularly deep discounting by both Pepsi and Coke in San Antonio in 1987. 341 However, this is not surprising. given the level of antitrust scrutiny that has been applied to the relevant market since the acquisition. In September, 1984, the Texas Attorney General' s Office filed suit to challenge the transactions whereby CCSW acquired the Dr Pepper and Canada Dry franchises, alleging that the transactions violated Texas antitrust law. 34' On July I, 1986, CCSW, DPUSA, and the Texas Attorney General entered into a settlement agreement applicable until July I , 1993 which prohibited CCSW from certain activity -- such as seeking or accepting more than 65% of the shelf space "regularly allocated for the sale of soft drinks" in any store -- during that period of time 343 In 1987, the Federal Trade Commission began its investigation of the acquisition, and its original complaint was fied on July 29, 1988. 344 In light of the intensive antitrust scrutiny at both the state and federal levels, it would be most surprising to find anything other than competitive conduct.

Moreover, although there is no evidence of express collusion there is some evidence of the kind of price leadership that typifies an 309J; that the Shircliff Report, plus other evidence (Campbell, Tr. 1950-5 I; Turner, Tr. 979; Trebilcock Tr. 5874-75), establish that soft drink prices in Texas are among the lowest in the United States (IDJ-'F paragraph 313- 14J; and that fierce price competition in San Antonio drove CCSW into financial difficulty rIDFF paragraph 322).

340 See ABR-A at 59. 341See Section IV. supra. 342 ex 2 A-B; IDFF paragraph 67.

343 ex 2 E. Among other things, the Senlement Agreement also prohibited CCSW from "seeking or consenting to participate in. on the average, more than 65% of' promotional ads during any calendar year, or seeking Of accepting "exclusive end-oF-aisle display space" For "more than 65% of the weeks in any given calendar year." IDFF paragraph 68. The ALl found that the provisions of this Settlement Agreement imposed constraints on eesw' use of marketing programs and practices in the San Antonio area, and that the Texas Attorney General' office had the authority and incentive "to dcter any collusive price increase by eeSW " IDFF paragraph 462. In light of this, as we\! as his assessment of other evidence, the AU found that co!1usion seemed unlikely. ID 77.

We do not rely on the Settlement Agreement to constrain eesw' s market conduct, because it expired on July 1 , 1993; although the Texas Attorney General is entitled to seek an extension of the order for a period of up to three years, ex 2-H, VII, there is no record evidence to indicate that the Attorney General has sought and obtained such an extension. 344 IDFF paragraph 43.

), , . . , , .. Opinion 118 F.T. oligopolistic market susceptible to tacit collusion.34' That is, as we discuss in detail below, it appears that each branded C5fbQttler in San Antonio, acting individually, has copied the price of the price leader in the market at certain times. As then-Judge (now Justice) Stephen Breyer has explained:

Courts have noted that the Shennan Act prohibits agreements. and they have almost uniformly held, at least in the pricing area, that such individual pricing decisions (even when each firm rests its own decision on its belief that competitors will do the same) do not constitute an unlawful agreement under Section 1 of the Sherman Act. . .. That is not because such pricing is desirable (it is not), but because it is close to impossible to devise a judicially enforceable remedy for ' interdependent pricing. How does one order a firm to set its prices without regard to the likely reactions of its competitors?346 Based on the record before us, we have no reason to believe that the price leadership that we observe in the San Antonio branded CSD market -- as described below -- is anything other than legal. But we do not view such pricing as desirable, and an- acquisition that may substantially increase the likelihood of interdependent pricing in a market that already appears susceptible to such pricing may have anticompetitive consequences. 347 As CCSW itself recognizes Coke is typically the price leader in 348the San Antonio market." CX 3806 G David Davis of Pepsi agrees: "Coke is usually the leader in the market. They go up, and then we usually follow, depending on our pricing structure. " Davis 345 See. e. , Clamp-All Corp. v. Cast Iron Soil Pipe Institute, 851 F.2d 478 (1st Cir. 1988) (Breyer, J. cat. denied 488 U.S. 1007 (1989) (oligopolistic pricing, including price leadership, is not competitively desirable) ("Clamp-A!!"). One of the purposes of the Clayton Act Section 7 is to prevent markets from becoming oligopolistic and thus susceptible to coordinated interaction, which "includes tacit or express collusion, and mayor may not be lawful in and of itself. " Merger Guidelines Clamp-All. 851 F.2d at 484 (emphasis In angmal) (c1tatlons omlttedj. 347346The Merger Guidelines explain that a merger may diminish competition by enabling finns more likely, more successfully, or more completely to engage in coordinated interaction that harms consumers." Section 2. 1. The Merger Guidelines define coordinated interaction as "actions by a group of firms that are profitable for each of them only as a result of the accommodating reactions of the others. Id. This behavior includes tacit or express collusion, and mayor may not be lawful in and of itself." ld. (emphasis added). Thus, the Merger Guidelines make clear that a merger may violate the FTC and the Clayton Acts because, among other things, it substantia1Jy increases the likelihood of tacit collusion that may be legal in and of itself. 348 This statement appear in a Texas Bottling Group ("TBG") presentation to its Credit Committee. CX 3806. While noting that ''TEG encounters aggressive competition from Pepsi " the document notes Coke s price leadership as a "mitigator, " CX 3806 G. The reference to price leadership is particularly telling, given the AU' s observation that statements in this document were likely influenced by !litigation considerations. ID 67.

THE COCA COLA BOTTING COMPANY OF THE SOUTHWEST 60 I 452 Opinion Tr. 4532. Emery Bodnar of Grant-Lydick reports the same: "I would say from where I sat, my price increase was pretty much dictaton when Coke increased, I followed as quickly as possible." Bodnar, Tr. 1356.

The record contains examples of such price leadership. For example, in February 1989, CCSW initiated a 6% wholesale price increase.'49 Big Red Bottling matched immediately in mid- February, followed by Pepsi on March 1.350 Texas Bottling Group, CCSW' owner, projected that CCSW' s 1989 price increase would bring increased sales: "A 6.0% increase in net price per case coupled with a shift in production mix wi! yield a 10.8% increase in sales in 1989. " CX 3806 Z- 351 It appears that such price leadership may have taken place even before this acquisition. Mr. Bodnar s testimony indicates that DPSA also followed Coke s lead on price increases.352 After the acquisition, there was a substantially increased probability that Big Red Bottling Company would play the "follower" role, since it had lost the Dr Pepper franchise (and thus Dr Pepper sales volume) to CCSW. Indeed, the market might have become more competitive if the Dr Pepper franchise had remained combined with the RC franchise, the 349 ex 3806 2-56.

350 ex 3806 Z-56.

351 The AU stated that CCSW tried to raise list prices in ) 989, but was forced to discount prices back to former levels due to lost sales. ID 69. We have not found any record evidence to show that CCSW rolled hack its 1989 price increase. The AU also stated that the Pepsi COBO "unsuccessfully tried to raise its prices in ! 989. allegedly losing /o of its Nielsen share during the first seven months of 1989. ID 69; IDt"F paragraph 410. Again. we can find no record evidence that Pepsi ever rolled back its 1989 price increase. Thus, we have no basis on which to regard the price increase as "unsuccessful" indeed, we presume that, if the Pepsi CaBO stayed with the price increase for an extended period of time, it did so because it was prot1table, despite any volume loss that might have been associated with it.

Finally, the AL looked to the profitability ofCCSW' s 1989 price increase as evidence indicating that the CUlTent market is competitive. The AU stated that, in 1989, CCSW raised its list price by $. per case, but over the year had a net profit increase of only $.01 per case, ID 69; IDFF paragraph 409, and that CCSW was unable to raise its prices as much as would have been necessar to account for cost increases. rDFF paragraph 409. But the issue is not whether the current market is competitive. Given the intense and ongoing antitrust scrutiny of this market, we would be surprised if it were not competitive, The issue here is whether this acquisition has taken place in a market sllsceptible to collusion. The price leadership shown in the 1989 price increase is one piece of evidence indicating that the market is sllsceptible to collusion.

352 Bodnar, Tr. 1356, 353See Section I.C.2 supra. Opinion 118 F.T. combination that Mr. Bodnar described as gIvmg DPSA critical mass. 354 In sum, we find that the price leadership by CCSW evident in the relevant market supports an inference that the market is susceptible to interdependent pricing -- that is, tacit collusion -- and that the evidence concerning Big Red Bottling s diminished ability to compete with respect to ad features demonstrates that CCSW' s acquisition of the Dr Pepper franchise substantiilily increased the likelihood that Big Red Bottling would continue to follow CCSW' s price leadership.

c. Collusion by Branded CSD Bottlers There have been over 40 price-fixing cases involving branded CSD bottlers in a number of local geographic markets. 355 Complaint counsel offered evidence relating to these collusion cases, but the AU rejected it as irrelevant356 We find the evidence to be relevant to the likelihood of collusion by branded CSD bottlers in the San Antonio market, because such cases suggest that there are local or regional branded CSD bottling markets that are conducive to collusion.35 The cases suggest that, in markets structured similarly See Section Vr.C.2. supra.

355 See. e. Convictions: United States Mid Atlantic Coca- Cola Boltling Co. 6 Trade Reg Rep. (CCH) paragraph 45,090 (ED. Va. 1990); United Stales v. A!1egheny Bottling Co. 695 F. Supp. 856 (E.D. Va. 1988). ajfd 870 F.2d 656 (4th Cir. 1989); United Slates v. Harrford (1988- 21 Trade Cas. (CCH) paragraph 68, 386 (4th Cir. 1989); United Simes v. Gravely, 840 F.2d 1156 (4th Cir. 1988). Guilty picas: United States Pepsi-Cola BOffling Co. of Walla Walla No. CR-89-394-0J (E.D. Wash. Jan. 16, 1990); Uniled States v. Coca- Cola Botllina Co. , Yakima and Tri Cities. No. CR 89-372-01 (E. Wash. Jan. 16, 1990); United Slates V. Ehw MOimlain BorTling Co. of Walla Walla No. CR 89-392- (E.D. Wash. Jan. 16, 1990). arf'd, 929 F.2d 526 (9th Cl"r. 1991); Ullited States v. Rice Bottling Enterprises. Inc. No. 3-89-72 (E.D. Tenn. , Oct. 16. 1989); United States v. Pelpsi- Cola Bottling of Petersburg, Inc. No. 89-00062 (E.O. Va., Del. 11, 1989); United States v. At/antic Soft Drink Co., No. 88-77 (E.D. Tenn., Dec. 23, 1988); United States v. Beverage SOl/th. Inc. No. 88-451 (D. C., Dec. 1988); United Stales v. All.Americcln Baffling Corp., No. 88-00038 (W O. Va., Apr. 12, 1988); United States v. Coca-Cola Bottling Co. of Roanoke. Va. No. 88-00012 (W.D. Va., Apr. 15 , 1988); United States v. Akron Coca- Cola Boaling Co. No. CR 88-044 (N.D. Ohio, March 15 , 1988); United Stale. v. Seven- UplDr Pepper BO/tling Co. Beckley, W. Va., No. 88-00012 (W O. Va., Feb. 1. 1988); United States v. NEG Holding Co" No. CR 87- 16-01 (N.D. Ga., Nov. 24 1987); United States v. Mid Atlantic Coca- Cola Bottling Co. No. 87-0420 (D. C., Oct. 14, 1987); United Slares v. General Cinema Beverages of Washington, D. No. CR 86-0352 (D. C.. Del 15, 1986). 356 RCX 3323-52; 3354; 3356-57; 3359-65; 3367-68; 3788; 3950; Tr. 114- 470- 4101-05. 4141 6089- . 6176. 6341-45, 6937- . 84! 7-22. 357 See Coca-Cola, slip op. at 48. In that case, we found evidence of branded CSD bottler collusion relevant to an assessment of the likelihood of collusion by a carel of branded CSD concentrate companies, because it suggested that, if such a care! raised concentrate prices nationally, botters could successfully pass on the price increase. Id. ,, . . THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 603 452 Opinion to the San Antonio market 358 branded CSD bottlers have perceived that "the number of competitive dimensions involved posed finsuperable obstacle to collusion. " Coca-Cola Co., slip op. at 48. 359 The bottler price-fixing cases also are relevant to and reinforce our conclusion that the relevant market in this case is branded CSDs in the San Antonio market. J6D The branded CSD bottler collusion cases provide evidence of actual collusive conduct that negates the hypothetical difficulties in colluding that respondent raises. Respondent argues that "the variety of brands, packages, flavors, sweeteners, and advertising support" for soft drinks complicates the market sufficiently to deter collusion 361 However, the same type of variety exists in the markets in which branded CSD bottler collusion took place and apparently did not 362 deter that collusion Indeed, the bottler collusion cases and the bottler documents in the record here suggest that factors such as 358 As we found in Coca- Cola Co. (mJost local markets for carbonated soft drinks have a Coca- Cola bottler, a Pepsi-Cola bottler, and a so-called ' third bottler ' which caIies various brands of soft drinks other than Coca-Cola or Pepsi-Cola brands. " Slip op. at 57. The record here similarly supports this finding. See also Lydick, Tr. 2937, 2943. "lAJ record of price fixing or other antitrust violations is some evidence that the structure of the market is favorable to collusion." R. Posner, Antitrust Law: An Economic Perspective 55-61 (J 976).

359 Under the Merger Guiddines (pJrevious express collusion in another geographic market will have the same weight fas express collusion in the same geographic markets when the salient characteristics of that other market at the time of the collusion are comparable to those in the relevant market. " Section 2 !. Here, the bottler collusion cases arose from a variety of areas in the Cnited States suggesting that the salient characteristics that facilitate collusion among branded CSD boulers are not unique, but instead are present in typical local branded CSD bottling markets. This is not surprising, since the basic structure of local branded CSD bottling markets in the united States tends to be only three branded CSD bottlers using DSD delivery. Lydick, Tr. 2937 , 2943. This trend follows a significant period of bottler consolidation. In 1960 there were 4 519 soft drink bottling operations in the United States, in ! 970 there were 3 054, and in 1980 there were only 1 960 (CX 996 A), and in 1983 there were only 1 500. CX 3218 M.

Moreover, the branded CSD bottler price-fixing cases are far more relevant to this case than the discussion of the OPEC calle! permitted by the ALl. See Strickland, Tr. 8283-85. Given the direct relevance of the bottler collusion cases to this market. we find that the ALl erred in refusing to admit this evidence.

he coUuders In t ese cases were bottlers of branded DS, and the actual pnce IIcreases typically were maintained for over one year. See cases in note 355 supra; e. , Allegheny Bottling Cu., 695 F. Supp. at 858. The ca. es usually identified discrete, local geographic markets of no more than twelve counties and as few as one, far less than the 107 counties proposed by respondent as the relevant geographic market in this matter. ex 4131; see IDFF paragraph 246. 361 ABR-A at 63 362 See generally Section IV supra (documents and testimony from national concentrate companies indicate same general competjtive conditions in terms of brands, packages. Oavors, . sweeteners, and advel1ising support for all of their bottlers). See also Lydick, Tr. 2937, 2943. Opinion 118 FTC. standard packaging ease price comparisons, which can facilitate collusion. 363 Respondent also argues that "(tJhe putative colluders, in addition to devising a complicated set of list prices, net prices, and net, net prices, would also have to control promotional programs so that volume changes would not disrupt each colluder s expectation of bottom-line profit." ABR-A at 64. But participants to some of the collusive schemes have fixed prices suceessfully simply by agreeing not to offer discounts on various products.364 In sum, the branded CSD bottler price-fixing cases reinforce our previous conclusions that collusion need not be perfect to be successful and that the relevant market in this case is susceptible to collusion.

4. Respondent s Arguments Against the Likelihood of Collusion Respondent presents a variety of additional arguments that supposedly negate any inference of an increased likelihood of collusion, tacit or express, following CCSW' s acquisition of the Dr Pepper franchise. As we discuss below, we find these arguments unconvincing.

a. Difering Profit Incentives Among Bottlers Respondent argues that CCSW, the Pepsi COBO, and Grant- Lydick all have differing profit incentives, and that such differing incentives could hamper collusion. ABR-A 61. Respondent points out that the Pepsi COBO is owned by Pepsi USA, and that Pepsi By contrastUSA makes a 96% gross profit on concentrate sales.36s respondent states that CCSW makes no profit on CCUSA' s concen- 363 For example, RX 582. entitled " 1987 Pricing Summary," shows that price comparisons arc relatively easy, given standardized packaging. For the periods of November and December. we are at parity on in-store pricing and at parity on ad feature pricing for the 2 liter and 3 liter non-holiday and one price unit disadvantaged versus Pepsi on cans non-holiday. For the holidays of November and December, we were at paritx on 2 liter. We are one price unit disadvantaged on cans. This document reflects quite simple comparisons, not complexity. Hartford. 1988-2 Trade Cas. (CCH) paragraph 68 386 at 60 13!; Allegheny Borrling Co. , 695 F. Supp. at 857.

365 ex 3913.

THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 605 452 Opinion Even CCSW and trate sales.J66 Grant-Lydick are dissimilar, in that Grant-Lydick purchases its cans from an independent packer 36? whereas CCSW provides its own cans. J08 Respondent is correct that such differing profit incentives may operate. to make collusion more diffcult. However, we must evaluate the evidence as a whole, and we are not convinced that such differing profit incentives, even in combination with other factors present here would be sufficient to deter collusion in this market. For example Mr. Davis of Pepsi has acknowledged that Pepsi' s strategy in San Antonio is now focused on profitability rather than on increasing market share, as was the case during the deep discounting period of 1987- 369 An emphasis on profitability rather than market share may increase the likelihood of collusion.370 As to Grant-Lydick, the fact that Grant-Lydick' s higher can costs provide it with a greater incentive than CCSW has to keep can prices high only suggests that it would favor a collusive agreement on can prices rather than another type of agreement. This argument alone does not demonstrate that collusion is unlikely in the relevant market; in fact, it might make collusion more likely. b. Difering Size Firms Respondent also asserts that the range of firm size in this case -which may produce different cost structures for each firm -- renders We agreecollusion "highly improbable."37 that, in theory, differing cost functions among firms may make it more difficult for firms to agree on a consensus collusive price.37 However, it would be a leap 366 Respondent cites R. Hoffman at Tr. 5577- , but the citation does not support respondent claim.

367 Turner, Tr. 1117; Bodnar, Tr. 1526-27. 368 Summers, Tr. 6403-04.

369 Davis, Tr. 4527- 28.

370 Respondent notes that none of the bottler collusion cases rendered by complaint counsel involved a Pepsi COBO. ABR-A at 62 n.53. We do not find that this absence renders those cases irrelevant, however. We note that San Antonio is onc of Pepsi' s worst markets -- a market in which Pepsi' s share increased only from 15% to 19% after a year and one- half of losing millions of dollars from offering extraordinarily low prices. Davis, Tr. 4548-4565. After such an experience, the Pepsi CaBO could well become more interested in collusive - and profitable -- price increases than in contint.ing vigorous price competition.

371 ABR-A 60-61.

- See B. F. Goodrich 110 FTC 207 , 321 (J 988). Opinion 118 FTC. of faith, given a lack of supporting analysis in the record, to decide that the different cost structures present here constitute a significant obstacle to collusion in this highly concentrated market. In addition CCSW' s apparent role as price leader and the possible benefits to reaching and maintaining a collusive agreement suggest that this dominance may be an offsetting force acting for rather than against collusion.

D. Unilateral Anticompetitive Conduct An acquisition may diminish competition by making it profitable for a firm to alter its behavior unilaterally by elevating price and/or This phenomenon may occur in markets wheresuppressing output37J products are differentiated by flavor, among other things.J74 Thus, an acquisition may enable the acquiring firm to raise the price of either its original product, or the acquired product, or both above the premerger level. As explained in the Merger Guidelines, "(sJome of the sales loss due to the price rise merely will merger partner and depending sales loss through merger may even though it would not have success of this strategy will significant share of sales in be diverted to the product of the on relative margins, capturing such make the price increase profitable been premerger. " Section 2.21. The success of this strategy will require that "there be a significant share of sales in the market accounted for by consumers who regard the products of the merging firms as their first and second choices and that repositioning of the nonparties' product lines to replace the localized competition lost through the merger be unlikely." Merger Guidelines, Section In this case, CCSW2.21.may have been constrained from taking some anticompetitive actions due to concern about ongoing antitrust litigation and certain restrictions imposed as part of CCSW' s settlement with the Texas Attorney General.J7 Nonetheless, there is some evidence of unilateral etfects that have occurred since the acquisition of the Dr Pepper franchise. We begin by examining this evidence - 373 Merger Guidelines, Section 2.

374 Merger Guidelines, Section 2.21.

375 For example, among other things, the Settlement Agreement prohibited CCSW from "seeking or consenting to participate in, on the average, more than 651'10 of' promotional ads during any.calcndar year. ex 2 E; IDFF paragraph 68. See also n. 343 supra THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 607 452 Opinion which involves the elimination of take-home sales of Mr. PiBB and a lessening of competition in the vending channel -- and theIY15efl.y discuss the potential for further effects with respect to ad features based on CCSW' s increased market power.

1. The Elimination of Competition Between Mr. PiBB and Dr Pepper CCUSA and Dr Pepper Company are the only firms in the soft Thedrink industry that have a viable "pepper" category soft drink.376 Dr Pepper Company sells Dr Pepper concentrate; CCUSA sells concentrate for Mr. PiBB.

Mr. PiBB was introduced by CCUSA in 1973. RX 888 D. CCSW has admitted,378 CCUSA defined the consumer role of Mr. PiBB as the "Alternative to Dr Pepper.,,379 In 1984, CCUSA defined the business role of Mr. PiBB as "Competitor to Dr Pepper " CX 1895B , designed to "combat" the brand:

Mr. PiBB represents the only viable alternative to Dr Pepper in its flavor category. The brand is necessar, especially in cold drink, to enable bottlers to combat Pepper where it is strong.

CX 791 C. CCUSA targeted Dr Pepper consumers with its Mr. PiBB brand. 380 Prior to CCSW' s acquisition of the Dr Pepper franchise, Mr. CCSW'PiBB was sold in San Antonio, as was Dr Pepper.J81 S business records reveal that CCSW viewed Mr. PiBB as the closest and consideredsubstitute to and a direct competitor of Dr Pepper,"2 Mr. PiBB to be one of its major sugar brands."3 San Antonio was a 376 ex 791 c; ex 790.

17 ex 790 B; ex 791 B, s; RX 888 e- 378 RRCCPFF paragraph 2098.

379 ex 1895 A; ex 790; ex 79L Mr. PiBS is perceived in the marketplace 8.-; a "me-too brand. ex 791 E.

380 ex 1885; ex 1898 B; ex 1896; ex 1894; RX 888 D; Turner, Tr. 954; Clarke, Tr. 4278, 4400-01; ex 1893.

381 Turner, Tr. 996; Bodnar, Jr. 1361; Schwerdtfeger, Tr. 2327, 2344; Anderson, Tr. 3850. 382 ex 596.

383 ex 5 10 R; ex 3480 E; CX 3481 E.

Opinion 118 F. priority market for Mr. PiBB 384 CCSW' s sales of Mr. PiBB were above the national average 385 and San Antonio accountecrfDr 3% of all Mr. PiBB volume in the United States. 386 In 1983, Mr. PiBB' s market share was 2. , and Dr Pepper share was 8.4%.387 After the acquisition, CCSW no longer sold Mr. PiBB in bottles and cans in the territory in which Dr Pepper was sold388 Mr. Hoffman testified that the elimination of Mr. PiBB occurred because it was a competing flavor with Dr Pepper, and flavor restrictions from the Dr Pepper Company prohibited CCSW from selling a competitive flavor. 389 Although there was testimony that CCUSA would consider licensing another distributor to distribute Mr. PiBB in San Antonio this has not happened. "" After CCSW stopped distributing Mr. PiBB in the take home market in San Antonio, Dr Pepper s market share began increasing; but only in 1987 did Dr Pepper s market share come close to the combined 1983 share of Dr Pepper and Mr. PiBB 391 Most significantly, after the acquisition, CCSW raised the wholesale price of Dr Pepper to parity with CCSW' s other products. 392 Other data show that retail prices of Dr Pepper in San Antonio, which prior to the acquisition had been below the national average of Dr Pepper prices, after the acquisition rose to above the national average 384 ex 792 G.

385 ex 3837 s, G.

386 ex 792 L; ex 1897 E. San Antonio also accounted for 9.5% of the CCCSA PiBS brand funding in 1983 and 7.6% of funding in 1982. ex 792 L. In (act, Mr. PiBS' s 80l (Brand Development Index (CX 591 Cl) in San Antonio was the highest in the nation. ex 792 O. 387 ex 1681 C. These market shares are based solely on sales of bottles and cans, since that is the channel of sales that was eliminated. See Hilke, Tr. 6030, 6033. (CX 168! uses Nielsen data for soft drink sales in food stores in Bexar County, which includes San Antonio). 388 Anderson, Tr. 3879. 3859; ex 596 A-I; ex 2192; Atchison, Tr. 5252; ex 3221 A. CCSW has continued to sell Mr. PiBS postmix syrup. eesw also sells Mr. PiBB outside of its Dr Pepper fraochise area.

389 E. Hoffman, Tr. 324, 421; ex 122.

390 See Atchison, Tr. 5252- Data show the 1983 combined share of Dr Pepper ( .4%) and Mr. PIBB (2. 1 %) for bottle andcan sales391in San Antonio as about 10.5%. ex l681 C. Dr Pepper s share for bottle and can sales reached 1% in San Antonio in 1987. ex 1681 c.

392 ex 563 E.

. . . . . . .

THE COCA COLA BOTTLING COMPANY OF THE SOUTHWST 609 452 Opinion In 1989, in a tellng memorandum from for Dr Pepper retail prices.J9 Mr. Summers to Messrs. E. and R. Hoffman, Mr. Summers state We are pricing Dr Pepper one increment above other brands on instores, since it has no competition in its flavor segment." CX 2261; Summers, Tr. 686870.

Dr Pepper had no competition in its flavor channel -- and therefore was priced higher in 1989 than it otherwise would have been -- because Mr. PiBB had been eliminated as a competitive option for consumers in San Antonio, as a result of the acquisition at issue in this case. Consumers in the San Antonio area who preferred Mr. PiBB to other bottler or canned soft drinks were placed in the position of having to switch to less-desirable alternatives and, as a result, were made less well-off.

2. CCSW' s Increased Market Power Over Vending Machine Sales The record demonstrates that, post-acquisition, the choices available to consumers from vending machines were reduced, and the prices charged to third-pary vendors increased. There was testimony that, prior to the acquisition, third-party vendors had been able to resist any attempt by a branded CSD bottler to force a vendor to take unwanted allied brands along with the desired brands.394 About three or four years after the acquisition, however, CCSW imposed a requirement that a third-party vendor cannot qualify for the best available discount unless 20% of its purchases are allied brands such Ladd Little of LVas Sprite, Sunkist, and Hires.395 Vending attributes CCSW' s ability to impose the requirement to its acquisition of the Dr Because of this requirementPepper franchise.'96 , he purchases Sprite, Sunkist, and some other flavors from CCSW, while he would prefer to purchase 7-Up and Crush from Grant-Lydick and Slice from Pepsi. 39? In addition, Mr. Little testified that, post-acquisition, Dr 393 ex 1685 A. E-H; Hilke, Tr. 6252-53, 6288-89.

Pnor to the acqlls!tlon, CCSW did not require third-party vendors to accept allied brands In order to get the desired brands; Pepsi had attempted to impose such a requirement without success. Little, Tr. 667- , 705.

395 Little, Tr. 665-66.

396 Litte, Tr. 665, 705.

397 Little, Tr. 668- . 704.

.

Opinion 118 FTC. Pepper case prices increased to the level of the Coke case prices. The unilateral effect in this instance appears to be CCSW"-ility to increase price either directly (by raising case prices of Dr Pepper) or indirectly (by tying purchases of other, less attractive products to discounts on attractive products).

3. CCSW' s Increased Market Power Over Ad Features As noted above, the acquisition of the Dr Pepper franchise increased CCSW' s ability to obtain ad features and thus increased The evidence suggests that CCSW obtain-CCSW' s market power.39 ed more feature ads after its acquisition of the Dr Pepper franchise The increasedthan it had previously.40o "pull" of all of CCSW' brands gives CCSW the potential power to extract more favorable deals from retailers and to disadvantage both the Pepsi COBO and 01 InBig Red Bottling in their attempts to obtain ad features addition, CCSW' s increased market power may- have contributed to its ability to raise Dr Pepper s price. See also Section Vr.C.2 supra. E. Power Buyers Respondent argues that there are power buyers who could constrain any collusive or unilateral attempt by branded CSD bottlers to raise price. The ALl agreed, stating that, in the face of a price rise among national CSD brands, retailers such as H.E. , Kroger, and others who stock their own private label brands "could easily promote those brands in place of national brands. " ID 76 402 398 Little, Tr. 669-70.

399 ar ct power me u es tea I try to essen competition on ImenSlOns ot er t an pncc, sue as product quality, service, or innovation." Merger Guidelines, Section O. ! & n 400 See ex 2954 H (in 1984. some of Dr Pepper s feature ads took place before the acquisition whereas at! of Dr Pepper s feature ads are attributable to CCSW). Drusa recognized the advantages of being advertised with Coke and advised that Dr Pepper should be advertised with Coke to build sales, RX 2825 The record shows that Pepsi already generally has to Dlfcr more ad feature payments to aretailerC.401than Coke because Pepsi doesn t sell as much product. Kaiser, Tr. 3210; see also ex 129;' ex 3814 at 28-29 (Adams).

402 The AU also found that concentrate companies such as eeUSA, Pepsi USA, and DPUSA had "the power and the incentive to deter collusion at the bottler level." ID 76-77. We find that the numerous bottler collusion cases listcd earlier see note 355 supra provide sufficient evidence to undermine any hope we might have that concentrate companies could prevent collusion in thi market; the concenlrate companies did not prevent collusion by the bottlers in those cases. g., THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 611 452 Opinion both the In analyzing the competitive effects of a merger, Commission and the federal courts have considered the possib power of buyers in deterrng anti competitive effects. 403 The relevant market here does contain large buyers who are large food retailers. , the largest buyer, accounts for approximately 25% of CCSW' s take-home sales and approximately 20-25% of Pepsi' s takehome sales 04 Kroger is the second largest customer of CCSW, purchasing from 9-12% ofCCSW' s total unit sales.405 Sam s Wholesale Clubs purchase 7-8% ofCCSW' s total unit sales 406 In addition to the leverage that may be provided by such sales volumes,'07 retailers have some leverage over branded CSD bottlers because the retailers can control the availability of their own ad features and in-store displays, which can be important to the marketing of the branded CSDs of the bottlers Just to note these facts does not demonstrate that retailers in this market could constrain any anti competitive price increases, however. Rather, we must analyze the extent to which retailers facing an anticompetitive price increase could avail themselves of options other than paying the price increase and thereby force the branded CSD bottlers to return to a competitive price: Consideration of large and sophisticated buyers generally focuses on the buyers ability to exert countervailing power, even against a seller s oligopoly, by (1) shifting a large proportion of business to any finns that are willing to deviate from the coordinated behavior; (2) inducing new entry into the oligopolized market; or (3) through vertical integration.

403 See, e. Adventist Health Syslcm/est, Dkt. No. 9234 (Apr. I , 1994), Concurrng Opinion of Commissioner Owen and Commissioner Yao, slip op. at 16- J 9; Owens-1!inois. Inc" Dkt. No. 92! 2 5 Trade Reg. Rep. Rep. (CCH) paragraph 23 ! 62 (FTC 1992): Olin O;rp., Okt. NO. 9! 96 5 Trade Reg. Rep. (CCH) paragraph 22 857 (f"lC 1990); United Slates v. Baker HURhes, Inc., 908 F.2d 981 (D.C. Cir. 1990); Unired Stales Ii. Syu.fy Enterprises 903 F.2d 659 (9fh Cif. 1990); United States v. Archer- Daniels- Midland Co. 1991-2 Trade Cas. (CCH) ! 69 647 (S.D. Iowa 1991); United Slate.l. v. Coullry Lake Foods, Inc. 754 F. Supp. 669 (D. Minn. 1990). 404 CX 3806 Z37; Summers, Tr. 6589; Davis, Tr. 4525; IDFF paragraph 432. 405 Summers, Tr. 6589; IDFF paragraph 433. 406 Summers, Tr. 6638; IDFF paragraph 435. 407 We note, however, that the size of these alleged "power buyers" faUs far short of that in Country Lake Foods, where the three largest distributors accounted for more than 90% of sales. 754 F. Supp. at Coyne, Tr. 3449- , 3487; Turner, Tr. ! 130-31; IDFFparagraph 17 J , 445. As we discussed ad features and in-store displays are extremely important to increasing sales of branded CSDs. earlier,674.408 See Sections IV.C.3, VLC.2 supra.

Opinion 118 FTC. Adventist Health System/West, Dkt. No. 9234 (Apr. 1,--4), slip op. at 16 (Concurring Opinion of Commissioner Owen and Commissioner Y ao ):09 As discussed below, we have considered these possibilities and have concluded that none appear to be realistic options for the retailers in this market. Moreover, we find that the instances of supposed buyer power cited by respondent and the'ALl do not suggest that the buyers in this market could successfully counter a collusive or unilateral price increase by branded CSD bottlers to retailers. I. Shifting Purchases to Others The ALl found that H. B. and other retailers who sell their own private label soft drinks could switch to promoting those soft drinks instead of branded CSDs if confronted by a collusive price increase:1O Our finding that private label soft drinks are not in the relevant market militates against this conclusion. As we have explained, the evidence shows that retailers depend on branded CSDs I and would notas a promotional item to draw in customers4I switch to purchasing non branded CSDs in the face of an anti competitive price increase: 12 The question then becomes whether H. B. and other retailers would switch to any firms within the market that would be willing to deviate from cartel conduct or undermine unilateral anti competitive conduct. In this market, there are only three main bottlers making sales of branded CSDs to retailers: CCSW, Pepsi COBO, and Grant- Lydick (Big Red Bottling). The branded CSD products of these firms are differentiated, however, and are not exact substitutes for each other. Thus, we would expect that switching among branded CSDs would not always be costless for a retailer, and that under certain circumstances retailers might be reluctant to try to substitute exclusive ad features on Pepsi or Big Red for all ad features on Coke, for example.

409 See also Baker Hughes. Inc. 908 F.2d at 986- 87; Country Lake Foods, Inc. 754 F. Supp. at 679; Olin Corp. 5 Trade Reg. Rep. (CCH) paragraph 22 857 . at 22 553. 410 ID 76.

41! E.g. ex 3806 Z37 , ZSO; see Sections IV C.3, VI.C.2 supra. See cetlon supra.

, , ). THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 613 452 Opinion In fact, the evidence indicates that, particularly with respect to Coke, retailers do not always regard branded CSDs as pect substitutes. The assessment of TBG, owner of CCSW, was that although CCSW was dependent on the retail chains for increased volume of sales the chains are dependent on soft drinks as a promotional item to draw customers into their stores." CX 3806 Z37. And not just any "soft drink" would do. TBG noted that, although H.EB. has significant negotiating power " a "mitigating" factor is that " B. must buy Coke products from TBG in its franchise territories." CX 3806 G. According to TBG' s own assessment (wJhile TBG may lose an occasional major ad to Pepsi, they believe that it is not in H. B.'s best interest, long term, to promote Pepsi products due to Pepsi' s relatively weak market share (20% vs. 60% Other for TBG)." CX 3806 ZS 5.41 evidence is consistent with TBG' s analysis.'14 This market share dominance of Coke over Pepsi also applies to Big Red, whose market share in food stores in 1984 was roughly 413 Similarly, in a 1987 antitrust suit against CCSW' s parent, TBG, by Oneta Company, an independent Pepsi bottler in Corpus Christi, Texas, Toby Summers testified that CCSW made basically "take it or leave it" offers to accounts such as Albertson Q. But I can t tell what the (cons of a counter offer or negotiated ultimate agreement was from the A. Well,terms whateverof your proposal.the ultimate agreement was would have been within the parameters of the proposal. These were not subject to negotiation. They were subject to this is the offer. ' And then it became incumbent on the account to - They either took it, or they didn -- avail themselves of the offer, or no! to avail themselves of the offer. It was not a matter of negotiation.

CX 4021 at 678 (SummersJ. See Oneta Company v. Texas BOfflin!; Group, Inc., No. C-87-97 (S. Texas - Corpus Christi Div.

James Nlcho\son of RC believes that the Coca-Cola brand IS so Important to retailers such as H. B. and Kroger that the retailers are reluctant to take-fctions that concern Coca-Cola bottlers. Nicholson, Tr. 3813- 15.

The ALl pointed to evidence that, in 1988, rather than risk relribution from H. , CCSW had complied with H. B.'s notice that it would not accept price increases for four months. Summers, Tr. 6769; IDFF paragraph 449. Bul this notice from H. B. went to vendors for all products (not just soft drinks) and was applicable for only a short time, Summers, Tr. 6769, so we do not find that it constitutes convincing evidence of H. B.'s ability to disrupt a branded CSD bottler cartel. Perhaps more telling is the history of CCSW' s negotiations with Stop- , a leading convenience store chain in the San Antonio area, regarding promotional programs. In 1986, Stop- Go refused to feature Coca-Cola products in South Texas for six to nine months, because CCSW would not agree to Stop- s terms for promotional programs. Howe!!, Tr. 4061-63; IDFF paragraph 450. The terms that CCSW wanted were exclusive promolions for Coke products. Bodnar, Tr. 1381 . In 1987 , Coke received exclusivity in exchange for adhering to Stop- s promotional terms. Moreover, in 1988 , Coke persuaded Stop- Go to drop its promotional terms while maintaining the Coke ad features schedule for 1988. Bodnar Tr. 1381. This history indicates that Stop- Go gave in to CCSW' s demands, not the reverse, and may reflect the leverage that Coke s market dominance in the San Antonio market gives to CCSW. Opinion 118 F.T.c. comparable to but smaller than Pepsi's"I5 and even more compellingly to the other branded CSD products sold by Grant- dick and the Espinoza companies, none of whose shares reach even the 20% mark416 In the face of such market share dominance by Coke, we are skeptical that H.E.B. (or any other retailer) would switch all purchases to another branded CSD, since such a switch might well have a large impact on the retailer s overall sales of branded CSDS 417 Nor does it appear that H. B. (pr any other retailer) has sufficient leverage over either Pepsi COBO or Grant-Lydick to force them to deviate from a possible collusive agreement. Mr. Davis of ' Pepsi COBO testified that H.E.B. does not have the clout to demand that Pepsi bottlers uniformly price their branded CSDs throughout B.'s sales territory, and that Albertson s had been unsuccessful in its attempts to convince Pepsi bottlers to price their branded CSDs Emery Bodnaruniformly throughout Albertson s sales territory.41 testified that Grant-Lydick has never rolled back a wholesale price increase at the request of H.E.B., and that H.EB. does not have the clout to force Grant-Lydick to rollback wholesale prices. In addition, this market does not feature the types of sporadic large, and not immediately observable orders that encourage cheating on a cartel.420 Although some retailers negotiate a promotion schedule of advertisements for an entire year, other large retailers -such as H. B. -- decide on promotions in much smaller time periods 42I Thus, the offers that branded CSD bottlers would make would involve a smaller profit potential and less incentive to cheat 415 ex 1681 C. See Hilke, Tr. 6030, 6033 (CX 1681 used Nielsen data for food stores in Bexar County. which includes San Antonio).

417 This situation contra. ts sharly with thaLin County Lake Foods, Inc., in which the three largest distributors accounted for 90% of sales and the product involved -- milk -was not differentiated so that distributors could credibly assert that a substantial increase in milk prices would prompt aggressive negotiations to seek a price reduction or an alternative supplier. See Country Luke Fouds, 754 F. Supp. at 679 ("Fluid milk processors face no significant product differentiation barrier. Therefore, a food distributor could change its supplier of fluid milk without losing sales due to brand !oyalty.

418 Davis, Tr. 4495-97, 4499-50l.

419 Bodnar, Tr. 1488-90.

420 See, e. . Baker Hughe. 908 F.2d at 986 (power buyers could decrease the likelihood of collusion where awards of lumpy orders -- sometimes ex.ceeding $1 million -- were made through confidential bidding by sophisticated buyers). 421 Davis, Tr.4512- 13.

..

THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 615 452 Opinion than if promotions were contracted on a long-term basis.42 In addition, changes to ad features and low-priced ad features wil quickly observable by other branded CSD bottlers, whose DSD delivery personnel can easily observe new promotions. 2. The Ability to Induce New Entry or Vertically Integrate There is no record evidence that any of the retailers in this market would vertically integrate into the production of branded CSDs in order to avoid payment of a collusive price increase. In order to do , a retailer would need a branded CSD franchise for a product such as Coke or Pepsi or Big Red, and there is no evidence to show that a retailer could wrest those franchises away from their current holders. Nor is there any evidence that retailers would induce new entry by another branded CSD bottler as a remedy to anticompetitive price increases. Indeed, as we discuss below, the evidence demonstrates that entry into the bottling of branded CSDs is extremely difficult because of the difficulty of obtaining a branded CSD franchise and associated problems. Thus, we find that this case is not comparable to those in which power buyers could decrease the likelihood of collusion because they could induce new entry or vertically integrate themselves to avoid succumbing to a collusive price increase.'24 422 Courts have noted that the possibility of a single large sale that is unlikely to be detected may tempt cheating by a carel member. E.g. , FTC Elders Crain, Inc. 868 F.2d 90 1 905 (7th Cir. 1989). It has also been noted that excess capacity can make it possible for a carte! cheater to supply a large quantity at little cost, thereby making the cartel cheating even more tempting. Elders Grain 868 F.2d at 905-06.

In this case, the ALl also found that the presence of excess capacity in this market reduced the likelihood of collusion. ID 74. We find that the existence of excess capacity in this particular market with its own set of distinctive market conditions would not significantly reduce the likelihood of collusion. If excess capacity were a major factor here, we would expect the record to show some pricing pressure effect from it; we have not seen any. In addition, product differentiation may mitigate the effect of excess capacity, since retailers would not necessarily find it profitable to substitute all of one branded CSD for sales of two others. See Section IV supra. In addition, we find the assertions of excess capacity to be somewhat inflated. Although the AU found that Grant-Lydick operates with 20-40% of llnllsed capacity during the busiest time of the year, IDFF paragraph 134, the AU failed to note that this applies only to bottles, since Grant-Lydick contract packs its cans. Turner, Tr. t 117; Bodnar. Tr. 1526- 27. The excess capacity listed for the Pepsi COBO -- IDFF paragraph 136 -- fails to note that sales of branded CSDs are highly seasonal, and that therefore excess capacity in February may be used capacity in July or December. Davis, Tr. 45 13- 14. Certain other citations to excess capacity involve bottlers that we have determined fat! outside of the relevant market. IDFF paragraph 135, 137, 138, 139. Thus we are not convinced that there is a great deal of excess capacity in the relevant market in any case. See CX 465 A.

424 Cf. Country Lake Food, tnc 754 F. Supp. at 679-80 (3 largest distributors had capability to vertically integrate, but court noted that possibility of vertical integration alone would not be sufficient to rebut presumption of market power). .

Opinion 118 FTC. 3. Conduct by Retailers Finally, we have examined whether conduct by any of the retailers suggests an ability to undermine a cartel among branded CSD bottlers. Although the evidence shows that H.E.B. and other large 42 they do not add up to theretailers have some bargaining power type of conduct indicative of retailer s ability to turn to alternatives and thereby defeat a branded CSD bottler .cartel. Indeed, the evidence is consistent that neither H. B. nor Kroger have attempted the type of market conduct that might indicate oligopsony power over branded CSD bottlers.'26 In any case, even if H. B. as a retailer accounting for significant portions of the sales of CCSW and Pepsi COBO could defeat a collusive price increase from branded CSD bottlers, that action may only protect H.EB., not other retailers. The discounts (including payments for ads and displays) negotiated between branded CSD bottlers and retailers are individualized, so the fact that H. continued to receive a competitive price would not necessarily protect other retailers from supracompetitive prices. That an anticompetitive effect may pertain only to some portion of the market does not immunize it from antitrust liability. 425 For example, H. B. and Kroger each have cancelled scheduled ads because they determined that the price was not competitive. Summers, Tr. 6626-27; Kaiser, Tr. 3218. H. B., Kroger, and Albertson all require that bottlers offer them their lowest net wholesale price. Brinkley, Tr. 2234; Bodnar, Tr. 1660-61; Chapman, Tr. 7245; Turner, Tr. !200; Summers, Tr. 6646, ex 3700- 0; Donald Tr. 5320- 5327-28; Kaiser. Tr. 3264. These events reflect the ability of the large retailers to ensure that they are getting prices that are comparable to those offered other retailers, but they do not show that the retailers could counteract a branded CSD bottler carlel. - 426 H. B. has never dictated the terms or conditions under which branded CSDs are sold in San Antonio or any other Texa. market. Brinkley, Tr. 2235-36; Chapman, Tr. 7242; Gonzaba, Tr. 2100-0 I. Specifically, H. B. has never dictated or attempted to dictate package sizes or product lines, prohibited or attempted to prohibit any bottler from running a branded CSD o;dvertiscment with one of H. competitors, asked that a bottler stop selling a particular package size to an H.E.B. competitor, or used its advantage in one market to gain an advantage in another market. Brinkley, Tr. 2236-40; Chapman Tr. 7242-44; Gonzaba, Tr. 2101-02.

Kroger ha. never dictated the terms and conditions under which branded CSDs may be sold in San Antonio or any other Texas market. Kaiser, Tr. 3215- 16. Indeed, Kroger has threatened not to run ads unless they got an equal deal on price, but never got a better price than others. Kaiser, Tr. 3216. 427 See, e. , United States v. United Tote 768 F. Supp. 1064 (D. Del. 1991) (liability found where 52% of market would be affected); FTC v. Ba. J BmJ. Enters.,- 1984- 1 Trade Cas. (CCH) paragraph 66 041 at 68 605 (N.D. Ohio 1984) (liability found where less than 1/3 of industry would havebeen affected). , , THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 617 452 Opinion VII. ENTRY We have held that a "primary consideration in evaluating the likely competitive effects of a merger or acquisition is the ease or difficulty with which new competitors might enter the market in response to supracompetitive pricing." Owens-Ilinois, slip op. at 27- 28. Under the Merger Guidelines, we recognize that if entry is " easy that market participants, after the merger, could not profitably maintain a price increase above premerger levels " then the merger is unlikely to lead to the exercise of market power. Merger Guidelines, Section 3.0. In such circumstances, the absence of barriers to entry "makes it highly unlikely that a merger or acquisition will have anticompetitive effects, because any effort to extract supracompetitive prices and profits will induce new entry, which will reduce prices to competitive levels. F. Goodrich 110 FTC at 295-96. On the other hand if prompt, effective entry is unlikely, customers may be exposed to sustained periods of anticompetitive harm." Owens- Ilinois, slip op. at 28. 428 In this case, the issue is whether a new bottler of branded CSDs could enter or whether an existing branded CSD bottler could expand sufficiently to remedy the anticompetitive effects that we have As we haveidentified as likely from the acquisition at issue.42' recently noted (tJhe Commission traditionally has assessed ease of entry by looking for identifiable barriers or impediments that could foreclose entry or prevent expansion by existing smaller tirms sufficient to forestall anti competitive conduct within the relevant market. Coca-Cola Co., Dkt. No. 9207, slip op. at 54. Entry barriers include any condition that necessarily delays entry into a market for a significant period of time and thus allows market power to be exercised in the interim. Echlin Mfg. Co. 105 FTC41O, 486 (1985). We have 428 See also United Slares v. Baker Hughes, Inc. 908 F.2d 981 (D. C. Cir. 1990); United Stales v. Waste Management, Inc. 743 F2d 976, 982 (2d Cir. 1984). 429 Under Section 1.32 of the Merger Guidelines, certain firms that participate in the market through supply-side response arc included as participants in the market, and are therefore treated separately from other firms that may enter the market. Here, following the Merger Guidelines avdroach would lead to the same conclusion.

We have found that expansion by CCUSA and DPUSA in sales of post-mix fountain syrup would be unlikely to prevent or disrupt tacit collusion by branded CSD bottlers. See Section VI.C.L supra. We also find that the record does not show that CCUSA and/or DPUSA would be likely to enter into bottling in order to disrupt price increases by branded CSD bottlers; the long list of cases in which collusion by branded CSD boulers was not prevented or disruptcd by entry by CCUSA and/or DPUSA supports our conclusion on this issue. See Section VLC.3 supra. Opinion 118 FTC. pointed out that "(bJarriers or impediments need not be absolute; rather, they are assessed 'in terms of the amount of time required for a motivated outsider to effect entry. '" Coca-Cola Co. , slip op.-at 54 citing Olin Corp. 113 FTC at 612; Owens-Ilinois, slip op. at 28.430 We find that the evidence in the record demonstrates that entry by a new branded CSD bottler would be difficult. The ALl agreed. Although the ALl found that entry "as a soft drink distributor is easy," he noted that, if the product and geographic markets asserted by complaint counsel were accepted, th n entry barrers existed. CCSW, the respondent, agrees. CCSW management has stated that the bottling business is characterized by. . . high barrers to entry.,,43 TBG, the owner of respondent, also agrees. A TBG presentation to its Credit Committee stated that TBG operates in an industry with strong barriers to entry/franchise monopolies/few competitors.,,433 Some aspects of the soft drink bottling businesses do not present any obstacles to entry. We agree with the ALl that the costs to lease delivery trucks and a warehouse are relatively small, and that a startup distributor could purchase contract-packed bottled and canned soft drinks without any capital expenditures for equipment. IDFF paragraphs 378, 380. If we had included private label and warehousedelivered CSDs in the relevant product market, we most likely would have agreed with the AU that entry into such a market would not be difficult.

But sales of branded CSDs are what concern us here, and entry as a branded CSD bottler is significantly more difficult. A branded CSD bottler must have a sufficient line of brands to be large enough to take advantage of various scale economies relating to the production, distribution, and marketing of CSDs. In Coca-Cola Co., we found that "(aJ bottler needs at least 8% to 15% of the local market for carbonated soft drinks to achieve minimum efficient scale. " Slip op. at 57. The record here indicates- that even a higher market share - perhaps over 20% where the bulk of the market is attributable to a single "flagship" brand -- may be necessary where one bottler such 430 The Merger Guidelines use a comparable analysis, assessing entry as "easy" if it is "timely, likely, and sufficient in its magnitude, character, and scope to deter or counteract the competitive effects of concem. " Section ID 72 & 3.0.22; IDFF paragraph 396.

432 ex 1406 Z9.

433 ex 3806 , .. THE COCA COLA BOTTING COMPANY OF THE SOUTHWEST 619 452 Opinion as CCSW dominates the market. 434 Conversely, branded CSD companies look to place their franchises with bottlers that have-rge enough operations that they can take advantage of such economies. Certain of the most important requirements for successful operation as. a branded CSD bottler interact, creating a situation in which each element is necessar in order to obtain the others. For example a branded CSD bottler must have a suffcient line of brands to gener- As weate enough volume to justify the costs of DSD delivery.436 noted previously, the testimony is consistent that DSD delivery is critical for the success of a branded CSD bottling operation. Conversely, in order to obtain a branded CSD franchise, a bottler would need to show that it intended to use DSD delivery. Moreover, to provide effective competition sufficient to thwart any unilateral or collusive anticompetitive activity, a new branded CSD bottler would need a line of brands with name recognition and volume sufficient to induce retailers to agree to ad features, not just As Mr.in-store or other, less effective promotional activities.43 Kaiser of Kroger explained, in selecting a brand for an ad-buy program (tJhe most important consideration we have is how strong the brand is (i.e., name recognitionJ and how many cases we can sell of it (i. volumes." Kaiser, Tr. 3231-32. As this discussion of the evidence makes clear, a key to competitive effectiveness as a branded CSD bottler is to obtain a line of brands suffcient to generate volume that will support the use of DSD delivery and the achievement of minimum efficient scale, and a volume and market share sufficient to provide the name recognition and throughput necessary to "grow the brand" through ad features and other significant promotions. In particular, a "cola" is necessary 434 Both r. Bodnar an r. Turner testified that P- A had Just reached cntlca ma. s In tenns of ability to obtain ad features in 1983. when DP-SA had just reached a market share in food stores of 22.6%. ex 1681 C; see also Section VLC.2. supra. 435 See, e. Amicus Brief ofDPUSA at 6- 436 ex 3941 at 288 (Schmidt; see al. Sections IV.C.2, VI.C, D supra 437See Section IV.C.2, I.C. D .fpra.

438 See Section IV.C.2 supra.

439 See Section VI.C.2 supra (Grant-Lydick has not provided significant competition to CCSW and Pepsi CaBO where Grant-Lydick could not obtain ad features). , Opinion 118 FTC. 0 The fact that the branded CSD bottlersto generate such volume that obtain ad features are those whose concentrate companies invest millions of dollars in advertising for their brands also indicates that a brand backed by substantial advertising by its concentrate company is necessary to achieve a level of competitive effectiveness sufficient to prevent an anticompetitive price increase.44I But it would be very difficult for a new entrant to obtain such a brand, much less a line of such brands. As we observed in Coca-Cola Co. (mJost local markets for carbonated soft drinks have a Coca- Cola bottler, a Pepsi-Cola bottler, and a so-called ' third bottler,' which carres various brands of soft drinks other than Coca-Cola or Pepsi-Cola brands." Slip op. at 57.442 The concentrate companies for branded CSDs are most interested in placing their brands with incumbents who have proven track records, not with new entrants who mayor may not be able to reach minimum effcient scale.443 In light of these facts, it is not surprising that expansion by an incumbent branded CSD bottler to defeat an anticompetitive price increase would also be very difficult. The pattern of franchise transfers in the relevant market has been that branded CSD concentrate companies seek to move their franchises to the largest bottler that is Just asnot prohibited from having them due to flavor restrictions.44 DPUSA moved its franchise to CCSW, so Dr Pepper/7-Up moved the Up franchise from Texas Bottlers -- with a 3.2% total branded market share in 1986445 -- to Grant-Lydick, with approximately a 14.3% total branded market share in 1986.446 This pattern reveals franchise moves that cause increasing concentration in this market 440 Bodnar, Tr. 1253- 441 See Section IV. C.3 supra. 442 See also Lydick, Tr. 2937, 2943.

443 See ex 3989 at 36 (Shanks).

44 Bottling franchises prohibit a bottler from selling more than one brand in a "flavor segment. IDFF paragraph 105; ex !668; RX 2938 C.

445 ex 1681 446 See ex 168! C (adjusting Grant-Lydick' s 1986 market share in food stores of 16. , ex 1681 C, fof fountain based on an interpolation of .85 from data in ex 4146 H results in an approxi"mate market share of 14.3% for Grant-Lydick in 1986). When Grant-Lydick acquired 7-Up, 7-Up had a market share of about 3%. (See 8% share in food stores shown in ex 1681 D. adjusted for fountain by . 23 factor set forth in ex 4146 H, results in overall market share of 10) The addition of this market share still was not suffcient to enable Grant-Lydick to reach the critical" mass that DPSA had just achieved in 1983 with the combination of Dr Pepper, RC, Canada Dry, and other branded CSD franchises. Bodnar, Tr. 1253-54.

THE COCA COLA BOTTING COMPANY OF THE SOUTHWEST 621 452 Opinion not expansion that would defeat an anti competitive price increase. Accordingly, we find that expansion by an existing incumbentas-well as entry by a new branded CSD bottler would be unlikely to defeat anticompetitive conduct in this market. 448 VII THE SOFT DRINK INTERBRAND COMPETITION ACT In a separate argument, respondent maintains that the Soft Drink Interbrand Competition Act ("SDICA"), 15 U.se. 3501-03, governs this proceeding and mandates dismissal of the complaint.449 We disagree.

It is apparent from the very language of the SDICA that the statute is a narrow one that does no more than legalize exclusive terrtorial restrictions and transshipping prohibitions.45!) The SDICA is thus solely concerned with legitimizing these vertical non-price restraints; it does not address horizontal acquisitions, which remain exclusively within the purview of the existing antitrust laws. Because the present case involves a horizontal acquisition and in no way 447 This pattern is consistent with a long-standing trend to bottler consolidation throughout the United States. See note 359 supra.

44R The Answering Brief of Respondent-Appellee did not assert any efficiencies that allegedly would outweigh any anticompetitive effects of the acquisition. Nonetheless, Respondent s Proposed Findings of Fact contain certain facts labelled as effciencies. See, e. RPFF paragraphs 527-531. To the extent that respondent relies on these facls on appeal, we find that such alleged effciencies do not outweigh the likelihood of a substantial lessening of competition due to CCSW' s acquisition of the Dr Pepper franchise, and that respondent made no showing that its alleged efficiencies could not be achieved by means other than the acquisition at issue in this case. See Merger Guidelines, Section 4 449 The SDICA provides as follows, in per1inent pari: Nothing contained in any antitrust law shall render unlawful the inclusion and enforcement in any trademark licensing contract or agreement, pursuant to which the licensee engages in the manufacture . , distribution, and sale of a trademarked soft drink product. of provisions granting the licensee the sale and exclusive right to manufacture, distribute and sell such product in a defined geographic area or limiting the licensee, directly or indirectly, to the manufacture, distribution, and sale of such product only for ultimate resale to consumers within a defined geographic area: Provided, that such product is in substantial and effective competition with other products of the same general class in the relevant market or markets.

15 U. c. Section 3501 (emphasis in original). 450 As the legislative history explains: "The Committee intends that (the SDICAJ provide necessar relief (that is, legitimizing exdusive territorial agreements when not anticompetitiveJ without Ilrantinll antitrust imml!nitv and without establishim:; anv oreeedent t t would weaken our bel ?guereq antitrust laws " H.R. Rep. No. 96- 1118, 96th Cong., 2d Sess. 2, 5 (1980) (emphasis added), reprinted in 1980 U. C.C.A.N. 2373, 2378. See also Commonwealth of Pen my Iv ani a v. Pep. ico, Inc. 836 F.2d 173 , 175-79 (3d Cir. 1988); Pepsi-Cola Metropolitan Bottling Co. v. Checkers, Inc. 754 F.2d 10, 18 (1st Cir. 1985); Coca- Cola Co. v. FTC, 642 F.2d 1387 , 1389-90 (D.C. Cir. 1981) (recognizing that the SDICA' s sole purpose is to legitimize, under cer1ain circumstances, the CSD industry s system of ex.clusive territorial distributorships). ,!.

Opinion lis F.T.c. challenges the existence of vertical territorial limitations and customer restraints 451 the SDICA is completely inapplicti. . IX. APPROPRIATE RELIEF Complaint counsel sought an order requiring divestiture of the Dr Pepper and Canada Dry franchises and prior approval by the Commission of any future acquisition by CCSW in the relevant market for a period of ten years from the date the Commission s order in this matter becomes tinal. The Commission has "wide discretion in its choice of a remedy," and "the courts will not interfere except where the remedy selected has no reasonable relation to the unlawful practices found to exist. Jacob Siegel Co. v. FTC 327 U.S. 608, 61 , 613 (1946). The Commission has the authority to impose prior approval requirements in merger cases. Abex Corp. v. FTC, 420 F. 2d 928 (6th Cir. 1970), cert. denied 400 U. S. 865 (1970). See also Coca-Cola slip op. at 63-64.

(l)t is industry market strcture and market conditions, not whether a ' knowing and deliberate violation ' or a ' likelihood of repeated unlawful conduct' has been shown that determnes the appropriateness of imposing a prior approval requirement in a particular case.

American Medical International, Inc. 104 FTC 1 224 (1984). We find that CCSW' s acquisition of the Dr Pepper franchise in the San Antonio market is likely substantially to lessen competition among branded CSDs in that market, and we therefore order divestiture of the Dr Pepper franchise to a Commssion-approved purchaser. Finding no anti competitive effects from the acquisition of the Canada Dry franchise, we decline to order its divestiture. In light of the highly concentrated market structure and the particular significance of increased market share in the branded CSD market in this case, we further order that CCSW must obtain Commission approval for any additional acquisitions in the relevant market for a period of ten years from the date on which the Commission s order in this matter becomes final. 45! In reaching this conclusion, we reject respondent's efforts to characterize the horizontal acquisition of assets (e. franchise agreements) from a competing bottler as a vcrtical transaction merely because licenses from concentrate companies are involved. If this argument were accepted, it would immunize vinually all acquisitions by bottlers. including the acquisition of a major competitor. fromantitrust scrutiny. THE COCA COLA BOITING COMPANY OF THE SOUTHWEST 623 452 Statement STATEMENT OF COMMISSIONER DEBORAH K. OWEN CONCURRING IN PART AND DISSENTING IN PART I agree that the acquisition of the franchise to produce and distribute Dr Pepper by the Coca-Cola Bottling Company of the Southwest CCSW") was likely to substantially lessen competition in the San Antonio market for branded carbonated soft drinks ("CSDs ). I therefore concur in the order to divest this franchise and to require prior approval for certain future acquisitions. I must nevertheless dissent from some of the reasoning accompanying the opinion of two Commissioners, which speculates on issues neither presented to the Commission, nor necessary to a decision.

The record is replete with evidence indicating a strong presumption that this merger created or enhanced market power or facilitated its exercise in the San Antonio market for branded CSDs, accompanied by a strong anti competitive effects story and difficult entry. The discussions in the opinion of two Commissioners relating to (I) the unilateral exercise of market power and (2) certain pricing behavior are, given the strength of the basic case, unnecessary to a just resolution .of this matter, and therefore contrary to accepted notions of judicial construction.

Concurring Statement 118 FTC. CONCURRING STATEMENT OF COMMISSIONER DENNIS A. Y AO I concur with the opinion of the majority that branded CSDs are an antitrust product market. The record supports both this conclusion and the existence of strong product differentiation between the takehome and cold drink segments of that market. With respect to the latter point, Section IV.C.I of the opinion discusses evidence that (i) Coca-Cola bottlers divide their busines es into take-home and cold drink markets, (ii) bottlesJcans are handled and marketed very differently than fountain products, and (iii) substantial price differences exist between equivalent-sized take-home versus cold drink branded CSDs. Such evidence of differentiation suggests the possibility that take-home branded CSDs also comprise an antitrust product market. My deliberations in this matter have led me to question whether, in the face of a price increase by branded CSD bottlers, retailers (other than convenience stores) could substitute cold drink individual can or fountain cup sales for take-home sales in 3-liter PET bottles or 6packs of 12-ounce cans, or whether fountain vendors could substitute I If we had foundsales in 3-liter PET bottles for individual can sales. a smaller relevant antitrust product market (take-home sales of branded CSDs) within a larger one (branded CSDs) in this case, that would not have been unique.' A take-home branded CSD market in the San Antonio area would be even more concentrated than the branded CSD market that we found However, since neither com- 1 We had no need to consider this issue in Coca- Cola Co., Dkt. No. 9207 (June 28. 1994), where we were examining whether branded CSD bottlers could substitute concentrate or syrup for each other in the face of a price increase by a concentrate company. There, the evidence compelled the conclusion that branded CSDs were the smallest relevant product market, since concentrate and syrup are linked in that syrup can be manufactured from concentrate. Indeed; CCSW manufactures fountain syrup from concentrate. Summers, Tr. 6508-09.

2 See Olin Corp., 113 FTC 400, 598- 600 (1990), afrd, 986 F.2d 1295 (9th Cir. 1993), cert denied, 114 S. Ct. 1051 (1994) (competitive effects analyzed within both a broader antitrust product market including the premium-priced and less expensive products, and a smaller antitrust product market consisting of only the premium-priced product). The pre- and post-acquisition HHIs would be: Pre-acquisition HHI 3841 Post-acquisition HHI 4554 HHI Increase 7 13 CX 1681 A; Hilke, Tr. 6033. These HHls are based on Nielsen data for soft drink sales in food stores in BexarCounty (which includes San Antonio), comparing Ocl/Nov. 1983 (pre-acquisition) with Aug.lSept. 1984 (post-acquisition) sales. Hilke, Tr. 6030. Since Nielsen data automatically exclude fountain and vending sales of branded CSDs, and since Bexar County accounts for 86% of the population in the IO-county relevant geographic market (Hilke, Tr. 6030, 6262; ex 4J31 A), these data provide a reasonably accurate measure of take-home branded CSD sales in the San Antonio. area. THE COCA COLA BOTTING COMPANY OF THE SOUTHWEST 625 452 Final Order plaint counsel nor respondent directly considered or briefed this possibility, we do not have a full record on which to decid this-pnt nor is it necessary, given the solid evidence of strong product differentiation within the branded CSD market.

FINAL ORDER This matter having been heard on the appea of complaint counsel from the initial decision, and on briefs and oral argument in support , and in opposition, to the appeal; for the reasons stated in the attached opinion, the Commssion has determined to grant the appeal in part, and reverse the initial decision. Accordingly, It is ordered That the following order be and the same hereby is ordered:

I. DEFINITIONS It is ordered That for the purposes of this order, the following definitions apply:

A. CCSW" means Coca-Cola Bottling Company of the Southwest, its directors, officers, employees, agents and representatives, its successors and assigns, its predecessors, subsidiaries, divisions groups and affiliates controlled by CCSW, directly or indirectly, and their respective directors, officers, employees, agents and representatives, and their respective successors and assigns. B. Affiliate means any firm in which there is 10% or more ownership or control, directly or indirectly, between firms. C. Concentrate means the base element, flavors or essences mixed according to a formula which, when added to carbonated water and nutritive or non-nutritive sweetener, is a carbonated soft drink. D. Syrup means the concentrate and nutritive or non-nutritive sweetener which, when added to carbonated water, is a carbonated soft drink.

E. Carbonated soft drink" means a carbonated beverage that does not contain alcohol and is produced by combining carbonated water with a sweetener and concentrates or with syrup. F. Branded carbonated soft drink" means a carbonated soft drink identified with any nationally or regionally recognized label; Final Order 118 P. name, or trademark that is, in general, heavily advertised, widely available, and ordinarly distributed by the direct-store-door delivery method. This definition does not include a label, name, or trademark associated solely with a single grocery or restaurant retailer, or with a generic flavor.

G. Branded concentrate or syrup means concentrate or syrup used to produce branded carbonated soft drinks. H. Direct-store-door delivery means a method of distribution whereby the producer or distributor delivers product directly to the retail outlet and ordinarily positions the product for sale to the retailer s customers.

I. Acquired Dr Pepper assets means the franchise to produce and distribute Dr Pepper products acquired by CCSW from San Antonio Dr Pepper Bottling Company on or about September 1984 and any franchises to produce and distribute Dr Pepper products in the San Antonio area acquired by CCSW after September 1984. II.

It is further ordered That within twelve (12) months after the date this order becomes final, CCSW shall divest the acquired Dr Pepper assets absolutely and in good faith, at no minimum price. The divestiture shall be only to an acquirer, and only in a manner, that receives the prior approval of the Commission. Pending any divestiture required by this order, CCSW shall take all measures necessary to maintain the acquired Dr Pepper assets in their present condition and shall not cause or permit impairment of the marketability or viability of such assets. The purpose of the divestiture is to remedy the lessening of competition found in the Commission s decision. II.

It is further ordered, That:

A. If CCSW has not divested the acquired Dr Pepper assets absolutely and in good faith and with the Commission s prior approval, within twelve (12) months after the date this order becomes final, CCSW shall be subject to the appointment by the Commission of a trustee to effect the divestiture. In the event the Commission or the Attorney General brings an action pursuant to Section S(I) of the THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 627 452 Fina! Order Federal Trade Commission Act, 15 D. C. 45(1), or any other statute enforced by the Commission, CCSW shall consent to the appoiflnt of a trustee in such action. Neither the appointment of a trustee nor a Commssion decision not to appoint a trustee under this paragraph shall preclude the Commission or the Attorney General from seeking civil penalties and any other relief available, including a courtappointed trustee, pursuant to Section 5(1) of the Federal Trade Commission Act, 15 D. C. 45(1), or any oth.er statute enforced by the Commission, for any failure by the CCSW to comply with this order.

B. If a trustee is appointed by the Commission or a court pursuant to this order, CCSW shall be subject to or, in the case of a court-appointed trustee, shall consent to the following terms and conditions regarding the trustee s powers, authority, duties, and responsibilities:

(I) The trustee shall be selected and appointed by the Commission or, in the case of a court-appointed trustee, bithe court. The trustee shall be a person with experience and expertise in acquisitions and divestitures. The appointment shall be effective fifteen (IS) days (the "effective date ) after CCSW' s receipt of written notifications of such appointment or, in the case of a court-appointed trustee, at such time as the court may order, unless CCSW has, on or before the effective date, presented substantial grounds for disqualification of the trustee. In the event of such objection to the appointment of the trustee, the effective date shall be stayed pending a determination by the Commission or, in the case of a court-appointed trustee, by the court.

(2) The trustee shall have the exclusive power and authority, subject to the prior approval of the Commission, to divest the acquired Dr Pepper assets. The trustee shall have twelve (12) months from the date of appointment to accomplish the divestiture. If however, at the end of the twelve-month period, the trustee has submitted a plan of divestiture or believes that divestiture can be accomplished within a reasonable time, the divestiture period may be extended by the Commission or, in the case of a court-appointed trustee, by the court.

(3) The trustee shall have full and complete access to the personnel, books, records and facilities of CCSW concerning the acquired assets, and CCSW shall develop such financial or other Final Order 118 F. information relevant to the property to be divested as the trustee may reasonably request. CCSW shall cooperate with the trust nd shall take no action to interfere with or impede the trustee s accomplishment of the divestiture. Any delays in divestiture caused by CCSW shall extend the time for divestiture in an amount equal to the delay, as determined by the Commission or, for a court-appointed trustee by the court.

(4) Subject to CCSW' s absolute and'unconditional obligation to divest at no minimum price and to the purpose of the divestiture as stated in paragraph II of this order, the trustee shall use his or her best efforts to negotiate the most favorable price and terms available in each contract that is submitted to the Commission for approval. The divestiture shall be made in the manner set out in paragraph II provided, however, that if the trustee receives bona fide offers from more than one prospective acquirer, and if the Commission approves more than one such acquirer, then the trustee shall divest to the acquirer selected by CCSW from among those approved by theCommission. (5) The trustee shall serve, without bond or other security, at the cost and expense of CCSW on such reasonable and customary terms and conditions as the Commission or a court may set. The trustee shall have the authority to employ, at the cost and expense of CCSW such consultants, attorneys, investment bankers, business brokers accountants, appraisers, and other representatives and assistants as are reasonably necessary to carry out the trustee s duties and responsibilities. The trustee shall account for all monies derived from the divestiture and for all expenses incurred. After approval by the Commission and, in the case of a court-appointed trustee, by the court of the account of the trustee (including fees for his or her services), all remaining monies shall be paid at the direction of CCSW, and the trustee s power shall be terminated. The trustee compensation shall be based at least in significant part on a commission arrangement contingent on the trustee s divesting the acquired assets. CCSW shall indemnify the trustee and hold the trustee harmless against any losses, claims, damages, or liabilities arising in any manner out of, or in connection with, the trustee duties under this order. Within forty-five (45) days after the appointment of the trustee and subject to the prior approval of the Commission and, in the case of a court-appointed trustee, of the court, CCSW shall execute a trust agreement that transfers to the THE COCA COLA BOTTLING COMPANY OF THE SOUTHWEST 629 452 Fina! Order trustee all rights and powers necessary to permt the trustee to effect the divestiture required by this order.

(6) If the trustee ceases to act or fails to act diligently, a substitute trustee shall be appointed in the same manner as provided in this paragraph.

(7) The Commission (or, in the case of a court-appointed trustee the court) may on its own initiative or at the request of the trustee issue such additional orders or directions as may be necessary or appropriate to accomplish the divestiture required by this order. (8) The trustee shall report in writing to CCSW and to the Commission every sixty (60) days concerning his or her efforts to accomplish divestiture.

IV.

It is further ordered, That, within sixty (60) days after the date this order becomes final and every sixty (60) days thereafter until CCSW has fully complied with the provisions of paragraphs II and II of this order, CCSW shall submit to the Commission a verified written report setting forth in detail the manner and form in which it intends to comply, is complying or has complied with those provisions. CCSW shall include in its compliance reports, among other things that are required from time to time, a full description of all substantive contacts or negotiations for the divestiture of the acquired Dr Pepper assets, including the identity of al1 parties that either contacted CCSW or were contacted by CCSW. CCSWalso shall include in its compliance reports copies of all written communications to and from such parties, all internal memoranda, and all reports and recommendations concerning divestiture. It is further ordered, That CCSW, for a period of ten (10) years from the date this order becomes final, shall not acquire, directly or indirectly, without the prior approval of the Commission: A. The whole or any part of the stock, share capital, or equity interests in any company or firm:

, Final Order 118 FTC. (I) Engaged in the manufacture, distribution, or sale of branded concentrate or syrup or branded carbonated soft drinks;- (2) Engaged in the franchising or licensing of any brand, name or trademark used in connection with the manufacture, distribution, or sale of branded concentrate or syrup or branded carbonated soft drinks; or (3) Holding an exclusive franchise or license of any branded concentrate company in any geographic area in which CCSW is engaged in the manufacture, distribution, or sale of branded concentrate or syrup or branded carbonated soft drinks; or B. Any franchise, license, brand, label, name or trademark associated with, or any assets engaged in, used for, or previously used for (and stil suitable for) the manufacture, distribution, or sale of concentrate, syrup or carbonated soft drinks in any geographic area in which CCSW is engaged in the manufacture, distribution, or sale of branded concentrate or syrup or branded carbonated soft drinks. Provided, however, that this provision shall not apply to the purchase or acquisition of any assets worth less than $100,000. One (I) year after the date this order becomes final, and annually thereafter for the following nine (9) years and at such other times as the Commission or its staff may request, CCSW shall file with the Commission a verified written report of its compliance with paragraph V of this order.

VI.

It is further ordered That, for the purpose of determining or securing compliance with this ords:r, and subject to any legally recognized privilege, upon written request and on reasonable notice to CCSW made to its principal office, CCSW shall permit any duly authorized representatives of the Commission: (A) access, during office hours and in the presence of counsel, to inspect and copy all books, ledgers, accounts, correspondence, memoranda and other documents in the possession or under the control of CCSW relating to any matters contained in this order; and (B) upon five (5) days notice to CCSW and without restraint or interference from CCSW, to interview officers or employees of CCSW, who may have counsel present, regarding such matters.

THE COCA COLA BOTIING COMPANY OFTHE SOUTHWEST 631 452 Final Order VII.

It is further ordered That CCSW shall notify the Conuission at least thirty (30) days prior to any proposed change in CCSW such as dissolution, assignment or sale resulting in the emergence of a successor corporation, the creation, dissolution or sale of subsidiaries or any other change that may affect compliance obligations arising out of the order.

Commissioner Azcuenaga and Commissioner Starek recused. I Prior to leaving the Commission fOff Commissioner Owen registered her -vote- in the affirmative for the Opinion of the Commission and the Final Order in this matter, with the notation that she dissented in part. as to discussions in the Opinion of the Commission relating to the unilateral exercise of market power and certain pricing behavior. Complaint 118 P.