The Great Atlantic & Pacific Tea Company, Inc

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

IN THE MA TIER OF THE GREAT ATLANTIC & PACIFIC TEA COMPANY INC., ET AL.

ORDER OPINION ETC. , I?- REGARD TO ALLEGED VIOLATION OF THE FEDERA TRAE COMMISSIOI' ACT AND SEe. 2 OF THE CLAYTON ACT Doket 8866. Complaint, Oct. 1971 Final Ord€r, Aprl, 1976 Order requiring a New York City operator of a large chain of retail grocry stores among other things to cease knowingly inducing, accpting or receiving net prices below that of its competitors in the purcha.'\€ ofmilk and other dairy products.

Appearances For the Commission: John J. Mathias, Edwin R. SO€ffing, and Andrew G. Stone.

For the respondents: Denis McInerny, Rayrrnd L. Falls, Jr. Thomas F. Curnin, Joseph P. Foley and Ira J. Dembow, Cahill Gordon, Sonnett, Reindl Ohl New York City, for The Great Atlantic & Pacific Tea Company, Inc. H. Blair White, Tlwodore N. Miller and Nathan P. Eirner, Sidley Austin Chicago, Ill. and Walter W. Koclwr Columbus, Ohio, for Borden, Inc.

COMPLAINT The Federal Trade Commission, having reason to believe that respondent The Great Atlantic & Pacific Tea Company, Inc. (hereinafter referred to as A&P), has violated and is now violating the provisions of Section 5 of the Federal Trade Commission Act (15 D. 45) and the provisions of subsection (f) of Section 2 of the Clayton Act (15 D. C. 13f) as amended by the Robinson-Patman Act, approved June 19, 1936, and that respondent Borden, Inc. (hereinafter referred to as Borden), has violated and is now violating the provisions of Section 5 of the Federal Trade Commission Act (15 D. C. 45), and it appearing to the Commission that a proceeding by it would be in tbe public interest, hereby issues its complaint charging as follows: COU T I PARAGRAH 1. Respondent A & P is a corporation organized, existing and doing business under and by virtue of the Jaws of the State of :Iaryland with its principal office and place of business located at 420 Lexington Ave., New York, New York.

GREAT ATLANTIC & PACIFIC TEA CO., INC., ET AL. 963 962 Complaint PAR. 2. Respondent A&P is now, and for many years has been engaged in the operation of a large chain of retail grocery stores. There are presently more than 4500 stores operated by respondent in 34 States of the United States, the District of Columbia and in Canada. Respondent A&P also operates a number of plants for the manufacture and processing of food and other products handled in its stores including plants which process and manufacture milk and other dairy products. A & p' s annual sales for the fiscal year ended February 22 1969, totalled over $5 436 325 000 and for the fiscal year ended February 28 1970, they totalled $5 753 692 000. Included among A&P's retail grocery chain stores are approximately 260 or more stores located in the States of Ilinois (including the Chicago metropolitan area), Indiana and Iowa, which stores comprise the Chicago Division of A&P. This division was formerly identified as the Chicago Unit of A&P' s Middle Western Division. PAR. 3. In the course and conduct of its business, respondent A&P has been and is now engaged in commerce, as "commerce" is defined in the Federal Trade Commission Act and the amended Clayton Act. In the course of that commerce A&P has been purchasing grocery and household products, including milk and other dairy products, for resale within the United States, from sellers also engaged in "commerce " as commerce is defined in the Federal Trade Commission Act and the amended Clayton Act.

In connection with such transactions, A&P is now, and has been, in active competition with other corporations, partnerships, firms and individuals also engaged in the purchase for resale and resale of grocery and household products, including milk and other dairy products, of Jike grade and quality which are purchased from the same or competitive sellers.

The aforesaid sellers are located in the various States of the United States, and respondent A&P and such sellers cause the products when purchased by said respondent, to be transported from the place of processing, manufacture or purchase, to A&P's retail stores and warehouses located in the same State and various other States of the United States. Further, in many instances the sellers must purchase or obtain raw materials, supplies and finished products from States other than the State in which such products are processed, manufactured or purchased as aforesaid, in order to fulfil the obligations of said sellers in their commitments to supply respondent A&P. PAR. 4. Respondent Borden (formerly The Borden Co.) is a corporation organized, existing and doing business under and by virtue of the laws of the State of New Jersey with its principal office and place of business located at 350 Madison Ave. , New York, New York. 964 FEDERAL TRADE COM:.ISSIOK DECISIONS Complaint 87 F.

PAR. 5. Respondent Borden is a holding and operating company having on December 31, 1969 , a 100 percent voting power in approximately 22 subsidiary corporations.

Respondent Borden has approximately 200 plants in the United States and Canada that are managed by four operating divisions. A diversificd dairy business including virtually all branches thereof, is conducted by Borden s Dairy and Services Division (formerly the Milk and Ice Cream Division). For this division, Borden s chief trade name is Borden.

Respondent Borden owns, maintains and operates a large number of receiving stations, processing and manufacturing plants and distribution depots located in various States of the United States from which it sells and distributes its said prcducts to purchasers. Borden s net sales amounted to approximately $1 740 184 687 in 1969 and approximately $1 827 341 000 in 1970.

PAR. 6. Respondent Borden sells milk and other dairy products of like grade and quality to a large number of purchasers located throughout the States of the United States, including the States of Ilinois, Indiana and Iowa, for use, consumption or resale therein. PAR. 7. In the course and conduct of its business, respondent Borden is now, and for many years past has been, transporting raw milk, or causing the same to be transported, from dairy farms and other points of origin to said respondent's receiving stations, processing and manufacturing plants and distribution depots located in States other than tbe State of origin.

Respondent Borden is now, and for many years past has been transporting milk and other dairy products, or causing the same to be transported, from the State or States wbere sucb products are processed, manufactured or stored in anticipation of sale or shipment to purchasers located in other States of the United States. Respondent Borden also sells and distributes its said milk and other dairy products to purchasers located in the same States and places where such products arc manufactured or stored in anticipation of sale. All of the matters and things, including the acts, practices, sales and distribution by respondent Borden of its said milk and other dairy products, as hereinbefore alleged, were and are performed and done in a constant current of commerce, as "commerce" is defined in the Federal Trade Commission Act.

PAR. 8. Respondent Bordcn sells its milk and other dairy products to retailers. Borden s retailer-purchasers resell to consumers. :vany of said respondent's retailer-purchasers are in competition with other retailer-purchasers of Borden.

PAR. 9. In the course and conduct of its business in commerce GREAT ATLANTIC & PACIFIC TEA CO., INC., ET AL. 965 962 Complaint respondent A&P has entered into an agreement with respondent Borden for the supply of milk and other dairy products, packaged under A&P's own private label, to the majority of the stores of A&P' Chicago Division (Chicago Unit at the time of initiation of the agreement). At the time Borden tendered its final offer to A&P, the offer which was accepted to form the said agreement, Borden informed respondent A&P that the offer was being granted for the purpose of meeting competition in the form of an existing offer offers which A&P then had in its possession. A&P accepted the said offer of Borden with knowledge that Borden had granted a substantially lower price than that offered by the only other competitive bidder and without notifying Bordcn of this fact. By the term private label, it is meant that such products were packaged under labels bearing brand names owned by A&P or peculiar to the retail operations of A&P, its divisions and subsidiaries, rather than under labels displaying brand names owned by Borden or peculiar to the operations of Borden.

PAR. 10. The aforegoing act and practice of respondent A&P violates the policy of Section 2 of the Clayton Act, as amended, is to the prejudice of the public and constitutes an unfair method of competition in commerce and an unfair act or practice in commerce within the intent and meaning and in violation of Section 5 of the Federal Trade Commission Act. (15 U. C. 45).

COLJ' T II PAR. 11. Paragraphs One through Ten of Count I hereof are hereby set forth by reference and made a part of this count as fully and with the same effect as if quoted herein verbatim. PAR. 12. In the course and conduct of its business in commerce respondent A&P has knowingly induced or received discriminations in price which are prohibited by subsection (a) of Section 2 of the Clayton Act, as amended.

Respondent A&P, in its negotiations with respondent Borden, before and after November 1 , 1965 , for the supply of milk and other dairy products under private label to the stores of A&P's Chicago Unit (now Division), knowingly induced prices which were and are discriminatory under the provisions of Section 2 of the amended Clayton Act. Further prices fromrespondent A&P has knowingly induced or received Borden in the purchase of such products for said stores which said prices were and are discriminatory under the provisions of Section 2 of the amended Clayton Act.

PAR. 13. When respondent A&P knowingly induced or received the discriminatory prices from its supplier, as alleged, A&P knew or should 966 FEDERAL TRADE CO:.MISSION DECISIONS Complaint 87 F.

have known that such prices constituted discriminations in price prohibited by subsection (a) of Section 2 of the Clayton Act, as amended by the Robinson-Patman Act.

PAR. 14. The aforegoing acts and practices of A&P are in violation of subsection (f) of Section 2 of the Clayton Act, as amended. COL T II PAR. 15. Paragraphs One through Thirteen of Counts I and II hereof are hereby set forth by reference and made a part of this count as fully and with the same effect as if quoted herein verbatim. PAR. 16. During the negotiations and dealings between respondents before and after November 1 , 1965, there developed a course of conduct to be followed by respondents. Pursuant thereto, A&P, for its part, introduced and sold private label milk and other dairy products at the prevailing retail price level for vendor label milk and dairy products, in spite of the fact that it was paying as much as 11 cents per gallon less for private label milk than it was paying for vendor (Borden) label milk. Thus, for its part in said course of conduct, A&P sustained and maintained existing retail prices for milk and other dairy products, and did not pass on to the consuming public any of the substantial price savings on private label milk and other dairy products which it received by reason of the wholesale price given it by Borden. Further, pursuant to said course of conduct, Borden, for its part failed to pass on, at the wholesale level, price reductions, similar to the reductions granted to A&P, to other purchasers who compete with A&P in the market areas covered by A&P' s Chicago Unit (now Division).

The foregoing course of conduct constitutes a combination between respondents which had the tendency or effect of stabilizing and maintaining prices for milk and other dairy products. Said combination had the further effect of permitting A&P to retain the substantial monetary and competitive benefits of the discriminatory price advantage which it had obtained as a result of its private label agreement with Borden.

PAR. 17. The aforegoing acts and practices of respondents A & P and Borden are to the prejudice of the public and constitute unfair methods of competition in commerce and unfair acts or practices in commerce within the intent and meaning and in violation of Section 5 of the Federal Trade Commission Act. (15 U.sC. S45). GREAT ATLANTIC & PACIFIC TEA CO., INC. , ET AL. 967 962 Initial Dccision INITAL DECISION BY HARRY R. HINKES, ADMINISTRATIVE LAW JUDGE SEPTEMBER 24 , 1975 PRELIMI ARY STA EMn.a (1) This proceeding began with the issuance of a complaint by the Federal Trade Commission on October 8 , 1971, against The Great Atlantic & Pacific Tea Company, Inc., (hereafter A&P) and Borden Inc., (hereafter Borden). The complaint contains three counts. Count I of the complaint charges that A&P violated Section 5 of the Federal Trade Commission Aet as well as the policy of Section 2 of the Clayton Act as amended, in the manner and method by which it negotiated (2) its price on private label milk. 1 The complaint charges that when Borden tendered its final offer to A&P it informed A&P that the offer was being granted for the purpose of meeting competition but that A&P accepted the offer with knowledge that Borden had granted a substantially lower price than that offered by the only other bidder and without notifying Borden of this fact.

Count II of the complaint charges A&P with a violation of subsection (f) of Section 2 of the Clayton Act as amended by knowingly inducing or receiving discriminations in price which are prohibited by subsection (a) of Section 2 of the Clayton Act as amended, from its supplier Borden.

Count III of the complaint charges both Borden and A&P with a violation of Section 5 of the Federal Trade Commission Act by a course of conduct constituting a combination between them which had the tendency or effect of stabilizing and maintaining prices for milk and other dairy products.

AI1 of the violations alleged in the complaint center about the private label supply agreement entered into between respondents A&P and Borden under which Borden began private label sales to A&P' Chicago Unit's stores on or about November 1 , 1965. Such sales continued under the agreement until February 1972. The agreement covered sales of milk and dairy products to more than 200 A&P stores located in a multistate area including portions of Ilinois and Indiana. The ilegal conduct described in Count I is alleged to have occurred during A&P's course of dealings with Borden leading up to the agreement, which dealings took place from late 1964 up until November 1, 1965. The il1legal conduct described in Count II is alleged to have occurred during the same period and also during a later time period when the agreement was in force. In connection with the ilegal combination alleged in Count III, complaint counsel limited their main 1 Prehewing conference of :May 15, 1973 (Tr. p. 5) 968 FEDERAL TRADE COMMISSION DECISIO)(S Initial Dccision 87 F. period of proof to include November 1 , 1964 through December 31 1968.

(3 J After a number of prehearing conferences, hearings commenced on June 11, 1973 and continued with short intcrruptions until March 27 1975 when the record was closed for further evidence. Not only is this case one of great complexity but its dimensions as well arc extraordinary. In addition to the more than 110 hearing days expended, the scores of witnesses heard, the stipulations received, the hundreds of exhibits aggregating about 10 000 pages received, and the more than 11 000 transcript pages of testimony and argument, counsel for the parties have submitted for my consideration briefs totaling more than 2 000 pages which, considering the limited time given for such endeavor, are most commendable.

Any motions not heretofore or herein specifically ruled upon, either directly or by necessary effect of the conclusions in this initial decision are hereby denied.

The proposed findings, conclusions and briefs as submitted by the parties have been given careful consideration and to the extent not adopted by this decision in the form proposed or in substance are rejected as not supported by the evidence or as immaterial. Refercnces to the record are made in parenthesis using the following abbreviations:

CX - Commission s Exhibit; RAPX - A&P' s Exhibit; Ans. Answer; Tr. - Transcript of the Testimony: DTR - Malone Deposition Transcript, CX 262; Compo -Complaint; Adm. - Admission.

Having reviewed the record in this proceeding and having considered the demeanor of the witnesses as they testified, together with the proposed findings, conclusions and briefs submitted by the parties, 1 make the following:

FI:\DIl'GS OF FACT A. A&P\ IDE TITY AND BLJSIKESS NATCRE 1. Respondent A&P is a corporation organized, existing and doing business under and by virtue of the laws of the (4J State of Maryland with its principal office and place of business located at 420 Lexington Ave., New York, New York (Comp. Par. 1; Ans. Par. 1). 2. Respondent A&P is now and for many years has been engaged in the operation of a large chain of grocery stores. During the middle 1960' s A&P ranked first nationally in terms of sales among retail grocery chains (CX 218G; Smith, Tr. 1340). It operatcs approximately GREAT ATLANTIC & PACIFIC TEA CO., INC., ET AL. 969 962 Initial Dccision 329 stores in 36 States of the United States, the District of Columbia and Canada. It also operates a number of plants for thc manufacture and processing of food and other products handled in its stores including plants which process and manufacture milk and other dairy products. Its annual sales for the fiscal year ending February 22, 1969 totaled over 5.4 billion dollars and, for the fiscal year ending February 1970 , totaled over 5.7 billion dollars (Comp. Par. 2; Ans. Par. 1). A&P stores carry a full range of grocery and related products, many of which arc purchased by A&P from other national sellers of goods and are shipped to A&P across State boundaries (CX 218E, F; Smith, Tr. 1387-1388; Schmidt, Tr. 1677-1678). A&P also manufactures and sells many private label products in its stores and did so even before 1965 (Smith, Tr. 1380-1381). It was A&P policy to promote private label products where feasible because their sale usually resulted in larger gross profit, b'Teatcr attraction of the public to A&P stores and more flexibility for A&P in its dealing with suppliers (Corbus, Tr. 7325). 3. At the time of the private Jabel ab'Teement which is involved in this case A&P' s Chicago Unit encompassed the States of Iowa, Indiana and Illinois (including the Chicago metropolitan area). It operated approximately 260 stores. In 1969, the Unit became a Division which was compriscd of 236 stores (Comp. Par. 2; Ans. Par. 1; Schmidt, Tr. 1675; Bartels, Tr. 1853- 1854).

4. Products sold through A&P's Chicago Unit stores were procured under the supervision and control of the Head Buyer located in Chicago who, for most of the 1964-1969 period, was Mr. EJmer Schmidt (Schmidt). He was also personally responsible for the Unit' s purchases of dairy products (5) except for cheese (Schmidt, Tr. 1667-1669). During the middle 1960's Bordcn was the principal supplier of dairy products to A&P's Chicago l'nit and supplied about 95 percent of that Unit s requirements (Schmidt, Tr. 1678). Sales of dairy products by Borden to A&P's Chicago Unit amounted to milions of dollars annually with purchases of eleven of the most important items amounting yearly to somewherc between 5 and 6 million dollars (CX 263B; Schmidt, Tr. 1679).

5. It was A&P's Chicago l'nit which was involved in the private label agreement with Bordcn here at issue. And although A&P Headquarters in New York gave final approval to that private label agreement, and reviewcd the reliability of the supplier chosen (Smith Tr. 1347), the private label negotiations themselves were handled primarily at the l'nit level (Smith, Tr. 1355-1356; Archer, Tr. 1232- 1233). Indeed, it was A&P' s l'nit Buyer, Schmidt, who was the principal company official involved for A&P in thc Chicago private label negotiations (Bartels, Tr. 1865). Schmidt undertook these 216- 969 O- I.':' - ,7 - 62 Initial Decision 87 F.

negotiations as part of a planned move toward private label originating at A&P's New York Headquarters (BarteJs, Tr. 1862-1863; Smith, Tr. 1351-1352).

6. The A&P Chicago Unit was one of four Units comprising the A&P Middle Western Division which included among other areas Milwaukee, Wisconsin (Smith, Tr. 1344; Bartels, Tr. 1857). Mr. Ira Bartels (Bartels) was A&P' s Director of Purchases for the Middle Western Division from 1964 until 1969 and supervised the activities of Schmidt (Bartels, Tr. 1853-1854, 1858; Schmidt, Tr. 1673-1674). 7. During the middle 1960' s A&P was composed of between five and seven divisions consisting in all of 32 L'units. In charge of purchases of each Unit was the L'nit Buyer such as Schmidt of the Chicago Unit. The Unit Buyer reported directly to the Divisional Purchasing Director, such as Bartels of the Middle Western Division, who in turn reported to the Division President and to the National Purchasing Director located in New York City headquarters. Similarly, the Unit Sales Manager reported directly to the Divisional Sales Director who in turn reported to the Division President and to the National Director of Sales located in the New York City Headquarters (Smith, Tr. 1336- 1337; CX 258). In 1969 the Unit-Division setup was abolished and A&P created 32 Operating Divisions so that the Chicago Unit became a Division. Bartels was designated Purchasing Manager when the Chicago Unit became a Chicago Division and Schmidt was designated as Buyer and (6) later as Administrative Assistant (Bartels, Tr. 1853- 1854; Schmidt, Tr. 1663-1664). When Bartels left the employment of A&P in 1970, Mr. Edmund Bayma (Bayma) became Purchasing Director of the Chicago Division (Bayma, Tr. 5918). B. BORDEN S IDENTITY AND BCSfNESS NATURE 8. Respondent Borden (formerly The Borden Co.) is a corporation organized, existing and doing business under and by virtue of the laws of the State of New Jersey with its principal office and place of business located at 277 Park Avenue, New York, New York (Comp. Par. 4; Ans. Par. 15-4).

9. Respondent Borden does business throughout the United States and in Canada (Minkler, Tr. 88; Archer, Tr. 1203). As of December 31 1969 Borden held a 100 percent voting power in about 22 subsidiary corporations and possessed about 200 plants that were managed by four Operating Divisions (Comp. Par. 5; Ans. Par. 15-5). Its Dairy and Services Division conducts a dairy business in which the chief trade name is Borden. Borden s net sales in 1969 as well as in 1970 amounted to over 1.8 billion dollars (Comp. Par. 5; Ans. Par. 15-5). 10. In the course of its business Borden owns, maintains and , p.

GREAT ATLANTIC & PACIFIC TEA CO. , IKC. , ET AL. 971 962 Initial Dccision operates receiving stations, processing and manufacturing plants and distribution depots located in various States from which it sells and distributes its products. Borden sells its milk and other dairy products to various purchasers including retailers. Most of its retailer-customers resell to consumers and certain of its retailer-customers are in competition with other of its retailer-customers (Comp. Par. 8; Ans. Par. 15-8). A&P, whom Borden served in the Chicago and Northwest Indiana area during aJl of the time involved in this case, is now and has been in active competition with certain other corporations, partnerships, firms and individuals also engaged in the purchase for resale and resale of grocery and household products including milk and certain other dairy products (Comp. Par. 3; Ans. Par. 15-3; A&P' s Adm. dated March 12, 1973 , p. 4, 5; Gintert, Tr. 244; Barney, Tr. 3688). 11. From the early 1960's through the end of 1966 Borden managed certain of its dairy operations through its Chicago Central District or Division which included Chicago, Illinois, the northern part of Ilinois part of Iowa, part of Minnesota, most of Wisconsin, part of Michigan and the (7J northwestern corner of Indiana (Minkler, Tr. 88). The Chicago Central District Management Group consisted in part of Mr. Ralph Minkler, President (Minkler); Mr. Joseph Malone (Malone), Vice President, Regulatory Controls (and formerly Comptroller); :Ir. Gordon Tarr (Tarr), Sales Manager for Chain Stores; and Mr. Orville Gose (Gose), Regional General Manager (Minkler, Tr. 88, 111-113; Tarr Tr. 843; Gose Tr. 1037). After 1966, Borden s Chicago Central District was merged into its Midwest District in Columbus Ohio. :Iinkler retired in December 1966 but Malone, Tarr and Gose remained (Minkler, Tr. 87, 279, 331; CX 262 Malone, DTR. 6, 7; Tarr, Tr. 842; Gose, Tr. 1067).

12. Minkler reported to the President of Borden s Milk and Ice Cream Division in Borden s New York Headquarters. That post was first filed by Mr. Harry Archer (Archer) who was succeeded in the middle 1960's by Mr. Pentz (Minkler, Tr. 98, 126; Archer, Tr. 1203-1205 1207-1208; CX 292 p. 2). After Chicago CcntraJ District' s merger into the Midwest Division, Borden s Chicago personnel reported to Mr. Patterson, President of Borden s Midwest District in Columbus, Ohio who in turn reported to Borden s New York City headquarters (Gose Tr. 1059).

C. COMMERCE 13. A&P admits that it has been and is now engaged in "commerce as that term is defined in the Federal Trade Commission Act and amended Clayton Act (Ans. Par. 2; Adm. dated September 27, 1972 across State lines goods, Sec. III , 1-2). A&P also admits transporting Initial Decision 87 F. which it had produced or processed in one State for sale through its stores in other States. Its Chicago Unit included stores in the States of Ilinois, Indiana and Iowa (Schmidt, Tr. 1665 1674) and advertised in newspapers distributed in both Ilinois and Indiana. 14. Borden also admits that it has been and is now engaged in commerce" as that term is defined in the Federal Trade Commission Act (Adm. filed March 19, 1973, p. 11 Sec. II-3). It also admits that it regularly transported raw milk from one State to another and that its plants received in excess of 10 percent of their raw milk supply from dairy farms in States other than those in which the plants are located (Adm. filed May 5 , 1972 , p. 4, 5). Borden advertises nationally (CX 212M; CX 214E). Its Woodstock, Ilinois, plant obtains much of its raw milk from Wisconsin (Minkler, Tr. 104). After processing, the milk (8 J and dairy products are shipped outside the State and outside the Federal Milk Marketing Order areas. The same Woodstock, Ilinois plant produced substantially all of the private label products sold to A&P under the private label agreement involved here, a great many of which products were then distributed to A&P stores in Indiana (CX 62; CX 263; Adm. filed May 5, 1972, p. 5). In addition to the Woodstock Ilinois, plant, Borden s Chicago Central District operated milk plants in Indiana, Iowa and Wisconsin (:Iinkler, Tr. 88). 15. A&P, pursuant to the private label agreement, purchased milk and dairy products from Borden for its grocery stores both within and without the State of Ilinois (CX 62; CX 182; CX 187; CX 2m). Borden to fulfill the private label agreement, obtained part of its milk from out-of-State and processed certain milk and dairy products in its plant at Woodstock, Ilinois (Adm. filed :Iay 5, 1972, PI'. 4- 5). Borden then shipped said products across State lines to be delivered to A&P' s outof-State stores.

16. Woodstock, which opened in 1964 and was in full production by 1965 or early 1966, received its raw milk from the Chicago Milk Shed which extends through Wisconsin, Ilinois and Indiana (Minkler, Tr. 104- 105, 110; Adm. filed September 29, 1972, p. 8, Par. II-5; CX 211B , G; Minkler, Tr. 99- 100).

17. Borden s practice in connection with its Woodstock operations was to purchase raw milk as needed for the expected demands of its customers, process it and deliver it to its customers as quickly as possible. Milk was not stored at Borden s Woodstock plant more than approximately 24 hours during this process (Borden Adm. dated March 1973 111-28; Minkler, Tr. 108-109). At the processing plant milk was inspected, cooled, standardized and in some cases fortified, pasteurized and homogenized (:Iinkler, Tr. 107- 109; Tise, Tr. 5834-5837). Processing of the major fluid milk items - milks and creams which accounted for GREAT ATLANTIC & PACIFIC TEA CO., INC., ET AL. 973 962 Initial Decision over 70 percent of Borden s sales (RAPX 75) - is a nearly instantaneous mechanical procedure which does not alter the product's chemical composition (Tise, Tr. 5834-5837; Graham, Tr. 7078-7079; Minkler, Tr. 107-108). Pasteurization is required before milk can be shipped in interstate commerce (Tise Tr. 5829-5830; Graham, Tr. 7072). Borden sells milk to a wide range of customers including retail stores. Such customers are also (9 J located in States other than the State of Ilinois in which Woodstock is located (Minkler, Tr. 114; Malone DTR. 23-24). A substantial amount of the milk and dairy products processed at Woodstock was distributed through the Hammond, Indiana Distribution Depot (:Iinkler, Tr. 102-103; :Ialone DTR. 26-28). The Hammond Depot primarily served customers in the State of Indiana (Minkler, Tr. 103). Among Borden s customers in this area were a number of A&P stores.

18. Alj of the acts and practices described above are and were performed in a constant current of commerce with raw milk flowing across State lines into Woodstock and processed milk and dairy products flowing across State lines to Borden s various customers including its largest chain store customer in the area, the retailer A&P. D. THE 1'TGOTIATIONS 19. In late 1964 A&P decided to explore the possibility of selling private label milk and dairy products in the stores of its various divisions. :Ir. Herschel Smith (Smith), A&P' s Headquarters Dairy Buyer, held a meeting in New York City at which he explained A&P' new New York private label milk program and instructed the company s divisional purchasing managers to discuss with 1J nit Buyers the possibility of instituting similar private label programs in their various areas (Smith, Tr. 1350-1352; Bartels, Tr. 1862-1864). A&P' Middle Western Division Purchasing Director, Bartels, attended this meeting. Shortly thereafter and pursuant to these instructions he instructed A&P's Chicago Head Buyer, Schmidt, to initiate negotiations with Borden for private label milk (Bartels, Tr. 1862-1865). 20. Schmidt, A&P' s Head Buyer for its Chicago L'nit, was in charge and conducted the ensuing negotiations. Bartels, A&P's :Iiddle Western Division Purchasing Director, supervised and was kept informed of Schmidt s over-all activities (Schmidt, Tr. 1668-1669 , 1683; Bartels, Tr. 1865- 1866: CX 12A).

21. Minkler, the President of Borden s Central District, was in charge of the ensuing negotiations for Borden. Minkler made all the decisions as to what to offer A&P and how to proceed. He communicated with Schmidt or Bartels on important occasions I njtial Decision 87 F. (Minkler, Tr. 123- 125, 201-203 , 225- , 247-248; CX 7; Tarr, Tr. 885- 887).

(10 J 22. Minkler was assisted during the negotiations by Malone Borden s Central District Vice President for Government Controls and Borden s expert cost accountant. Malone was very knowledgeable concerning all aspects of Borden s production and distribution expenses in the Chicago area and counseled Minkler concerning Borden s costs of serving A&P during the entire course of the negotiations (CX 262 Malone DTR. 6-8, 10- 29- 39; Minkler, Tr. 111-113 309-310; CX 42 54). Malone prepared all of Borden s costs analyses and most of Borden s private label proposals to A&P (CX 14, 18, 21 , 23- , 36, 42, 54 , 87). He was extremely well respected and was relied upon by both Borden and A&P officials (Minkler, Tr. 309-310). A&P' s Bartels testified: "The Borden company' , 'had' , ' very sophisticated cost accounting and legal departments" (Bartels, Tr. 1894A). 23. Tarr was Borden s Ccntral District Chain Store Sales Manager and handled most of the line contact with A&P during the negotiations. Tarr spent about 80 percent of his time servicing the A&P account (Tarr, Tr. 845-846, 848-850, 876-878). 24. Around November 1964, A&P asked Borden to submit a private label offer for Chicago. :Iinkler and Malonc immediately prepared a Borden proposal which granted A&P a one cent discount per half gallon of private label milk provided A&P would accept drop delivery and limited service, that is, store door delivery with no in-store service no special deliveries, and no returns except for inferior or defective products (CX 5, 6; Minkler, Tr. 133-134; CX 262; Malone, Tr. 32-33). The offer was transmitted to Borden s New York office in late November 1964 and turned over to A&P in New York on December 2 1964. A&P refused this offer (CX 5, 6, 7F, 12; :\!inkler, Tr. 133-137, 140- 142; Archer, Tr. 1209- 1212).

25. Although Borden s profit margin in the Chicago area might be cut by a reduction in its price on private label products it could not refuse to make such a reduction because it was dependent on continuing to serve A&P (CX 7D-F).

26. Borden was also concerned with A&P' s possible retailing plans for private label milk. A&P was in the (11 J initial stages of implementing a nationwide plan to install private label milk in its stores (Smith, Tr. 1350-1353). In some areas, including Columbus, Ohio Dayton, Ohio, and Dallas, Texas, A&P had introduced private label milk at a lower out-of-store price than advertised label milk (CX 8G- Miller, Tr. 5975-5976, 5978). Smith of A&P had discussed at least some of the cities with Archer in connection with A&P's private label plans for Chicago (CX 7F). Borden felt that if A&P followed this policy and GREAT ATLANTIC & PACIFIC TEA CO., INC., ET AL. 975 962 Initial Decision introduced private label milk at such a differential in Chicago a price war would follow and its profits would be destroyed (CX 7D- Minkler, Tr. 154-157).

27. Minkler prepared a new private label offer to A&P which granted a 2-cent per-one-gallon price reduction which Borden tendered to A&P in late January but which was also refused (CX 7G- , 9 12). 28. On February 5, 1965, Smith of A&P requested A&P's Chicago Office to take over the negotiations. Bartels turned the matter over to Schmidt and he in turn caJJed Tarr of Borden to request an offer (CX 11; Schmidt, Tr. 1688; Tarr, Tr. 849-851).

29. Schmidt told Borden that "he very much appreciated the sensitiveness of this market." He added that Borden "had to be right" if A&P was not to put its private label business out for bids (CX 12). 30. Borden thereupon prepared a fuJJ and complete offer to seJJ private label milk and dairy products in the Chicago-Calumet area of Ilinois and Indiana (CX 14). The Chicago-Calumet area encompassed the A&P stores in the Chicago area, its immediate Ilinois suburbs and the Calumet area of Indiana, including Gary and Hammond, Indiana. This area was served by a number of Borden distribution branches including Borden s O'Hare Branch at Irving Park, Ilinois, and Borden s Hammond Branch at Hammond, Indiana (CX 262, Malone DTR. 72; Gose, Tr. 1072-1073; RAPX 68). Tarr delivered the offer to Schmidt on February 9, 1965 and discussed it with him (Tr. 850-852). Borden s new offer granted A&P a discount of 2.55 cents per-halfgallon off its then current price and commensurate reductions of the other five main items taken by A&P which products accounted for percent of A&P' s total sales of milk in the Chicago-Calumet area. Borden also quoted a private label price on five by-products. This offer indicated that Borden would realize only .45 cents per-quart gross profit on sales of the main homogenized Vitamin D milk products (CX 14D).

(12) 31. In May, Schmidt asked that Borden come up with a private label program for areas outside of A&P's Chicago-Calumet area. Such outside areas extended north to the Wisconsin State line west into Iowa, south almost as far as Peoria, Ilinois and east beyond South Bend, Indiana to Goshen, Indiana (Tarr, Tr. 856-857; CX 18C, E, G, L). Borden prepared such an offer setting different prices for each zone of A&P' outside areas for each product offered under private label for Chicago (CX 18, 17B , C).

32. A&P was not interested in having only some of its stores in Ilinois or Indiana supplied with private label milk. Later, when Bowman was invited to bid, a list of all the A&P stores in Indiana Iowa and Ilinois was given it (Cannon, Tr. 6126-6127). The Dean Dairy Initial Decision 87 F. was rejected when it bid only on Indiana stores (Schmidt, Tr. 1711). It is thus clear that the private label arrangement of A&P with its supplier was unavoidably interstate in character and that sales of milk to A&P stores in llinois would be me.de only if A&P stores in Indiana were also served by the supplier.

33. Borden s May offer (CX 18-0) shows that the dairy expected to lose money on its sales of gallon containers of milk and that its gross profit on half gallons to be only .11 cents. Its total gross profit from sales of the higher volume containers (gallons, half-gallons and quarts) would be only $76.89 per week. The gross profit referred to takes into account only the cost of raw mj1k, containers, production, route delivery expense and other distribution expense. 34. A&P's business represented more than 55 percent of Borden wholesale route business in the Chicago area and 25 percent of its Woodstock plant production. Borden could not, therefore, afford to lose the A&P business by refusing to lower its prices to A&P (CX 7F, 42 A- D). At the same time Borden feared that A&P might cut its out-ofstore price for private label milk below the market price for advertised label milk (CX 7- , E; Minkler, Tr. 155- 157). Actually, A&P was selling private label milk at a reduced out-of-store price in two out of the three areas studied (CX 8 G-J) and had introduced private label milk at a reduced price in Dayton, Ohio in February 1965 (Miller, Tr. 5975- 5976). Consequently, Borden believed that if A&P created such differential the market price for advertised label milk would break and a price war would result (CX 7E). Such a price war had been experienced by :Iinkler and Malone in Wisconsin a year earlier and in (13) Chicago during the 1950's (Minkler, Tr. 155-156; Malone DTR 97- 99; Soberg, Tr. 2753 , 2754-2764). Such a price war would not only reduce profits of all dairies in the area but Borden particularly, because of its investment in the new Woodstock plant which cost over five milion dollars (CX 13J). Many of the other dairies in this area were older and fully depreciated and thus had substantially lower fixed costs (Schaub, Tr. 3Ilo-3111). Thus, if Borden reduced its price and A&P followed suit at retail level for private label milk, a price war might ensue and reduce Borden s profit. If Borden refused A&P' request for lower prices Borden would lose A&P's business and the volume necessary to operate Woodstock efficiently. Neither position was desired.

35. On May 25, 1965, Tarr and Malone delivered Borden s offer for the outside areas to A&P and discussed Borden s position with Schmidt (CX 19A). Borden pointed out that it expected almost no profit from the A&P contract but that it was presenting such a bid because A&P' business was vital to Borden s efficient use of its Woodstock plant. GREAT ATLANTIC & PACIFIC TEA CO., INC., ET AL. 977 962 Initial Dccision Schmidt pointed out that A&P's goal in its private label program was to increase its gross margin on milk. The Borden representatives however, warned that this could not be accomplished if A&P reduced the price on private label milk, citing a price war that had occurred in Wisconsin as the result of private label pricing (CX 19E, F; CX 262 :Ialone DTR 97-98).

36. In July 1965 A&P asked Borden to revise its May quotations for the outside areas to take account of possible savings in delivery costs. On July 26, Tarr delivercd the revised quotation to Schmidt (Tarr, Tr. 863-865; CX 21).

37. On July 26, 1965, Bartels of A&P met with Minkler to discuss Borden s costs and to push for lower prices. Bartels wanted to know why Borden s Chicago costs were higher than its New York costs and Minkler agreed to review the matter (Tarr, Tr. 863-869; CX 21, CX 22A- , ex 25B).

38. Malone prepared a written response to Bartels' questions (CX 23-25A). This response indicated that the primary reason for Borden higher costs in Chicago was that Borden s Chicago delivery costs were higher in that the Chicago dairy drivers were paid substantially more than their New York counterparts; that union limitations on types of delivery and stricter health regulations limited efficiencies in delivery in Chicago; that A&P stores in New York made larger average purchases; and that there was a greater density (14) of A&P stores in New York. Malone concluded that Borden s gross margin per quart would be .36 cents not .45 cents as previously calculated and pointed out that this margin figure did not inelude anything for either overhead or burden (CX 2. , 24D). This cost study was submitted by Minkler and Malone to Bartels on August 6, 1965, and discussed with him (Minkler, Tr. 202-203; CX 262, Malone DTR 114). 39. Again, in August, A&P requested Borden to lower its offering price on some of the by-products. On August 13 , Tarr presented Schmidt with a new revised bid for both the Chicago-Calumet area and A&P' s outside areas (CX 31, 36, 37). The new bid focused on A&P' profit as a result of the private label contract indicating that A&P' costs would be reduced some $410 000 (CX 37A- , CX 36; Tarr, Tr. 869- 871).

40. Kevertheless, Schmidt told Tarr that A&P was not satisfied and had decided to put the matter out for bids (Tarr, Tr. 871-873). Tarr had told Schmidt at various times that some dairies in the area might bid under their cost to get A&P's business (Tarr, Tr. 872). On August , 1965, :Iinkler told Bartels that other dairies might submit bids which included only out of pocket or direct costs leaving out all fixed charges, on the theory that these costs were already covered by their . .

Initial Decision 87 F. present business (:\inkler, Tr. 213-214). Nevertheless, Bartels indicated that he was proceeding to get other bids.

41. Malone calculated that if Borden lost A&P business Borden gross profit would be reduced more than 1.6 million dollars per year compared to the private label price reduction of some 400 thousand dollars (CX 42A- , CX 262, :Ialone DTR 122-123). 42. During the latter half of August A&P asked Bowman Dairy, Dean Milk Co. and Sidney Wanzer & Sons to submit bids for A&P' business. Wanzer was not interested and the Dean Milk Co. submitted a bid for a part of Northern Indiana only, an area which A&P did not want served separately from Chicago (CX 45, 51A-C; Schmidt, Tr. 1711).

43. Bowman Dairy submitted an offer to A&P (CX 50). Schmidt immediately telephoned Tarr and told him that Borden s bid was " far out of line that it is not even funny. You are not even in the ball park* * * this bid is so far different from yours that there isn t any comparison " (Tarr, Tr. 873-874). Schmidt added that even if Borden promotions and other assistance were valued at $50 000 per year "this would not be a drop in the bucket" (Tarr, Tr. 874). (15 J 44. The next day, September 1 , 1965, Schmidt met with Tarr and told him that the new development was "terrible" and that Borden s bid was " nowhere near " to which Tarr replied that some bids might "be screwy" and that he didn t see how anyone can sell any cheaper than Borden (Tarr, Tr. 873-874).

45. Malone had computed what a bid based upon direct costs only would be. Borden also knew that losing the A&P account would cost it about 1.6 million dollars per year. As Minkler explained (at Tr. 22- 226):

This became Hut a matter of logic in many respects. It was a matter of struggling to try to guess again what might have been put in there by a competiwr. must tell you that it was just largely (pure) guess. We had the $410 00 (bids in, we were clearly wid that an additional $50 00 wouldn t begin to cover the difference, there are (the J facts that we had to consider.

In addition to that, we had this figure of one million six hundred some thousand dollars that it would cost us if we lost the business. We had to weigh that, and at some point, I said that it seemed to me that we had to consider doubling this $410 00 figure and you could ask me all night how I arrived at that, and I wouldn t be able to tell you except that we were just determined that we had to save this volume for this brand-new plant that we were just bringing on stream, We were really desperate and by that time, we were pretty we!1 convinced that the approach by the competitor had been on an anyhow accounting basis, they were going to get it and correct their problem later on, we rationalized, so I again got New York on the phone and got their clearance if I thought this was alj that could be done to go ahead. I was told, save the business GREAT ATLANTIC & PACIFIC TEA CO., INC. , ET AL. 979 962 Initial Decision 46. On September 8, 1965, Minkler and Tarr met with Schmidt to discuss a further Borden bid (CX 55E). Schmidt (16) reiterated the conversations almost verbatim that he had with (Tarr), that (Borden offer J wasn t even in the hall park, and how shocked and surprised he was at our company trying to go along with such a low bid when someone else was bidding and' , 'the whole thing" (Tarr, Tr. 876-887; Minkler, Tr. 226). Minkler offered to submit a new quotation which would yield A&P a saving of $820 000 per year, double the saving of Borden s then current bid. :Iinkler stated that this offer was made " the basis of meeting competition" (Minkler, Tr. 227-22). Minkler also stated his belief that the bid submitted by Borden s competitors probably took only direct costs into account and were based upon anyhow accounting" but that Borden could not lose A&P' s business (Minkler, Tr. 227-228, 237-238 , 376-377). Schmidt said "now you are in the ball park" and asked Minkler and Tarr to go back, spread the new savings figure to Borden s product line and write up a form of service (Minkler, Tr. 24-24; Tarr, Tr. 878-879).

47. During the following week Borden worked up a formal quotation including the 11 items previously offered under private label as well as glass gallons of homogcnized Vitamin D milk and 2 percent milk. Tarr told Schmidt of this change in the offer and Schmidt said such change would be finc (Tarr, Tr. 880-882; Minkler, Tr. 24-245). The new bid was submitted to Schmidt's assistant by Minkler and Tarr on September 17, 1965 (CX 56, 57, 58; Tarr, Tr. 883; Minkler, Tr. 245). 48. Shortly thereafter, Schmidt called Tarr to reject the offer telling Tarr that Borden s bid was not fair to the other bidders who did not bid on glass gallons " (Tarr, Tr. 883-884; Minkler, Tr. 245). However, the Bowman Dairy offer ha specified glass gallons of milk (CX 50). Actually, A&P did not want to sell private label milk in glass containers (Schmidt, Tr. 1777-1780). Schmidt also told Tarr that Borden should "sharpen your pencil a litte bit because you are not quite there as a result of which Borden reduced its price somewhat (Tarr, Tr. 884-885 , 963-965).

49. The Bowman bid was apparently based upon volume and A&P did not ascertain how a different volume would affect that Bowman bid except to determine that if Bowman sold less volume to fewer stores its prices would go up (Schmidt, Tr. 1760-1761). The Bowman bid also assumed certain possible days of delivery in thc Gary-Hammond area. A&P did not ascertain how a change in such assumption would affect this offer.

(17 J 50. Borden prepared a corrected bid spreading the 820 thousand dollars to the original 11 item private label product line (CX 62). Using the raw milk prices as of the first half of 1965, the savings to A&P , , , , , I nitial Decision 87 F. would have been over 880 thousand dollars but since the raw milk prices had risen in the meantime the actual savings would be about 820 thousand dollars (CX 66B). The new bid reduced Borden s price for a half gallon of milk by 5.75 cents below its then current advertised label price to A&P and by about 3.2 cents below Borden s July 26 prices, the last price Borden had made any effort to cost justify (CX 62, CX 21). 51. Minkler and Tarr presented Borden s final bid on September 21 1965. Minkler told Schmidt:

. . . I said Elmer, there is certainly something here that I want you to very definitely understand. " I said This price is given to you by us on the feeling and belief that we are meeting a competiti've bi. We krw afrw other way to justify thu,. YO' have to accept it on that oo'!is. You must make that clear to your superiors and to your legal! people. 1 don t know what may come of this in the future, but I want you to understand this (: J we are going to say always that we felt we were meeting a competitive offer that you had received from someone else. " And he said to me, words to the effect Ralph, you don t need to worry. I read you loud and clear. I understand what you are saying. My superiors and our legal department wil understand it." And he says Just don t woIT' about it." (Minkler, Tr. 247 24; emphasis supplied). 52. Mr. Tarr, a witness to the meeting, testified to the same effect: At the last of it Mr. Minkler said, and I can t give you his exact words, but in effect he said Now, Elmer, look, we have met a competitive situation here on this bid, on this quot€. You know that. This is right, isn t it " and Elmer said "Yes." (Tarr, Tr. 886; also see Tr. 1020).

(18) 53. Minkler also told Schmidt that the new bid was based upon anyhow accounting" (CX 63A-D). Schmidt admitted that Borden told him on this occasion and in many other conversations that the final bid was based upon "meeting competition" (Schmidt, Tr. 1782- 1801). 54. Schmidt rcquested a letter from Borden stating that its prices were proportionally available to others (CX 66). Instead, Borden supplied a letter which stated only that Borden felt its prices were legal and that it was prepared to defend them (RAPX 2). This letter was not in Borden s usual form relating to offers which were available to others (Archer, Tr. 1254-1255 , 1258). Herschel Smith, A&P' Headquarters Dairy Buyer, understood that Borden s lcttcr did not say its prices were available to others (Smith, Tr. 1427- 1429 , 1438-1440). 55. Schmidt submitted Borden s new bid and Bowman s bid to Bartels noting that Borden s hid offcrcd a savings of 820 thousand dollars while Bowman s savings amounted to some 737 thousand dollars. Schmidt recommended retention of Borden as did Bartels who forwarded it to Smith in "'ew York (CX 66A- , CX 65A-C; Schmidt Tr. 1713-1716; CX 263). Smith noted that Borden s offer was substantially better" than the one from Bowman and approved Borden on October 14, 1965 (Smith, Tr. 1369, 1413-1414; CX 263 , 70). GREAT ATLANTIC & PACIFIC TEA CO., INC., ET AL. 981 962 Initial Decision Borden commenced serving A&P in its Chicago area in November 1965 (CX 75A- , 255B).

56. During mid-November 1965 Borden brought its service and delivery terms to A&P on Borden label products in line with those granted on A&P label products but at a substantially lower discount (RAPX 66). Thus, during 1966 and thereafter Borden s maximum discount to A&P taking the same service on Borden label products was 30 percent while A&P' s private label discount was 35 to 38 percent. 57. On May 24, 1966, Borden increased its Borden label list price hy 5 cents per point' to reject its increased container, labor and social security costs. At Borden maximum discount of 30 percent this increase raised Borden s prices by .35 cents per point (CX 124, 125, 139 140 , 123, 138; Gose, Tr. 1110). Borden informed A&P of the price increase (19 J and asked that A&P accept a commensurate increase on A&P label products (CX 79; Gose, Tr. 1050-1054; CX 262, Malone DTR 178-180). A&P accepted the increase on Borden label products but refused to accept it on private Jabel products (CX 81; Gose, Tr. 1053- 1054; CX 255B). The contract on private label provided that prices were to be subject to change to reflect increased container and labor costs (CX 62H- , M).

58. Again, in March 1967, Borden asked A&P to accept the price increase to reflect cost passed on to Borden s other customers in 1966 pointing out that all of its other customers had accepted the price increase (Gose, Tr. 1056- 1062: ex 102-103, 105, 106). A&P refused to accept more than a .2-cent per-point price increase on private label products (CX 102B , 106C, 110, 225B; Gose, Tr. 1058). 59. After the institution of the private Jabel agreement with A&P Borden reorganized its Chicago Central District merging it with another district. Minkler explained that this was necessary because of the reduction in income occasioned by the lowering of its price to A&P on private label products by some 800 tbousand dollars (Minkler, Tr. 276-280). Job positions involving almost all faeets of the Borden operation were eliminated within the Chicago region following the beginning of the Borden-A&P private label arrangement due, at least in part, to that new arrangement (Gose, Tr. 1069-1070). E. GRADE AND QLALITY 60. Respondent A&P admits:

(8 Jubstantially all of the fluid milk product. Borden supplied under A&P' s private label to A&P stores Lin the area of complaint counsel's proofs during the period ovember 1965-1970 were physically and chemically identical to the equivalent products , A point is equivalent to a quart of whole milk. Thus a gallon iB meaured a. four point. and a half gallon as two points. Some other dairy prod\lcts have different point vaiues. Initial Decision 87 F. contemporaneously sajd under the Borden label these A&P stores and to other stores in the "Chicago and suburbs " Gary, Hammond and Valparaiso areas (A&P's Adm. Filed March 12, 1973, Par. II-6).

Borden admits substantially the same thing (Borden Adm. filed September 29, 1972, Par. II- 10; :Iinkler, Tr. 2'"0; Schmidt, Tr. 1727- 1729).

F. CONTEMPORANEOL:S SALE 61. Respondent A&P admits that the fluid milk and dairy products sold by Borden under private label to A&P stores (20) in the area in question and contemporaneously under Borden label to competing retailer-purchasers, were commodities sold for use, consumption or resale within the United States (Comp. and Ans. Pars. 8, 9; A&P Adm. Par. II -6).

G. THE DlSCRlMINATIO TIw Gary-Harnrrnd and Valparaiso Areas of Indiana 62. Burger s Super Markets, Inc. , operated a supermarket in Hammond, Indiana, which competed with the A&P stores in Hammond, Indiana (A&P's Adm. Par. II-9(a)-(e); White, Tr. 2064-2065; Kristoff, Tr. 2285-2286). Burger s store purchased about two to five times as much milk from Borden as its A&P store competitors (CX 182; RAPX 234 p. 19) and was substantially larger than its A&P store competitors in Hammond (Kristoff Tr. 2270-2271 , 2381 , 2385). 63. Model Food Center, Inc., operated a supermarket in Hammond Indiana, also. A&P admits that the Model Food Center in Hammond Indiana, competed with its A&P store # 306 in the Hammond suburb of Munster, Indiana (A&P Adm. Par. II-9(a)-(e)). Model Food Center purchases of fluid milk products from Borden were commensurate with the A&P stores in the Gary-Hammond area with which it competed (CX 182; Barney, Tr. 3682-3688).

64. Wallies Market in Hammond, Indiana, purchased fluid milk and dairy products from Borden from 1966 until December 1, 1969. A&P admits that Wallies Super :Iarket competed with one of its stores (A&P' s Adm. Par. II-9(a)-(e)). Wallies bought approximately the same amount of milk from Borden as its A&P competitors during the period 1966-1969 (CX 182).

65. Wilco Food Center is a supermarket in Gary, Indiana. A&P admits that during the period 1964-1970 Wilco Food Center competed with one of its stores in Gary (A&P' s Adm. Par. II-9(a)-(e)). Wi1co bought substantially more than twice as much milk from Borden as the average A&P store in the Gary-Hammond, Indiana area (CX 182). GREAT ATLANTIC & PACIFIC TEA CO., INC. , ET AL. 983 962 Initial Decision 66. Joseph Tittle & Sons, Inc. (Title) operates and has operated a chain of supermarkets in Northwestern Indiana for many years. Between October 1966 and March 1967 Tittle owned and operated five supermarkets in this area, two in Gary, Indiana (Lake Street and Village Court), one in Hammond (Hohman), one in a suburb of Hammond (Highland), and one in Valparaiso, Indiana. Between 1967 and 1969 (21) Title sold its Hohman store and moved its Vi1age Court store to another location. Consequently, between 1969 and 1970 Title had two stores in Gary, one in Highland and one in Valparaiso. A&P admits that the Title stores competed with its A&P stores in such areas (A&P Adm. Par. II-9(a)-(e); Minard, Tr. 3136-3138, 3141-3148). The Title stores in Gary bought approximately the same volume of milk from Borden as the A&P stores in the area. In Hammond the volume of the Title stores was lower than the volume of some A&P stores but higher than others. In Valparaiso the Title store s volume was at least three times more than that of the A&P store (CX 182). 67. Burger s :Iarket bought slightly more than 28 thousand dollars worth of milk products from Borden between October 1966 and March 1967. Had it been charged A&P's prices under A&P' s private label agreement with Borden, Burger s :Iarket would have paid $3 468. less. It, therefore, sustained a price discrimination of 14 percent. Between June 1969 and March 1970 it bought over 84 thousand dollars of milk products from Borden but would have paid $11 506. 15 less had it been charged A&P's prices under the private label agreement sustaining a 15.7 percent price discrimination (CX 187, 188A). 68. During the same two periods :Iodel Food Center was charged $12 496 in the earlier period and $29 519 in the later period. Had Model been charged A&P prices under its private label agreement with Borden, :Iodel would have saved $2 297 in the early period and $4, 138 in the later period, thus experiencing a price discrimination of 22. percent in the early period and 16.3 percent in the later period (CX 191A , 187).

69. Wallies Market bought more than $14 000 worth in the early period and a similar amount in the later period. Had it been charged A&P prices under the private label arrangement it would have saved 193 in the earlier period and $2 203 in the later period reflecting a price discrimination of 17.5 percent for the early period and 17. percent for the later period (CX 200A, 187). 70. The Wilco Food Center paid more than $35 000 for milk products in the same earlier period and more than $71 000 in the later period. Had it been charged the A&P prices under the private label agreement with Borden it would have (22) saved $4,550 in the earlier period and $4,433 in the later period. The price discrimination r nitial Decision 87 F. amounted to 14. 5 percent in the earlier period and 6.6 percent in the later period (CX 202A , 1S7).

71. Title s Lake Street store paid more than $13 000 for milk products in the earlier period and more than $32 000 in the later period. Had it been charged the A&P prices under the private label agreement with Borden it would have saved $1 571 in the earlier period and $1 939 in the later period. The price discrimination was 12.9 percent for the early period and 6.4 percent for the later period (CX 196A, 1S7). 72. Title s Village Court store paid Borden more than $12 000 for fluid milk products in the earlier period and more than $3S 000 during the later period. Had it been charged the A&P prices under the private label agreement with Borden, it would have saved more than $1400 in the earlier period and more than $lS00 in the later period, sustaining a discrimination of 12.5 percent in the earlier period and 5 percent in the later period (CX 197A , 1S7).

73. Title s Supermarket on Hohman Avenue in Hammond, Indiana, paid more than $15 000 to Borden for fluid milk products during October 1966-March 1967. Had it been charged A&P' s private label prices from Borden, it would have paid more than $lS00 less, reflecting a 13 percent discrimination during that period (CX 195A, 1S7). 74. Title s Supermarket in Highland paid more than $12 000 for fluid milk products bought from Borden during the early period or almost $1400 more than it would have paid had it been charged A&P' private label prices from Borden. This amounted to a 12.4 percent discrimination. During the later period of proof this store paid more than $23 000 for fluid milk products but would have paid $1 335 less had it been charged A&P's private label prices from Borden, thus experiencing a 5.9 percent discrimination (CX 194A, 1S7). 75. Title s Supermarket in Valparaiso bought approximately three and onc- half times as much milk from Borden as the competing A&P store (CX 1S2). Nevertheless, during the first period it paid more than $51 000 for fluid milk products bought from Borden which was $7 603 or 17.5 percent more (23) than it would have paid had it been charged A&P' s private label prices from Borden. During the second period of proof, it paid more than $111 000 for such products or $6 984 or 6. percent more than it would have paid had Borden charged it A&P' private label prices (CX 19SA, 1S7).

76. Combined, the Title stores paid $13 S04 during the earlier period in excess of what it would have paid had it been charged A&P' private label prices from Borden. This was an average of 15 percent more. During the later period the Title stores paid $12,104 more or 6. percent more than it would have paid had it been charged A&P' private label prices (CX 187).

GREAT ATLANTIC & PACIFIC TEA CO., INC., ET AL. 985 962 Initial Dccision 77. The unfavored competitors described above purchased from Borden at lower discounts and at higher prices than enjoyed by A&P as a result of Borden s formula for pricing A&P private label products. This situation commenced with the institution of the private label program in 1965 and continued until it ended in February 1972 (RAPX 66; CX 187, 120-136 288 289; Bayma, Tr. 5932-5933; Kristoff, Tr. 2279; Barnes, Tr. 3852, 3914; Minard, Tr. 3155-3156). II. THE CHICAGO A:\D SUBURBS AREA 78. A&P sold fluid milk and dairy products in all of its stores of the Chicago area as did all of Borden s other store customers in that area (CX 256, 183). Two of such other Borden s store customers were Mayfair Foods and Jim s Groceries, who competed with A&P stores in that area during February and March 1966.

79. A&P admits that its store No. 137 located approximately two blocks from :vayfair Foods drew customers from the area served by this independent food store (A&P's Adm. filed March 12, 1973, Par. 111- 11). This was corroborated by the store manager (Lasorso, Tr. 264 2646 2686) who testified that he regularly compared his prices with the prices charged by A&P, and adjusted his prices in response to those of A&P. Customers of Mayfair and A&P shopped both stores. 80. Jim s Groceries was a member of Central Grocers Cooperative and was also known as Centrella store. It was located in Oak Park llinois, where it competed with A&P for the sale of fluid milk and dairy products during 1965 (24J and 1966 (A&P Adm. Par. 1I-11). This is confirmed by the owner of Jim s Groceries (Cox, Tr. 3946-3948) who testified that he checked A&P's prices in setting his own prices and that A&P was a competitor.

81. Borden s price to A&P on A&P label products in the Chicago and suburbs area in February and March 1966 was equivalent to a discount from list price of about 36.6 percent (CX 183). In May 1966 Borden imposed a price increase on Borden label products but A&P refused to accept it on the private label products (Gose, Tr. 1052- 1054) and thus A&P's effective discounts from Borden s list prices rose to 37.8 percent as of August 1966 (CX 180). This continued until April 1967 when A&P accepted a partial price increase. 82. Borden s discount schedule in January 1966 established a maximum discount on Borden label products of 30 percent off of list (CX 138) which discount was available only to customers taking 600 points or more per delivery and the same service as granted to A&P. Although CX 138 states Borden s maximum discount to be 28 percent for such customers, Borden also offered an additional 2 percent discount if the store was ccntrally biled (Gose, Tr. 1110). 21fi- 9B9 1' - 77 - 63 986 FEDERAL TRADE COMMISSION DECISIO:-S Initial Decision 87 F. 83. During February 1966 Mayfair Foods purchased milk and dairy products from Borden in the amount of $2 736 and received a rebate of $818 or 29.9 percent of the list price. During March 1966 Mayfair Foods purchased $3 056 of such products and received a rebate of $915 or 30 percent of the list price. During the same two months the A&P store with which :Iayfair competed paid a net price which reflected a discount of about 36.6 percent of its private label purchases or 6. percent additional discount above and beyond that granted to Mayfair Foods (CX 183).

84. Jim s Groceries' purchases of milk and dairy products in February and March 1966 totaled $2 269 and $2 661 , respectively. Borden s rebates totaled $522 and $607 or 23 percent and 22.8 percent respectively for those months (CX 183), which was in accordance with Borden s published discount schedule since Jim s Groceries took only about 400 points per delivery and did not take limited service or central biling (RAPX 173, pp. 82, 84). The A&P stores in Oak Park paid a net price which reflected an effective (25 J discount of about 36.6 percent during this period. These A&P stores were about the same size or smaller than Jim s Groceries, taking about 251 points and 419 points for each of the two stores (RAPX 173, pp. 109, 116, 118). Even if the A&P stores qualified for an extra 2 percent discount for limited service and another 2 percent for central billing, these stores should have been receiving discounts of only 23 and 28 percent, respectively, not the maximum 30 percent discounts set out on Borden s discount schedule or 36.6 percent they actually received. The discrimination thus experienced by Jim s Groceries exceeded 10.4 percent. 85. The five highest volume items under the private label contract were homogenized milk quarts, homogenized milk half gallons homogenized milk gallons, 2 percent milk half gallons and skim milk half gallons. These items accounted for 80 percent of Borden expected sales to A&P pursuant to the private label contract (CX 65). As wi1 be seen from the attached Appendix, Borden s discriminations in favor of A&P on these items ranged from a low of 8.3 percent for quarts of milk to 18 percent or more on half gallons of skim milk. Borden s discrimination on half gallons and gallons of milk which accounted for more than 55 percent of Borden s sales to A&P range from 9. 7 percent to 14 percent.

H. JURY TO COYtPETITIO;- The attempt to discern the competitive effect of price discrimination has been said to be "roughly akin to explaining how to avoid harm in a snake pit" (Keck, Lawful Price Discrimination, 8 Antitrust Bull. 381 (1963)). Here, however, the task is simplified. , GREAT ATLANTIC & PACIFIC TEA CO. , INC., ET AL. 987 962 Initial Dccision 86. As A&P admits fluid milk is one of the most important commodities carried in retail grocery stores' . . (IJt is a perishable high-volume, fast turnover item sold by most food retailers. Milk is also a staple and is purchased (by consumers J more frequently than other products (sold in grocery stores 1" (A&P Adm. filed March 1, 1973, Par. III- , 24). This is confirmed by the testimony of non-favored Borden customers (Minard, Tr. 3149; Kristoff, Tr. 2274; Gintert, Tr. 2456; Barnes, Tr. 3861; Barney, Tr. 3686; Lasorso, Tr. 2650; Cox, Tr. 2951- 2952). Consumers are familiar with the retail price of fluid milk products and consequently competing retail grocers price milk competitively (A&P Adm. filed :Iarch 1, 1973, Par. II-21; Minard, Tr. 3149; Kristoff, Tr. 2277; Gintert, Tr. 247-248; Barnes, Tr. 3861-3862; Barney, Tr. 3690; Lasorso,. Tr. 2650; Cox, Tr. 3947-3951). :Iilk products are sometimes used as (26J price leaders which are priced below the normal market price to draw customers to a store where it is hoped the customer will purchase additional products (Minard, Tr. 3153-3154; Kristoff, Tr. 2277-2278; Gintert, Tr. 2458; Barnes, Tr. 3862; Barney, Tr. 3691-3692; Lasorso, Tr. 2652-2653). Selling milk at higher prices than those charged by one s competitors would cause customers to shop other stores and would hurt business (see foregoing citations). 87. A&P admits that profit margins have been notoriously low the retail grocery business (A&P Adm. filed March 1 , 1973, Par. III-23). This is confirmed by sales and profits summaries of the unfavored competitors of A&P (CX 184, 189, 192, 199A , J- , 201, 203). All of the un favored competitors had pretax net profits of less than 5 percent of sales. Four of them had net profits of 2.3 percent or less and one had a net profit of less than 1 percent (see Appendix IV(in COrYra ) submitted in complaint counsel's proposed findings). 88. In view of the consumers' consciousness of fluid milk prices, the importance of milk to the grocery store, and the low level of their profit on sales by grocery stores, the purchase price of fluid milk items has been important to the ability of food retailers to price these items at competitive levels. The unfavored competitors of A&P, had they received a larger discount from Borden on milk products, would have increased their net profits and permitted them to be more competitive. Some of them would have been able to lower the retail price of milk to the consumer, increase advertising or otherwise give better service (Minard, Tr. 3154; Kristoff, Tr. 2273, 2278; Gintert, Tr. 2455-2460; Barnes, Tr. 3860-3861; Barney, Tr. 3688; Lasorso, Tr. 2653-2657; Cox Tr. 3950-3951). The price discriminations experienced by the unfavored competitors on their purchases of milk and milk products from Borden when compared with A&P's prices on private label was at least several times larger than those stores' level of net pretax profit. In many cases 988 FEDERAL TRADE COMMISSIOK DECISIONS Initial Decision 87 F.

the percentage of discrimination was as high or higher than the stores level of gross profit on sales. Consequently A&P enjoyed a competitive advantage. Since it did not reduce the retail price of its private label milk, its lower costs were not passed on to the consumer but resulted in higher gross profits for itself (CX 255B; White, Tr. 2059). (27) I. A&P DUCEME OR RECEIPT OF BORDEN S PRIVATE LABEL PRICII"G 89. There can be litte doubt that A&P induced Borden into granting the private label prices. Its behavior as outlined in the section of this decision dealing with the negotiations clearly demonstrates that A&P maintained unwavering pressure upon Borden to arrive at the prices finally agreed upon. It solicited a private lahel arrangement first only for its Chicago-Calumet area, then later for its outside areas it hammered down Borden s successive bids and even when it was informed that the bid was at or close to Borden s costs put the matter up for outside bids. It then informed Borden that its bid was not even in the ball park." Finally, Borden, desperate to retain A&P business for its new Woodstock plant, made a bid which A&P could not refuse and the deal was closed to begin in November 1965 and to last until February 1972.

J. A&P S K.1\'OWLEDGE TH.A.T BORDE7\ S BID WAS NOT COST JUSTIFIED 90. As dctailed earlier in the section of this decision entitled Negotiations, A&P was made aware that Borden s price could not be cost justified. Borden repeatedly told A&P that its price was based solely on meeting competition and that it was intended to meet a competitive bid based on incremental costing. Moreover, Borden first offered A&P an $820 000 saving covering 13 private label items with an annual volume of $6.5 million. When A&P insisted that Borden drop its glass gallon feature, Borden revised its bid to cover 11 private label items with an annual volume of only $5.6 million (CX 56B- , 62, 75F- Tarr, Tr. 884-885; Schmidt, Tr. 1777). A&P should have known that the same cost savings applicable to different products with different volumes would not likely be related to Borden s cost savings. 91. The study of Borden s Chicago-Calumet area costs in July 1965 was made by Borden cost accountant, Malone, and was submitted then to A&P. This showed Borden s cost on gallons, half-gallons and quarts of milk to be higher than the final price per unit to A&P (CX 24D, 62C). Even when these costs were computed in February 1965, the GREAT ATLAl'TIC & PACIFIC TEA CO. , INC., ET AL. 989 962 Initial Decision final prices offered A&P for the two highest volume items, halfgallons and gallons of milk, were lower than Borden s cost (CX 14D). (28 J 92. Borden also told A&P that its cost of serving A&P stores outside of A&P' s Chicago-Calumet area were higher than its prices. Borden reported its costs for the major whole milk items as follows: Half-gallon 3167 Gallon 6260 Quart 1584 (CX 21U) Borden s final offer was substantially below these costs, particularly for half-gallons and gallons (CX 62D-G). A&P should have known that Borden s costs were higher than the final price quotation. In May 1965 Borden s !Vr. Malone spoke with A&P's Mr. Schmidt about Borden costs and showed Schmidt that Borden s gross profit per quart was less than one mill (CX 19E). At that time Borden s cost for paper halfgallons of milk was $.3162. Borden s final price, averaged, was $.2866 for half-gallons in the outside areas.

93. Mr. Bartels also knew of Borden s costs having been presented with Malone s calculations (Bartels, Tr. 1869; Minkler, Tr. 203; Malone DTR p. 114).

94. It is also of si!,TJificance that when A&P asked Borden for a letter stating that its private label prices were granted to others on a proportional basis, a statement which would have been in conformity with A&P's policy of obtaining such a written statement from suppliers (Smith, Tr. 1428), Borden replied: .Our prices are proper under applicable law and \' 8 are prepared to defend these prices (CX 263H).

95. A&P's Chicago officials made no effort to determine if Borden s private label prices were to be made available to others (Schmidt, Tr. 1729; Bartels, Tr. 1877).

96. It does not appear that A&P's New York office made any such inquiry (Archer, Tr. 1251). A&P's headquarters buyer for milk and dairy products who reviewed thc Borden offer for legality (Smith, Tr. 1347) testified that he understood that Borden s letter did not state that Borden s prices were available to others on proportionately equal terms and was not the letter of availability that A&P (29J had asked for (Smith, Tr. 1406). He further testified, however, that he telephoned Mr. Archer, and held a conference with him about the Borden letter but that Archcr assured him that the letter was a letter of availability and even more (Smith, Tr. 1407). Whereupon, he sent the Borden offer 990 FEDERAL TRADE COMMISSION DECISIO Initial Decision 87 F. their letter and a memorandum to A&P's legal department for review and that the legal department responded in writing (Smith, Tr. 1413 1442). Archer, however, testified that he had no discussions with Smith regarding Borden s final Chicago quotation (Archer, Tr. 1246). Neither Smith' s memo to the legal department nor the alleged legal departments' approval were offered in evidence. On balance, I do not credit Smith' s version of the actions taken after the receipt of Borden s final offer.

97. A&P's trade experience also should have given it reason to believe that Borden s final price could not be cost justified. It was quite familiar with the dairy industry and methods of doing business in the industry at the time of the private label negotiations. Their officials had discussed private label milk for Chicago, the pricing of private label milk, Borden s pricing in the market, dairy pricing in the Chicago area as compared to pricing in other cities, private label pricing in other markets, competitive conditions, Borden s costs, etc. (Minkler, Tr. 201, 213, 227; Tarr, Tr. 857, 871, 897; Schmidt, Tr. 168 , 1687 , 1692; Bartels, Tr. 1865; Smith, Tr. 1350; CX 7D- , 12, 19, 22, 25B- , 30, 31 43). Admittedly, one of its purposes in seeking private label milk for the Chicago area was to cut its cost of milk below the price it paid for Borden label products (Corbus, Tr. 7325: Smith, Tr. 1400). A&P also purchased fluid milk and dairy products in the Chicago area from other dairies in addition to Borden and must have been aware of the price level for these dairies (CX 256). It also knew that Borden s final private label price to A&P was substantially below Borden s advertised label prices to A&P since it was buying both from Borden. 98. Even A&P's costing guide lines indicated that Borden s private label prices had to be below Borden s cost. In 1964, A&P negotiated for the supply of private label milk by Glen & :Iohawk Milk Associates for areas in upstate New York and Massachusetts. The president of that dairy told (30 J A&P how a store customer could estimate a dairy cost. He said that it could sell milk to A&P stores at a cost of approximately 6 cents per quart above the dairy s cost of raw milk. Two of such 6 cents went for cartons or fillers. Two more cents were attributable to plant costs and profit and the last two cents would cover delivery (Abrams, Tr. 6398-6399). Nevertheless, Glen & Mohawk' s prices ranged from .28 cents to . 65 cents per quart more than the prices Glen & Mohawk told A&P it might expect (RAPX 137A, 62, 63). Moreover, the 2 cents figure for plant costs would not nccessarily apply in New York City because wage rates were higher there than those used in the formula (Abrams, Tr. 6402-6404, 6423). In addition, Glen & Mohawk' offer did not include milk delivery but was a dock pick-up price and did not guarantee delivery costs the customer taking the risk of delivery GREAT ATLANTIC & PACIFIC TEA CO., IXC., ET AL. 991 962 Initial Decision (Abrams, Tr. 6388). Glen & :Iohawk' s delivery cost estimate would not necessarily apply in areas with different union wage rates and was based on an average per store delivery of at least 40 to 50 cases (Abrams, Tr. 6399-6401, 645-646). Finally, delivery costs were for sidewalk drop delivery and not for putting milk into the store (Abrams Tr. 6396). In sum, the " 2 formula" was merely a reference point from which to work and had to be adjusted for different conditions in different areas. A&P was told that in-plant wage rates in Chicago were substantially higher than in New York (CX 25C, 26). The average A&P store was taking less than 600 points or 25 cases per delivery (RAPX 233, p. 8). Union work rules and Chicago Health Regulations forbade sidewalk drop delivery and Borden, not A&P, took the risks of delivery (CX 23D- , 62). Nevertheless, the final Borden offer granting prices to A&P for half-gallons and gallons of milk were .6 to .8 cents per point below Glen & Mohawk's prices (RAPX 137). 99. After the commencement of the private label contract, it was obvious to A&P that its private label milk costs were below its Borden label milk costs for which it got an effective discount of only 30 percent and not the 35 percent to 38 percent effective discount available on the A&P label products (CX 255B; RAPX 66). This was so, although Borden s services to A&P in connection with its Borden label sales were the same as those rendered in connection with Borden s private label sales. Products of both labels were delivered in the same trucks and in the same manner. Products of both labels were pre-ordered. Borden did not accept returns of either product after expiration of the code date and Borden did not provide any in-store or promotional services (CX 75A- , RAPX 66, Supplement C, pp. 238-239). Thus as mid-November 1965, and thereafter, A&P should have known that its average (31) store in the Chicago Unit took less than 600 points per delivery (CX 24B , 27, 20H; RAPX 233, p. 8) but it was purchasing Borden label milk at a price of $.3430 per half-gallon and A&P label milk at a price of $.3124 per half-gallon. ln January 1966, Borden offered other customers the same limited service option provided to A&P, provided such other customers took 600 points of milk per delivery. Such other customers, however, had a maximum discount of 30 percent off of Borden s list price. Actually, the discount was 26 percent off Borden s list price plus 1 percent additional for advance ordering and limited service plus 1 percent additional for limitation on rcturns plus an additional 2 percent for those customers which were centrally biled, for a maximum of 30 percent (CX 123 , 138; Gose Tr. 1110). A&P' s private label cost was the equivalent of a discount ranging between 35 percent to 38 percent without regard to the volume taken by any of its stores. Borden services to all were the 992 FEDERAL TRADE COM:\ISSIOK DECISIOKS Initial Dccision 87 F. same (CX 75A- , 123, 138). Although Borden s private label contract did not offer salesman s services to handle promotions, its limited service to non-A&P stores also stated that Borden would not provide promotions (CX 123B; Lasorso, Tr. 2699; Barnes, Tr. 3910; Cox, Tr. 3992; Havemeyer, Tr. 9137-9138).

100. It thus is apparent that the only benefit lost by A&P in connection with the private label contract was advertising. But dairy advertising was not a big expense - 2 to 3 mills per point or less (CX 54: Soldwedel, Tr. 3387, 3430). Indeed, Bowman told A&P that it could not justify any price difference between private label and advertised label products (Cannon, Tr. 6136).

101. A&P also knew or should have known that its price preference on private label products was increased in :Iay 1966. Then Borden raised its price by $.0035 (at maximum discount) to cover increased costs of labor, containers and social security costs (Gose, Tr. 1047, 1110; ex 79 , 123 , 124 , 125 , 138 , 139, 140). A&P refused to accept the price increase except on its Borden label purchases and Borden repeated its request in March and April of 1967. At that time A&P accepted only a 002 increase (CX 81 , 102, 103 , 105, 106, 110, 255B; Gose, Tr. 1052-1058; Schmidt, Tr. 1723).

K. THE BOWMA AND BORDEN COMPETITIVE BIDS 102. As noted earlier, A&P had only two bids under consideration in 1965, the Borden bid and the Bowman bid. It compared the two bids on the basis of their prices on the private label products offered by Borden which was the only factor in which it was interested. (32) 103. The Bowman bid was based upon a 3.5 percent butterfat content of milk. The Borden offer was based upon a 3.4 percent butterfat content. There is nothing in this record to indicate by what amount Borden would have increased its bid to supply milk with a butterfat content of 3.5 percc," ' ,lead of 3.4 percent. Similarly, there is nothing in this record to indicate how much Bowman would have reduced its price to supply 3.4 percent butterfat milk instead of 3. percent butterfat milk. The two bids are not comparable with respect to butterfat content.

104. The Bowman bid, unlike the final Borden bid, included glass gallons of milk, but any comparison between the two bids based upon glass gallons of milk is improper inasmuch as A&P did not want glass gallons under private label and when Borden attempted to include such product, A&P told Borden to take it out. Glass gallons were therefore not included in Borden s final offer (CX 62, 66; Schmidt, Tr. 1777; Tarr Tr. 883-884).

105. A&P argues that any comfarison of the Bowman and Borden GREAT ATLANTIC & PACIFIC TEA CO. , IKC., ET AL. 993 962 Initial Dccision bids should include by-products which were not offered under private label by either Bowman or Borden. This is also improper inasmuch as A&P was not interested in such items (CX 62; RAPX 50). 106. The Bowman bid was actually inoperative. It was based upon a volume that A&P would not and could not provide. It was based upon an estimated total dollar volume of approximately $1 million per month using Chicago list prices. Bowman s letter to A&P stated: A substantial increase or decrease in the size of your order (compared to the assumption of dollar volume set forth above) would affect these prices . (CX 50A; Cannon, Tr. 6149). A&P' s actual purchases using list prices totaled about $754 000 per month (CX 13A , 17B , C, 13H). This calculation is arrived at by taking the unit volume for one week, annualizing such figure and multiplying the result by Bowman s list prices for such units. Taking into consideration the fact that these calculations were for the early part of 1965 and that A&P's purchase volume declined later in 1965 and thereafter, it is reasonable to infer that Bowman s bid was subject to upward revision by reason of such substantial decline in volume (CX 75Z; Scbmidt, Tr. 1684 , 1731, 1737). This is confirmed by Bowman table of price (33 J adjustments showing that its prices would be about 7 percent higher if A&P bought from Bowman s for only 70 percent of its stores and 3.1 percent bigher if it bought from Bowman for only 50 percent of its stores (CX 50R; Schmidt, Tr. 1761). The Borden offer was not based upon volume (CX 62).

107. The Bowman bid was also inoperative with respect to deliveries. The Bowman bid was based upon three-day-per-week deliveries in the Gary-Hammond area of Indiana (CX 501,). Actually, it delivered milk six days per week in tbe Gary-Hammond area and could not reduce the days of service because of its union contract (KaJchbrenner, Tr. 9295; Banaski, Tr. 9534: CX 23D, 27, 31A; RAPX 246A). Although the Route Foreman for Borden testified that the union contracts permitted a 4-day delivery, he stated that the practice was &-day delivery (Szczepaniak, Tr. 1287-1288). The 1965 contract provided that no change shall be made in days of route operation (CX 299i) and the 1967 contract states: "The present delivery system of &day operation shall continue in effect" (CX 300k, emphasis supplied). It is clear that &-day delivery was the union practice and the requirement here.

108. In connection with the Bowman bid, A&P admits that it did not ask for and did not receive a letter or statement from Bowman stating that its prices were legal or proportionately available to others (Schmidt, Tr. 1848- 1849).

initial Decision 87 F. L. COST STIFICATION 109. Joseph Malone, Borden s cost accountant, made several studies to determine Bordcn s direct costs of serving A&P with private label products. Reference has heretofore been made to CX 24D made in July 1965, in which Borden s direct cost for half-gallons, gallons and quarts of milk is shown to be higher than the final price per unit in CX 62C. CX 87 was prepared in June 1966 after the commencement of the private label program. It shows that while Borden was making a profit on Borden label sales, it was losing from $2 200 to $4,870 per week on the A&P private label business. CX 206 analyzes Borden s Chicago area sales for the 12 months of 1965 and the first eight months of 1966 and shows that Borden ovemll profits in the Chicago metropolitan area fell precipitously in November 1965 and remained at almost a break even point in 1966 , but it did 134 J not segregate private label sales which represented more than 25 percent of Bordcn s Chicago area sales from Borden label sales (CX 42D).

110. Considering that the unfavored customers competing with A&P purchased as much as, if not more than, the A&P store purchased from Borden, that the cost of processing A&P's private label products was no less than its cost of processing the equivalent Borden label product (RAPX 232 Attachment II) and that, except for advertising, no differences in service or delivery terms existed between the labcls but that advertising expense was insufficient to relate meaningfully to the difference in prices Borden charged A&P compared to competing retailers, no price discrimination can be justified under the Robinson- Patman Act. There is no differcncc "in the cost of manufacture, sale or dclivery resulting from differing mcihods or quantities in which such commodities arc sold or delivered. " Indeed, Malone s calculations demonstrate the necessary resultant difference in net income because of the unjustified price differentials.

A&P' s F ACTUAL LEGAL CO?'TEl'TlOr-S A. COC T I 111. A&P contends that it made no misrepresentations to Borden during its private label negotiations. 1 t notes the testimony of Borden witnesses to the effect that A&P's Schmidt asked Borden to eliminate glass gallons from its bid because other bidders did not bid on glass gallons (Tarr, Tr. 883-884; Minkler, Tr. 245-246). Although Bowman the "other bidder had offered a price on glass gallons, A&P contends that there was no misrepresentation to Borden. A&P contends that Schmidt was referring to a comparison he was preparing between the GREAT ATLANTIC & PACIFIC TEA CO., INC. , ET AL. 995 962 Initial Decision Borden and Bowman bids which was based on the earlier Borden bid and which did not include glass gallons. The testimony of Mr. Schmidt however, is not persuasive. He admits that "my recollection is hazy . . . 1 believe I recall that I was at that time reflecting the opinion of the Chicago unit, members on my level of management, that we did not want to carry glass private label milk because of the problems at the store level" (Schmidt, Tr. 1777-1778). On balance, I find the testimony of Tarr and Minkler more persuasive. A&P further contends that the reason for not desiring glass was irrelevant (35 J because it would not have resulted in the slightest change in Borden thinking or calculations. But that is not the point. A&P rejected Borden s bid to meet competition on a ground that was baseless, that is, for the reason that Borden s bid included glass gallons, an item which A&P falsely represented to be excluded from the alleged competitor s (Bowman bid.

112. A&P also contends that the bid was not induced or altered by a sharpen your pencil" comment, arguing that it is doubtful that this comment was made at all and, if it was made, it almost certainly was made at a stage in the ncgotiations when it did not affect Borden final quotation. This ignores the credited testimony of Mr. Tarr who said:

This thing that I say about sharpning the pencil was just toward the very end of the thing. It was after all these things had been talked about and before we brought over the final papers. (Emphasis supplied, Tr. 985). 113. A&P further contends that there was no misrepresentation by it even if it convcyed the impression to Borden that Borden s bid merely met Bowman s bid rather than actually bcating it. It argues that, even if the Borden bid was lower by some $82 000, the difference was less than 1.5 percent of the total business involved and that therefore comments by A&P like "in the ball park" and "substantially met the other quotes " would appcar to be quite accurate. The difference in savings between the Bowman and Borden bids amounted to approximately 3145 000. With the adjustment for the increase in raw milk costs, Borden s bid offered A&P a savings of some $820 000 while the Bowman bid' s savings amounted to some $737 000. The difference in these two savings amounted to some $82 000 or 11 percent of the Bowman bid savings which is not a de minimis amount particularly in the grocery or dairy business. Even a comparison of the differential in savings to the total business involved, the 1.5 percent referred to by A&P, would not necessarily be de minimis in the grocery store industry where net profits of some retailers did not exceed such percentage figures. Here 1.5 percent represented about 996 FEDERAL TRADE CO:.MISSION DECISIO"'S Initial Decision 87 F.

one-third or more of the unfavored competitors' pre- tax net profits (See Finding 87 s1J,pra).

(36) 114. Perhaps the most meaningful misrepresentation by A&P was that it had a competitive offer with which it could compare the Borden offer. The Bowman bid, unlike the Borden bid, was conditioned upon a certain minimum dollars' worth of sales to A&P. Moreover, the Bowman bid was based upon its assumption that it could deliver three days a week in the Gary-Hammond area of Indiana, but this was incorrect. Bowman delivered milk to stores in this area six days per week and could not reduce its days of service because of the union practice and requirement that milk be delivered six days per week to stores (See Finding 107 surra). It follows, therefore, that A&P actually did not have a competitive bid from Bowman to compare to the Borden bid. Although A&P attempts to minimize these differences and to explain them away, the fact remains that Borden was asked to meet an unrelated bid which was therefore not competitive. A&P' calculations to arrive at a price Bowman would have bid if conditions were identical are conjectural and not factual. They cannot be accepted.

115. There can be no doubt that when A&P accepted Borden s offer it understood that Borden had granted a substantially lower price than that offered by Bowman, the only other competitive bidder if, indeed there was an offer which could be compared with Borden s (See Finding 55 upra). When it told Borden that its offer was being accepted after Borden had told A&P that it was submitted only to meet competition, it misrepresented the situation to Borden. 116. A&P argues that complaint counsel seek to enact "a rule requiring all buyers to tell all sellers who indicate that they are bidding against competitors what the competitors' price is " citing Beatrie Foods Co. v. Kroger Co., Inc. 76 F. C. 719 aff'd 438 F.2d 1372 cen. denied 404 U. S. 871 (1971), as well as Forster Manufacturing Co. FTC 335 F.2d 47 (1964). This is just. not so. As the court held in Beatrice:

The controlling point here is not the " hard bargaining" nor the price levels but the mi.srepresentatum of the Broughton bid, in order to induce a discriminatory price (438 F.2d at 1378; emphasis in oribrinaJ). (37) Here, A&P misrepresented the situation to Borden. It knew or should have known that there was no operative bid from Bowman. Nevertheless, it told Borden that it had such a bid and when Borden made it quite clear that it was submitting its bid to meet the competitive offer A&P made no effort to disabuse Borden of that belief, although it understood that Borden had beat that bid. There GREAT ATLANTIC & PACIFIC TEA CO., INC., ET AL. 9!7 962 Initial Decision was no obligation upon A&P to divulge the name of the competitor or the actual price offered by a competitor. It was sufficient if it had informed Borden that the meeting competition defense was not available. In failing to do so it misrepresented the situation to Borden and it was this misrepresentation that was a violation of Section 5 of the Federal Trade Commission Act. As the Commission concluded the Health Hazards Statement (29 F.R. 8354-8355 (1964)): In the last analysis, the Commission s responsibility in this area is enforce a sense of basic fairness in business conduct.

A&P failed to meet that standard. As the Health Hazards Statement further points out:

No enumeration of examples can define the aukr limits of the Commission s authority to proscribe unfair acts or practices, but the example should help to indicate the breadth and flexibility of the concept of unfair acts or practices and to suggest the facwrs that determine whether a particular act or practice should be forbidden on this ground. These factors are a., follows: (1) whether the practice, without necessarily having been previously considered unlawful, offends public policy as it has been established by statutes, the common law, or otherwise -- whether, in other words, it is \\within at least the penumbra of some common Jaw, statutory, or other established concept of unfairness; (2) whether it is immoral, unethical, oppressive, or unscrupulous; (3) whether it causes substantial injury to consumers (or competitors or other businessmen). If all three factors are present, thc challenged conduct will surely violat. Section 5 even if there is no specific precedent for proscribing it. The wide variety of decisions interpreting the elusive concept of unfairness at least makes clear that a method of selling violates Section 5 if it is (38) exploitive or inequitable and if, in addition to being morally objectionable, it is seriously detrimental to consumers or other. A&P' s misrepresentation was within an established concept of unfairness (see Beatrie, supra): it was unethical and it did cause substantial injury to Borden as well as to A&P' competitors (See Findings 59, 62- supra).

117. A&P contends that its conduct caused no injury to Borden or to Borden s employees arguing that Borden s merger of its Chicago Central district with the Midwest District in Columbus, Ohio, was a normal part of Borden s operations in order to improve efficiency levels. This disregards the testimony of :Ir. Minkler (Tr. 279-280) which attributed the reorganization to the loss of Borden s income occasioned by the private label contract. That the reorganization had other justification referred to by A&P to improve efficiency does not necessarily invalidate the fact that I have found supra (Finding 59), that at least in part the reorganization was prompted by the reduction in income caused by the private label contract. Nor does Borden s 1966 annual report making no reference to the loss of income from the private label contract necessarily invalidate such a finding inasmuch as 998 FEDERAL TRADE CO:\MISSION:: DECISIONS Initial Decision 87 F. a company might not regard an annual report to be the most auspicious place to speak of income losses due to a private label contract. B. COU1\T II Count II of the complaint charges A&P with knowingly inducing or receiving price discriminations in violation of subsection (f) of Section 2 of the Clayton Act, as amended.

The Discrimination 118. A&P attacks the calculations made by complaint counsel for the prices paid by non-A&P store customers to Borden for products purchased by A&P under its private label program, arguing that complaint counsel has failed to properly authenticate the underlying Borden rebate information relied upon. It states that the rebate checks relied upon by complaint counsel were incomplete, citing three instances. In Burger s Supermarkets, A&P notes that a second check from Borden was never included in complaint counsel's calculations. The amount of this second check was $207.63. Even were it to be included, (39) it would scarcely diminish the discrimination suffered by Burger s Supermarket which experienced thousands of dollars in higher costs incurred. A similar inconsequential discrepancy is mentioned in connection with Model Food Center where the rebate check does not tally exactly with the method of calculation. This rebate check amounted to $79.36 and should have amounted perhaps to $8. more. A&P also objects to complaint counsel's failure to include certain payments by Borden to Wi1co and Title. These appear to be checks for intercst payments with no relation to the volume of purchases by these companies and would have insignificant impact upon the price discrimination calculations.

119. In addition, A&P contends that the documents relied upon by complaint counsel were incomplete. But the subpoena for this information to Borden required all documentation necessary to show the net price to customers and it appears that such was provided. Moreover, the rebate checks used by the Commission s accountant tallied almost exactly with Borden worksheets. Under these circumstances it appears that the data used by complaint counsel were authenticated sufficiently.

120. A&P also attacks complaint counsel's calculations for non- A&P customers in Ilinois, arguing that discounts greater than those shown on Borden s discount schedule were offered on a customer-bycustomer basis. Although complaint counsel's accountant was not able to reconcile the price book amounts with the price lists in some of his GREAT ATLA:-TIC & PACIFIC TEA CO., INC., ET AL. 962 Initial Decision spotchecks, it was apparently due to such things as the i1egibility of entries and the inclusion of butter which was not on the price list and did not necessarily indicate that these customers were not charged these prices (Bitting, Tr. 4882). Moreover, such additional discounts were apparently given on small volume by-product items and would not substantially alter the overall pricing pattern (CX 238 D Bitting, Tr. 4225-4231, 4234-4236; Soldwedel, Tr. 3421). II. TIw Representativeness of tlw Stoes 121. A&P further contends that while complaint counsel may have shown that A&P's discriminatory price caused injury to those Borden customers discussed by complaint counsel, this showing is not sufficient because these stores were not "representative " of all purchasers from Borden. It is doubtful that complaint counsel has such a burden.

(40) In Beatrie, supra the Commission said: The Robinson Patman Act requires consideration of secondary line injury on a location-by-loction ba."is. Each local competitive area, each Kroger store loction becomes the competitive environment within which the potential effect on competition with Kroger must be examined. (76 F. C. at 805). The Commission noted that in Forerrst Dairis, Inc. v. Federal Trad Commission 348 F.2d 674 (5th Cir. 1965), cert. denied 382 l.. S. 959 (1965), the only competitive environment examined was the city of Albuquerque, New Mexico, in which were located some eight stores a small chain which received favored prices. Only one of these stores was located in proximity with an un favored Foremost customer. Here local competitive areas were examined. In the Gary-Hammond area of Indiana, A&P's five largest independent grocery competitors paid substantially more than the A&P stores in that area during the six month period from October 1966 through March 1967 and the later ten month period from June 1969 through March 1970. Similarly, the Title store in Valparaiso paid substantially more than its competitor A&P store. On the O' Hare route Borden s cash store customers purchasing $800 or more of milk per month paid substantially more than A&P for fluid milk products sold under private label (See Findings 62 through 77 supra; also CX 183). These price discriminations have been found to be meaningful and productive of injury to such non-favored competitors of A&P. In a secondary hne competition case, the Supreme Court has said that it is self evident'" '" . there is "a reasonable possibility, " that competition may be adversely affected by a practice under which manufacturers and produc.rs sell their goods to some customers substantially cheaper than they selJ like goods to the woo FEDERAL TRADF CO:.MISSION DECISIO IniiiaJ Decision 87 F.

competitors of these customers (Federal Trad Commissi v. Mor Salt Co., 33 S. 37 (1948).

As the Commission stated:

It seemed self evident that where a producer is selling a homogeneous product such as salt, automotive part, or gasoline, where competition is extremely keen among retailers, and where margins of profit or markups arc small, a lower (41) price to one or some of the competing retailers not only "may" but must have the effect of substantially lessening competition (Sun Oil Co. 55 F. C. 955, 962 (1959)). Although the Seventh Circuit seemed to inject a necessity to make realistic appraisals of competitive impact" in its decision in Anwn:Can Oil Co. v. Federal Trade .Commission 325 F. 2d 101 (7th Cir. 1963), cen. denied 377 U.S. 954 (1964) even that court assumed competitive injury upon a showing of different prices charged competing customers where the differential was " substantial and sustained" (Bargain Car Wash, Inc. v. Standard Oil Co. , (Ind.) , 466 F.2d 1163 , 1174 (1972)). Assuming nevertheless, that the representativeness of the non-favored stores selected by complaint counsel here must be shown, I find that burden to have been met. The discriminations and injuries shown here involved local competitive areas of Gary-Hammond, Indiana, where A&P' s five largest independent grocery competitors' costs were compared with the A&P stores' costs in the same area. In the Valparaiso, Indiana, area thc Tittc store was compared with its competitor A&P store. In the Chicago and suburbs area Borden cash store customers were compared with A&P stores. In the Indiana area the purchases of the unfavored competitors represented more than two-thirds of aji of Borden s non-A&P store wholesale sales in the area (RAPX 2:J4, p. 19; CX 182). In the Chicago and suburbs area, the unfavorcd competitors choscn represented more than 40 percent of Borden s total sales on its O'Hare wholesale routes, which in turn represented a major part of Borden s total sales in the Chicago area (Compare RAPX 233, p. 69 Co1. (1), with p. 8, 1 , Cols. 7, 13; RAPX 76; Bitting, Tr. 4788-4789, 4815-4816; CX 211D; Gose, Tr. 1072). 122. A&P contends that the cost comparisons are defective because they do not compare a11 of the allegedly injurcd competitors' dairy purchases from Borden against a11 of A&P' s purchases but only against A&P' private label purchases. In the Beatrice case, respondent Kroger raised the same argument. There, however, Kroger, in addition to its private label purchases, was an unfavored purchaser of Beatrice labeled products and arb'ued that its purchases of both lines should be considered in assessing injury. (The situation here is even stronger in that there is no claim that A&P was an unfavored customer in its purchases of Borden label products. ) The Commission held: GREAT ATLA:-TIC & PACIFIC TEA CO., INC., ET AL. )00) 962 Initial Decision We think the proper inquiry is whether Kroger by obtaining cheap milk under its own label was given a competitive advantage which it used to potential injury of its competitors (at 806).

(42) The Commission went on to hold that only the Kroger purchases of private label products were the correct measure of competitive injury and discrimination since Kroger s purpose in purchasing private label products was to obtain an overall advantage over its competitors since it used the private label milk "for the benefit of its entire grocery business and not merely to sell against name brand milk in the stores of its competitors." In that case Kroger, as did A&P, purchased private label milk at a cheaper price but sold it at the higher price of brand label milk, thereby reaping higher profits. It did not pass any of its cost savings on to the consumer. Moreover, Kroger continued throughout the period of discrimination to purchase the various items under both private label and brand label. A&P , however, carried the products in question under both labels for only a short introductory period. From October 1966 on, it purchased litte or no Borden label milk products, so that it enjoyed a preferential price on all its milk purchases. Thus except for a lew months, whether or not A&P's Borden label purchases were aggregated with its A&P label purchases would make litte difference in the calculation of A&P dollar advantage over its competitors.

123. A&P also contends that complaint counsel' s proof of injury was insufficient, arguing that Borden s discrimination in favor of A&P lasted only briefly. This ignores the evidence of record that the discrimination lasted for years. A&P also cites evidence to the effect that the unfavored competitors could purchase milk at prices below those charged by Borden. There is no evidence, however, that such competitors were offered prices which even came close to A&P' substantial discount at the times in question (Gintert, Tr. 2492-2496; Minard, Tr. 3256-3258; Kristoff, Tr. 2329). II. Knowledge 124. A&P also argues the issue of knowledge. I have already found that Borden s Minkler told A&P that the Borden bid was to meet a competitive bid only and that A&P' s Schmidt said that he understood (Findings Nos. 51 , 53 supra). Moreover, I have found that A&P' officials understood that Borden s bid was "substantially better" than supra). Nor did A&P have anythe Bowman offer (Finding 1'0. 54 reason to distrust the representations 01 Borden since it admittedly had a high regard for Borden s cost accounting and legal (43) departments (Finding 22 supra). A&P' s Schmidt admitted that he considered Tarr, Minkler and Malone of Borden "honest and forth- 216- 969 O- LT - 77 - 61 1002 FEDERAL TRADE CO:\MISSION DECISIONS Initial Decision 87 F.

right" (Schmidt, Tr. 1843). A&P's argument also ignores the fact that it knew Borden s cost savings were made without spreading it to individual products and that when it did make such a spread, it changed the products covered and the prices without changing the savings which could hardly be the result of cost differences attributable to differing methods of sale or different quantities sold. Moreover A&P realized that Borden removed the statement "these prices are available to all on a proportionally equal basis" from its private label price change tables. Since it was A&P' s policy to require such statements A&P should have known or surely been put on notice that Borden s new private label prices were not available to others. 125. A&P also attacks the reliability of Borden s cost studies arguing that they do not reflect savings Borden could expect from the introduction of the private label program. It attacks the Malone studies which show Borden s direct costs to be higher than the final price agreed upon to A&P, arguing that certain inconsistencies and irregularities make these studies unreliable. This position, however, is difficult to reconcile with the testimony of A&P's witnesses who considered Mr. Malone to be expert, forthright and honest. If so, A&P then must have known at that time that the price it had wrested from Borden was below Malone s calculations of cost. 126. A&P relies on Smith' s testimony that he was assured by Borden s Archer that Borden s letter was a letter of availability and even more, which testimony I have not credited, preferring to believe Borden Archer who denied such conversation. A&P contends that Archer must have seen Borden s letter and lied when he said he didn see it, because, as A&P contends, Archer had all the other materials sent to New York on the negotiations. The evidence shows, however that Minkler forwarded two sets of Borden s final proposal to Mr. Pentz in New York (CX 64A). Minklcr forwarded one copy of the legal letter to Pentz but this transmittal contains a penciled note which does not indicate that Archer was sent a copy of that document (RAPZ 3). Moreover, since Archer had been promoted Executive Vice President of Borden earlier and would not have been as deeply involved in the affairs of the milk division as previously, it is not likely that he would have reviewed the legal letter from Borden. A&P also notes that in 1967 with reference to another bid (44) by Borden it asked Borden w hat was meant by language similar to that on Borden s legal letter here. On that occasion Borden replied that it meant the price would be available to others. No explanation is offered, however, why similar inquiry was not made by A&P at this time. The fact that Bordcn gave it such a meaning at another time, later, cannot relate back to an earlier period of time without supporting evidence. GREAT ATLANTIC & PACIFIC TEA CO. , INC., ET AL. 1003 962 Initial Decision 127. A&P disputes that its trade experience would have led it to believe that Borden s final prices could not have been cost justified. 1 have found, however, that A&P was quite knowledgeable about the dairy industry (See Findings Kos. 97 supra). A&P argues also that Borden expected reduced costs as a result of its new Woodstock plant. But we are here concerned with differences in the costs, not expectations which mayor may not be realized. Nor is it relevant that A&P could get better prices on private label in other parts of the country. We are here not concerned with other parts of the country. Nor is A&P' s reliance upon the costing guidelines of the formula appropriate where, as I have found, this formula was a mere reference point and was subject to adjustments for local differences which would make it inapphcable to the Chicago area.

IV. Meeting Competition 128. As respects the issue of meeting competition A&P argues that although its officials thought tbe Borden bid was "substantially better" than the Bowman bid in 1965 its review in 1973 shows that in fact the Bowman bid was better. I have found, however, that the Bowman bid was not comparable to the Borden bid. Unlike the Borden bid it was based upon a certain anticipated volume of sales with a higher cost if the volume dcchned, which it did. It was also based upon a certain delivery schedule which was impossible of performance due to union practice and requirements. Although A&P attempts to develop the change in the Bowman bid with a corrected delivery schedule, the correction is A&P's estimate and not Bowman s. A&P also adjusts both the Bowman and Borden bids to reflect the same butterfat content of the milk, inasmuch as the Bowman bid allegedly specified a 3.5 percent butterfat content compared to the Borden milk which was 3.4 percent butterfat. Again, this is a correction by A&P and not by either Borden or Bowman. But it doesn appear that A&P was interested particularly in the butterfat content as long as it met the legal minimum requirements (Schmidt, Tr. 1727-1729). (45) A&P uses Borden calculations to raise Borden s price for 3.5 percent butterfat milk using Borden s adjustmcnt of its raw milk cost. This does not necessarily mean that Borden s price would be similarly adjusted by Borden had A&P asked for a different butterfat level. In addition, since the Borden butterfat level was about 3.45 percent anyway (Walker, Tr. 5513) and Borden was desperate to keep the A&P business, Borden would not necessarily make any adjustment in its prices to raise the butterfat content to 3.5 percent (CX 62X; Havemeyer, Tr. 6779; RAPF 130, p. 134).

129. Above and beyond the foregoing obstacles to an adjustment of 100 FEDERAL TRADE COM:.ISSIO DECISIONS Initial Dccision 87 F. the Borden bid to conform to what was felt to be the Bowman bid, that 5 percent butterfat content, there appears to be serious doubt that the Bowman bid actually called for 3.5 percent butterfat content. Such was not the impression of the negotiating individuals in 1965 but was only the result of second thoughts generated by a review of the bids at the request of A&P counsel in 1973. The testimony of the witnesses however, indicates that the Bowman bid was based upon a 3.5 percent butterfat content of the Class I raw milk and not the butterfat content of the processed milk (Smith, Tr. 1415-1417 , 1445-1450; Schmidt, Tr. 1815; Bartels, Tr. 1897). Indeed, a Bowman official, Mr. Cannon testified that the Bowman offer did not contain any specifications compelling Bowman to supply at any particular butterfat level (Cannon, Tr. 6187-6188). This was confirmed by the Commission expert dairy economist (Walker, Tr. 5450-5453). In fact, this expert found that Bowman s butterfat level in 1965 was actually lower than Borden s butterfat level judging from the handler s reports which he examined (Walker, Tr. 5461).

V. Cost Justificatio71 a. Attacks upon Malone s calculations 130. With respect to cost justification, A&P first attacks the computations by Malone as erroneous and unreliable. Specifically, CX 54 purporting to show Borden s cost to permit Borden to make an additional savings proposal to A&P is attacked because Malone allocated the costs on a point basis rather on the individual packages involved. But higher priced and lesser volume byproducts are more costly to handle so that a cost computation by the number of units could not clearly reflect actual cost whereas a cost computation by points in which these higher priced and lower volume byproducts are given higher point values would be more nearly accurate (See CX 7A). Moreover, since r 46 J Malone was seeking to ascertain the direct costs of A&P's business it was sufficient for his purposes apparently to ascertain overall costs and relate them to the A&P volume. That Malone was familiar with costs can hardly be disputed and Malone did consider A&P' s volume of business (Malone, DTR pp. 148-151). A&P also attacks CX 54 because the costs therein are higher than CX 4, a Borden calculation of costs pertaining to Borden s operations in Wisconsin in August 1964. But CX 54 relates to Chicago and to 1965. A cost study which is neither contemporary nor based on the same market areas is hardly relevant and cannot be the basis for discrediting CX 54 which was prepared for the correct time period and locality, particularly when it appears that the highest labor rates are in Chicago GREAT ATLANTIC & PACIFIC TEA CO. , INC., ET AL. )005 962 Initial Decision as against aJl other dairy rates in the l:united States (Schaub, Tr. 3305) and where costs were increasing during 1964-1965 (See Resp. A&P's Proposed Findings, pp. 380-381).

131. CX 54 is supported by CX 87, another calculation prepared by Malone some 10 months after CX 54. In CX 87, Malone calculated costs experienced selling to A&P by two different accounting methods and in each case found that Borden was selling private label to A&P at a loss. Although A&P argues that CX 87 does not relate specifically to Borden s cost of serving A&P, citing Malone, DTR p. 240, it appears that Malone was speaking of another exhibit, CX 13 , at that point. With respect to CX 87, Malone testified:

The three documents representing 87 A, Band C were developed in an attempt to reconcile our position as to which would be the best course to follow. Number one, we had the experience with the A&P private label operation and wanted to see on a general basis, on a general application of the elements of cost what our position would be, either lose or gain, and relate to where we would be if we lose all of the business. (Malone DTR 184185). 132. A&P also argues that ex 87 be disregarded because "the data used to determine the Borden costs shown on that exhibit were unavailable to A&P at the time of trial." This objection is now being raised for the first time in connection with CX 87. A&P's motion to strike CX 87 on November 30 , 1973, objected to the admissibility of that document on the ground that it was neither past recollection recorded nor a business record. These objections were overruled (see my order of :Iarch 5 1974). It ill becomes A&P to show concern for the (47J absence of underlying data to CX 87 literally years after such objection would have been timely and appropriate, that is, when Malone gave his deposition (Malone DTR pp. 229-250), or at least on November 30, 1973, when it moved to strike CX 87. 133. A&P also disputes CX 87 in that it compares the prices of the products sold to A&P on the O'Hare branch. Inasmuch as the O'Hare branch is the highest price area covered by the Borden A&P contract (CX 62) whereas Malone used an average price for aJl A&P stores including stores with lower prices than those from the O'Hare branch it necessarily follows that the average price arrived at for all will be lower than that at the O'Hare branch.

Burden of Proof 134. A&P contends that complaint counsel have not proven that the price discriminations were not actually cost justified and that they have not carried the burden of proof required under Section 2(f) of the v. FTC 346 U.amended Clayton Act, citing Automati", Canteen Co. .. .. ..

100 FEDERAL TRADE COMMISSIOJ\ DECISIONS Initial Decision 87 F. 61 (1953). Complaint counsel must come forward with evidence which proves both that the prices paid by the respondent buyer were not in fact cost justified and that the buyer knew or should have known this fact. Where, however, the buyer knows he is receiving a substantially lower price although thc quantities in which hc buys and the manner and amount of exertion with which he is served by a seller is the same as for his competitors the Commission need only show, to establish its vrrnJacie cae, that the buyer knew that the methods by which he was served and the quantities in which he purchased were the same as in the cae of his competitors (346 COS. at 80). Even under Canteen complaint counsel have met their burden here. The competitors were charged a higher price for Borden s milk products despite the fact that they took at least as much and often more than the individual A&P store with which they were competing. Further, although their service requirements were in some cases greater than that required for A&P they could have had the identical services but still at a higher cost than A&P was paying. Thus with all differences in quantity and service eliminated there were no cost differences with the possible exception of advertising which I have found to be negligible in this particular situation. (48) 135. !'or can it be denied that A&P knew or should have known that the methods of service and the quantities it purchased were the same as its competitors. A&P, not a novice in the dairy retailing business, certainly knew that Borden s prices were set according to its price lists and discount schedules. Moreover, it appears that A&P was purchasing Borden label products at the same prices that were charged its competitors. A&P officials and Borden offcials discussed pricing patterns in the Chicago area as compared to pricing in other cities and even if A&P did not see Borden price lists it certainly knew the general contents and could have solicited further information (Tarr, Tr. 857- 859; CX 12B, 19). Under these circumstances and in view of A&P' other knowledge of Borden s prices and costs A&P clearly had the duty to inquire to determine whether its prices were legal. In Fred Meyer Inc. v. Federal Trade Commission 359 F.2d 351 , 365-366 (9th Cir. 1966), it was held that "Meyer had a knowledge of at least sufficient facts to create a reasonable suspicion that the payments it received were probably illegal* * * the duty to inquire (thus) arises. The Commission has held:

Where a buyer knows that he buys in the same quantities as his competitor and is served by the seller in the same manner or with the same amount of exertion as the other buyer the Commission need only show to establish its prma facw cae that the buyer knew that the methods by which he wa.'i served and the quantities in which he purchased GREAT ATLANTIC & I'PACIFIC TEA CO. , I"C., ET AL. 1007 962 Initial Decision were the same as in the cae of his competitor . However, assuming the matter to involve different methods or quantities if complaint counsel show such facts and circumstances as would have given the buyer reason to believe based on the knowledge available to him, including knowledge of the methods of doing business in the particular industry, that the different methods or quantities could not have resulted in cost savings sufficient to justify the differential allegedly accorded to him, they would have met their initial burden (Subrban Propam Gas Cor. 71 F. C. 1695 , 1698-1699 , n. 2 (1967)). (49) A comparison with the Beatrice case is also appropriate. In that case complaint counsel were held to have met their burden of proof relying upon Kroger s market knowledge, the fact that it did not seek evidence of cost justification from the various bidders, and the fact that Beatrice s pricing to Krogcr was on a cost-plus basis, whereas Beatrice was seHing to competitors on a list-price-minus-discount basis which was not based on actual costs (supra at Pl'. 820, 821). In that case there was no communication between the parties concerning cost justification. Here, however, Borden submitted detailed cost data and informed A&P that its bid could only be justified on a meeting of competition. By any approach, therefore, complaint counsel have met their burden of proof to demonstrate that Borden s price to A&P was not cost justified and that A&P knew or certainly should have known that such was the case.

Havemeyer s Cost Studies 136. in rebuttal to the contemporaneous Malone Studies (CX 54B CX 87A-B) and other cost data (CX 2lU, CX 23-25A, CX 206), A&P submitted three basic cost calculations, RAPX 232, 233 and 23. RAPX 232 purports to demonstrate that Bordcn could have expected to earn a profit on its private label price proposal to A&P. This exhibit purports to calculate the total cost of processing, transportation to branches and delivery to the store, of A&P private label items based on costs of Borden for May 1965 or for other periods in determining some of these costs. it compares such cost to the private label price granted to A&P in September of 1965 to arrive at a calculated weighted margin per point for Borden in its sales of private label dairy products to A&P. It breaks down the total cost into seven categories: Direct material costs: processing costs; trailer loading; hauling to branches; truck loading at branches; delivery; and indirect marketing general and admir.istrative costs. RAPX 233 compares Borden s cost of serving A&P stores with that of 28 competitors on Borden s O' Hare routcs in March 1966. RAPX 234 compares Borden s cost of scrving A&P stores with that of five competing purchasers in eight store locations in Hammond and Valparaiso, indiana, in October 1969.

137. These studies were prepared by Mr. Robert G. Havemeycr Initial Decision 87 F.

(Havemeyer), called as a witness by respondent A&P. Havemeyer is a management consultant and a partner in the management (50 consulting firm of Case & Co. , and has been with that company and its predecessor firm, Stephenson, Jordan & Harrison, since 1953. His professional career has been very largely devoted to cost accounting in the dairy industry (Havemeyer, Tr. 6770-6775). He did not, however base his studies upon any personal observations of Borden s operations and based his study upon only a portion of Borden s records which portion had been selected by someone else (Daube, Tr. 6704-6705). He did not partake in the process of selection or even review all of the available documentation to determine the completeness and representativeness of the selection (Tr. 7520). His experience included only one other Robinson-Patman cost justification study presented in a litigated case (U.s. v. Borden (Bowman), 370 U. S. 471 (1962)). (Tr. 6769-6775 9131-9132).

i. Delivery Costs 138. Of critical importance is his calculation of delivery costs, being the most important and largest expense item with the exception of direct material costs. Havemeyer originally calculated them using Case & Company s wholesale milk delivery time standards. Complaint counsel objected to the admission in evidence of such time standards on various grounds. By order dated September 11 , 1974, I overruled complaint counsel' s objections insofar as they were based upon alleged unreliabiIity, immateriality or irrelevance. However, by order of September 16, 1974 , I sustained complaint counsel's objections to Havemeyer s studies insofar as they were based upon Case & Company s time standards because data underlying the conclusions , 76;reached on these studies were not produced (RAPX 232, PI'. 89 RAPX 233 , pp. 39-66; RAP X 234, pp. 3-10). Consequently, the ultimate conclusions of Havemeyer for the cost of this function have been exeluded and no meaningful comparison is possible between his studies and the cost calculations prepared by Malone. (See my order dated November 20 1974).

139. A&P nevertheless contends "there is amplc evidence in the record to calculate the costs that the Borden Company could have anticipated if A&P accepted the significantly reduced delivery service which was an explicit condition of the private label price proposal" It argues that delivery costs can be calculated using the official Wisconsin Manual. It contends that Wisconsin is within the same region of the Borden Company as the Chicago Metropolitan area and that operating conditions within the region will be roughly comparable.

GREAT ATLANTIC & PACIFIC TEA CO. , I , ET AL. 100 962 Initial Dccision (51) 140. The so-called Wisconsin Studies are RAPX 9. Dr. Solverson testified on behalf of complaint counsel in June 1973. The thrust of his testimony was to the effect that a differcntial in delivery times per unit for different types of delivery services narrowed as quantities per delivery increased. In cross-examination counsel for A&P offered several exhibits purporting to contradict Dr. SoJverson s testimony and these were received in evidence (RAPX 8 through 15). As I noted in my order dated November 22, 1974, these were received in evidence for impeachment purposes.

141. A&P argues that Dr. Solverson testified that the time requirements for effecting full and limited service deliveries in Chicago would be the same as full service in Milwaukee (Solverson, Tr. 1556). I do not so read his testimony. Rather, it appears that Dr. Solverson was speaking of a pattern or relationship between modified drop and full service in delivery time savings and that it was this pattern of decreasing delivery time savings when volume goes up that is true for a large number of geographical areas. He did not testify that time standards from the Wisconsin Manual or his thesis were applicable anywhere else. Indeed, the Wisconsin Manual and Dr. Solverson s thesis were admitted to show disparities between the two. Moreover, his thesis contained time observations from one additional company not included in the Manual figures (Solverson, Tr. 1601-1602 1604-1605). There were also differences in thc definitions of the various service terms as between this company and those developed in the Wisconsin Manual (Solverson, Tr. 1606). Thus, RAPX 9 shows that one of the functions perform cd under full service in Wisconsin was stamping merchandise, a service that was not part of full service in Chicago (Pergler, Tr. 5651-5654; RAPX 9, p. 2; RAPX 35). 142. It therefore, inappropriate to attempt to establish time standards from RAPX 9. It fails to be precisely applicable to the State of Wisconsin, let alone the Borden experience in the Chicago market area. Even Havemeyer admitted that there were very substantial differences in both the fixed time per stop and the variable time per case as between his New Jersey Study and his Bowman Dairy Study (Havemeyer, Tr. 8183-8187). Thus, even considering RAPX 8 through 15 as not merely admitted for impeachment purposes, they constitute insufficient basis for a determination of Borden s cost in the Chicago area for the periods of time involved in this proceeding. (52) 143. Nor does it appear appropriate to consider a cost justification defense on the basis of standardized costs. We are involved in actual costs, not what the costs should have been. As Havemeyer reported to thc Virginia Milk Commission on a study conducted in that state:

Initial Decision 87 F. The cost of the three principal functions - plant processing, distribution and instore handling - are independent. That is, a low cost processor could be either a low or high cost distributor; so, also, could a high cost processor. The costs of each of these functions are independent of those incurred by the customers (stores) to which merchandise is delivered. (RAPX 24A, p. 11-4). Following this rationale, it is obvious that a particular dairy such as Borden may be a low cost processor but a high cost distributor or vice versa. Standards would arrive at an average or a mean and would not necessarily be descriptive of any particular dairy without some evidence to associate the activities of such a dairy with the average dairy, which evidence is lacking here.

144. Even if Havemeyer s computations rc service-times were acceptable, his calculations re delivery costs appear to be unreliable for other reasons. Some of the activities were improperly computed on a time basis. Thus, Havemeyer allocated driver commissions on a time basis rather than to the volume the driver delivers. But commissions were earned and paid on the basis of volume delivered (Pergler, Tr. 5731-5732; Banaski, Tr. 9560; CX 119J-K; ex 298H-J; CX 299H-K; CX 300H-J). Pergler testified:

The commission that the routemen received was based on their sales not on the time that they put in at an individual store on an entire route. and Banaski testified that wholesale drivers were paid both a salary commission. The union contracts specified that the salary is determined on a hourly, daily or weekly basis but that the commission is s calculations based on the number of points delivered. Havemeyer result in a higher cost for such activity performed for non-A&P s cost customers and inasmuch as this is apparently incorrect, Borden saving on this function in connection with private label sales (53 J to A&P is improper (Lemberg, Tr. 9405, 9408). Similarly, Havemeyer allocated total building expense to the wholesale delivery function and on the basis of the time spent at the store by the deJiveryman. In the - wholesale first place, since a11 four sales categories were involved of the total building bulk wholesale, vendor and retail - only a part expense should have been allocated to wholesale. Moreover, such expense has no apparent relationship to the time spent at the store by as loading, storage the deJiveryman but to the function involved such etc. (Lemberg, Tr. 9409-9412). Again, Havemeyer allocated the automotive expense of the delivery trucks according to the time spent serving this store by the deliveryman. The proper approach would be to allocate such costs to the products carried since the trucks are used merely to transport the products to the stores (Lemberg, Tr. 9408-9409 9447 9451).

GREAT ATLANTIC & PACIFIC TEA CO., INC., ET AL. 1011 962 Initial Decision 11. Processing Costs 145. But even independent of delivery costs, Havemeyer s calculations arc not persuasive. Arguing thlt Malone s costs as shown on CX 54B are unreliable, Havemeyer calculated Woodstock processing costs by multiplying the unit cost experienced by Borden at its two Wisconsin dairies (the so-called Milwaukee operation) by the units produced at Woodstock arriving at the total expense that Woodstock woud have incurred in producing each product had it experienced :Iilwaukee unit costs (RAPX 232, pp. 10- , 21). He then adjusted the costs to reflect the fact that Woodstock's labor costs were higher than Wisconsin s while its non-labor costs were lower (Havemeyer, Tr. 6990 6998-6999). This approach does not, however, appear to be reliable. We are not here concerned with the processing costs in Wisconsin but in Borden Woodstock lllinois plant. Moreover, we are concerned with the period of May 1965 rather than August 1964, the month used by Havemeyer. There is no record substantiation for the propriety of comparing the two. Havemeyer was not shown to be familiar with either of the Wisconsin plants at the time in question and, in fact, was not familiar with thc operations of the Woodstock plant. One of the two Wisconsin plants used by Havemeyer was inadequately identified although the calculations of that unidentified plant appeared to be in some respects (54 J referrablc to Woodstock. There was no evidential showing that the Woodstock and the Wisconsin plants had similar product lines, similar labor forces, similar equipment, similar utilization of that equipment, similar management problems, similar volume ratios among the product lines, similar filling machines, similar efficiencies, etc. (Tr. 8100-8111). Moreover, Malone was one of the Borden officials who submitted this study upon which Havemeyer relies. Malone stated, however, that "there is no relationship between the Chicago factors and the Wisconsin factors" (Malone DTR p. 36). 111. Havemeyer s Unfamiliarity 146. Havemeyer s lack of familiarity with Borden s activities also diminishes the value of his calculations concerning rental costs of the of thisEx-Cell-O machine, which is a filing machine. The allocation cost was originally made in accordance with a formula which, however was rejected for failure to submit substantiating evidence (Tr. 7397 7402, 7393). Havemeyer then submitted an alternate method of allocating these costs (RAPX 154; Tr. 8114-9118). Hc testified however, that his original method was supcrior (Tr. 8117) because the alternative was " of an averabring nature." Moreover, since the rental consists of 2 parts - a base rental and a production rental - (Tr. 8131initial Decision 87 F. 8132, 8133) and the base rental was paid on a monthly basis but the production rental may have been on a quarterly basis or some other basis other than monthly, we cannot be sure that all machine rental costs are shown in Havemeyer s calculations. His lack of familiarity with Borden s operations makes it impossible for him to supply such information. Malone s informed allocation of this cost in CX 54B and RAPX 65 must be considered superior to Havemeyer s averaging method.

147. There were a number of other areas in which Havemeyer calculations of cost would not be as reliable as Malone s calculations when one takes into consideration Havemeycr s lack of familiarity with Borden s operations compared to Malone s intimate knowledge of such operations. Thus, in connection with trailer loading expense Havemeyer discredits Malone s computation where Malone allocated the same expense to gallon glass containcrs as he allocated to paper gallons (Tr. 8160-8161). Although there were more paper gallons in a case than there were glass, there were also more cases of glass gallons on a pallet (Perglcr, Tr. 5632) and there may have been other economies involved L55J in loading the glass gallon pallets of which Malone was aware but of which Havemeyer could not have been aware.

148. A similar difficulty arises in connection with truck loading expense. There Havemeyer attempted to make a determination of the average points of each product that would be contained in a loaded case relying on his experience of normal practice in the dairy industry (Havemeyer, Tr. 7010-7011). Since, however, he was not familiar with Borden s practices, his computations arc not necessarily indicative of Borden s experiences without a determination that such experiences were the normal in thc dairy industry. In addition, Havemeyer calculated such costs in contradictory ways. In connection with loading costs in the Chicago area, he included total purchases by thc unfavored competitors (the 11 items involved in the A&P private label agreement with Borden and all other Borden dairy products purchased by such stores). In the Indiana study, he considered only the private label items and private label equivalents (RAPX 233, 234). In the latter study, the loading costs were virtually identical. Since this matter involves discrimination on private label items only, the latter study was the only proper approach and would eliminate much, if not all, of the cost differential in this function.

149. Havemeyer s expertise is insufficient to remedy other apparent irregularities of his calculations. Havemeyer attempted to calculate product waste cost but in contradictory ways. He admitted that Icakcrs and product spoiling within date would be product waste GREAT ATLANTIC & PACIFIC TEA CO., INC. , ET AL. 1013 962 Initial Decision (Tr. 7749). He added, however, that this would be a very small part of the product waste which included a privilege to return out-of-date merchandise that was not defective. Consequently, in his profitability study (RAPX 232) he aJ10cated 5 percent of "other delivery costs " to A&P private label (Tr. 7753). In his studies of O' Hare costs (RAPX 233) and Hammond (RAPX 234) Havemeyer took the position that leakers and product spoiling within date are not product waste but were accounted for in some other manner in Borden s bookkeeping system (Tr. 8490). He was unable, however, to cite any record evidence for that statement or from whom that information was obtained. Moreover there is serious doubt that leakers and product spoiling within date would be only a smaJ1 pc)ration of the route returns. Competing retail grocers generally testified that they had very few returns (Minard, Tr. 3258; Kristoff, Tr. 2431; Barney, Tr. 3838; Cox, Tr. 4002-4003). Minard (56 J testified that what returns his company did have were usually leakers or product spoiling within date. Havemeyer himself indicated that Borden probably experienced a high number of gaj10n leakers in this period due to the fact that the paper packaging equipment frequently had problems during this period (Tr. 8101). IV. Indiana Irregularities 150. Another apparent irrq,rularity in Havemeyer s calculations concerns Valparaiso sales where he used RAP X 81 and 87 as the source for cost data on the assumption that Mr. Marquardt, (the independent distributor who delivered in Valparaiso for Borden) picked up his milk at the Hammond dock. He was unable to substantiate his position by documentation (Tr. 8641). Marquardt himself never testified as to where he picked up his milk (Marquardt, Tr. 1315-1331). Record evidence associates Valparaiso more closely to South Bend than to Hammond (CX 17K; ex 621). There is, therefore, insufficient justification for the use of RAPX 81 and 87. But even if there were, the expenses thereon are Borden s expenses dealing with vendors, not with stores served by vendors who sold on their own terms (Marquardt, Tr. 1323).

151. A&P argues that Havemeyer should be permitted to apply Hammond branch costs for vendors for his analysis in Valparaiso. It bases its argument simply on the additional cost it would have to sustain to make still another study for this single comparison. But that is not the point. A correct study should have been made in the first instance. Havemeycr s cost studies' deficiencies render his conclusion unpersuasive.

152. The VaJparaiso study is also misleading. It attempts to calculate Borden costs in its dealing with :\Iarquardt who was Initial Decision 87 F.

independent distributor. Borden records, therefore, show its expenses of dealing with such independent distributor and not of its dealings with the distributors' customers with whom it had no relationship. Accordingly, if Havemeyer were to apply Valparaiso s expenses to stores served by those vendors, al1 or part of those expenses would have to be shown as applicable in that way. This has not been done. Some of the expense allocated to the non-A&P store in Valparaiso (Title) is made up of "provision-bad loans. (57) Any bad loans however, made by Borden to its vendors would not be attributable to the stores to whom these vendors deliver, inasmuch as the Borden customer involved in the vendor account is the vendor and not the vendor s customer, a result reinforced by the fact that Title received no loan from Borden (Minard, Tr. 3223-322). Nor can sellng expense be ascribed to the Title store inasmuch as Valparaiso is not on a Borden route, Borden salesmen did not visit Title s store there (Minard, Tr. 3207-3208) and Title was not even a customer of the vendor which delivered to it (Marquardt, Tr. 1317). Similarly, there is no basis for allocating credit and collection costs to the Title store for the additional reason that it was not responsible for any credit loss (:\inard, Tr. 3157). Tittle did not pay through a distributor (Marquardt Tr. 1317). Borden s provision for bad accounts with milk distributors could not be applicable to Title. Further, since Tittle was never overdue (Minard, Tr. 3157) there could be no interest on overdue accounts.

v. Branch Selling Expense 153. Havemeyer s calculations in connection with branch selling expense - the expense related to the activities of Borden s salesmen - are also questionable. In this connection he estimated the number of points sold to independent stores, thus eliminating A&P sales and sales to schools and other non-store customers from consideration, and arrived at a branch selling expense pcr independent store point (RAPX 233 , pp. 23, 16; Tr. 8492-8494). Such allocation was based on the assumption that the entire time of the salesman was spent in soliciting these particular store accounts and serving these accounts. But Mr. Daube, a Borden executive, testified that the salesmen "contacted new customers. . . . (h )is basic function was to develop and obtain new business for the company" (Daube, Tr. 9226-9227). Such missionary work is part of the selling expense of doing business on the part of Borden and it should be borne equally by al1 customers, A&P as well on a volume basis (Lemberg, Tr. 9419-9420). Moreover, there is no record evidence to justify Havemeyer s elimination of non-store customers but only his belief that Borden would have special bulk GREAT ATLA TIC & PACIFIC TEA CO. , I!(C. , ET AL. 1015 962 Initial Decision wholesale salesmen who would call on bulk customers. He admitted however, that there is nothing in the Borden expense account that shows that the salesmen do not call on non-store customers who are served by the wholesale routes (Tr. 8496-8497). In addition, the Borden- A&P contract provided that "Borden s salesmen s services will be supplied at the store level only, to follow (58) through on a complaint covering such matters as service, product or container" (CX 62Q). But no time has been computed by Havemeyer for such activities. The absurdity of this conclusion is demonstrated by the amount al10cable to Jim s Groceries where Borden s salesmen visited maybe once a week or maybe once a month (Cox, Tr. 3992). But Havemeyer s allocations would run about $165 for March 1966 (RAPX 233, pp. 23, 31). Nor did Havemeyer give any consideration to the fact that Borden incurred some selling expense in connection with the functions of Tarr and the salesmen who assisted him in servicing the A&P account (Tr. 8503). Yet Tarr spent almost al1 of his time in serving the A&P account (Tarr Tr. 845-846, 897, 899). Finally, Havemeyer s approach on this function contradicts the treatment in his profitability study where he allocated some selling expense to A&P private label.

vi. Failure of Classification 154. Havemeyer s cost calculations are also flawed by his failure to classify customers in accordance with size, type of store, volume of purchases and other characteristics. In his profitability study he treated al1 independent stores as having one set of characteristics and A&P stores as having another. Thus, in RAPX 232 p. 89, he established various volume brackets for "other stores" but made no volume analysis for the A&P sales, despite the fact that 10 of the 88 A&P stores listed in the March 1965 breakdown of O'Hare sales took less than 300 points per delivery (RAPX 74). In calculating the number of points per case in May 1965, for al1 "other stores" Havemeyer examined the route books for 3 out of 17 routes in March 1966 and admitted that "there was not sufficient data to be certain of a variation of the points per case by size of store" (Tr. 7845). While admitting that it was highly unlikely these various size stores purchased the same number of points per case, he explained that "the data was not there unless you went through a tremendous analysis (Tr. 7850). His resulting fihrure of 22 points per case for "other stores is obviously quite arbitrary. Similarly, in connection with RAPX 233 and 234, Havemeycr gave no recognition to different classes of customers with different cost experiences in connection with his allocations of delivery expense, clerical expense, product waste, selling expense, etc.

. . .

1016 FEDERAL TRADE COM:\ISSJON DECISIONS Initial Dccision 87 F.

155. This failure of classification is hard to understand in view of Havemeyer s prior experience. The one (59) prior Robinson-Patman type cost study made by him was for Bowman Dairy in the case of U.S. V. Borden 370 U. S. 460 (1962). His study was rejected by the Court at 471:

In the (Bowman J study the experts charged ali independents and no chain store with these costs. Yet it was not shown that all independents received these services daily or even on some lesser basis. Bowman s study indicate only that a large majority of independents took these services on a daily basis. Under such circumstances the use of these cost factors across the board in calculating independent store costs is not a permissible justification for it possibly alloct€s costs to some independents whose mode of purchasing does not give rise to them. The burden was upon the profferer of the cla.o:sifieation to negate this possibility and the burden has not been met here.

There can be no doubt that there are vast differences between the characteristics of the various customers on Borden s routes, differences in physical size, layout of stores, purchase volume, services received and so on. Even the A&P stores had such differences. Havemeyer studies here merely repeat the crror made in his earlier Bowman study. 156. In the O'Hare and Hammond studies Havemeyer determined costs on an average basis for A&P stores and then compared this average with specific competitors. Thus, in his delivery cost analysis in RAPX 232, he proceeded on the basis that A&P was one single customer. This was condemned, however, by the Commission in National Dairy Products Corp. 70 F. C. 79 , 193- 196, 207 (1966). The Commission stated:

The Hearing Examiner rejected respondent' s cost justification defense, in part for the reason that discount.o; to multi-unit purchasers must be cost justified on a stare-bYRstore basis. We agree . . . the averaging af purchases has no relationship to the cost actually incurred in dealing with each store (at p. 194). Thus, the largest independent customer respondent has in this area received only a 6 percent discount whereas a competing chain store unit of (GO) the same volume receives 12 percent, not by virtue of any savings in cost to the store but solely by reason of its membership in the chain . In practice therefore respondent takes the over-cost justification in delivery to the large stores of a chain and credits this time saved to the smaller stores. In our view this is not a valid costing procdure under the Robinson-Patman Act (at pp. 194- 195).

Yet this is exactly what has been done here. A&P stores in the Hammond area range from a low of only 17 plus cases per delivery to a high of 52 plus cases per delivery (RAPX 234, p. 19). Thus, although Burger s :Iarket took more than 78 cases per delivery, the competing A&P store #325 took only 17.68 cases per delivery but this small A&P store was averaged out with larger A&P stores to hide the fact that there was such a hugh disparity in size and purchase volume. Of course GREAT ATLA:\TIC & PACIFIC TEA CO. , I , ET AL. 1017 962 Initial Decision by such avera6ring out, the fixed costs applicable to this store operation become much less than they would be had the true volume been known since fixed costs are usually allocated on the basis of volume.

VII. Countervailing Costs 157. Nor did Havemeyer give adequate consideration to countervailing costs - costs necessary for the A&P business but not for other business. Reference has already been made to Havemeyer s failure to consider the expense of Tarr, the Borden employee whose activities were in great measure devoted to the A&P account. But Minkler Malone, Gose and others in Borden s Chicago office also spent a very great amount of their time in getting and keeping the A&P private label business (Minkler, Tr. 125-279; Tarr, Tr. 849-889; Gose, Tr. 1045- 1063; Malone DTR pp. 29-204); some of Borden s home office personnel in New York spent considerable time on that same account (Archer, Tr. 1209- 1212, 1228, 1280) but Havemeyer made no special study of these expenses (Tr. 7757-7758, 9699-9710).

158. :Ioreover, it was often necessary to mix A&P label and Borden label products together on a pallet to make up a full pallet for an A&P store (Pergle, Tr. 5697-5698, 5700 , 5702-5704). This mixing had to be done by hand (61J which would involve extra effort and expense which was however, ignored by Havemeyer who stated that the mixing of products on pallets was done equally for everyone. But Havemeyer has not personally observed this procedure (Tr. 7980), and Pergler testimony relates primarily to the A&P deliveries. 159. There may have been additional costs for A&P stores involved in connection with clerical expense. Thus, there was a daily delivery ticket for each individual store and a summary billing sheet for a number of stores in an area (Tr. 8484-8485). This additional summarization was not considered by Havemeyer (Tr. 8478-8486, 7716-7730). In addition, Borden had to include the retail price and extend the retail value for each product on a total basis at A&P's request (RAPX 170 pp. 87-91; A&P Adm. March 12 , 1973, p. 2, second par.). Despite this indication of countervailing costs Havemeyer made no study of the effect without which a true cost difference could not be ascertained (Lemberg, Tr. 9421-9422).

160. Furthermore, Havemeyer determined the clerical expense for A&P private label on the basis of the fact that private label represented 11/38'5 of the number of lines on a summary billing to A&P (RAPX 233 , p. 11; Tr. 8478-8479). This ignores the fact that an A&P billing had 38 lines but the competitor s stores billing had at the most 27, the difference, 11, being the 11 private label items which A&P 216- 969 O- LT" 77 - 05 Initial Decision 87 F.

was purchasing but not the competitor stores. Thus, it would appear that clerical costs should have been determined by comparing the number of lines on the billngs required for the two different stores which would result in a higher clerical cost in this connection for A&P stores than for the competing stores. In effect, A&P's purchases involved clerical expense in connection with the biling of the private label items, an expense which was not incurred by the competing stores which did not buy any private label items.

viii. General Selling and Administrative Costs 161. Havemeyer s calculations concerning general selling and administrative expense is also questionable. He first ascertained the overhead cost" which he had selected from Borden s central office accounts (RAPX 233, pp. 5, 6). He then took the sum total of all other functional expenses which he had allocated to O'Hare customers divided this total into the overhead cost, arrived at a resulting percentage figure and applied the percentage figure to the (62) total costs for the other seven functions which he had calculated for the Hare customers. These seven branch functions, however, did not include production or transportation costs which, according to Havemeyer, totaled about $270 000 in :Iay 1965 (RAPX 232, pp. 24, 33 , 39), whereas the branch expenses in Yrarch 1966 were calculated as being about $82 000 (RAPX 233, p. 6). It is only reasonable to assume that the central office would be concerned v,'ith the totality of Borden operation and would spcnd some time concerning itself with production and transportation costs but none of such costs have been considered in this connection by Havemeyer. Instead, a very substantial amount of the overhead expense is allocated on the basis of the time the driver spends at the individual store since delivery expense is allocated on this basis. Thus, Jim s Groceries was a daily cash basis customer (RAPX 233, p. 69). His monthly purchase volume would be about 10 000 points (RAPX 233 , p. 31) and Havemcyer would have this customer credited with having generated about $41 in expense for the Borden home office staff in :Iarch 1966 (RAPX 233, p. 41). Yet it is hardly likely that the home office would have any contact with Jim s Groceries. Similar results would apparently be rcached for other cash customers (RAPX 233, pp. 60- , 66, 69).

ix. Credit Loss 162. Still another questionable area with respect to Havemeyer calculations arises in his treatment of credit loss. Havemeyer estimated the amount related to the wholesale business and allocated it cqually GREAT ATLANTIC & PACIFIC TEA CO., INC., ET AL. 1019 962 Initial Decision per point among the 17 independent store purchasers on the Hammond wholesale routes. The "milk regular" account by Borden included both wholesale routes and retail home delivery routes (Tr. 8818-8819), Havemeyer allocated 12 percent of the expense to retail home delivery customers and 88 percent to wholesale customers (Tr. 8818), but admitted that if done a volume basis retail sales would account for about 25 percent (Tr. 8819). Havemeyer admits that this is a "bad account" expense record (Tr. 882&-8828; RAPX 234, p. 90) and there are substantial amounts attributed to "legal expense" and "interest overdue accounts. " It would appear, therefore, that this account is primarily related to particular customers who arc in arrears. With the exception of Wallies, all the independents had been extended credit for a maximum of 4 weeks (comparing the average weekly sales figure with the largest balance shown for each store). This would hardly constitute a bad account nor was it so regarded by Borden who continued to grant rebates to these stores (CX 288-289; CX 188, 191 194-197 , 200, 202). There was then (63) no reason to allocate that account' s expenses to customers who bad not contributed to such expense and if A&P was not so charged there was no reason for charging the independents, some of whom may have been as financially sound, if not more so, than A&P (CX 189, 192, 199, 203 carMra; CX 21&-219 , 221).

x. Advertising Expense 163. Still another area of doubt involves advertising expense. Four products were involved in this advertising (RAPX 90, p. 3) only one of which was involved in the Borden-A&P private label agreement (Bitting, Tr. 9334). Two of the other three products were sold only under the Borden label and the fourth item was not sold to any of the store customers in this area (Bitting, Tr. 9335). The one pertinent item represented only some $3 000 out of a total expenditure for all four of $114 000. Although Havemeyer argues that these advertisements are really advertisements of the entire line of Borden products, not just the named product, it was shown that each of the products involved was a new product which was being introduced and Havemeyer did not know what was said in any of the advertisements (Tr. 9638-9655). Without such information there is no basis for assuming that the advertisements referred to more than the four products named. Even if such advertising represented institutional advertising the expense should be applied to all sales including those of private label since it represented a general expense of doing business and expanding sales (Lemberg, Tr. 9419-9420).

, p.

Initial Decision 87 F. T. Xl. The Availability Argument 164. Finally, A&P contends that complaint counsel seem to argue that even if Borden s prices to A&P were cost justified, A&P does not have a defense unless it can be also shown that those prices were offered on some kind of proportional basis to others. It bases this construction of complaint counsel's arguments on complaint counsel's language to the effect that the cost difference in this case is not due to differing methods or quantities of sale or delivery, which is a requirement under Section 2a of the Robinson-Patman Act, but the difference in availability, that Borden offered all stores the same type of service spelled out in the Borden-A&P private label contract but at (64J a much higher price (CX 138; ex 62P-Q). Indeed, I have found that virtually the same service was offered by Borden to its non- A&P store customers (provided they had a certain volume of purchases) but at a maximum discount from list prices of not more than 30 percent compared to A&P's effective discount of more than 35 percent. A&P's logic would rcquire competitors to accept the same service program at much less discount in order to preserve their rights under the Robinson-Patman Act, even if this meant compounding their Joss or disadvantage by requiring them to incur more in-store service cost as the result of reduced service from Borden, without an adequate discount to compensate them for it. As complaint counsel view this argument a buyer may legally discriminate in price jf he wraps up the price package in service terms and offers the same service to others at a much higher, unattractive price, which the competitors cannot accept. The difference in servce t€rms will then justify the price difference (complaint counsel's Proposed Findings, Vol. II 29).

Such an obvious flout of the Robinson-Patman Act cannot be tolerated. Non-A&P stores were offered the same service terms enjoyed by A&P but were required to pay more. There was, however, no difference in Borden s cost savings as between the A&P stores and non-A&P stores receiving the same services. It follows, therefore, that no price differences could be justified thereby. It may be argued that the non- A&P stores did not take the limited service that A&P had and that therefore, Borden actually experienced cost savings in its dealings with A&P that it did not experience in its dealings with the non-A&P stores. But Borden s preferential discounts were not "functionally available to the non-favored customers who had to meet certain minimum volume purchases which A&P did not (Morn Salt Co., 334 U. S. 37, 42 (1948)). It appears only proper to make cost comparisons in a seller dealings with competing purchasers where such dealings are com para- GREAT ATLANTIC & PACIFIC TEA CO., INC. , ET AL. 1021 962 Initial Decision ble and functionally available. The Act recognizes the need for like grade and quality of items sold. Certainly the same conditions of sale must exist before cost comparisons can be made. With the same conditions of sale offered all purchasers, no differences in costs could arise. A&P's inducement of a preferential price under such circumstances would violate the Act, particularly where, as here, the buyer knows that the price its competitors are obliged to pay for their purchases is one which it, the buyer, refused to pay and succeeded in forcing the seller to reduce to that preferential level. (65) XII. Conclusjon 165. I conclude, therefore, that A&P' s cost justification studies RAPX 232, 233 and 234, are so defective and inadequate as to furnish no evidentiary basis for justifying A&P' s prefercntial price for private basis for justifying A&P' s prcferential price for private label items on the basis of Borden s savings in costs.

VI. TIw "Borden Defense 166. A&P argues that since Borden "admittedly" has a defense to a Section 2(a) charge by reason of meeting competition, and complaint counsel have not charged it with a violation, A&P cannot be guilty of violating Section 2(f) It should be noted, however, that Borden has not been adjudged to have a defense to a Section 2(a) charge by reason of meeting competition. Bordcn simply has not been cbarged by the Commission which has such discretion in issuing complaints. Jt cannot be said with certainty that, had Borden been charged with a 2(a) violation, it would have been able to defend such a charge successfully. It has been held that a seller cannot accept a buyer s story (such as one about a competitive offer) without more (National Dairy v. FTC, 395 2d 517, 528 , (1968) where the court distinguished decisions in Forster Manufacturing Co. v. FTC 335 F.2d 47, 54 (1964), Beatrie Food Co. Inc. 68 F. C. 286 , 350 (1965) and Continental Baking Co. 63 F. 2071 , 2164 (1963) in which efforts to verify were found). But even had Borden been able to defend such a charge successfully, the result would nevertheless be insufficient to exonerate A&P under the circumstances. Thus, in the Kroger case the supplier, Beatrice, was found to have not violated 2(a) of the Act. Nevertheless, Kroger, the buyer, was found to have violated 2(f) of the Act: Kroger lcontendsJ that as a matter of law the discharge of Beatrice requires the acquittal of Kroger because there cannot be a violation of Section 2(f) v,rithout there being onc under Section a). While ordinarily this may be true - a matt€r we need not pass upon - it is not true under the peculiar circumstances here . . . . . ), Initial Decision 87 F. where Kroger was found by the Commission to have given "false price information " to Beatrice as to Broughton s competing bid (438 F.2d at 1374). (66) Here, too, A&P gave false price information to Borden as to Bowman s competing bid. Although A&P argues that Kroger actively misrepresented thc other bid but that A&P had a lower bid from Bowman, the fact rcmains that A&P represented Borden that it had a bid which could be compared to Borden when, as a matter of fact, it did not, in view of the volume conditions and delivery conditions. Even if those differences could be ignored and the two bids compared, A&P represented to Borden that Bowman s bid was lower than Borden when, as a matter of fact, A&P itself understood Bowman s bid to be higher than Borden VII. Commerce and By-Products 167. A&P cites the provisions of thc Robinson-Patman Act which make it unlawful for any person engaged in commerce, in the course of such commerce, either directly or indirectly, to discriminate in price between different purchasers of commodities of like grade and quality, where either or any of the purchases involved with such discrimination are in commerce It is not sufficient that the parties be engaged in commerce, the purchases involved in such discrimination must be in commerce. A&P' purchases from Borden were under a contract which covered sales and deliveries not only to its stores in Ilinois where Borden s Woodstock plant was located but sales and deliveries to its stores in Indiana and elsewhere as well. It was not severable. The contract, therefore covered purchases in commerce because it was multi-State in nature. 168. Moreover Borden supply of raw milk came from outside the State of Ilinois and its processing into milk products did not interrupt the interstate flow of milk from farm to retailer. Forerrst Dairns Inc. 62 F. C. 1344 (1963) aff'd. 348 F.2d 674 (5th Cir. cert. denied, 382 S. 959 (1965): Dean Milk Co. 68 F. C. 710 (1965), aff'd. 395 F.2d 696 (7th Cir. 1968).

169. A&P argues that certain of the milk products were so processed as to make them separate and distinct commodities, thus breaking thc interstate flow. It refers to products such as whipped cream, sour cream, eggnog, and onion dip, which allcgedly were so physically, (67) biologically and chemically different from raw milk that the interstate flow of commerce was terminated at Borden those products were a processing plant in Ilinois. 1\ evertheless relatively insignificant part of the contract entered into which was not severable. Whole milk constituted the great bulk of the sales under the GREAT ATLANTIC & PACIFIC TEA CO., INC., ET AL. 1023 962 Initial Decision contract and did not interrupt the interstate flow from the raw milk supply to the retailer. There is no basis for splitting the contract where the parties have not done so. To allow these insignificant products to control the legality of private label purchases under such an indivisible contract would have the tail wag the dog.

C. COuST II 170. Count III of the complaint alleges that A&P sold private label milk and other dairy products at the prevailing retail price level for vendor label milk and dairy products in spite of A&P's lower cost for the private label milk, thus maintaining the existing retail prices for milk and other dairy products. Borden for its part failed to pass on at the wholesale level price reductions similar to the reductions granted to A&P to other purchasers who compete with A&P. Respondents' course of conduct, it is alleged, constitutes a combination between respondents which had the tendency or effect of stabilizing and maintaining prices for milk and other dairy products in violation of Section 5 of the Federal Trade Commission Act. 171. Tbe area involved for purposes of proof of this Count is the Chicago Standard Consolidated Area (SCA) consisting of the counties of Cooke, DuPage, Kane, Lake, McHenry and Wil in the State of Ilinois and the counties of Lake and Porter in Indiana. 172. A&P ranked third in total sales among the retail grocery chains operating in the Chicago area, its sales in 1965 amounting to 203 milion dollars (CX 222A). It was a significant factor in the market and , Tr.its competitors were careful to monitor A&P' s prices (Lasorso 2646; Cox, Tr. 3947-3948, 3986; Barney, Tr. 3686; Minard, Tr. 3147 3152, 3153). Had A&P created a price differential on its private label milk, Scott Lad (a competing grocery) would have had to meet the lower price (Schaub, Tr. 3330).

173. Borden ranked either second or third in Class I milk sales wasamong the dairy companies operating in the Chicago area. It important market factor (HoJin, Tr. 2965-2966, 2982-2987; Soldwedel Tr. 3373-3374; Hitchner, Tr. 3596). Prior to the private label agreement Borden (68J was A&P' s principal milk supplier for its Chicago unit and provided approximately 95 percent of A&P' milk supplies (Schmidt Tr. 1678).

174. A&P and Borden competed with one another in the sale of milk and milk products at retail but the record in this proceeding does not establish clearly the dates of such retail deliveries by Borden nor the extent. Witnesses referred to Borden s retail sales as having lessened and ceased somewhere between 1963 and 1968. Thus, Borden Gose testified that Borden had "some" retail routes when he left , , 1024 FEDERAL TRADE COMMISSIOI; DECISIONS Initial Decision 87 F.

Borden s employ in 1968, but he added "I don t remember exactly, but we had considerably less " (Tr. 1041). Later, however, Gose said that retail home deliveries were phased out "somewhere" between 1963 and 1968 (Tr. 1113). Thc President of Wanzer Dairy testified that Borden "could have had 21 or 22 routes which means nothing in retail" (Soldwedel, Tr. 3406). The record does not establish with any degree of certainty that meaningful horizontal competition between A&P and Borden at the retail level existed at the time critical to this proceeding, 1965- 1968.

175. Borden competed indirectly with A&P at retail by selling at wholesale its Borden brand milk to non-A&P stores competing with A&P at retail (Kuhlman, Tr. 5273). If A&P were to reduce its retail price competing retailers might be forced to lower their price of the Borden label products and seek cost reductions from Borden. 176. A&P had strong motivation to stabilize and maintain prices on milk and milk products. As noted earlier, as the result of its private label agreement with Borden there was an expectation on the part of A&P to realize an estimated annual saving of approximately 820 thousand dollars. This saving could be realized only if A&P' s profi t margin was protected by the maintenance of the retail price (Kuhlman Tr. 5230). A reduction of A&P' retail price would necessarily lower its net savings unless Borden reduced its price to A&P even further. This A&P could not expect. It had no better competitive bids and there was always the realization that there was some point beyond which Borden could not possibly recede despite its need for volume. A&P was also anxious that Bordcn not grant others the same lower price enjoyed since if this were done, the A&P competitor might lower his retail out-of-store price and A&P might have been obliged to meet such lower price by reducing its retail price, thus losing some of the expected savings.

(69) 177. Borden similarly had strong motivation to maintain milk prices. Its newly completed processing plant at Woodstock, Ilinois, had a very large capacity and A&P' s Chicago Metropolitan Rcgion alone accounted for over 55 percent of Borden s wholesale sales volume (CX 42D). If Borden was to maintain an efficient level of production at Woodstock it had to retain the A&P business (Minkler, Tr. 217; Kuhlman, Tr. 5230). A break in milk prices if A&P would reduce its out-of-store private label prices would harm Borden seriously: We have said all this before many times, but, we do hope A&P will price this private label and the advertised brand Jabc1 the same, certainly to begin with. If private label out-of store is dropped to reflect this Ie per point, this market will rock immediately, we think. We believe such a move on their part would destroy the present price level for advertised brands and would add explosive fuel to a war over the differential question. GREAT ATLAKTIC & PACIFIC TEA CO., INC., ET AL. 1025 962 Initial Decision We know some competitors will be strongly opposed to a diffcrcnc. in price for a private label. One big decision we have to make ha.'J to do with how low we wiB go to force acceptance of a differential. We folded up in Wisconsin bceause of a "no sales below cost" stipulation in the Wisconsin Law. We have no such "out" in Ilinois. This is not exclusively a matter in our interest. We believe retailer s margins would eventually get caught in whatever upheaval takes place. With such a move on A&P' part - instead of an improvement in their overall margin, one point I am making is that their dairy department spread could collapse. If that should happen, they will not have gained anything by bringing in private label and, certinly, we wi1 have been knocked to our knees. (CX 7B) (70 J 178. In the Wisc:onsin situation referred to above, Borden was supplying milk to the Kroger stores both under the Borden brand and two secondary labels but at a lower price to Kroger for the milk products under the secondary labels. Kroger then lowered the retail price on these secondary labels which resulted in a chain reaction of price breaks in the prices of other dairies and chains forcing Borden to lower both its wholesale and retail prices on Borden label (Soberg, Tr. 2753-2756, 2759). A&P was also involved in this price war (Soberg, Tr. 2764-2765).

Complaint Counsel's Contentions 179. Complaint counsel contend that to stabilize and maintain the price of milk in Chicago all that was necessary in light of the market structure was for each party to agree or arrange to assure the other that neither would do anything which would tend to disrupt the Chicago market or any milk prices therein. The agreement and/or arrangement of mutual assurances were comprised of and/or were supported by, inter alia respondents' communications with one another, their exchanges of information, and their market knowledge and market conduct. (Complaint Counsel's Proposed Findings, You, p. 22).

180. With respect to respondents' communications with one another, complaint counsel cite record evidence indicating some discussion of market sensitivity between them. Thus, on February 9 1965, Borden s Minkler wrote to Borden s Pentz about a conversation between A&P's Schmidt and Borden s Tarr:

fWJhen Schmidt talked to Tarr yesterday he made it very plain that we had to be right. He said he very much appreciated the sensitiveness of this market and that he did not want to put this private label out for bids (CX 12). Nor did Borden want this business put out for bids. As noted earlier in Minkler s internal letter (CX 7A, B) Borden was concerned about a market upheaval" as result of private label bids and believed that retailers' margins would eventually get caught in whatever upheaval 1026 FEDBRAL TRADE COMMISSIO:- DECISIO:-S Initial Dccision 87 F.

takes place* * *their dairy department spread could conapse win be knocked to our knees.

(71 J 181. Complaint counsel also cite the meeting between A&P and Borden officials on or about May 26, 1965. Schmidt expressed a hope that the private label program might be productive of 500 thousand donars annuany. CX 19F states:

In this regard, it was stressed that this type of product contribution could only result from the continuance of present Qut-oi-store margins on fluid milk products and referred to the Wisconsin price war. Complaint counsel contend that this comment, taken together with the discussion which took place concerning Borden s narrow profit margins, nwas clear warning that A&P should not reduce the retail price on private label milk" (Complaint Counsel's Proposed Findings, Vol. I, p. 227). It should be noted, however, that CX 19F does not indicate whether it was Borden or A&P which "stresscd" the continuation of present margins. 182. Complaint counsel also cite the language of Borden in its price quotation to A&P in May 1965 (CX 18P). In it Borden says: In certain areas the primary basic price has reco!-'li7.€d the depressed competitive situation prevailing in such markets. Such situations must eventually be corrected. It is considered that any economic improvement that occurs subsequently will be mutually recognized in accordance with the situations. Complaint counsel' s economic expert, Dr. Kuhlman, contends that the parties were discussing prices in the retail market (Kuhlman, Tr. 5232- 5233 5308-5309). According to Dr. Kuhlman:

This constitutes in my mind a very clear and very explicit correspondence regarding the price of milk, and the fact that one of these companies is a supplier of milk to the other one can in no way be used to explain that this is not price fixing (Tr. 523).

183. Complaint counsel also argue that Borden s right to terminate the private label agrccment at any time helped (72J insure that A&P would not choose to alienate Borden as its supplier by reducing the retail price (Kuhlman, Tr. 5234).

184. Complaint counsel also cite the fact that the 820 thousand dollar annual saving offered by Borden to A&P under the private label agrecment was not made on any straight discount basis but spread across eleven different product lines and argue that Borden influenced A&P' s retail price because Borden determined how much of the savings A&P should realize on any individual product (Kuhlman, Tr. 5233-5234).

185. Complaint counsel also argue that A&P must have known that Borden was not prepared to give the same deal to competitors of A&P GREAT ATLAKTIC & PACIFIC TEA CO., INC. , ET AL. 1027 962 Initial Decision inasmuch as it had a constant flow of information regarding Borden costs in connection with the private label pricing (Kuhlman, Tr. 5231). Similarly Borden knew about A&P's retail prices since A&P generally supplied Borden with its retail prices (A&P Adm. dated March 12, 1973 II D , p. 2).

186. Complaint counsel also cite the fact that when the private label program began and thereafter, A&P sold private label milk and dairy products at the prevailing retail price level for vendor label milk and dairy products. Similarly, Borden failed to pass on at the wholesale level price reductions similar to tbe reductions granted A&P to other purchasers who compete with A&P in the Chicago area, at least for two or three years after the commencement of the private label agreement with A&P (Gose, Tr. 1095-1099).

187. Finally, complaint counsel cite the testimony of Dr. Kuhlman its expert economic witness, who found a combination between the respondents to stabilize the price of milk (Kuhlman, Tr. 522-5235). The evidence consisted of "mutual assurances " their post-contract behavior, their spreading of the savings across eleven different product lines and the fact that the agreement was terminable at will. II. Respondents ' Reply 188. It appears, however, that Borden s failure to offer the private label prices it granted to A&P to other customers immediately was was instructed by Bordenbecause its Chicago sales manager, Gose law department that Borden could not make such offers to others and stil rely on the meeting competition defense of tbe Robinson-Patman Act (Gose, Tr. 1076). Later, however, when it appeared that (73) Borden s competitors were selling at prices comparable to its private label prices to A&P, Borden solicited business at such prices (Dischner Tr. 7165).

189. Respondents also cite the fact that John White was sales manager of A&P' s Chicago unit at the time and that it was he who established the out-of-store prices without consultation with Schmidt Bartels or Smith, whose authority was limited to purchasing the products. White did not know of the Wisconsin experience nor had he had discussions with Minkler, Tarr or Malone and was not involved in the private label negotiations (White, Tr. 2058-2063). 190. With respect to the internal Borden letter of December 23 1964 (CX 7), Borden notes that Minkler spoke of the effect of an outof-store differential between private and brand labels but notes that it neither called for nor requested any action and that neither the letter nor its contents were disclosed to A&P. 191. With respect to the "sensitivity of the market" which was Initial Decision 87 F. brought up during the February 1965 meeting between Tarr and Sehmidt, Borden argues that it reflected Tarr fear that competitors might disregard overhead costs in bidding for A&P's business (Tarr Tr. 872). Minkler also mentioned this to A&P's Bartels later (':inkler Tr. 213-214).

192. With respect to the notation in Borden s May 1965 price quotation to the effect that "the primary basic price has recognized the depressed competitive situation prevailing" and " any economic improvement that occurs subsequently will be mutually recognized" Borden contends that this does not refer to prices in the retail market hut to the fact that in some areas the prices bid by Borden were not based on Borden s costs due to competitive wholesale situations then existing and would have to be adjusted later to a cost basis. Even if the so-called depressed competitive situation referred to prices in the retail market the critical question is whether the "mutual recognition" to be given to any economic improvement thereafter does not necessarily refer to A&P's out-of-store prices but to acceptance of a higher price from Borden by A&P based upon cost rather than competitive situations. In any event, Schmidt told Tarr thereafter that A&P did not want the depressed market in the outside areas to be given special consideration in the bids bd that al1 prices should be based on costs (RAPX 132A-C).

(74J 193. With reference to the May 26 1965 meeting between A&P and Borden officials where A&P expressed its hope for a saving in the amount of 500 thousand dollars using the private label agreement and it was stressed that that amount could result only from the continuance of present out-of-store margins on fluid milk products (CX 19F), Borden rejects complaint counsel's interpretation of this language as a "clear warning " that A&P should not reduce its retail price. Borden s Malone testified that the statement was made to disabuse A&P of any possible belief that Borden would guarantee those savings and support A&P in a price war (Malone DTR pp.217- 218). See also Smith' s testimony on rcjecting Borden s July bid: (VJe didn t have any protection in cae of a price war. If a price war broke out and we had to meet competitive retails for some period of time, that would be our expense instead of Borden s expense (Tr. 1361). 194. With respect to the spread of the savings over 11 different private label items, Borden notes that each item was ordered separately on a demand basis and that a single price for one year purchases of allll items could not therefore be quoted. 195. Borden also disputes complaint counsel's conclusions on the terminable-a1.-will provisions of the private label agreement, arguing GREAT ATLA1iTIC & PACIFIC TEA CO., INC. , ET AL. 1029 962 Initial Decision that such contracts are pro-competitive. It notes that the National Stores shifted from Dean Dairy to Hawthorn-Mel1ody, A&P from Bowman to Borden, Jewel from Borden to Dean, Hillman from Capitol to Wanzer, Wieboldt from Bowman to Wanzer, K-Mart from Beatrice to Borden and Eagle to Borden (Schaub, Tr. 3109; Soldwedel, Tr. 3392; Dischner, Tr. 7170-7175; Douglas, Tr. 3508-3510; Schmidt, Tr. 1765- 1766; Gose, Tr. 1102). It also cites Borden s attempt to secure Kroger milk business in 1965 , 1966 and 1968 as well as its proposal to supply Consolidated Food, K-Mart, White Hen, National and Eagle (Gose, Tr. 1089-1102, 1131, 1170; Malone DTR p. 224-226; Markham, Tr. 7041). Moreover, Borden notes the inconsistency in complaint counsel's argument that Borden was dependent on A&P's business for the efficient operation of its Woodstock plant and Borden s situation was so precarious that it was compelled to grant whatever concessions A&P demanded, while at the same time complaint counsel argue that Borden was independent enough to use the terminable-at-will provisions of the contract with A&P to terminate the contract immediately if A&P lowered its retail price on private label milk. (75) II I. Cond usion 196. It is well settled that al1 combinations having any tendency or effect to tamper with price structure are unlawful. Us. v. Container Cor. 393 U. S. 333 (1969); AlIYrecht v. Herald Co. 390 U. S. 145 (1968); US. v. Parke-Davis Co. 363 U.S. 29 (1960). The essence of a combination is the existence of concerted action between the parties US. v. Contairwr Cor. , supra. The test is not so much the manner in which the parties combine as that there be a "collaborative element present Pearl Brewing Co. v. Anheuser-Busch, Inc. 339 F.Supp. 945 951 (S.D. Texas 1972).

197. Complaint counsel cite the decision in Sunny Hill Farms Dairy Co. v. Krafto Corp. 381 F. Supp. 845 (E.D. :10. 1974), arguing that the facts there were remarkably similar to this situation. Complaint counsel argue that Sunny Hill was a case in which the supplier, Seal test, was asked to provide the buyer, :Ialone & Hyde with private label milk at a lower cost than Seal test sold its own brand label to it. Subsequently Seal test and Malone & Hydc engaged in discussions concerning price wars, depressed markets and limited retail price differentials; the court found an a!,rreement between the defendants to fix prices at both the vertical and horizontal levels. Complaint counsel cite the testimony of one of the witnesses to the effect that Seal test advised the buyer to guard against getting into a price war by lowering the price of milk. They also cite testimony in the Sunny Hill case to the effect that there was conversation between the Initial Decision 87 F. defendants about the out-of-store price. Counsel for A&P, however point out that by order dated July 19 1974 , the judge in the Sunny Hill case withdrew and vacated his memorandum and dismissed the complaint. K evertheless, complaint counsel urge the logic of that decision. A&P points out, however, that in Sunny Hill the court found that the defendants were independent and competing suppliers of milk to certain grocery stores which followed the recommendations Malone & Hyde. Thus, Malone & Hyde had certain limited control over the marketing programs of such stores. The court then found that the record is replete with evidence indicating that Sealtest and :Ialone & Hyde a,qreed upon the into-store prices of their competitive products and found ilegal horizontal price fixing ir such agreement between competing milk suppliers Seal test and Malonc & Hyde. As respects the fixing of out-of-store prices, (76) the court found that Section 1 of the Sherman Act was violated by a commitrnent to a common scheme or to some type of joint action ' , , stated in another fashion the quintessential yardstick for conduct deemed violative Section 1 is whether or not there is a collaborative clement present. " The court added:

An agreement to fix the maximum resale price is also proscribe because "such agreements, no less than those to fix minimum prices, cripple the freedom of traders and thereby restrain their ability to sell in accordance with their own judgment. ". . . the eclectic aiteria that can be gleaned from these cases for the ase-rtainment of a violation is whether the agreement or conduct interferes with the freedom of sellers or traders in such a manner as to prohibit or restrain their abi ity to seu in accordance with their judgrrnt and not what particular effect the agreement or conduct has on the actual prices (Emphasis added). Thc court then not€d that :.alone & Hyde sought a price differential in the two different labels of milk whereas Scaltest was desirous that the two labels be priced the same. The parties, however, finally agreed to a price differential. Thereafter, the court found that there was the necessary implementation to effectuate actual retail price maintenance by :.aone & Hyde s control over its affiliates saying there is direct evidence that Malo Hyde dictawd the out-oj-state pre to its affiliates. " The court found in Sunny Hill that there was " more than a mere non-biding resale price recommendation but a subtle mode of dictating resale price. 198. Thus, aside from the fact that the judge in Sunny Hill withdrew his memorandum and vacated it, the logic of his earlier position is apparently inapplicable here. Lnlike Sunny Hill there was no express or explicit abrrecment between A&P and Borden concerning A&P' s price differential for private label milk. The agreement, instead was limited to the Borden price to A&P for such private (77) label milk and the differential in such cost compared to the Borden label cost to A&P. Although there was some discussion of price both in this case as well as in Sunny Hill there is no reason to assume that the court in Sunny Hill would have reached the same conclusion absent a finding GREAT ATLANTIC & PACIFIC TEA CO., INC., ET AL. 1031 962 Initial Decision of an express agreement between the parties concerning a pnce differential on resale.

199. Here Borden as well as A&P had the same objectives in mind: The maintenance of the Borden label level of prices for the private label products. For Borden this was preferable because it avoided the likelihood of a price war in which retailers would seek financial help from Borden, their supplier. For A&P this was also preferable because it maximized their profits on dairy items by lowering the costs of some of them. Under such circumstances had the parties then agreed to the maintenance of such level of prices there would of course, have been a violation of law. There is, however, no evidence in this record which would suggest that Borden ever spoke to A&P about A&P' s resale prices on private label items except to note that a differential might create problems. A&P never gave Borden any assurances that it would not create such a differential. The "mutual assurances " on which complaint counsel rests Count III of the complaint are not found in this record. It must be borne in mind that A&P approached Borden about private label milk in an effort to maximize its profits in dairy products by securing the private label at a lower cost. It was only natural therefore, for A&P as well as for Borden to note the mathematical fact that if A&P lowered its price on private label milk it would not realize the increase in profit that it sought since the demand was inelastic and lowering the price would not increase sales (Markham, Tr. 7038). This realization constitutes no assurance from either party. Similarly, it was an open secret that Borden needed the A&P business or something its equivalent for the efficient operation of its new Woodstock plant. Consequently Borden would go to great lengths to avoid losing such a customer including a reduction in price to a levcl it felt its ccmpetitors might offer in an effort to lure the A&P business away from Borden that is, to a level which barely covered, if at all, the mere direct costs. Borden s failure to offer similar savings to other chain stores immediately thereafter has been explained and there appears nothing unreasonable in this point of view. In short (78) the behavior of the respondents in pricing the private label items was the behavior that one would expect of any rational businessman independently pursuing only his self-interest in markets such as those found in this case. The circumstantial evidence of agreement is equally consistent with a finding of independent action.

To sustain a finding of fact the circumstances proven must lead to the CGnc1usion with reasonable certainty and must be of such probative force a. to create the basis for a legal inference and not mere suspicion. Circumstantial €\idcnce, even in a civil cae, is not sufficient to establish a conclusion where the circumstances are merely consistent with . .

initial Decision 87 F.

such conclusion or where they give equal support to inconsistent conclusions (Pe1)ely Dairy Co. v. UniUid 8taxi. 178 F.2d 393 (9th Cir. 1949), cen. Mnwd 339 U. S. 942 (1950). See also V,:rginia- CaTulina Peanv.t Association 51 F. C. 1156 , 1188 (1955).

200. A&P's failure to pass on to the consumer its savings on private label milk are not necessarily indicative of any collusion. As complaint counsel's witness Kuhlman put it:

The non-collusive oligopolist in an industry having a relatively inelastic demand will be unlikely to pass the benefits of cost reduction programs on the consumer (Kuhlman, Tr. 5312).

201. I conclude, therefore, that complaint counsel have not met their burden of proof to show that the respondents herein have combined to effect a stabilization and maintenance of prices for milk and other dairy products in violation of Section 5 of the Federal Trade Commission Act.

D. THE DENIAL OF DUE PROCESS A&P's Contentions 202. A&P contends that it has been denied procedural due process in the presentation of its defenses and in its ability to cross-examine complaint counsel's witnesses due l79) to the loss of witnesses and documentary evidence that would have been available but for the Commission s delay in issuing the complaint. The Commission ruled on January 19, 1973 r82 F. C. 213 at 216J:

the time taken to conduct the investigation leading the issuance of the complaint herein is neither atypical nor inordinate, particularly in view of the complexity of the investigation and the Jimited manpower available to the Commission for conducting investigations jn connection with its varied and broad statutory responsibilities. Moreover, it seems clear from the pJeadinbTS and affidavits before us that respondents were tirneJy alerted to the fact that the Commission was interested in the private label arrangement between them in the Chicago trading area from the standpoint of possible violations of the amended Clayton Act and Section 5 of the Federal Trade Commission Act. And there is no indication that either A&P or Borden was ever advised that the Commission had abandoned its investigation. On the contrary, the papers before us clearly indicate that attempts to pursue and complete the investigation were made by Commission s staff on a continuing and regular basis during the period "March 1967 December 1969. The Commission went on, however, to note that the question of alleged unfair prejudice to A&P by possible denial of adequate opportunity to defend is one which cannot b€ answered now, but only at the conclusion of hearings.

GREAT ATLANTIC & PACIFIC TEA CO., INC., ET AL. 1033 962 Initial Decision The Unavailability of Documents 203. A&P contends that its defense in this case was hampered because its subpoenas proved unfruitful and that the lack of response thereto was occasioned by Joss or destruction of documents due to the passage of time. In this connection, it offered RAPX 135 and 136, being summaries of the subpoenas directed to retail stores and dairies respectively. In these exhibits there are repeated references by the deponents of the subpoenas involved that "no documents are available in response to (80) certain specifications of the subpoenas. It does not appear, however, that the unavailability of the documents was necessarily due to the passage of time. Thus, at page 728 of the transcript Mr. Dembrow, an attorney for A&P, was asked why documents not produced were " not available." The following testimony then took place:

THE WITNESS: No, I would have no other information as to why they did not supply us with the document.,;. All I know is we never received any documents that were responsive in whole or in part 1A the specifications in question. JUDGE HINKES: And that we are not tv assume from that typ of a response that the documents ever existed or were destroyed earlier or were gimpJy not found by the Respondent." or any other reason for their oem-response? THE WITNESS: Well, I think in general, Your Honor, that is correct .. . JUDGE HINKES: Alj I am saying is do we have any evidence-I don t want the assumptions now. Does this document contain reasons for the non responsc in evidence? THE WITNESS: No, the reason for non response is not indicated except in those few instances where I think Complaint counsel pointe to one that was not in the Chicago metropolitan area. Something Jike that or one that was not seiling to wholesale stores. JUDGE HINKES: I am only speaking now of the responses where nothing was said except "No documents available" and was signed by the dairy. (81 J THE WITNESS: That simply means that we physically .by " " I mean the respondents cannot receive documents or information in letter form that pertined tospecific items in our subponas. JUDGE HINKES: And this exhibit contains no explanation for that failure, is that correct? THE WITNESS: That would be correct, Your Honor. 204. In many instances the records sought were never kept or, if kept, were only so maintained for a brief period of time and then discarded. See, for example, the affidavit of Bruce Spear, an attorney for A&P, constituting an attachment to A&P' s motion of October 13 1972, to dismiss the complaint. In it, Mr. Spear relates his interviews with certain retail store operators. Some of them spoke of discarding records after maintaining them for only two to six months. Others discarded records after their replacement by more current schedules. 2JU- "8 O- LT - 77 - 66 . . . .

Initial Decision 87 F.

Still others discarded them after completion of the tax audit. Another said that he doubted that records were kept other than on a current basis.

205. It is thus dear that even had the complaint been issued witbin a reasonably short period of time following the commencement of the investigation such records would have hecn unavailable even then and their unavailability to A&P when the complaint was issued years later cannot be the result of the passage of such time. The Unavailability of Witnesses 206. A&P also contends that as a result of the delay by the Commission in instituting this proceeding the testimony of certain individuals who died in the interim became unavailable. Thus, Joseph Malone, Borden s vice president, regulatory controls, and formerly comptroller, died prior to the start of hearings in this matter. His deposition, however, had been taken by complaint counsel and it was received in evidence (CX 262). A&P contends, however, that it had not yet begun its program of discovery when the deposition was taken nor had it begun its cost study of the private label (82) prices received by A&P. Consequently, "detailed probing examination of Malone which was essential to the proper resolution of the various issues surrounding Borden s cost of serving A&P and the cost of serving its non-A&P customers was not possible, 207. Complaint counsel filed their application for taking the deposition of Mr. Malone on November 4, 1971. Said application was granted by the administrative law judge by order dated November 8 1971 , in which the judge noted:

The application states the complaint tD involve a private labe! milk arrangement between respondenl A&P and Borden beginnjng in 1965 and continuing to date. The negotiations giving rise to the agreement are stated to extend back to 1964 with the three principal negotiators for Borden during 1964 and 1965 to be now retired from said company. The three negotiators are described as now being elderly and their testimony to be of importance to proof of the charges of the complaint. (One of the three negotiators was identified as Joseph Malone. The order went on to describe the nature of the expected testimony considerable detail, specifically referring to "Borden s relationships and dealings with A&P and competing retail grocery purchasers in the Chicago area * * * . The witnesses' own roles in the dealings with A&P concerning the Chicago area private label arrangement* * *the pricing and sales practices of Borden in connection with its sales to A&P and competing purchasers in the Chicago area, etc." Pursuant to said order Malone first took the stand on December 3 1971. Counsel for each of the parties attended said appearance but due to a dispute GREAT ATLANTIC & PACIFIC TEA CO., INC. , ET AL. 1035 962 Initial Decision regarding Malone s request for immunity his testimony was then postponed until February 9, 1972. At that time Malone was examined by complaint counsel and cross-examined by counsel for both Borden and A&P.

208. Considering that A&P was on notice of Malone s testimony as early as November 8 , 1971, attendee! Malone s first appearance on December 3 1971, as well as his actual testimony on February 9 and 10 1972, at which time it participated in the examination of the witness; considering (83) also that A&P under the Commission s Rules was also entitled to take the deposition of Mr. Malone at any time either for discovery purposes or to preserve evidence but failed to do so; and considering also that A&P was able to use Mr. Daube who worked for Borden in the Chicago central district from 1958 until 1966 as chief accountant and assistant district controller and then became assistant to the district controller when the Chicago central district was merged with the mid west district in Columbus, Ohio (Daube, Tr. 5864-5870), I conclude that A&P' s failure to examine Mr. Malone prior to his death attributable to its neglect or failure to take advantage of the opportunities and rights it had as a party to this proceeding and in any event, was not prevented by any delay in the institution of this proceeding.

209. A&P also claims prejudice to its case resulting from the unavailability of David Parmalee, Bowman s senior cost accountant who died prior to the start of hearings. It appears, however, that A&P interviewed Mr. Parmalee after the complaint was filed and it is only reasonable to assume that if Mr. Parmalee testimony would be helpful, A&P would have followed complaint counsel' s example with Mr. Malone and taken his deposition as well (affidavit of Denis McInerney, October 13 , 1972, page 13). Moreover, A&P admits that Parmalee s testimony would have merely supplemented Cannon testimony who did appear and testify (A&P ,Proposed Findings, Volume 1 , p. 508). Here, too, therefore, A&P's failure to have Parmalee s testimony is due to its own oversight rather than any delay in the institution of this proceeding.

210. A&P also cites the unavailability of the testimony of L. Edward Hart, an attorney who worked very closely with Bowman management in connection with a court order prohibiting discriminatory prices by Bowman. Mr. Hart died prior to the start of the hearings but it is doubtful that his testimony would be received in evidence inasmuch as a claim of privilege in connection with his work as an attorney for Bowman could be expected. Moreover, the testimony of Mr. Hart, like that of Mr. Parmalee, would have been only supplemen- Initial Decision 87 F.

tary to that of Mr. Cannon who did testify (A&P Proposed Findings, Vol. 1, p. 508).

211. A&P notes that testimony of two other witnesses was unavailable due to their deaths prior to the start of hearings. One of these was Walter Roney, Chief Executive Officer for High-Low Foods. His death, however, occurred in 1969 (A&P Proposed Findings, V 01. 1 p. 501). The (84) unavailabiliy of his testimony, therefore, can hardly be attributed to any unreasonable delay in the institution of this proceeding. The other witness referred to by A&P is Fred Nonnamak- , a former Bowman Company employee who died prior to the filing of the complaint. It was his death that prompted complaint counsel to seek the depositions of Messrs. Malone, Minkler and Tarr on November , 1971. Presumably, complaint counsel was seeking testimony which could be used against A&P and this could explain A&P' s failure to seek Nonnamaker s deposition any earlier. In any event, his unavailability is a Joss to an of the parties in this proceeding. A&P has failed to demonstrate any particular prejudice to it as a result of his death prior to the start of hearings.

212. The record in this proceeding makes it clear that both A&P and Borden were aware, even before service began under their contract, that their private label arrangement would probably result in litigation (Smith, Tr. 1429). Certainly, the Commission s investigation of this case was brought to the attention of Borden and A&P as early as February 1967 when an attorney-examiner for the Commission contacted both respondents requesting certain data and supporting interviews with company personncl concerning the private label agreement between the respondent and possible violations of the Clayton Act and the Federal Trade Commission Act. An of this has already been caned to the attention of the Commission in connection with A&P' s motion to dismiss the complaint filed on October 13, 1972 and complaint counsel' s answer thereto filed on October 20, 1972. If indeed, A&P failed to prepare itself for its defense to the complaint that was issued later, its failure must be deemed the result of its own oversight and neglect, having been repeatedly put on notice and baving obviously realized even before the commencement of the private label al,'1eement that some litigation was likely in connection therewith. The Commission took steps to prepare for such litigation although it, too, experienced the loss of some witnesses and data. For example, potential witness Nonnamaker died before his deposition could be taken. He had testified in the Dean-Bowman case, Dkt. 8674 to the effect that Bowman s relationship with A&P was such that it could not expect to obtain any A&P business under any circumstances (Dkt. 8674, Tr. 2207-2208). ("Bowman couldn t get (the A&P account) . . ..

GREAT ATLANTIC & PACIFIC TEA CO., INC., ET AL. 1037 962 Initial Decision if the Lord Himself were their solicitor ). Such testimony in this proceeding would have demonstrated that the Bowman offer to A&P was a sham and could not be used legitimately by A&P for a meeting competition defense.

(85) 213. I conclude, therefore, that there has not been an unjustifiable delay in the filing of the complaint considering thc difficulty the Commission expericnced in marsha11ing its evidence from the respondents. As the Commission noted in its order of January 19 1973 (supra), the time taken to conduct the investigation Jeading to the issuance of the complaint herein is neither atypical nor inordinate, particularly in view of the complexity of the investigation and the limited manpower available to the Commission.

In addition, any prejudice experienced by the respondents herein by the passage of time between the commencement of the investigation and the issuance of the complaint herein, appears to be minimal and the result of the respondents' indifference to the likelihood of litigation resulting from their private label agreement which they themselves anticipated and 'which was brought to their attention by the activities of the Commission as carly as February 1967. CONCLUSIONS OF LAW 1. The Federal Trade Commission has jurisdiction of the subject matter of this procecding and of respondents The Great Atlantic & Pacific Tea Company, Inc., (hereinafter A&P) and Borden, Inc. (hereinafter Borden).

2. The complaint herein states a cause of action and this proceeding is in the public interest.

3. Respondent A&P in the course and conduct of its business in commerce has acted unfairly and deceptively in purchasing milk and dairy products from Borden in that it accepted a price offer from Borden tendered on a meeting competition basis when in fact such meeting competition defense was not available and without informing Borden of this fact in violation of the policy of Section 2 of the amended Clayton Act and in violation of Section 5 of the Federal Trade Commission Act.

4. Respondent A&P in the course and conduct of its business in commerce has knowingly induced or received a discrimination in price in the purchase of fluid milk and other dairy products in violation of subsection (f) of Section 2 of the Clayton Act as amended by the Robinson-Patman Act.

(86) 5. Complaint counsel have not satisfied their burden of proof to Initial Decision 87 F.

show that respondents A&P and Borden in the course and conduct of their business in commerce entered into a combination which had a tendency or effect of stabilizing and maintaining prices for milk and other dairy products in violation of Section 5 of the Federal Trade Commission Act.

THE REMEDY As a result of the failure of proof in connection with Count III of the complaint, a dismissal of said Count will be ordered. As respects Count I of the complaint, complaint counsel propose that A&P be directed to inform prospective suppliers who submit prices a meeting competition basis if the price offered beats existing competitive offers. It is contended that such requirement does no more than obligate A&P to comply with its statutory duties. The order proposed however, would cover not only milk and other dairy products but any product. Moreover, if A&P informs a supplier that a competitive offer has been made to A&P and a supplier makes an offer based on the meeting of the competitive offer, A&P would be required under the proposed order to notify the supplier that the price being offered is below the existing competitive offer to A&P if such be the case.

With respect to Count II of the complaint, complaint counsel's proposed order would direct A&P not to violate Section 2(f) of the Robinson-Patman Act again in any product Jine rather than merely in milk and other dairy products. Moreover, the proposed order would prohibit A&P's inducement, receipt or/and acceptance of a net price which it knows or should know is below the net price at which said products of like grade and quality are being sold by such seller to other purchasers competing with A&P unless A&P slwws that it knew or had reason to believe that the price difference in its favor was cost justified on the basis of differing methods or quantities or said price was granted to meet a lawful equally low price of a competitor. With respect to the proposed coverage of a11 products in both Count I and Count II of the complaint, complaint counsel contend that A&P' violation was not an isolated incident relating to milk. Rather it was an action in the context of A&P' s entire business; " and that without such coverage "A&P could easily evade the order simply by turning its attention to other areas in its diverse product line where similar anticompetitive practices would be equally fruitful" (87) It is well settled that the Federal Trade Commission has wide discretion in its choice of the remedy adequate to cope with the unlawful practices disclosed. FTC v. Ruberoi Co. 343 U.S. 470 (1952). The validity of the Commission order is gauged by its reasonable GREAT ATLANTIC & PACIFIC TEA CO. , INC., ET AL. 1039 962 Initial Decision relation to the unlawful practices found to exist. FTC v. Colgate- Palrrlive 380 U. S. 374 (1965); National Lead Co. v. FTC 352 U.S. 419 (1957). A&P points out, however, that the Commission itself has recognized situations where a broad order would not be appropriate Oats Co. 60 F. C. 798 (1962). In that case the order wasQuker limited to "cat food and products related to cat food. " Similarly, in National Dairy Prodw;ts Cor. v. Fedeml Trade Commission 412 F. 605 (7th Cir. 1969) the order was Jimited to the product line of the single division found to have made the unlawful discriminations. See also American Horn.€ Prodw;ts Cor. v. Federal Trade Commission, 402 2d 232 (6th Cir. 1968).

It is particularly noteworthy that complaint counsel have drawn a parallel between this case and the Kroger case, 76 F. C. 719 (1969). There, as here, the buyer went beyond the bounds of permissible bargaining by misquoting the competitive offer or failing to convey correct information. The order issued in the Kroger- case was limited to milk and other dairy products despite the fact that Kroger, a national grocery chain, obviously had other products which could be subject to the same abuse. In that case, however, the order was not Jimited to the Charleston division of Kroger where the discriminations took place but to "outlets operated by respondent. " Consequently, the order herein will be limited to fluid milk and other dairy products but wil apply to A&P's outlets wherever operated by A&P.

As respects Count II, A&P argues that it cannot comply with the order proposed by complaint counsel because of the burden on A&P. It contends that it would be enjoined from receiving a net price on any product which it knows or should know is below the net price at which the product is being offered to any A&P competitor unless A&P is able to demonstrate that it knew or had reason to believe that the lower price to A&P was cost justified or that the lower price was given to A&P to meet a lawful, equally low price offered to A&P by another supplier. It argues that this would require A&P to cost justify its suppliers prices (88) a burden which cannot be properly placed on a buyer who cannot obtain the costing information required (Automatic Canteen Co. v. FTC, 346 U. S. 61 (1953)). A&P, however confuses the burden of proof requirement with the obligation of law. A&P cannot induce or accept a discriminatory price from a supplier where it knows or has reason to know that the lower price to it cannot be cost justified or that the lower price given to it was not to meet a lawful equally low price offered by another supplier. The burden of proof, however, remains with complaint counsel under the doctrine of Automatic Canteen. This is not unlike the situation in Count I where A&P is forbidden to engage in misreprescntations but Initial Decision 87 F.

the burden of proving the misrepresentations still rests upon complaint counsel. A&P's objection to the proposed order with respect to Count , therefore, is reasonable only to the extent that it suggests shifting the burden of proof to A&P. The order herein corrects that error. A&P also objects to the proposed order arguing that other statutory defenses available to it are excluded by the language of the order since it does not mention such defenses as Jack of knowledge of the discrimination or lack of injury to competition. This argument was also raised in the Ruberoid case supra; (The absence of such provisos J cannot preclude the seiler from differentiating in price in a new competitive situation involving different circumstances where it can justify the discrimination in accordance with the statutory provisos. In that case the order simply prohibited Ruberoid from discriminating in price by selling its products to any purchaser at prices lower than those granted other purchasers who compete with the favored purchaser in the resale of the products. The court found it unnecessary to specify the statutory defenses. Here, too, unmentioned statutory defenses available to A&P have not been negated nor lost to A&P by reason of their omission. (89 J ORDER It is ordered That respondent, The Great Atlantic & Pacific Tea Company, Inc., a corporation, and its officers, representatives, agents and employees, directly or through the use of any corporate or other device in or in connection with the offering to purchase or purchase of fluid milk and other dairy products in commerce as "commerce " is defined in the amended Clayton Act, for resale in outlets operated by the respondent, do forthwith cease and desist from: (a) misleading any supplier of fluid milk or other dairy product by misrepresenting a material fact which could reasonably be expected to affect thc suppliers' price to A&P; and (b) directly or indirectly inducing, receiving or accepting from any seller a net price it knows or should know is below the net price at which said products of like grade and quality are being sold by such seller to other purchasers where A&P is competing with the purchaser paying the higher price, unless A&P knew or had reason to believe that the price difference in its favor made only due allowance for cost differences resulting from differing methods or quantities in which such products are sold (90) or delivered to such purchasers or said price was granted to it by the supplier to meet a lawful equally low price of a competitor. For the purpose of determining the "net price" under thc GREAT ATLANTIC & PACIFIC TEA CO., INC., ET AL. 1041 962 Initial Decision terms of this order there shan be taken into account an discounts or other terms and conditions of sale by which net prices are effected. It is furtlwr ordered That A&P shan forthwith distribute a copy of this order to each of its operating divisions. It is furtlwr ordered That A&P shan notify the Commission at least thirty days prior to any proposed change in the corporate respondent such as dissolution, assignment or sale resulting in the emergence of a successor corporation, the creation or dissolution of subsidiaries or any other change in the corporation, which may affect its compliance obligations arising out of the order.

It is furtlwr ordered That A&P shan, within sixty (60) days after this order becomes final, file \\ th the Commission a report in writing form in which it has compliedsetting forth in detail the manner and and win comply with this order.

It is furtlwr ordered That Count III of this complaint be, and the same hereby is, dismissed.

Initial Decision 87 F. Appendix PRICES PAID FOR MILK BY A&P COMPARED TO LOWEST PRICE A V AILABLE TO ITS COMPETITORS IN THE CHICAGO & SUBURBS AREA" 11/65-12/68 HOMO - HALF-GALLONS - PAPER Non-A&P Stores A&P Stores Maximum Borden Discount Prvate Discount Lit Minimum From Li Label From List Percent Date Prce Prce Prce Prjce Prce Overchage 11/65- $0.49 $0.3479 29. $0. 3124 36. 11. 1/66 1/66- 3430 30. 3124 36. 2/66 2/66- 357 30. 3254 36. 5/66 5/66- 364 30. 3254 37.4% 11.9% B/66 8/66- 378 30. 3376 37. 12. 9/66 9/66- 392 30. 3514 37. 11.6% 10/66 10/66- 392 30. 3534 36. 10. 11/66 11/66- 392 30. 3514 37. 11.6% 3/67 3/67- 385 30. 3376 38. 14. 4/67 4/67- 385 30. 3416 37. 12. 5/67 5/67 399 30. 3502 38. 13. 5/67- 399 30. 3534 38. 12. 9/67 9/67- .413 30. 3664 37. 12. 5/68 5/68- 427 30. 3762 38. 13. 7/68 7/68- 434 30. 3828 38. 13. 12/68 SOUIce: CX 137, 139-150.

Sources: ex 120, 138; Gase Tr. 11l0.

Source: CX 255B.

Calculated by dividing the difference between the minmum price paid by others and the private label price by the private label price paid by A&P. GREAT ATLA.1\TIC & PACIFIC TEA CO., INC., ET AL. 104 962 Initial Decision Appendi HOMO - GALLONS - PAPER Non-A&P Stores A&P Stores Maximum Borden Disount Prvate Diount Lit Minimum From Li Label From List Percent Date Prce Prce Prce Prce Prce Overcharge 11/65- $0. 50.6958 29. 0'/0 50.6248 36. 11.4% 1/66 1/66- 686 30. 0'/c 6248 36. 2/66 2/66- 1.02 714 30. 6508 36. 5/66 5/66. 1.04 728 30. 6508 37.4% 11. 8(66 8/66- 756 30. 0'/0 6752 37. 12.0% 9/66 9(66- 1.2 784 30. 0'/, 7028 37. 11. 10(66 10/66- 1.2 784 30. 7068 36. 10. II/66 11/66- 1.2 784 30. 7028 37. 11. 3/6 3/67- 1.0 770 30. 0'/, 6752 38. 14, 4/67 4/67- 1.0 770 30. 6832 37. 12. 5/67 5/67 I.4 798 30. 7004 38. 13. 5/67- I.4 798 30. 7068 38. 12. 9/67 9/67. 1.8 826 30. 00/0 7328 37. 12. 5/68 5J6B- 1.22 854 30. 7524 38. 30/0 13. 7(68 7/6B- 1.24 868 30. 0'/, 7656 38.3% 13. 12/68 Initial Decision 87 F.

Appendix HOMO - QUARTS - PAPER Non-A&P Stores A&P Stores Maximum Borden Discount Prvate Disount List 1\finimum From List Label From List Percent Date Prce Prce Prce Prce Prce Overcharge 1I/65- $0. 265 80.1882 29. 80. 1712 35. 1/66 1/66- 265 1855 30. 1712 35.4% 8.4% 2/66 2/66- 275 1925 30. 1777 35A% 8.3% 5/66 5/66- 1960 30. 1777 36. 10. 8/66 8/66- 2030 30. 1838 36. 10. 9/66 9/66- 2100 30. 1907 36. 10. 10/66 10/66. 2100 30. 1917 36. 11/66 11/66- 2100 '.0. 1907 36. 10. 3/67 3/67- 295 2065 30. 1838 37. 12. 4/67 4/67- 295 2065 30. 1858 37. 11.1% 5/67 5/67 305 2135 30. 1901 37. 12. 5/67. 305 2135 30. 0"/0 1917 37. 11.4% 9/67 9/67- 315 2205 30. 1982 37. 11.3% 4/68 4/68- 325 2275 30. 2031 37. 50/c 12. 7/68 7/6fJ 2310 30. 2064 37. 11.9% 12/68 GREAT ATLANTIC & PACIFIC TEA CO., INC., ET AL. 104 962 Initial Decision Appendix 2% MILK - 1/2 GALLON - PAPER Non-A&P Stores A&P Stores Maximum Borden Discount Prvate Discount Ust Minimum From Ust Label From List Percent Date Prce Prce Price Prce Prce Overcharge II/65- $0. SO. 3266 29. 0'/, SO. 2914 36. 12. 1/66 1/66- 3220 30. 0'/0 2914 36. 10. 2/66 2/66- 3360 30. 3044 36. 10. 5/66 5/66- .19 3430 30. 0'1, 3044 37. 12. 8/66 8/66- 3570 30. 3166 37. 12. 9/66 9/66. 3710 30. 3304 37. 12. 10/66 10/66- 37JO 30. 0'/0 3324 37. 11. II/66 II/66- 3710 30, Orc 3304 37. 12. 3/67 3/67. 3640 30. 0'/, 3166 39. 15. 4/67 4/67- 3640 30. 3206 38. 13. 5/67 5/67 3780 30. 3292 39. 14. 5/67. 3780 30. 3324 38.1% 13. 9/67 9/67- 3920 30. 00; 3454 38. 13. 5/68 5/68- 4060 30.0', 3552 38. 14. 7/68 7/68. .4130 30. 3610 38. 14. 12/68 Initial Decision 87 F.

Appendi SKIM MILK - 1/2 GALLON - PAPER Non-A&P Stores A&P Stores Maximum Borden Discount Prvate Dicount Lit IVlinimum From List Label From List Percent Date Prce Prce Prce Prce Prce Overchage 11/65- S0.46 50. 3266 29. 0'/0 50.2834 3804% 15. 1/66 1/66- 3220 30. 2834 38. 13. 2/66 2/66- 3360 30. 2964 38. 13. 5/66 5/66- 3430 30. 2964 39. 15. 8/66 8166- 3570 30. 3086 39. 15. 9/66 9/66. 3710 30. 3224 39. 15. 10/66 10/66- 3710 30. 3244 38. 14. 11166 11/66- 3710 30. 3224 39. 15. 3/67 3/67- 3640 30. 3086 40. 18.0% 4/67 4/67- 3640 30. 3126 39. 16. 5/67 3780 30. 0'/, 3212 10. 17. 5167. 3780 30. 3244 39. 16. 9/67 9167- 3920 30. 3374 39. 16. 3/68 3168 3920 30. 3000 46. 30. 3/68- 3920 30. 3374 39. 16. 5/68 5168- 4060 30. 3472 40. 16. 7/68 7/68- 4130 30. 3538 16. 12/68 GREAT ATLANTIC & PACIFIC TEA CO., INC. , ET AL. 1047 962 Opinion QpINIOt- OF THE COM.\1ISSlOJ' By NYE Commissioner:

(1) The complaint in this matter was issued on October 8, 1971. Count I of the complaint charged that The Great Atlantic & Pacific Tea Company, Inc. (hereinafter "A&P") violated Section 5 of the Federal Trade Commission Act (2) (hereinafter the " C. Act; " 15 C. ~45) as well as the policy of Section 2 of the amended Clayton Act (hereinafter the "Robinson-Patman Act" ; 15 1.S.C. S13) by misleading Borden, Inc. (hereinafter "Borden ) during negotiations concerning the sale by Borden of milk and other dairy products to A&P. The gravamen of the charge is that A&P failed to inform Borden that Borden offer, purportedly made to meet a competitive bid received by A&P from another potential supplier was in fact substantially lower than that other bid. Count II charged that A&P violated Section 2(f) of the Robinson-Patman Act (15 U. C. ~13(f)) by knowingly inducing or receiving discriminations in price from Borden which discriminations are prohibited by Section 2(a) of that Act. Finally, Count III charged that both A&P and Borden violated Section 5 of thc F. C. Act by combining to stabilize and to maintain the retail and wholesale prices of milk and other dairy products. Hearings were held before Administrative Law Judge Harry R. Hinkes, and an initial decision was rendered on September 24 , 1975. Judge Hinkes found that A&P had violated the law as charged in Counts I and II but dismissed Count III. Judgment was entered accordingly.

A&P has appealed from the decision, asserting that it did not violate the law as charged in Counts I and II. Complaint counsel has appealed from the decision, arguing that A&P and Borden violated the law as charged in Count III. The Commission heard oral argument.s on these appeals on February 25, 1976.

(3) The decision of Judge Hinkes is affirmed as to Counts II and III but reversed as to Count I. Except as qualified or altered by this opinion, we adopt his findings and conclusions. We have substituted a new order.

1 For convenicnr.€, the following abbreviation are u5€ in thi opinion: LD. Initial deci ion of a.rnini trative law jud!!", (Cite by paragph number) Tr. - Transcript of timony (Cite by page number). CX- Commission exhibit.

RX-- Respondent A&P exhibil Cit.tioll to "Appeal AJl wcr " and "Reply" refer the brief fied in reference to C.QunUJ I and II. Citation to " Appea!" ref..,. to complaint could,' s appeal brief filed in reference to Count III , 104 FEDERAL TRADE COM'IISSION DECISIONS Opinion 87 F.

I. FACTUAL BACKGROCND Although the record in this case is voluminous, exceeding 20 000 pages, the facts are generally not disputed. All of the violations alleged in this action concern the events surrounding a November 1, 1965 agreement between A&P and Borden under which Borden began supplying private label milk and other dairy products to more than 200 A&P stores in an area that included portions of Ilinois and Indiana. From the record, it appears that in November of 1964, Herschel Smith, A&P's headquarters dairy buyer in New York, advised A&P' divisional purchasing directors of the savings that could be realized by switching from advertised brand to private label milk (Tr. 1350- 1351 1862- 1863). Soon thereafter, A&P asked Borden to submit an offer to supply, under private label, certain of A&P' s milk and other dairy product requirements in the Chicago area, where Borden had served A&P since the late 1950's. From then until August 13, 1965, A&P and Borden engaged in extensivc negotiations which culminated in Borden s offer to grant A&P a discount for switching to private label provided that A&P would accept limited delivery service (CX 36, 37). Borden claimed that this proposal would reduce A&P's costs by $410 000 per annum compared to what A&P had been paying (CX 37; Tr. 869-871). A&P decided that it was not satisfied and solicited offers from other dairies (Tr. 871; Bee Tr. 1361-1362). (4) Thereafter, on August 31 , 1965, A&P received an offer from Bowman Dairy (CX 50) that was lower than Borden s August 13 offer (See CX 65A; A&P' Proposed Findings at 121-122). 2 On or about September 1 , 1965, Elmer Schmidt, A&P's Chicago unit buyer telephoned Gordon Tarr, Borden s Chicago chain store sales manager and stated I have a bid in my pocket. You lBorden J people are so far out of line it is not even funny. You are not even in the ball park." (Tr. 873). Although Tarr asked Schmidt for some details, Schmidt said that he could not tell Tarr anything except that a $50 000 improvement in Borden s bid "would not be a drop in the pocket." (Tr. 874-875; see 53A-B). Contrary to its usual practice, A&P then offered Borden the opportunity to submit another bid (Tr. 1248-1249, 1892-1893). In consequence, Borden offered to submit a new bid which doubled the savings to A&P from $410 000 to $820 000 (Tr. 227-228). In prescnting the offer, Borden emphasized to A&P that Borden needed to keep A&P' s business and was making the new offer solely to meet the described competitive bid (Tr. 22). Of course, as noted, Borden knew very little about the competitive bid that it was purporting to . Contrary to Judge Hinke. ' conclusion (1.0. 103 , 106, 114), we find that theBowmanbidwMo rative and could beeompar to the Borden offer.

GREAT ATLANTIC & PACIFIC TEA CO. , INC., ET AL. 1049 962 Opinion meet. Later, Borden submitted a detailed offer (CX 56), which it slightly revised to meet some of A&P's specific demands (CX 62).:1 (5) A&P concluded at the time that Borden s offer was "substantially better" in price' than Bowman s (Tr. 1413-1414, 189&-1997; CX 263), and therefore, in October of 1965, A&P accepted it (Tr. 1413; CX 71). The offer was, in fact, much better in price than Bowman (See 263B; compare CX 75F-J with CX 50).

A&P' s Schmidt requested a letter from Borden which would state that the prices Borden had quoted were proportionately available to competitors of A&P (Tr. 1783, 1786). Borden supplied a letter stating simply that the prices were legal and that Borden would defend them (RX 2). In fact, officials of Borden thought that the other bids were based on "anyhow accounting including only out-of-pocket or direct costs- and informed A&P of their belief (CX 63; see Tr. 213-214, 872- 873, 955).

II. COUT 1: FAILI:\G TO INFORM Count 1 charged that A&P violated Section 5 of the F. C. Act as well as the policy of Section 2 of the Robinson-Patman Act by misleading Borden during negotiations concerning the sale by Borden of milk and other dairy products to A&P. The thrust of the charge is that A&P failed to inform Borden that Borden s offer, made in an attempt to meet a competitive bid received by A&P from another potential supplier was substantially lower than that of the only other bidder. The count is directed to the question of what must legally be disclosed during contract negotiations.

We briefly state the relevant facts. A&P compared the Bowman bid with Borden s and concluded that Borden s bid was (6) substantially lower. Also, A&P knew that Borden was rely ng on the meeting competition defense (Tr. 248 , 886, 1789, 1792, 1801). Finally, A&P accepted Borden s bid without telling Borden that Borden substantially "beat" Bowman s bid as it clearly had (Tr. 1413; CX 71 , 263; see Answer at 42). Because A&P did not then inform Borden that the meeting competition defense was unavailable, Judge Hinkes concluded that A&P engaged in an "unfair" practice that violated Section 5 of the F. C. Act (l.D. 115 , 116).

At the outset we emphasize our belief that Section 5 of the F. 3 AfWr 3submission of the bid, Borden excluded g-Ia& gallon hcu9€ A&P t.te that "it jg 110t fair to the other bidders," (Tr. H.). In fact, however, the Bowman bid had includ d a quote on gla" gallons (CX 50). A&P also told Borden sharpn your pencil a little bit b"wu you are not quite there " meaning, acrding to Borren s Tai, that Bon:en har not yet met the competitive bid (Tr. 8S) . A&P was only intere te in obtaining the lowest pric€ possible (CX , 263B; Tr. 14014Dl, 17:?1729) , Although A&P made certir. misreprrscntations 8te note 3, 1m'J, such misrepresentations ar not relevant thechB.e 2J 6- LT - 77 - 67 , Opinion 87 F.

Act can reach actions that violate the policy of the Robinson-Patman Act cf. PTC v. Sperry Hutchinson Co. 405 U.S. 233, 24 (1972), and further, that Section 5 should be used to fill the gap in coverage caused by the manner in which Congress added the buyer liability provision contained in Section 2(f) of the Robinson-Patman Act. It is well established that Section 5 can reach buyers who induce violations of the promotional allowance and payment for services provisions of the Robinson-Patman Act. Fred Meyer, Inc. v. FTC 359 (7) F.2d 351 367 (9th Cir. 1966), rev d on otlwr grounds 390 U. S. 341 (1968); Gint Foo Inc. v. FTC 307 F.2d 184, 186 (D. C. Cir. 1962), cert. denied 372 U. S. 910 (1963); Grand Union Co. v. FTC 300 F.2d 92, 98-99 (2d Cir. 1962). According to the Court in Grand Union the application of Section 5 to buyers who violate specific provisions of the Robinson-Patman Act neither offends Congressional policies nor circumvents the criteria of illegality prescribed by the Robinson- Patman Act. Id. at 98. In addition, it is appropriate to use Section 5 to fence in activities which if permitted to mature, could violate the antitrust laws, including the Robinson-Patman Act, or which otherwise violate the policy of the antitrust laws. Id. at 99.

Having recognized this, however, and having measured A&P' conduct against these standards, we do not think that a violation of Section 5 has occurred here. The imposition of a duty of affirmative disclosure, applicable to a buyer whenever a seller states that his offer is intended to meet competition, is contrary to normal business practice and, we think, contrary to the public interest. As the Court in Forster Mfg. Co. v. FTC 335 F.2d 47 , 56 (1st Cir. 1964), cert. denied 380 U. 906 (1965), stated in describing normal practice: The scllcr wants the highest price he can get and the buyer wants to buy a.q cheaply as he can, and to achieve their antagonistic ends neither expets the other, or can be expeted, to lay all his cards face up on the table. Battle of wits is the rule. Haggling has ever been the way of the market pace.

(8) We recognize the need to curb undue pressure on sellers by powerful buyers such as A&P but do not think that changing the rules of commercial bargaining in this way is the answer. We are fearful that such a change would harm the freedom of buyers to engage in aggressive bargaining over price and would thereby affect competitive distribution. As the Attorney General's National Committee to Study the Antitrust Laws stated Legalistic impediments to the normal Although IWbiTlson-Patman Act prohibitions wc!' preduminant:y directe at discrimina.torycon ioru c.rcdb;,' poweriulhuye!" thelcgislativep=relegateSetion2if)toaminorrole Section2(f)first emerged a. an amendment offered in debate on the fJOor of the Senate. . . and repres€nt. a lllt minute afterthought addition" Att n.'8 !"at' Q,mm to Study tn", Antitrust Laws, Report 19:1 (1955) (hereinafter Report . .

GREAT ATLANTIC & PACIFIC TEA CO., INC. , ET AL. 1051 962 Opinion bargaining process * * * might well deprive the public of gains that under effective competition it has a right to expect." Report supra note 6, at 196. We fear a scenario where the seller automatically attaches a meeting competition caveat to every bid. The buyer would then state whether such bid meets beats, or loses to another bid. The seller would then submit a second, a third, and perhaps a fourth bid until finally he is able to ascertain his competitor s bid. We do not believe that the failure of buyers to engage in such disclosure should violate the F. C. Act. Therefore, Count I should be dismissed. (9) III. COCN'I II: KNOWINGLY INDUCI G OR RECEIVING Discriminations I:' PRICE Count II charged that A&P violated Section 2(f) of the Robinson- Patman Act by knowingly inducing or receiving discriminations in the price of milk and other dairy products from Borden, which discriminations are prohibited by Section 2(a) of that Act. Section 2(f) was enacted to make the prohibitions of price discrimination applicable to buyers. Automatic Canteen Co. v. FTC 346 U. S. 61, 70 (1953). In order to (10) prove a violation of Section 2(f), complaint counsel must not only establish all elements of a Section 2(a) violation, but must also meet special requirements with respect to showing the respondent' s "knowledge." Further, the buyer is entitled to raise any defenses that the seller might have raised, specifically those listed in Sections 2(a) and 2(b).

A&P admits that the A&P-Borden transaction contained many of the elements of a Section 2(a) offense. It argues, however, that the sales in Ilinois were not "in commerce" and that A&P did not receive a 7 We IIrc not concluding that !1ondioclooure during commercial bargaining never ri to II Section 5 offens. We simply hold that it should not be A&P' s responsibility to Uoll Borden whether Illegal defense is available. As disu!! lit p. 20 n. 19, II !!ller a r-oo faith meeting competition dcfcnJ! under the lWbinsn-Patman Act is dependent on imme factors that are more reily available tohim than to the buyer. For example, to have Ii meeting competition defense the seller who bR. knowingly discriminate in price muat show that he acte prudently in surmsing whether hill bid, in fact, met the equally low price of II competiwr.FTC Y. A. E. Sw.l.y Mfg Co.,324 U, S. 746, 759- (194): Vivi1W Macaroi Co. v. FTC 411,F.2d 2., 257-58 (3d Cir. 1969). tion2(f)read:

It shall be unlawful for any person cngaed in commerce, in the cou of such commerce, knowingly to induceorreceiveadiscriminationinpriecwhichi pruhibitebythi3section Section 2(11) read in pertinent par.

It shllil be unlawful for Ilny person engage in commerce, in the courne of such commerce, either diretly or indir tly, to discriminate in price between different purchllrB of commodities3 ofe likeand quality where the eff t of such dhwrimina.tion may be substanti811y to lesn competition or to injure, d troy, or prevent competition with IIny person who either grnts or knowingly reive3 the benefit of euch discrimination, or with cu tomers of either of them:Provide.d That nothing herein c-ntained shall prevent differentials which make only due allowance for difference in the cost of manufacture, Mle, or delivery ulting from the differing methoo or quantitie3 in which euch commodities3 ar to such purchasrn sold or delivered""

9 Section 2(b)reain pertinent par " . " nothing hereincontained shall prevent a seller rebutting the prima. facie C& thus made by showing that his lower price " " " to any pureh!lr or pureh!lrn was made ingoo faith to meet an equally low price of II competitor ), 1052 FEDERAL TRADE COMMISSIO DECISIONS Opinion 87 F.

discriminatory price that caused competitive injury. In addition, A&P asserts that complaint counsel failed to establish (1) that the price quoted by Borden was not made to meet a competitive bid and was not cost justified and (2) that A&P knew that tbe price received was not covered by one of those defenses.

We find that A&P has violated Section 2(f) and reject its proffered arguments.

In Com,'!nerce The Robinson-Patman Act forbids discriminatory pricing arrangements only "wbere either or any of the purchases involved in such discrimination arc in commerce. " 15 U. C. ~ 13(a). Therefore to conclude that jurisdiction exists to (11) prosecute A&P, either A&P' or its competitors' purchases must be " in commerce, " Recently, the Supreme Court has reiterated its position that at least one of the two transactions which result in the discrimination must cross a State line. Gulf Oil Corp. v. Cq:rp Failing Co. 419 U. S. 186, 199-201 (1974). Although A&P docs not dispute the finding that the sales to stores in Indiana arc "in commerce " for Robinson-Patman Act purposes, A&P argues that the "in commerce" requirement is not met for thc Ilinois sales.

We reject A&P' s claim. ln F,rrem()st Dairis, Inc. v. FTC 348 F. 674 (5th Cir. cert. denwd 382 L:.S. 959 (1965), thc respondent dairy claimed that none of its sales were made in interstate commerce although 20 percent of its raw milk came from out-of-State. The dairy sold fluid milk processed in its Albuquerque, Kew Mexico plant to stores in Albuquerque. In holding that the "in commerce" requirement was met, the Court stated:

lTJhc milk passed in a steady flow from the farms in Colorado through the Santa Fe processing plant, where it underwent a rather negligible processing operation, which did not change it.'J character appreciably, the shelves of retail gro ry establishmenls in Albuquerque * . . IV Jnder the stream of commerce doctrine, milk produced in one slate and processed and distributed in another state does not lose its inwrstate character because of standardization, and pasteurization. Id. at 677.

Other courts concur with thc rcasoning and results in Forerrst Dairies. E. , Dean Milk Co. v. FTC 395 F.2d 696, 714-715 (7th Cir. 1968) (25 percent of raw milk came from out-of-State). (12) Most of the raw milk used by Borden s Woodstock, Ilinois dairy- approximately 60 percent-came from Wisconsin (Borden Admission at 8, Sept. 29, 1972). This dairy supplied substantially all of the private label milk sold to A&P (Bordcn Response at 5, May 3 , 1972; A&P Admission at 7, Sept. 27, 1972). Finally, the processing in ), GREAT ATLANTIC & PACIFIC TEA CO., INC., ET AL. 1053 962 Opinion Woodstock did not interrupt the flow of milk from the farm to A&P' stores (J.D. 17 168; see Borden Additional Admission at 13- , March , 1973; Tr. 107-109). Therefore, the FOTerrst Dairis test is met. In addition, the negotiations leading up to the agreement between A&P and Borden were multi-State in nature (Tr. 1711) and therefore in commerce " (J.D. 167). See Beatrie Foods Co. 76 F. C. 719, 747 (1969). A&P' s Schmidt testified that A&P would not accept an offer to supply only a portion of its stores in Ilinois and Indiana (Tr. 1711; see Tr. 6126-27).

Discrimination in P1-ice and Competiti1Je Injury The Robinson-Patman Act requires both that a discrimination in price take place and that the possible effect of such discrimination be substantially to lessen competition or to injure, destroy, or prevent competition with a primary or secondary Ene competitor or with a direct or indirect customer (13J of the buyerIO In addition, to conclude that a Section 2(f) violation has taken place, it must be found that the buyer knowingly received thc illegal price discrimination. Automatic Canteen Co. v. FTC, supra 346 U.S. at 71. A&P argues that the evidence fails to demonstrate that it was a beneficiary of a price discrimination or that competition was injured. A&P criticizes the price comparisons of complaint counsel, the rcpresentativeness of the competitors who were allegedly injured, and the establishment of substantial injury. In addition, A&P states that it did not know that it was a beneficiary of an illegal price discrimination. Wc agree with Judge Hinkes' findings (J.D. 62- 88; 118- 123) that thc relevant comparisons demonstrate the existence of discriminations price. We note that a price comparison does not have to include the full line of dairy products nor A&P's purchases of Borden brand products. We faced and dismissed this same argument in Beatrie Foos Co., 76 C. 719 806 (1969), aff'd sub rwm. , Kroger Co. v. PTC 438 F.2d 1372 (6th Cir. cert. denied 404 l:.S. 871 (1971). Specifically, the price of private label milk and other dairy products sold to A&P may validly be compared with Borden s sales of such items to other groceries to ascertain whether (14J discrimination took place. The Kroger Court stressed "the competitive advantagc which results when disadvantaged competitors are denied a r private J label which the favored buyer receives. " 438 F.2d at 1379. In addition, although there were differences in the way A&P and its competitors were served, the distinctions were oftcn minimal and, in any event, are more properly considered in connection with the cost justification defense. '" Although one could conceivably argue that injury al the primary or seller level to place, complaint counsel limited theirCl proving.econdary or buyer level discrimination , ), Opinion 87 F.

We are unclear as to what A&P means when it argues that the sample of stores must be "representative. " As the Court in Kroger Co. v. FTC, id. at 1378, stated (EJvidence of discrimination in only one area is sufficient to meet the requirements of the Act. Accord Forerrst Dairis, Inc. v. FTC, S'upra 348 F.2d at 681. "(UJnlcss minuscule, the portion of the market that might be affected hy the charging of these discriminatory prices is immaterial" National Dairy Products Cor. v. FTC 395 F.2d 517, 522-23 (7th Cir. cert. denied, 393 S. 977 (1968). The area studies 11 presented by complaint counsel are quite detailed and are, in fact, representative of the areas allegedly affected (see, e. CX 182; RX 234 at 19). If anything, complaint counsel has exceeded any requirement of "representativeness (see LD. 121).

(15 J In addition, complaint counsel has shown that competitive injury occurred regardless of whether we applied the reasonably possible " test, FTC v. Mortn Salt Co. 334 U. S. 37, 46 (1948); Purolato Products, Inc. v. FTC 352 F.2d 874 , 881 (7th Cir. 1965), cert. denied 389 U. S. 1045 (1968), or the "probable" test Forerrst Dairis Inc. v. FTC, supra 348 F. 2d at 680. 12 The Robinson-Patman Act does not require proof that injury actually occurred but merely a showing that competition may be reduced because discrimination took place. National Dairy Products Corp. v. FTC, supra 395 F. 2d at 521. "The statute is designed to reach such discriminations 'in their incipiency before the harm to competition is effected. FTC v. Mor Salt Co. supra 334 U. S. at 46 n. 14.

We note that many courts have discussed the significance of differences in the prices charged to competing businesses in industries such as retail food, which are characterized by keen competition and low profit margins. , Kroger Co. v. FTC, supra 438 F.2d at 1379; Fred Meyer, Inc. v. FTC, supra 359 F. 2d at 367; see United Biscuit Co. of Arrria v. FTC 350 F.2d 615 , 621-22 (7th Cir. 1965), cert. denied 383 U.S. 926 (1966). Milk is one of the most important (16J commodities carried in retail grocery stores (A&P Admission at 3, March 1 , 1973). Therefore, discriminations in favor of A&P amounting to as much as 22.5 percent (Model Food Center, CX 182, 187, 191A)13 in this extremely competitive, price sensitive industry compel an inference of " These include Chicago and its suburbs, Gary-Hammond, and Valparaiso, Indiana 12 As the court in F(mmw. t Dairis note, 34 F.2d at 58) n. 11. "(TJhc distinction between the two standard ma:, bemoreapparentthanrr..l"

13 The KruflwT' Court, 43 F.2d lit 1379 n 4, listed examples of discounts in sendary line cao.. The dis('(unt. A&P are in line with those found w cali! injury in other cao.s. For example, the Court in Farml). Dairis 1m. FTC, l/pr, 34 F.2d at 675, found competitive injury where discunts were a. low a. 5 perc.nt , .

GREAT ATLANTIC & PACIFIC TEA CO., INC., ET AL. 1055 962 Opinion adverse competitive effect. See, e. , FTC v. Morton Salt Co. supra, 334 S. at 46-47; Kroger Co. v. FTC, supra 438 F. 2d at 1379. That A&P did not pass the discount on as a lower price to consumers does not rebut the inference. Forerrst Dairis Inc. v. FTC, supra 348 F.2d at 680. For example, in receiving the price discount, A&P enlarged its profit margin (See CX 185, 255; Tr. 2059-2060) 15 and thereby obtained a material capital advantage in this highly competitive field. See Kroger Co. v. FTC, supra 438 F. 2d at 1378; Forerrst Dairis, Inc. FTC, supra 348 F. 2d at 680.

(17) Finally, we address the issue of knowledge. A&P admitted that milk is one of the most important commodities in groceries and that milk is a staple purchased more frequently by consumers than other products (A&P Admission at 3, :Iarch 1, 1973). It is acknowledged that selling milk at a higher price than one s competitors will cause some consumers to switch stores. A&P also admitted that profit margins in the retail grocery business are very low (A&P Admission at 4, March 1 1973). Therefore, A&P, by virtue of its trade experience or simply common sense, knew or should have known that it was a beneficiary of a price discrimination having the requisite harmful competitive effects (see J.D. 86- , 97). Accordingly, we reject A&P' s arguments. Meeting Competition A seller who has discriminated in price is guiltless under the Robinson-Patman Act if the lower price was offered in good faith to meet an equally low price of a competitor. 15 U. C. 13(b). The Supreme Court has stated that the buyer is not liable under Section 2(f) if the lower price he induces is within the meeting competition defense or if the buyer is unaware that the price cannot be protected by that defense. Automatic Canteen Co. v. FTC, supra 346 U. S. at 74. The Court in Kroger Co. v. FTC, supra 438 F.2d at 1377, noted that when a buyer is charged with violating Section 2(f), the price he induces must come within the meeting competition defense not only from the seller s point of view but also from the buyer s. This is consistent with the Supreme Court's statement concerning the burden of going forward in a Section 2(f) case:

(18) OUf view that 2(b) permits consideration of conventional rules of fairness and convenience of course requires application of those rules to the particular evidence in question. Evidence, for example, that the seller s price was made to meet a competing seller s offer to a buyer charged under 9 2(f) might be available 1. The Court in Fowl.,. Mfg Co- v. Galick 415 F.2d 124, 1252 (9th Cir. 1969), r..n. denim 39 U.S. 1012 (1970), II private Robinson-Patman Act adion tate that a prir,€ differelltifll alone establishes competitive injury. There is no need to prove II les. ening of competition to obtain damages. The profit margins of the unfavored competitor. were very low, ranging from les than 1 percent to 5 percent (See Appendix 4 (in camera J; ex 11M, 11;9, 192, 199A, 199 J- , 201 , 203 (all in GamRTa J; J.D. 88). 1056 FEDERAL TRADE COMMISSIO DECISIONS Opinion 87 F.

to a buyer more readily even than to a selJe Autotic Canteen v. FTC, supra 346 U. S. at 79 n. 23.

Specifically, proof that the seller did not act in good faith is unnecessary to a finding that a buyer cannot assert the meeting competition defense.

In its reply brief, A&P argues that its contemporaneous comparison of the Bowman and the last Borden bid is not relevant (Reply at 35). We disagree. The comparisons demonstrate whether A&P acted in good faith. For example, if the Borden bid beat" thc Bowman bid yet A&P contemporaneously reasonably believed that it only met the bid we believe A&P (as well as Borden, if Borden prudently reached the same conclusion) could assert the meeting competition defense. See FTC v. E. Staley Mfg. Co. 324 U. S. 746, 759-60 (1945).16 However, we conclude that A&P contemporaneously concluded that Borden s bid was "substantially better" than Bowman s (Tr. 1413-1414, 189&-1897; CX 263), and A&P, therefore, cannot assert the seller s meeting competition defense.

We assume arguendo that A&P can assert the defense if the Borden bid did, in fact, happen only to meet the Bowman bid. (19) However while we believe that the Bowman bid was comparable and operative 1? we conc1udc that the Borden bid was substantially better than the Bowman bid. Specifically, we find that prior to adjusting the Bowman bid to make it reflect such items as A&P's true volume requirements and delivery requirements, Borden s bid substantially beat Bowman (See CX 263B; Comparc CX 75 F-J with CX 50). In addition, after adjusting the Bowman bid to make it responsive to A&P' s true situation, Borden s bid beats Bowman s by an even greater amount. Bowman would not be able to give A&P such a low price because A&P's volume was less than the $1 million per month stipulated in tbe Bowman offer (See citations in J.D. 106) and because Bowman would have to increase the days of delivery in the Gary- Hammond area (See citations in J.D. 107)18 Finally, A&P argues that because Borden has not been found in violation of Section 2(a), A&P cannot be charged under Section 2(f). In other words, A&P asserts that Borden must be convicted of a Section 2(a) violation prior to or at the same time as A&P is found in violation of Section 2(f). In developing this argument, A&P claims that complaint counsel' s case against A&P is premised on Borden s having a 16 A&P aJIIJeI! that in thig case, it contemporaneously b€lieved that Bordens bid heat Bowman s hut after properly" comparing the bidg, Bowman s bid, in fact, beat Borden s (Appel at 31-3) 17 Speifically, we revel1Judge Hinkes' oonc\usion to thc contrary (See FFI03 106 114) Ii The differ'ncr in stipu;ated butterfat content wou,d not greatly change the difference in pric. O:mtrnryto A&P' s study, BonJ"n s hutterfat level was probably about 3.45 percent (Tr. 5513), and the BO\lan bid may not have caliedfor3.5percent(Tr. 'iS4S3 S461 6187-6188;see1.D. 129) \ . , . GREAT ATLANTIC & PACIFIC TEA CO. , I:-C. , ET AI,. 1057 962 Opinion valid meeting l20J competition defense, and that therefore, Borden would be absolved from any Section 2(a) charge. We reject the argument.

The Commission alone establishes the enforcement policy which wid best accomplish the ends contemplated by Congress. FTC v. Universal- Rundle Corp. 387 D.S. 24 , 251 (1967); Moog Intl'ustris, Inc. v. FTC 355 U. S. 411 , 413 (1958); see Jewel Companies, Inc. v. FTC 432 F. 1155, 1160 (7th Cir. 1970). In deciding to issue a complaint against A&P charging a violation of Section 2(f), the Commission concluded that A&P' s conduct was egregious enough to warrant such action. It did not foDow that the Commission condones Borden s behavior. In addition complaint counsel notes (Answer at 50), A&P was the "principal malefactor" in the negotiations with Borden. It placed the pressure on Borden, not vi.ce versa. Although the Commission decided not to issue a complaint against Borden for violation of Section 2(a), the Commission has not concluded that Borden would have a valid mceting competition defense to such a charge l21J Furthermore, even though a seder has a defense to a Robinson- Patman Act charge, a buyer can stjJ violate Section 2(f). The court in Kroger Co. v. FTC, supra found that the buyer lied to the seder about the details of a competing bid, 438 F.2d at 1375-1376, and that the seder had no way of knowing that the buyer had lied. Therefore, in submitting the second bid, the seder had a meeting competition defense within the meaning of Section 2(b). However, this defense did not exonerate the buyer who knew that the bid was, in fact, not within the defense. "To hold otherwise ' , * would put a premium on the buyer s artifice and cunning in inducing discriminatory prices land) would violate the purposes of the Act, and frustrate the intent of the Congress. Id. at 1377 Cost J 1J.sf.:ji.cation A seder who charges different prices is exonerated if the differences are justified by a savings in cost. 15 U. C. 13(a). As in the case of meeting competition, a buyer is not liable under Section 2(f) if the Judge Hinkes did not reach the question of whether Bornen had a meeting competition defense(Ree J.D. 166) We believe that it is very prohahic that Borden dieNJI have such a defense- To have a meelin" compdition dcfcll the record must demonBtraUC the exist.nce of facts which wouid ieau a rea.';onahle and prudent peron to conc ude tha the lower price would, in fact, meet the comp€titors price FTC\" A, E, SmUiyMfg. Co. mpr, 3241., S, at 759-760; llamplm Graff Ve7/Ji71g Ca"478 1".2d ,S21, 5:J. (5th Cir. 1973). As note, BOrU,"n had gerious doubts con rning whether the comP'ting bid was legal SP'cifi('3l1y, it believed that the other bid only coflidere direct costs (Tr. 237- 239 376-77 878 961-62) !tshoulrl hav," asked A&P for more information about the comP'tingbid. By not milking the request, it wa. not acting pn;dently.See Standard Oil Cr, v. Broum 23 f' 2d M, 58 (5th Cir, 1956); f. Rowe, Prce Discriminatior. under the Robingon-Patman Act 226 (1962). AB th," record clearly indic8te, A&P had knowledge of Borden s belief that other dairies might submit bids that did not i clude a:1 co ts (Tr, 213-214, 872-873 , 955) 1058 FEDERAL TRADE COM:.ISSION DECISIONS Opinion 87 F.

lower price he induces is within the cost justification defense or if the buyer is unaware that the price is not protected by that defense. (22j The parties and Judge Hinkes disagree as to who has the burden of proof in demonstrating whether the cost justification defense is available (J.D. 134; Answer at 66-68; Reply at 48-49). A&P and Judge Hinkes are mistaken in concluding that the A utornatu: Canteen case dealt with the burden of proof issue. Rather, the Supreme Court stated twice that it was deciding only who had the burden of first coming forward with probative evidence. 346 U. S. at 65, 82. In promulgating its rule of "fairness and convenience id. at 78, the Court determined that the Commission did not satisfy this burden by showing only that the buyer had knowledge of a price differential" Rather, the Commission must present evidence to demonstrate that the buyer was reasonably aware that the price differential could not be cost justified. Id. at 78-80.

The Court believed, however, that the burden of showing this knowledge " should not be difficult." Id. at 79. As an example of the evidence which would satisfy this burden in a situation where methods of services and quantities delivered differ, the Court held that the Commission " must only show that the differences could not give rise to sufficient cost savings and that the buyer, knmving that these were the only differences, should have known that the differences were not (23 j cost justified. Id. at 80. The Court emphasized that it would not attempt to illustrate "what other circumstances can be shown to indicate knowledge on the buyer s part that the prices cannot be justified id. at 80, concluding that the Commission should do this. Id. at 80-81; occord, Suhurban Propane Gas Cor. 73 F. C. 1269 , 1272- 1274 (1968) (interlocutory order); 16D J. von Kalinowski, Business Organizations: Antitrust Laws and Trade Regulation ~~ 36.05(2j- 36. 05(3j (1971).

We have described on several occasions how complaint counsel can successfully meet the burden of going forward with the evidence where a favored buyer purchases in quantities or under methods differing from those accorded to disfavored buyers: (I Jf complaint counsel show such facts and cjrcumstance as would have given the buyer reason to believe, based on the knowledge available to him, including knowledge of the methods of doing business in the particular industry, that the different method or quantities could not have resulted in cost saving-'" sufficient to justify the differential allegedly accorded him, they would have met their initial burden. Beatrie Fo()ds Co. 20 In Auitu: Canwen the Commi$. ion only €8talJlished that the btiyer had obtained prices as much one-third belowitsc.mpetiwrs 21 The Commission\ ha. frequeT\tly decided what circumstance, dcmonstrllte thllt II buyer h!l the reuisite knowledge. , Mid-Soth IJtrilwtMS v. FTC, 287 F.2d ,S12, r:18-19 n 15 (5th Gir. ), wi. rknie 368 U.S. &'\ (1961) GREAT ATLANTIC & PACIFIC TEA CO., INC. , ET AL. 1059 962 Opinion supra 76 F. C. at 820 quti"1 Sulrban Prapne Ga Cor. 71 F. C. 1695 1699 n. 2 (1967) (int€rlocutory order).

We reiterate what we said in Suburban Propane Gas Cor., supra, 73 C. at 1273- the burden of establishing a prma facie case or first going forward with the evidence may (24 J be satisfied without introducing a formal cost study showing that the lower prices were not, in fact, cost justified. The Court of Appeals for the Sixth Circuit has affirmed our use of this rule. Kroger Co. v. FTC, supra 438 F.2d at 1378.

Once complaint counsel has satisfied their initial burden, the buyer must demonstrate that it did not know or could not reasonably have known that the price differential was not cost justified. If it fails to prove lack of knowledge, it must show that the price differential was in fact, cost justified. See Automatic Canteen Co. v. FTC, supra, 346 S. at 74. Therefore, if the buyer establishes that it neither had nor should be charged with such knowledge, the buyer does not have to submit any cost justification study. However, if the buyer cannot establish the requisite lack of knowledge, it must submit a cost study if it wishes to prevail, which study generally will be held to the same standard as are cost studies offered by sellers. In other words, it is possible for the buyer to be unable to respond adequately to complaint counsel's evidence demonstrating knowledge yet be able to demonstrate that, in fact, the price differential was cost justified. Lnlike the meeting competition defense, one does not need to demonstrate good faith in proving cost justification. Therefore, it is completely permissible to receive in evidence cost studies prepared after the alleged discriminatory sale and only for purposes of litigation. Rowe Cost Justification of Pre Differentials Undr tlw Romnson-Patman Act, 59 Colum. L. Rev. 584, 610 (1959).

(25 J Finally, if the buyer satisfies the burden of meeting complaint counsel's prma facie case, then complaint counsel must rebut the buyer s evidence. The buyer may then introduce other evidence responding to complaint counsel Although the burden of going forward may shift back and forth during the hearing and is frequently not so easily segregated, the burden of persuasion as to the issue of whether the buyer had the requisite knowledge that the cost justification defense is unavailable rests with complaint counsel while the burden of persuasion as to the issue of whether the prices are, in fact, cost justified rests with the buyer. Suburban Propne Gas Cor. supra 73 F. C. at 1274- 1275; see Beatrice Foos Co., supra 76 F. at 820.

We believe that the evidence submitted by complaint counsel in this case well exceeded its initial burden. Ylany competitors were charged a , 1060 FEDERAL TRADE COMMISSION DECISIO:-S Opinion 87 F.

much higher price than A&P for Borden s milk products even though they bought in greater quantities than the individual A&P stores with which they were competing (E. Burger s: CX 182, 187, 188A, RX 23 p. 19; Tr. 2270, 2381, 2385). Although many of these stores received more service than A&P, the service savings to Borden could not in any way justify the price differentials (See J.D. 109-110). In addition, even those companies that received 1limited service paid a higher price than A&P (See Complaint Counsel's Proposed Findings, Vo1. 1 at 150-151), a price differential that was not justified by the saving of such costs advertising (See CX 54; Tr. 3387 34,10 6136). A&P knew about the differentials, since it was being charged the "re"rular" price for branded items (See CX 75, 255; RX 66). (26) This evidence in itsc1f would probably be sufficient to shift the burden of going forward. However, complaint counsel introduced much more evidence to demonstrate that A&P knew or should have known that the prices it received were not cost justified.

First, Borden stated exp1icitIy that its final offer could only justified on the basis of meeting competition, and A&P accepted the offer on this representation (Tr. 248 , 886, 1789, 1792).2 As the Supreme Court stated in Automatic Cant en Co. v. FTC, mpra 346 at U. S. 73: " ~ 2(f) was explained in Congress as a provision under which a seller by informing the buyer that a proposed discount was unlawful under the Act, could discourage undue pressure from the buyer. Accord, 80 Congo Rec. 9419 (1936) (Statement of Congressman l.tterback prcsenting conference report to House); Report supra note 6 , at 193. The Borden statement put A&P on notice that it was likely that Borden did not have a cost justification defense. Cf. Beatrie Foods Co. supra 76 F. C. at 821 (because seller was told by two other dairies that price might not be cost justified, he had notice of possible cost justification problems).

(27) Second, although A&P requested the customary assurances from Borden that Borden s discounts were available to others on proportionately equal terms (Tr. 1783 , 1786- 1788, 1978-1979), A&P did not receive them. In fact, Borden specifically refused to comply, writing instead our prices are proper under applicable law and we are prepared to defend these prices. " (CX 263H; see Tr. 1406, 142, 1438- 1440)23 Third, Borden provided A&P with detailed cost data during the !Z When Ralph Minkler, Prc iuent of Borden Central UiBtrict, pregent. Bmxkn o final \Jid A&P' s &hmidt, he Btat.: "Thi price is given UJ you by us on the feiling amJ belief that we &1" Ineitin" a competitive hid. We know of no other way to justify thi"- You have to a.eceptit OI1 that basig, " (Tr. 248) " Herschel Smith, A&P'g Head uilrtr 8 Dairy Buyer jd that r.e reived oral !I!\.ourance from Harr Archer Borden s Executive Viii Pres,d nt in New York, that this Wa1 11 letter of availability (Tr. 1407). However, Archer denied having Buch a convcl"ation (Tr. 1246-1247). After viewing both witness IInd examining ali of the evidence Jud)i Hinke. did not credit Smith'g testimony (1. 0. %) );

GREAT ATLANTIC & PACIFIC TEA CO. , I/(C., ET AL. 1061 962 Opinion course of the negotiations, data which showed that Borden would be either losing money or making a minimal profit (CX 23 , 24, 25A, 26; Tr. 201-205 , 1866-1869). Although the data may not be completely accurate, it demonstrates that A&P at least had knowledge that the discounts could drastically affect Borden s profits, and therefore, A&P should have inquired whether the prices were available to others. See Fred Meyer, Inc. v. FTC, supra 359 F. 2d at 365-67.24 Of course, if the A&P-Borden contract was profitable for Borden before the (28J switch to limited service and Borden would lose money after the change, then A&P should have doubted whether the new prices could be cost justified (See J.D. 91-93).

Fourth, A&P' s trade experience should have given it reason to know that Borden s private label prices were not cost justified. See A automatic Canteen Co. v. FTC, supra 346 U. S. at 79-80 (Trade experience "can afford a sufficient degree of knowledge to provide a basis for prosecution. Kroger Co. v. FTC, supra 438 F.2d at 1378. As Judge Hinkes concluded (J.D. 97, 98, 127), A&P was familiar with the dairy industry generally and with the Chicago area specifically. Also A&P knew that Borden s private label price was substantially lower than its branded label price because A&P bought both types of milk from Borden (See ex 75, 255; RX 66). Therefore, A&P should have known that its competitors, some of whom received limited service were paying much marc than A&P for the same product. Finally, A&P' s conduct after the private label agreement was consummated convinces us that A&P knew or had reason to (29J know that any differences in service were not cost justified. First, in mid- K ovember of 1965, Borden decreased the price A&P paid for Borden label products but at a much lower discount than it granted A&P on private label products (RX 66). Except for the small amount devoted to advertising, there was no difference in the cost to Borden of private and branded products. Therefore, A&P should have realized that the price was not cost justified (see J.D. 99). Second, in May of 1966, A&P accepted an increase on Borden label products which reflected increased container, labor, and social security costs, an increase which Borden apparently charged al1 of its purchasers. However, A&P at first refused Borden s request that the price of A&P' s private label ,. CitingAulnmatic Canwen Vi. FTC' lang1agc aboutpuffinginabargllining situation 346L:. at80n. 24 A&P arueg that Bon:Jcn might very well have submitt.d a pul)(J8dy inacurate bid (Appeal at 45-7), At this point A&P' s evaluation of ,m"'I'h :\alone, the Vice President of GvvernmCf1lli Controls and formerly the Comptroller of Bordcll s Centrol District and the empluyee who prepared the data, i relevant: A&P believed that Malone WB. "honest anr! forthright- " (Tr- 184). 1n addition, if we r!i regared all tudieB in which a g(l inte!"st motive is parent, then, of COUI1, we mu t dismi the studies prep"_red for A&P's cost :ustification def.m"' " )J note at pares 23- , we do not belit'vt' that complaint couIlel ha.the bun:lcn of intrucing a formal study demonstrating that the price differential i8 not cost justified and do not cOIlider thel calculatiolls such a study ," Of cours, labei do not differentiate products for pUfPS of determining "like gre and quality FTC BrYry.k" Cu" 38 C.S, 637, 64 (196), The only benefit ioot hy A&P WB. advertising, which WB. a very smaH experu 1062 FEDERAL TRADE COMMISSIO:- DECISIONS Opinion 87 F.

products be raised commensurately (CX 79 , 81 , 255B; Tr. 1050-1054) and later accepted only a proportionately smaller increase (Tr. 105&- 1062; See ex 102, 103, 105, 106, 110).2 Because A&P readily accepted the increases on branded milk, we infer that A&P believed that Borden was selling branded milk to some of A&P's competitors at delivery terms similar to those at which Borden was selling branded and unbranded milk to A&P, and therefore, considering Borden s small advertising expenses, A&P's private label discount was not cost justified. See Automatic Canteen Co. v. FTC, supra 346 U. S. at 80. (30 J In conclusion, viewing all of these facts together, we find that complaint counsel fully established a prima flow case that A&P knew or should have known that the prices Borden charged it for private label products was not cost justified.28 The burden then shifted to A&P to demonstrate that it did not have knowledge that the prices it received were not cost justified or, failing to establish that, that the prices were, in fact, cost justified.

We do not find persuasive A&P's assertions that it did not know nor should have known that the prices it received were not cost justified. Its statements concerning the meeting competition language in the negotiations and the "legal letter " are not convincing (see A&P' Proposed Findings at 76-86). In a post-hoc analysis of the cost data that Borden supplied it, A&P argued that the Borden study was not completely reliable (Appeal at 45-47). However, A&P did not show whether it had contemporaneous criticisms of the quality of the data nor whether it communicated any of its criticisms, perhaps in the form of additional inquiries, to Borden. As noted at page 27, this data does not have to be completely accurate (31) but is useful in proving that A&P was on notice that the new contract could be unprofitable to Borden. Finally, A&P's post-agreement behavior is relevant to show what knowledge A&P had at the time the agreement was entered. A&P introduced evidence which showed that A&P' s Smith used a " 2 formula" to evaluate the reasonableness of a bid. The formula states that a dairy can profitably sell milk for approximately six cents more per quart than the dairy s cost for raw milk. Although we agree that this formula is relevant in considering whether A&P had knowledge, it does not rebut complaint counsel's prma flow case. " This action is wnc.ivahly 0. second inducement by A&P of di riminatory prices and therefore, a separate and distinct violation of the Rohin!In-Patman Act (&e LD. 57 , 89, 101; wmplaint Couni;I's Prpo Findingo, Vol. 1 at 119). Because we have found that A&P reived an illegal pricl preference prior to thes increa.'lS . we do not speific.lly address this claim.

We are 110t convince that bcuse Borden spread the saving- over eleven rather than thirtn it€rr, A&P WI! on notice that the bid Wa. not wst justified. Although generdlly we ag thllt "the same cost saving1 applicable different products with different volumes would not likely be relate to. . . cost saving'"(J.D. 9O), itigat llt oonceivahle that thc $820 00 1!vings reprcsente the amount Borden would gavc by switching to limited delivery service 2\ Two C€n eah ar allocte for canons, plant cost and profi, and delivery ros (Tr. 6.'!) GREAT ATLANTIC & PACIFIC TEA CO., INC., ET AL. 106 962 Opinion noted, the formula was simply a reference point from which to estimate costs (Tr. 6399). It was developed from A&P's experience in negotiating for its private label supply of milk in New York and Massachusetts. Because the costs and conditions in the Chicago area were different (e. in-plant wages: CX 25C, 26), it should have been adjusted. In fact, as A&P admits, the formula was never used to determine the exact measure of a dairy s cost but only to serve as a workable rule of thumb. " (Appeal at 49). Reliance on such a formula would be very irrational in light of all the other evidence (See J.D. 98).

(32) Since A&P did not present enough evidence on the issue of knowledge to shift the burden back to complaint counsel, it had to submit a cost justification study if it wished to prevail. The studies entered by A&P on the record (RX 233, 23) were prepared specifically for purposes of this litigation.

We affirm Judge Hinkes' findings (J.D. 136-165) that A&P' s cost justification studies "are so defective and inadequate as to furnish no evidentiary basis" to justify the price differential that A&P received for private label products on the basis of Borden cost savings (J.D. 165). Because the studies are so flawed, we would not find them persuasive under any reasonable standard. However, as we have stated, generally a buyer s cost justification study should meet the same standard that a seller s must. On the other hand, a buyer need not submit any study if complaint counsel fail to prove that the buyer knew or could be charged with knowledge that the prices were not cost justified. See (33) Automatic Canteen Co. v. FTC, supra 346 U. S. at 74. 32 We note in passing that we would be willing to lower the standard if it was demonstrated that the buyer could not obtain the necessary records from the seller to undertake the study. See id. at 79, 81. A&P alleges that hostility existed between Borden and itself (Appeal at 15- 16; A&P's Proposed Findings at 8-9) but does not confirm that these inherent conflicts" prevented it from obtaining records from Borden. In fact, the record shows that A&P was able to get the documents it needed.

A&P asserts that "the most egregious example " of the failure by Judge Hinkes to follow a liberal approach to its cost study is his rejection of the original derivation of Borden s delivery costs (Appeal 30 A&P' g other &rgumenL-- (or example, that A&P could get better price on private labels in other par of the country (Tr, 7326)-aJsodo notadeluately rebut complaint counsel's strong showing J! Because we do not find the studies sufficiently rciiable, we did !Jot decide whether Borden s interrHl 00t calculations (eX 54, 87, 2() are accurate, If A&P' studic.. adequately showed some typ of cot justificatiol1, the Borden studies would have been msjdere a. compl!lint counsel's rebuttal. Accrdingly, we did not !!rtin whether the A&P study (RX 232) which was offered to rebut Borden, inUcrnal Court calculations is a.urnw 32 Of COIU, if the seller is joined in the IWhinwn-Patman action and relies on the cot justification defer, he wil Buhmitthe cost justification study.

. , Opinion 87 F.

at 53). The calculations, based on "time standards " developed in 1949 from studies made of Bowman Dairy deliveries in Chicago and confirmed by additional work in the 1950's and 1960' , are summary in nature. Also, the underlying data necessary to conduct an effective examination of their reliability were destroyed (order, Sept. 16, 1974). In addition, Robert Havemeyer, the author of the A&P studies, did not participate in the collection of the data nor in the preparation of the standards (Id. at 2). (34 J Therefore, he could not be effectively crossexamined on the validity of the underlying data.33 For these, as well as other reasons, Judge Hinkes properly refused to admit the calculations. Id. This is not a matter of preferring one method of calculation over another. Judge Hinkes simply could not determine whether the calculations were even arguably reliable. Cf. FTC v. Standard Moto Products, Inc. 371 F.2d 613, 622 (2d Cir. 1967). A&P' s arguments concerning the relevance of the Wisconsin research (RX 9 , 11) are adequately dealt with by Judge Hinkes (See J.D. 139-142). We note, however, our disagreement with the Judge general rejection of standarized costs (J.D. 143). If it could be proven that standardized costs are reflective of actual costs-a showing not made in this case-then we would be willing to accept standarized costs as a substitute for actual costs.

The studies are unpersuasive for numerous other reasons. For example, the studies did not consider the fact that a portion of Borden s expenses were incurred only in connection with its dealings with A&P (see citations in J.D. 157-160). Without a study of these costs one cannot ascertain thc true (35 J cost differences between the costs of handling the A&P account and the costs of handling the accounts of other purchasers (see Tr. 9422). See Rowe supra note 19, at 287 (" would be hard to defend a cost study which lacked internal consistency by resort to analytic reasoning with respect to one factor that would vitiate another. ). Furthermore, the studies do not allocate costs according to the manner in which they were incurred (see citations in Complaint Counsel's Proposed Findings, Vol. II at 295-296), do not allocate overhead expenses according to accepted accounting principles (id. at 295-297; see FTC v. Standard Moto Products, Inc., supra, 371 2d at 622), and inaccurately assign costs to certain customers (e. citations in J.D. 162).

But the most important task in evaluating the studies lies in ascertaining the credibility, intelligence, and experience of Mr. 3J Hllvemeyer frequently WlU not helpful in describing the r€liltion hip of underlying data tG the summaries. For example, wilen a.ked which study was not included in making certin observations, he state J Clnnot tell you bec.usetheworkingpapcl' lir€nolon!!crllvailable, (Tr. 913o.9131) ,. Even if the time calculations were acptable, we would not find Hllvemeyer 8 ca1cullltion of delivery cooj, reliable. Some of the probiems with the Havemeyer ana!ysis are describe in the initial ded ion (tD. 144) GREAT ATLANTIC & PACIFIC TEA CO. , r""c. , ET AL. 106 962 Opinion Havemeyer. Judge Hinkes found (J.D. 137) that Havemeyer did not base his studies on any personal knowledge of Borden s operations. Because Havemeyer was unfamiliar with Borden s operations and activities, he did not consider many relevant factors (see discussion in J.D. 146- 152). In addition, Havemeyer did not review a11 of the available evidence to determine the complcteness of his studies. For example, when asked if he made any check to determine if a11 loading expenses were included on a certain ledger, he stated he did not but added that his staff had made "some tn,e of check." When asked about the type of check, he stated that he could not be specific (Tr. 7520). (36 His experience with Robinson-Patman cost studies was limited to preparation of a study for one other litigated case United Stat€S Borden Co. 370 U.S. 460 (1962), in which the study was rejected by the Court because Havemeyer failed to classify stores adequately. Id. 470-471.

A&P asserts that Judge Hinkes erred in finding that the studies failed to classify customers properly (Appeal at 55-56; see J.D. 154-155). The studies do, in fact, differentiate stores on the basis of "limited service " and " full service. " However, it is clear that stores differ in terms of all types of expenses, including delivery expense, clerical expense, product waste, selling expense, credit loss, merchandising expense, and overhead. lore important, the stores also differ in volume of purchases.

The Supreme Court in Unit€d Stat"'s v. Bffdcn Co., supra criticized requiring cost justification on a customer by customer basis. 370 U. at 468. 1t tried to strike a balance which would permit group averaging where the group members were sufficiently alike especially "on the essential point or points which determine the costs considered" to make the averaging a reasonable indication of the cost of dealing with individual members. Id. at 469. We believe that the A&P classification makes some sense and could meet the Borden test although we are troubled by the volume differences between stores. See Foremost Dairis, Inc. 62 F. C. 1344, 1361- 1362 (1963), aff'd, supra. However we (37) do not have to decide this issue because thc studies are so fundamentally flawed.

For a11 of these reasons, we find that the cost justification studies are inadequate to shift the burden back to complaint counsel. " Similarly we do not have torEaeh th"Guesti"" whdhcr th 8turiel 3re invalid tJuS€ theydewrmine c.' for A&P stores o ' an average ba..ig and comp re this avcrage with the spedfic costs of serving allegedly injure CDmpetitors (Ap &.1 ilt 5G-57; Ble J.D, 156) We nou' , however, that ,'Jch a prOCdu cou,d hid" largp diocrepancies :n Biz" among A&P storeg, And size, of caUl' , is often If,t.d u) co,t di:fenme('.s. &e A'atl Hw-1 m1ry Pruct: ():. FTC, supra 395 F.2d at 52526 atry 70 I"T.C. 79, 19&-19(' , 2m 0%3) Fina"". we ai odo not have toadd)' % complaint coun el' a5rtions that the ,tudie.' ohould cover the entire discrimination period(Bee OJmplair.t OJun!\l' Prorx e(t Yindings, Vol. II !lt3IXJO'J) We thini- , however, that such a requirement is unduly bUITen. onw 216- 9"5 0- 11' 77 - 68 1066 FEDERAL TRADE COMMISSJO)\ DECISIONS Opinion 87 F.

Therefore, we find that A&P did not prove that the differential it received was cost justified.

IV. COe-NT m: COMBINING TO STABILIZE PRICES Count III charged that A&P and Borden entered into a combination having the tendency or effect of stabilizing and maintaining retail and wholesale prices for milk and other dairy products. Judge Hinkes dismissed the count. Complaint counsel argue that Judge Hinkes applied the wrong standard of proof -. that he did not view the evidence in the case in tlJto but looked at each part separately. According to complaint counsel (C.C. Appeal at 3- 8), Judge Hinkes required that each element of the conduct of the parties be sufficient to (38 J demonstrate the existence of an unlawful combination contrary to the teaching of such cases as Twu Fire Equip. Co. United States 197 F.2d 489, 494 (9th Cir. 1952). Complaint counsel emphasize that the communications between A&P and Borden and the " needs " of these parties show that a combination took place (e. C. Appeal at 9 , 18).37 Concerning the communications, we agree with Judge Hinkes (J.D. 199) that the record contains no evidence suggesting that Borden and A&P discussed the price A&P would charge retail customers or the price Borden would charge A&P's competitors, except to note that a price differential at the retail level bctwcen branded and unbranded products might create problems. Viewed together as well as separately, we do not find that the statements relied on by complaint counsel show that a combination to stabilize price existed. " The statements do not in any way convince us that A&P ever gave Borden the assurance that A&P would not create a price differential at the retailleve1. (39J We also do not understand why A&P and Borden needed to enter into a combination to stabilize prices. Their pre-and postagreement conduct show only that th y were acting as economically rational businessmen. Although clearly adversaries in their ncgotiations- A&P wanting the lowest price possible and Borden the '" Although the c.unt charge that the wmbination had a tendency to ,tahiti,,, price at both the I'tail and wholesalelevel thethr\1BtofcolTplaintcuun"'nsg1J.'i8dired. eJwthe rl'tail level a.pect(oe I.D 179- 187) 3' Complaint counsel presenta other ar1(mentd wh;ch we filuj or.convincing. For example, they note that the $820 00 3avinga WM not made or. any straight ui1!count biWig but spread hy Borden over the 11 different product \in&! Thereby Borden could influence A&P' s retai price (Tr. 523523). But as A&P' s expert ecnumic witness, Dr Mnrkham, stated (Tr. 7041-4), a distribution of cost saving- limon" the various products W!l rationlll and in .(rdance with cuswmary industry practice. In addition, each il.m was ordered separately on a demand basis, and therefore asinglepriceforoneyear spuITha. ofallllitemscouldnotbequote(S€€l.D.194) '" These include the " sengitivenes. of the market (CX 12)," the " depre.se competitive situation (CX 18PJ," and oommenL at the :\a:' 25 196') meeting between A&P and Borden (CX 19). GREAT ATLANTIC & PACIFIC TEA CO. , INC., ET AL. 1067 962 Opinion highest-both A&P and Borden were strongly motivated by independent causes to maintain existing dairy pricesJ A&P hoped that by switching to private label milk, it could obtain lower costs and increase its gross margin (e. Tr. 1400-1401). This was its whole purpose in negotiating the private label agreement with Borden, not passing lower costs on to customers to increase volume. We note that throughout the A&P-Borden negotiations, no provision was ever discussed for increases in volume of A&P's previous sales. We infer from this that the parties never contemplated that A&P would use its lower cost to lower price and thereby increase volume. Were A&P to lower its retail price, A&P would not achieve its immediate goal: to maximize profits (see Tr. 5 10; 7038). Also, A&P feared that if it lowered (40) the retail price of milk, Borden might be forced to give a similar wholesale discount to A&P' competitors. In addition, A&P was aware that the price it obtained from Borden cut Borden s profit margin to a point at which Borden might be driven out of the market if the prices that Borden charged decreased further. Borden had recently completed building a large plant in Woodstock lIinois and was wiling to sacrifice some price to increase volume and thereby operate efficiently (Tr. 217, 5230). Especially, Borden needed to keep its major customer, A&P (see CX 42D). Borden did not want A&P to use a private label at all but was ultimately forced to offer to provide one. Borden then sought to minimize the substantial price concessions A&P requested. Unless such discounts were the only way to make use of idle machinery at Woodstock, Borden did not want to give other potential purchasers the same discount'O However, if A&P charged a lower price on private label products and thereby precipitated a price war 'l Borden would be forced to extend the discount to others. Since Borden s dairy business was in a precarious financial (41) position, such action could conceivably drive it to the point where it would have to withdraw from the Chicago market. Therefore, because it was in both parties interest for A&P to maintain the prevailing price on its private label dairy products, there was no comprehensible reason for A&P and Borden to construct an illegal combination to achieve that end (see J.D. 199). We decline to construct one for them. Of course, combinations having any tendency or effect of stabilizing prices are unlawful. See United States v. Container Cor. 393 U. S. 333 336-338 (1969); United Statl3s v. Socony- Vac"um Oil Co. 310 L. S. 150 3" The evidence submitted rc:ating to market structure(ooe c. Appel at 19-23) con incc that A&P lind Borden lacked the ability t. stabiliz prices(81" Bon.len s Answering Brief at 31-37). Therefore, this cvidcnc. doe not buttres. complaintc.un!\.,' spoditiontr.atacomhination Lok place .w In fact, becaus. Borden probably waokJ tD increase volume, it considered offering the diRC.count W others. Its counsel, however, advise Borden that such action would de troy Borden meeting compdition defense (Tr. 10'6) " Although note, A&P a.fld Borden lacked the ability stabilize the market, A&P w&; Ia.rge enou!'h to start a. pnce wa.r ), ), 1068 FEDERAL TRADE COMMISSIO DECISIONS Opinion 87 F.T.

221 (1940). Although complaint counsel has introduced only circumstantial evidence to prove the existence of a combination intended to stabilize the prices in a given market, we recognize that such evidence may well be sufficient. See Interstate Circuit, Inc. v. United States, 306 S. 208 , 221 (1939); Treasure Val. Potato Bargaining Ass v. Ore-Ida Foods, Inc. 497 F.2d 203, 208 (9th Cir. cert. rknied 419 U.S. 999 (1974). The problem is that complaint counsel's circumstantial evidence provides more support for the proposition that both parties acted independently in deciding their prc- and post-agreement strategies than for the inference that they entered into an illegal combination. And, even if the circumstantial evidence here gives equal support to inconsistent conclusions, it is insufficient to establish JiabiJity. Pevely Dairy Co. v. United St.at€S 178 F.2d 363 , 370 (8th Cir. 1949), cert. denied 339 U.S. 942 (1950).

(42) We concludc that both A&P and Borden acted as rational and law abiding businessmen in this regard. See Callaway Mills Co. v. FTC 362 F.2d 435, 441-42 (5th Cir. 1966). They did not need to engage in an ilegal combination because maintenance of the retail price of milk was in the independent interests of each. "An inference of conspiracy would only arise from similar business conduct if it appeared more to the interest of competitors to adopt different practices. Indpendnt Iron Wm'ks, Inc. v. United States Steel Cor. 177 F. Supp. 743 , 747 (N. D. Ca1. 1959), aff'd 322 F.2d 656 (9th Cir. cert. rknied 375 U. S. 922 (1963). That A&P did not lower the price it charged consumers and that Borden did not immediately lower the price it charged its purchasers " should bc deemed neither surprising nor illegal" North Carolina v. Chas. Pfizer Co. 384 F. Supp. 265, 28 (E. C. 1974). Therefore, we reject complaint counsel's arguments and affirm Judge Hinkes' dismissal of Count III.

Y. A&P S DCE PROCESS CLAIM A&P asserts that it was denied due process of law by the Commission s delay in instituting these proceedings. It makes this claim as to all three counts. Because Counts I and III are dismissed, we respond to the claim only as it relates to Count II. As early as February of 1967 , A&P was informed that the Commission had commenced an investigation of its private label agrecment with Borden and that thc purpose and scope (43) of thc investigation was to determine whether an ilegal price discrimination in violation of the Robinson-Patman Act and the Federal Trade Commission Act had occurred (see Affidavit of Richard A. Palewicz attached to Complaint Counsel's Answer to A&P' s Motion to Dismiss Oct. 20, 1972). At that point, less than 11/2 years after thc agreement ), . . .

GREAT ATLANTIC & PACIFIC TEA CO., INC., ET AL. 1069 962 Opinion and less than nine months after the renewed inducement see note 27 supra A&P was on notice of possible litigation, especially with regard to the Robinson-Patman Act. 42 If A&P was, in fact, indifferent to this possibility of litigation, the Commission is not at fault. Certainly, A&P' s lack of concern does not lead us to conclude that A&P was denied due process. In addition, we reiterate our position as stated in our Order Denying A&P's Motion to Dismiss, Jan. 19 , 1973, that considering the complexity of this case as it was presented and argued by complaint counsel and A&P , the time taken to conduct the investigation leading to the issuance of the complaint in 1971 was not excessive. Id. at 4.

Because we have concluded that the delay in notifying A&P of the investigation and in brir,ging the complaint is not unreasonable, we do not have to reach the question of prejudice to conclude that A&P has not been denied due (44J process'3 Unit/Jd States v. Wilson 357 F. Supp. 619, 620 (E. n. Pa. 1973); see Unit/Jd Stat€S v. Marin 404 U. 307, 324 (1971). However, because we invited A&P to renew this argument (Order at 5, Jan. 19, 1973), we address it now. We agree with Judge Hinkes (I.D. 213) that any prejudice that A&P has suffered from the passage of time is minimal and no more than that suffered by complaint counsel. Therefore, we reject A&P' claims. VI. TJ-Il; REMEDY Since we have decided to dismiss Counts I and III , the final order refers only to the Count II violation. We have made three substantive changes in Judge Hinkes' order. First, the ordcr states that in the future A&P has the burden of going forward with the meeting competition defense. Of course, this provision comes into effect only when A&P knows or has reason to know that prices offered to it arc lower than those offered to others. We believe that such a requirement is not inconsistent with Antmuatic Canteen Co. v. FTC, supra 346 u. at 79 n. 23. See Arrrian Motor Specialti/Js Co. 55 F. C. 1430, 1466-67 (1959), aff'd 278 F.2d 225 (2d Cir. cert. denied 364 U. S. 884 (1960)44 As (45 J demonstrated in this case, evidence that the seller s offer was .. In 1966, the Justice Department initiate 3n invegtigation of the agrement to a.rtin whether the ment violate the oroer illued against Borden for di. rrimination in fllvor of A&P and Jewel Te1i Company (Complaint wunsel's Answer at a, OdOlJr ZO, 1972) 4J It should be notel the.! the cases defining unr(asonable delay cited by A&P!lre criminal C1S where, unlih the situ!ltion in an F, C. proeding-, tbe Sixth Amendment and other re. rictive procures and safeguards arc appliClhle, See, e. n1li7U; Parts Co, FTC 445 F. 1382, 1387- 1388 (5th Cir. 1971); United Srat &"1Idtree 2d84 852(5their 19W) .. The Court in M1d-Smdh Disr,ri&ul(1, v, FTC, S'upra 287 F.2d al517 state that the Commis.ion h the lJUroen af going forwaro with evidenr.e to show thlt the buyer knew that tf. Iier muld not lirt the me€ling competition defense, Although we ;ire nol convinced that complaint Clun5€1 have this bunien, we note that they a..'\sumed it and Rustaineditinthisc""e Of cour.e, th., Commission can "forbid acl lawful in them."!eives lor J compel affinnat;ve act. of lX'mplianCi (QmtinW!d) . & Opinion 87 F.

made to meet a competing seller s offer is more readily available to the buyer than to the seller. Also, this new provision is not only reasonably related to the unlawful practice found to cxist, see FTC v. Colgate- Palrrlive Co. 380 U. S. 374 394-95 (1965); FTC v. Nat'wnallAad Co. 352 U. S. 419, 429 (1957), but is, in fact, directly related. Borden stated explicitly that it was relying on the mecting competition defense to justify the price it quoted to A&P. A&P knew that Borden had, in fact beaten the competitive offer yet did not communicate this to Borden. If it had, Borden might very well have withdrawn its bid. See Kroger Co. v. FTC, supra 438 F.2d at 1376-77. We believe that this provision will render it substantially less likely that A&P will violate the Robinson-Patman Act in the future.

Second, the order req ires A&P to distribute a copy thereof to A&P' s suppliers of milk and other dairy products as well as to all of its operating divisions. By so doing, A&P will make suppliers aware that A&P is subject to an order that prohibits it from violating the Robinson-Patman Act. The notification requirement is also related to the (46J unlawful practice. See FTC v. Colgate-Palrrlive Co., supra; FTC v. National lead Co. , supra. The Robinson-Patman Act was enacted in large part to prohibit large buyers, especially chain stores from using their purchasing power to pressure suppliers into granting unfair discounts. See, e. R. Rep. 1\0. 2287, 74th Congo 2nd Sess. 3 (1936): 80 Congo Rec. 6621 , 7324, 7887, 8104 (1936). The courts have also emphasizcd the need to prohibit prcssurc tactics from large buyers. , FTC V. Henry Brach 363 U. S. 166, 168 (1960). The record in this case shows A&P exerted such pressure on Borden. Suppliers should be made aware that if A&P exerts similar pressure on them such a practice is illegal.

Third, the wording of the cost justification proviso is changed to make it consistent with the holding in Au.tomatic Canteen Co. v. FTC supra 346 U. S. at 74.' Complaint counsel object to the wording of the proviso because it does not place upon A&P the burden of going forward with the cost justification defense. Pursuant to Automatic Canteen, id. at 79, we find that either the seller or complaint counsel (if the Commission has brought a Section 2(f) complaint) must retain this burden because both are better able to develop the relevant information than is the buyer.

(47J A&P objects to the ordcr because it is nationwide in scope. We to restore and rehabilitate comp€titive conditions.EM" Pnxfw:I- Coo C 116:-11216 (196); seeFTCv. Colgaw- Palmolit' e Co. 380 V. S .174394-95 (196). Therefore, even if this burden is more stringent than that de1crilJ in Mw- &uth Di tridon8 it C3n b€ impoRerl on A&P as ar. exercise of the Commil'ion B remedial powers.Eg., FTC Ruberol Qi, 343U.S. 470, 473 (1952) ., We concur with the Court in Mid-Smith I)ilri!ni.urs FTC, Irupm 287 F.2d at 520, that th..Auhnnti. Cantun Court :; use of double negatives is "awkward. " Nevertheles.s, we believe that such us. conceivably makes the Btandllrd more restrictive than thepositiveapprollch and therefore, we have decided to retain it in this order. GREAT ATLAKTIC & PACIFIC TEA CO., INC., ET AL. 1071 962 Opinion see no injustice in making the order nationwide. In finding the nationwide order appropriate, we have considered the frequency and duration of the violations, the business and competitive history of A&P, and the likelihood that A&P knew that it was violating the Robinson- Patman Act. See Joseph A. Kaplan Sons, Inc. v. FTC, 347 2d 785, 789 (D.C. Cir. 1965). A&P's private label contract was the result of a nationwide program. It was initiated and supervised by A&P' s New York dairy headquarters. This specific aspect of the nationwide program covered more than 200 A&P stores located in Indiana and Ilinois, lasted for more than six years, and resulted in substantial discriminations against competitors of A&P. The purpose of A&P's program was to increase its gross margin on milk products an extremely important line, and thus to improve its overall gross profit position.

Complaint counsel object to the order because it does not include all products purchased by A&P for resale. Although we find some merit in the position that the conduct challenged in this case related to A&P' entire business and that A&P could conceivably evade the order by turning its attention to other products, wc cannot ignorc the realities of the kinds of negotiations that go on with respect to a company with thousands of stores dealing in tens of thousands of products. We also note that only A&P's dairy headquarters was involved in this contract. Although we believe that an (48) all products order could be enforccd see FTC v. Ruberoid Co., supra 343 U. S. at 473; Jacob Siegel Co. FTC 327 U. S. 608 , 611 (1946); Forerrst Dairis, Inc. v. FTC, supra 348 F.2d at 681- , we have decided to exercise our discretion and limit the order to milk and other dairy products. We reject A&P' s claim that "it is impossible to infer that, unless an order is entered, A&P wil hereafter engage in the conduct alleged in the Complaint." (Appeal at 71). As noted, the purpose of A&P' program was simply to increase its gross margin on milk and other dairy products and thereby to improve its overall gross profit position. If merchandising methods now being relied upon by A&P to increase revenues fail, A&P might very well again feel the need to pressure small dairies to give A&P an unlawful discount and thereby increase its gross profits.

We also reject the contention that this case is stale. Cf Columbi Broadcasting System, Inc. v. FTC 414 F.2d 974, 981-82 (7th Cir. 1969), cert. denied 397 U. S. 907 (1970). In fact, A&P did not terminate its Final Order 87 F.

relationship with Borden until 1972, four months after issuance of the complaint and five years after commencement of the investigation. The dairy and retail food industries have not drastically changed in the last few years.

(49) Finally, we note that, although not specifically mentioned in the order, A&P is still entitled to all statutory defenses in any subsequent action enforcing the order. See FTC v. Henry Broch Co. 368 U. 360 , 366-67 (1962); FTC v. R"beroid Co., s"pra 343 U. S. at 475-76. Thus, for the foregoing reasons, we find that A&P violated Section 2(f) of the Robinson-Patman Act. We, therefore, enter the appended order.

FINAL ORDER (lJ This matter having been heard by the Commission upon the appeal of Thc Great Atlantic & Pacific Tea Company, Inc. (hereinafter A&P") from that portion of the initial decision dealing with Counts and 11 , and upon the appeal of c:complaint counsel from that portion of the initial decision concerning Count III; and The Commission having considered the oral arguments of counsel their briefs, and the whole record; and The Commission for reasons stated in the accompanying opinion having granted the appeal of A&P concerning Count I but otherwise denying the appeals of A&P and complaint counsel; accordingly (2 J It is ordered That, except to the extent that it is inconsistent with the Commission s opinion, the initial decision of the administrative law judge be, and it hcrcby is, adopted together with the opinion accompanying this order as the Commission s final findings of fact and conclusions of law in this matter: and It is furthrr ordered That the following order be, and it hereby is entered:

ORDER It is ordered That A&P, a corporation, and its officers, representatives, agents, and employees, directly or through the use of any other device in connection with the offering to purchase or purchase of milk and other dairy products in commerce, as "commerce " is defined in the Robinson-Patman Act, for resale in outlets operated by A&P, do forthwith cease and desist from directly or indirectly inducing, receiving, or accepting from any seller a net price that A&P knows or has reason to know is below the net price at which said products of like .6 A&P gt.te that it terrninilt€darrangements with &rden o.Glmsc it was offered lower prices by other dairy Bupplicrs(Appealalln.

GREAT ATLAKTIC & PACIFIC TEA CO., INC., ET AL. 1073 962 Final Order grade and quality are being sold by such seller to other purchasers with whom A&P is competing; provided, however that this prohibition shall not apply if A&P did not know and had no reason to know that the price difference in its favor did not make due allowance for cost differenccs resulting from differing methods or quantities in which such products are sold or delivered to such purchasers, or if A&P can show that said price was granted to it by the supplier to meet a competitor s equally low price, which (3 J price A&P reasonably believed to be lawful. For the purpose of determining the " net price under the terms of this order, there shall be taken into account all discounts and other terms and conditions of sale. It is furtlwr ordered That A&P shall forthwith distribute a copy of this order to its operating divisions and to its suppliers of milk and other dairy products.

It is furtlwr ordered That A&P notify the Commission at least thirty days prior to any proposed change in A&P's structure, such as dissolution, assignment, or sale resulting in the emergence of a successor corporation, the creation or dissolution of subsidiaries, or any other change in the corporation, which may affect compliance obligations arising under this order.

It isfurtlwr ordered That A&P shall, within sixty (60) days after the effective date of this order, file with the Commission a written report setting forth in detail the manner and form of its compliance with this order.

Not having participated in the oral argument in this matter Chairman Collier did not participate in the resolution of it. Complaint 87 F.