Full opinion text
I GENERAL BACKGROUND BOOTLE, Senior District Judge: This is a civil antitrust action for treble damages and injunctive relief under the antitrust laws. The plaintiff, a dairy processor, is a Georgia corporation with its principal place of business in Columbus, Georgia. The defendant is a milk marketing association with its principal place of business in Louisville, Kentucky. This court has jurisdiction over the subject-matter and parties, and venue is properly laid in the Middle District of Georgia. This action was commenced in August 1973 and was tried to the court without a jury in May and June of 1980. The trial itself lasted three weeks and resulted in a transcript in excess of 2,800 pages, plus uncounted pages of exhibits and depositions and transcript of testimony from trials of other cases. Originally, the plaintiff alleged violations of Sections 1 and 2 of the Sherman Act (15 U.S.C. §§ 1 and 2 (Supp. 1980)), Sections 2 and 3 of the Clayton Act (15 U.S.C. §§ 13 and 14 (1973)), Section 4 of the Agricultural Fair Practices Act of 1967 (7 U.S.C. § 2303 (1973)), and breach of contract. By time of trial, however, the scope of the complaint had narrowed considerably, the plaintiff charging only violations of Sections 1 and 2 of the Sherman Act. The plaintiff charges that the defendant, acting alone and in concert with other milk producers, and other milk marketing associations, attempted to and did successfully restrain trade and commerce in and monopolize the supply of raw Grade A milk. The plaintiff complains that the defendant entered into a series of anticompetitive arrangements with other cooperatives with the purpose and effect of obtaining or maintaining monopoly power, and undertook specific predatory acts against dairy processors, including the plaintiff. The defendant’s response to these charges is twofold. The defendant first contends that its conduct amounted to legitimate marketing behavior which did not violate the proscriptions of the Sherman Act. The defendant also contends that even if its behavior might ordinarily be in violation of Sections 1 and/or 2, nevertheless, as an agricultural cooperative, it is entitled to the benefit of the antitrust exemptions provided by Section 6 of the Clayton Act (15 U.S.C. § 17 (1973)) and the Capper-Volstead Act (7 U.S.C. § 291 (1980)) (hereafter referred to collectively as the Capper-Volstead Act or the C-V Act). The facts involved here do not lend themselves to a simple recitation. The case bristles with a multitude of intertwined issues. Because of this, it is extremely difficult to “compress them into an opinion that will not be of fatiguing prolixity.” To resolve the issues involved, it is necessary to have a knowledge of the parties to this action and to analyze each activity of the defendant which serves as the basis of the plaintiff’s charges. Accordingly, this memorandum opinion is intended to suffice as findings of fact and conclusions of law under Rule 52, Federal Rules of Civil Procedure, and its form is as follows: I GENERAL BACKGROUND II THE PARTIES p. A. The Plaintiff 613 B. The Defendant 613 (1) Formation and Structure 613 (2) Functions 613 III ACTIVITIES OF DEFENDANT 614 A. Joint Activities 614 (1) The Georgia Super Pool 614 (2) Great Lakes-Southern Milk, Inc. 615 (3) The Standby Pool 615 B. Unilateral Activities 617 (1) Premium Pricing Policy 618 (2) Disputes with Handlers 618 (a) Sealtest-Nashville 618 (b) Sealtest-Atlanta 619 (c) Sealtest-Chattanooga 619 (d) Broughton Co. 619 (e) PetJackson 619 (f) Mayfield Dairy Farms 619 (3) Written Supply Contracts Note 620 (4) Supply Cut-offs 621 (a) Kinnett 621 (b) Borden 622 (5) Hauling Cut-off 622 IV FACTUAL BACKGROUND AS TO SHERMAN ACT SECTION 2 CHARGE 623 A. Facts Relating to Product Market 623 B. Facts Relating to Geographic Market 624 (1) The Area in Which Seller Operates 624 (2) The Area to which Buyer Can Practicably Turn 624 (3) Down Allocation, Compensatory Payments and Base Plan 625 (4) The “Submarket” Contention 627 C. Facts Relating to Market Power 628 (1) Defendant’s Share of the FMO-7 Area 628 (2) Defendant’s Share in Each of its Four Proposed Alternative Markets 629 (3) Defendant’s Share in Plaintiffs Proposed Nine-Federal Order “Sub-market” 629 V CONCLUSIONS OF LAW AND APPLICATION OF THE FACTS TO THE LAW 630 II THE PARTIES A. The Plaintiff. The plaintiff, Kinnett Dairies, Inc. (Kin-nett) is a closely held Georgia corporation with its principal place of business in Columbus, Georgia. Kinnett is primarily engaged in the processing of Grade A milk for fluid consumption. To a lesser extent, it is also in the business of manufacturing ice cream. B. The Defendant. (1) Formation and Structure Dairymen, Incorporated (D.I.) is a nonprofit agricultural milk marketing association organized and existing under the laws of the state of Kentucky and qualifying as a Capper-Volstead marketing association. D.I. was formed in 1968 as a combination of and successor in interest to eight cooperatives located throughout the southeastern United States. Today, D.I. has ten separate divisions serving this area. These geographically distinct operating divisions are responsible for the day-to-day marketing operations carried on by the cooperative, including quality control, hauling arrangements with proprietary handler customers assigned to the division, and resolution of handler and/or producer complaints. The membership of D.I. consists solely of dairy farmers (“producers”) who produce raw milk to be marketed for commercial use in dairy products. Each D.I. member signs a Membership Marketing Agreement which gives D.I. authority, inter alia, to market the member’s milk. Since its formation, D.I. has had a board of directors elected under a delegate system from among its members. This board sets the policies and has final authority over all major decisions which may affect the cooperative as a whole. (2) Functions Some of the reasons the defendant offers for its existence are: (a) to give producers more representation in the marketplace; (b) to ensure that producers obtain a fair, secure, and guaranteed market for their product; (c) to provide its members some countervailing power with which to bargain with processors for prices above the federal order minimum price and more reflective of market conditions; (d) to provide members with improved and more efficient assembly and distribution of milk; and (e) to accomplish more easily by consolidated effort the intermarket alignment of raw milk prices. Perhaps the best “nutshell” statement of the function of D.I., and of marketing cooperatives in general, is that they exist to provide their members with the best return possible for their product. Raw milk can be utilized in one of two ways. It can be used for fluid consumption (Class I utilization) or it can be used for making certain “manufactured products” like butter, cheese, or ice cream (Class II and III utilization). A dairy farmer can get more money for his milk if the processor uses it for “fluid” purposes. Consequently, these farmer producers desire that as much of their milk as possible obtain a Class I utilization. Cooperatives like D.I. exist to help the producers obtain this goal. Of course, the producer would also like to avoid having to dump his surplus milk (milk in excess of the fluid milk needs of processors) down the drain. The cooperative aids the producer in this respect by finding Class II and Class III markets for the surplus milk. Ill ACTIVITIES OF DEFENDANT The plaintiff contends that the defendant engaged in a course of conduct designed to obtain and maintain the power to control prices and exclude competition in the supply of raw milk in the Georgia marketing area, and that this conduct exceeded any legitimate functions of a cooperative bargaining for the sale of its members’ milk. The conduct complained of can be divided into two categories: D.I.’s involvement with other marketing associations (joint activities), and its individual conduct toward Kinnett and other dairy processors (unilateral activities). A. Joint Activities In its most recent brief (Plaintiff’s Reply to Defendant’s Post Trial Brief) the plaintiff contends that the following listed incidents of joint activity by the defendant evidence a scheme of cooperative monopolization which includes predatory conduct and that such conduct is neither legitimate nor immune from suit. The defendant, on the other hand, contends that all this activity is a part of the “legitimate objects" of a Capper-Volstead (C-V) organization and is a part of the conduct of “processing, preparing for market, handling and marketing ... and hav[ing] marketing agencies in common ... [and] mak[ing] the necessary contracts and agreements to effect such purposes.” (1) The Georgia Super Pool The Georgia Super Pool is a group of C-V cooperatives in Georgia which appoints one of its number to serve as the marketing agent for the members. D.I. has been this common marketing agent since 1970. The pool has two main functions: disposal of the surplus milk of its members; i. e., milk not needed by the members for their own processing plants, and establishing a formula for distribution of milk payments received which are in excess of the federal order minimum price for the FMO-7 order. The plaintiff contends that both functions of the Super Pool operate to neutralize competition among the cooperatives in Georgia and to give D.I. effective control over the outside disposition of the bulk of locally produced milk. It contends that, since 1969, substantial amounts of surplus milk have been produced by the Super Pool members and, that in the absence of the Super Pool, these members would have been willing and able to sell surplus milk to Kinnett. The defendant’s response is that without the Super Pool, members would have had to make “distress” sales at prices below the cost of production in order to get rid of their surplus milk. It contends that the formula for distributing premiums among its member cooperatives does not stifle competition, but rather, provides for an equitable distribution which serves as an attempt to avoid retail price wars which had been so harmful to Georgia dairy farmers in the past. It contends further that the Capper-Volstead Act not only permits but encourages agreements among cooperatives to dispense with competition among themselves. (2) Great Lakes-Southern Milk, Inc. Great Lakes-Southern Milk, Inc. (GLSM) is a federation of dairy cooperatives covering a geographic area more or less reflected by its name. D.I. has been a member of GLSM since October 1968. The stated goal of GLSM at its formation was to put an end to processors “playing one market or one cooperative against another” in order to bring down the raw milk prices. Its functions include: improving the prices of the dairy farmer members of the member cooperatives, attempting to assure that the prices in the various markets served by member cooperatives are in alignment, providing a forum for discussion of Class I prices and alignment, and making recommendations for Class I pricing and price alignments. Kinnett charges that the federation effectively eliminates competition in the sale of raw milk and confers upon its members the power to control prices throughout the designated market served by each member. It complains that D.I., and the other member cooperatives of GLSM, coordinated their pricing changes and that this alignment eliminated the obstacle associated with raising Class I prices — the failure of cooperatives in neighboring markets to raise their prices as well. D.I. counters that GLSM is a necessary organization by which producers can meet processors on equal terms; that its marketing information and pricing recommendations aid in establishing an orderly market; i. e., one in which there is an economic movement of milk and the existence of a general understanding of what all markets are doing; and that because of milk’s perishability, this marketing information is necessary to facilitate decisions by members concerning supply, volume, alternate sources, where to dispose of excess, and other marketing needs. (3) The Standby Pool The Standby Pool (ARSPC) is a federated agricultural cooperative whose members are Capper-Volstead cooperative associations. It performs two related operations: the establishing of a “reserve supply” of milk by purchasing “options” on unregulated milk in the upper Midwest and, the purchasing and sale of milk from that reserve supply. It is possible to establish a reserve supply because of the chronic surplus of Grade A milk which exists in the upper Midwest area, primarily in Minnesota and Wisconsin. What ARSPC does is purchase, “options” on much of this surplus Minnesota and Wisconsin milk. This option requires participating manufacturing plants in those areas to hold this surplus milk for a certain length of time before using it for their own purposes. If ARSPC does call on the milk for use in a market to the South, they will pay for the milk itself. This payment would be in addition to the option payment. If ARSPC has not exercised its option on the milk within the specified time, then the manufacturing plant is free to use it in any way it wishes. For the purposes of this case, these manufacturing plants were all owned by farmer cooperatives. (Tr. 204, 207-208; Stip. ¶ 217). The plaintiff contends that ARSPC operates as a device by which cooperatives in the South can obtain control over the Minnesota-Wisconsin supply of unregulated surplus milk to keep it from being sold in markets to the South in competition with those cooperatives. In the absence of the Standby Pool, Kinnett argues, this great surplus of Midwest milk would be attracted to southern markets because of the premium prices for milk there, and that through ARSPC, southern cooperatives like D.I. paid producers in the high production states to keep and process their milk in that area and not to sell it in markets to the South in competition with these cooperatives, i. e., they protected premium prices by keeping milk off the market. The defendant contends and the court finds that there is a need to maintain a reserve supply of milk. Supply and demand in the southern milk markets are cyclical; i. e., there are times of peak and times of low supply and demand. Unfortunately, when the supply of milk for the market is lowest, the demand tends to be the highest, and vice versa. In these “short seasons”, when total milk supply is less than or equal to the fluid milk demand, the chronic surplus of Grade A milk in the upper Midwest can provide the needed reserve supply for the southern states. Membership in ARSPC would be beneficial because it would provide the means for obtaining dependable supplemental Grade A milk supplies at reasonable predetermined prices, thus eliminating the need to develop and maintain costly year-round local reserves. For ARSPC members, payments made to fund the option payments to the supply plants were like insurance payments to have milk available when needed. The defendant denies that the Standby Pool was intended to prevent Kinnett or anyone else from obtaining supplemental milk. As support, it notes that from the Standby Pool’s inception there have been at least eight billion pounds of Grade A milk in Minnesota and Wisconsin annually on which the pool has no options, which was not used for local fluid sales, and which could have moved freely to Kinnett. The pool’s options have rarely exceeded approximately 1.6 billion pounds of milk annually. The plaintiff’s response to this is that although ARSPC’s options on the total amount of Grade A milk in the M-W area were relatively small, it did have options on the bulk of the unregulated milk in the area. The significance of this, according to the plaintiff, is that it is this unregulated milk which is most likely to be attracted to the southern markets. The plaintiff also contends that the amounts of milk under contract with the Standby Pool greatly exceeded the total needs of its members for reserve supplies. The plaintiff points out that, even in the fall of 1973 when there was supposedly an “emergency” shortage condition, only 39.6 percent of milk committed to ARSPC was shipped in any one month, and never in the Standby Pool’s history up to 1976 did amounts shipped exceed 11.1 percent of the yearly amounts under option. The defendant, of course, disagrees with this assessment and points out that there have been times when the amounts of milk under option to ARSPC have been barely sufficient to meet the supplemental fluid needs of its members during the “short” seasons. As an example, in 1975 and 1976 ARSPC had the milk of twelve supply plants under option. Taking a sample of seven days, there were twenty-two occasions on which the ARSPC shipments exhausted the supplies available in certain of the supply plants: And during the period 1972 through 1975 the following amounts were shipped: 1972 — 122,278,000 lbs.; 1973 — 233,755,000 lbs.; 197^82,223,-000 lbs.; 1975 — 95,374,000 lbs. (total amount shipped equaled 533,630,000 lbs.) B. Unilateral Activities The plaintiff, in its most recent brief, lists at length its complaint against the following unilateral activities of the defendant: (1) announced non-negotiable premium prices unilaterally; (2) cut off Kinnett’s profitable hauling operation in retaliation for Kinnett’s having dared to compete legitimately for supplies from individual farmers; (3) repeatedly threatened to, and often actually did, cut off any customer who tried to replace any portion of DI’s supplies, regardless of the volume involved with the contemplated substitute, the time of year, the notice afforded to DI (oftentimes more than DI gave its customers prior to cutting them off), the reason for the proposed substitution or its source; (4) successfully warded off the entry of a potential competitor, NFO, into the Georgia market by cutting off its entire supply to Sealtest-Atlanta, some 75%-80% of that plant’s requirements in retaliation for Sealtest’s attempt to obtain some 20%-25% of its needs from NFO; (5) subsequently insisted upon full supply contracts and, following successful litigation against it, committed volume contracts which tended to lock processors in to dependence on DI and lock potential competitors out of the processor market for their milk; and used additional, thinly veiled, cutoff threats to induce processors to sign the contracts; (6) cut off Kinnett’s entire DI supply on no notice in the middle of the tightest season in years, in deliberate breach of the committed supply contract which DI itself had insisted upon; (7) continued, even after this action was filed, to take affirmative steps in an effort to stop Kinnett from receiving alternative sources of supply, including continued threats not to deal with Kin-nett (or to sell to it only at special supplemental milk prices) and renewed insistence on year-long committed volume contract, this time one which contained no ceiling or limit on the prices DI could impose unilaterally. The plaintiff recognizes that these actions, in themselves, would not violate Section 1 of the Sherman Act because a single cooperative should be treated as a single entity, the equivalent of a business corporation, and Section 1 does not touch unilateral activity of a single enterprise. The defendant contends that all of such activity on its part was lawful and supported by ample legitimate business justification and in fact was necessitated to protect the interests of the milk producers, and that moreover, all of it was a part of marketing and was covered by the Capper-Volstead exemption. (1) Premium Pricing Policy The plaintiff complains of the allegedly excessive prices the defendant charged for its milk. To understand the arguments surrounding this “premium” price, it is first necessary to understand the nature of the Federal Order System. Under this system, minimum prices are established by the United States government for Class I and Class II milk, with Class I milk having a higher minimum price. All handlers (like Kinnett) regulated on a particular order must pay this minimum price for the milk they receive from the producers. All the producers share equally in the order revenues from the combined higher priced Class I sales and the lower priced Class II sales through the receipt of an announced federal order “blend” price. The federal order does not establish the maximum price which producers may receive for milk which handlers purchase. Any price they charge which is in excess of this minimum price is an “over-order premium.” The plaintiff complains that the over-order premiums set by D.I. were artificially and unreasonably high, and contain what plaintiff refers to as a “monopoly increment.” It is the defendant’s contention, however, that it was necessary to charge the premium it did for its members’ milk because the minimum prices for Grade A milk in the Georgia federal order did not establish realistic prices. The Class I differential has remained basically the same since 1969 in spite of rapidly increasing transportation costs, which today are far above 15 cents per 100 miles. The defendant claims that because of these increased transportation costs and inflation, over-order pricing is necessary to attract an adequate supply of milk for Georgia handlers. (2) Disputes with Handlers (a) Sealtest-Nashville: Prior to their disputes, D.I. (and its predecessor) had been supplying Sealtest-Nashville with the majority of its fluid milk needs. At a February 11, 1970 meeting between Sealtest and D.I. officials, Sealtest attempted to persuade D.I. to reinstate special pricing provisions which had existed for sales made by Sealtest in the southern Kentucky area, but which D.I.’s predecessor had terminated. Sealtest, failing to get these provisions reinstated, undertook to replace about 10 percent of the milk D.I. was supplying by placing sales with National Farmers Organization (NFO), a rival cooperative. When D.I. heard about this, it put Sealtest-Nashville on an irregular basis in supplying it milk, and notified it that all milk sold would be at Class I prices. This first dispute was resolved in April 1970, when the defendant agreed to restore the special pricing provisions and Sealtest granted the defendant the right to supply it with all milk, except that supplied by certain NFO producers. Subsequently, other disputes occurred between the parties, the net result of which was a reduction in Sealtest-Nashville’s purchases of D.I. member milk from 100 percent in early 1970 to 37 percent in 1974. It should be noted that Sealtest often gave the defendant no notice as to when their sales were to be reduced, and used its NFO agreement as a wedge to secure concessions from the defendant, just as the defendant was using its strong supply position as a wedge to force Sealtest to accept a full-supply agreement. (b) Sealtest-A tlan ta : In the autumn of 1970, the defendant supplied some 60 percent of Sealtest-Atlanta’s raw milk needs. The defendant claims that Sealtest substantially reduced its purchases from D.I. with little notice. This reduction occurred around Christmas 1970, when Sealtest gave D.I. three days notice of a 17 percent reduction of milk purchases from D.I. In retaliation, D.I. informed Sealtest that if it replaced some milk it would have to replace all of its purchases from D.I. Sealtest did this; however, it experienced “quality” problems with its new supplies and consequently resumed its purchases with the defendant. However, this settlement was only temporary. Other disputes occurred, and today D.I. sells little milk to Sealtest-Atlania. (c) Sealtest-Chattanooga: Sealtest’s processing plant in Chattanooga purchased 100 percent of its raw milk needs from D.I. during 1970. In September of 1970, D.I. increased its price for Class I milk and Sealtest began to look for other sources of supply, to replace its purchases from D.I. On June 22, 1971 it informed D.I. that it had discussed an alternative milk supply of an unspecified quantity with NFO. D.I. informed Sealtest that it was not interested in being a partial supplier and warned that it would refuse to sell any milk to Sealtest if the NFO milk was purchased. In response to this Sealtest, on two days notice, replaced D.I.’s entire supply of raw milk, and purchased no more raw milk from D.I. before going out of business in 1973. (d) Broughton Company: Since 1968 D.I. and predecessor cooperatives had supplied Broughton the entire raw milk requirements for its Lexington, Kentucky plant. On or about April 20, 1971 Broughton advised D.I. that, beginning May 1, it would accept approximately 40 percent of its total raw milk requirements from NFO. Broughton apparently gave D.I. no prior notice of this substitution and, on May 1, returned three of the four tankers of milk it had ordered from D.I. On May 2, 1971 NFO began supplying all of the requirements of the Broughton-Lexington plant, and no subsequent purchases have been made from D.I. members. (e) Pet-Jackson: In 1969 the defendant, which was supplying approximately 55 percent of the fluid milk needs of the Pet-Jackson plant, notified that plant that it would cut off its supply if Pet-Jackson did not agree within thirty days to purchase its full supply from the defendant. D.I. claims that it made the threat because of the Pet plant’s tactics of periodically reducing its purchases from D.I. with little notice and of turning away split loads of milk tendered to it by the defendant. The threat was never carried out, since negotiations between the defendant and Pet-Jackson resulted in an oral agreement wherein Pet-Jackson was allowed to continue buying milk from certain other producers. Pet-Jackson eventually ceased purchasing any milk from D.I., and purchases none from them at present. (f) Mayfield Dairy Farms: The Mayfield incident primarily involved extended negotiations on service charges and on how May-field’s supply needs would be divided between the defendant and certain independent producers. Mayfield Dairies purchased approximately 50 percent of its fluid milk from D.I. and 50 percent from independents. In September 1970, D.I. gave Mayfield thirty days to consider entering into a full-supply contract, later extending this time to December 1970. As a result of negotiations between the parties, it was finally agreed that Mayfield would pay service charges equaling approximately two-thirds of the amount which had been asked for by D.I. In addition, Mayfield agreed not to take on any additional independent producers as sources of supply over the amount being supplied by the independent dairy farmers then selling to Mayfield. However, Mayfield maintained these existing independent sources under the agreement’s “grandfather clause.” (3) Written Supply Contracts The defendant contends that it was these frustrating experiences with processors that led its Board of Directors to authorize D.I. managers to seek written sales agreements for transactions between the members and fluid processor customers. D.I.’s marketing department drew up a form contract and, during the summer months of 1971, this contract was presented to each fluid processor customer regulated on a federal order. The plaintiff complains about several terms in the contract, but the main contention centers on the contract’s supply provision. The original proposed contracts were termed “full supply” contracts, requiring the customer to purchase from D.I. all of its needed supplies of raw milk, except for those then being provided by other “grandfathered” suppliers. The proposed contracts were to run for a term of one year, with the term beginning in the fall months when the farmers’ bargaining strength was highest. Doctor Albert J. Ortego, Jr., the defendant’s vice-president of marketing, explained that the “grandfather” provision was inserted because D.I. wanted to start off under the contract from where things were right then, meaning that generally speaking, D.I.’s customers to whom the contract was being presented were at that time receiving their full supply from the defendant, and D.I. wanted to go for a twelvemonth period. “We didn’t want to kick nobody [sic] else out but we didn’t want to sell him no [sic] less milk either.” Doctor Ortego testified further that this was a fully negotiable provision; that D.I. did not expect to “get that with everybody” and that D.I. fully expected that many would not accept the provision and that D.I. would have to sit down at the bargaining table and negotiate about it; that D.I. would ask for a full supply contract (modified by the “grandfather” provision) and failing to get that would settle for a committed supply agreement. A number of Georgia processors did sign the original proposed contract promptly; others, like Kinnett, refused to sign. Two processors, Holland Dairies and the Borden Company, sought and obtained Temporary Restraining Orders against the defendant for activities which occurred as a result of a dispute over this full supply requirement. Because of these problems with processors over the full supply terms, D.I. entered negotiations with several processors, and together they developed an alternative supply provision. This provision committed the customer to purchase, and D.I. to supply, specified quantities of raw milk each month for a year. The plaintiff never signed a full supply contract and did not sign the “committed supply” contract until January 12,1972, making it the last Georgia processor to sign a written sales agreement. (4) Supply Cut-Offs (a) Kinnett. In 1973, a number of factors converged to cause a substantial increase in the cost of milk production. At the same time, the quantity of milk production was decreasing. Because of this combination of circumstances, D.I. approached its handler customers, stating that higher prices were needed for milk at the producer level or a shortage of milk production was bound to result. D.I. proposed to amend the sales agreements it had with the handlers in order to obtain the suggested increase of $.73 per cwt, thus raising the price of Class I milk from $8.13 per cwt to $8.86 per cwt. All of D.I.’s customers agreed to this “Supplemental Agreement” when presented except Kinnett. Kinnett and D.I. had discussed the proposed price increase, and these discussions had led Frank McDowell, Jr., D.I.’s Georgia manager, to believe that Kinnett understood the difficult situation of the dairy farmers and would agree to the proposed price increase. However, at 3:30 p. m. on July 30, 1973, only a day before the scheduled increase was to take effect, Kinnett dispatched a letter by courier from Columbus, Georgia to Decatur, Georgia to inform D.I. for the first time that it intended to oppose the price increase. Kinnett refused to agree to this price increase unless D.I. would assure it that all of any price increases would be passed on to the cooperative members. D.I. considered this to be an attempt to intrude into the dairy farmers’ decision-making process and rejected it. The written sales agreements D.I. had with its customers gave any contracting customer the right to the most favorable terms of trade received by any other customer. Thus, Kinnett’s refusal jeopardized the supplemental agreement with other D.I. customers and the price increase they had agreed to allow. On August 2,1973, D.I. refused to deliver milk to Kinnett except at the increased price then being paid by its other customers. Kinnett refused to accept the milk at that price unless the acceptance was “without waiving any right as to the price of future sales and without waiving any right under the written Sales Agreements.” D.I. attached a notice to its delivery ticket requiring Kinnett to agree that acceptance of milk on that day would constitute acceptance of the higher price. When Kinnett refused to sign the delivery ticket unless D.I. signed the non-waiver statement, D.I. drove its truck away. As a result, Kinnett instituted the present litigation and obtained a Temporary Restraining Order requiring D.I. to continue selling milk to Kinnett at prices established by the contractual formula until the contract expired. During this period, the higher price was being paid by all other Georgia customers of D.I. (b) Borden. A more recent dispute occurred between D.I. and Borden, Inc. concerning supplies to Borden’s Macon, Georgia plant. It occurred at a time wdien Borden was a full supply customer and the parties were operating under an oral agreement and were engaged in a series of discussions regarding price credits which Borden was requesting and D.I. was considering for milk Borden was selling from its Georgia processing plants into South Carolina. On March 25, 1977, at the peak of the spring surplus season, the manager of Borden’s Macon plant advised D.I. that Borden intended to replace approximately 25 percent of the plant’s D.I. milk in three days. D.I. responded by advising Borden that it considered three days notice at the peak of the flush season to be unreasonable, and that if Borden elected to replace D.I.’s milk it could replace all of it at its Macon plant and its Augusta, Georgia plant as well. Borden then instructed D.I. to cease all deliveries at Borden’s Meridian, Mississippi processing plant. D.I. initiated litigation, charging that Borden’s three-day notice of the change in supply arrangements was not commercially reasonable notice as required by the Uniform Commercial Code. An injunction directing Borden to continue purchasing its milk from D.I. was issued. Forty days later Borden succeeded in having the injunction dissolved on the ground that a commercially reasonable time had passed since its initial notice to D.I. It immediately replaced all D.I.’s milk at its Macon plant. Borden-Augusta ultimately reduced its purchases from D.I. from 100 percent to approximately one-third of its needs between March 1977 and August 1978. (5) Hauling Cut-Off Until 1970, Kinnett hauled the milk of each of the farmers supplying it from their farms to the Kinnett plant. This included D.I.’s members as well as Kinnett’s independent processors. It preferred this arrangement because the hauling operation was profitable and because Kinnett believed it could better ensure the quality and flavor of the milk in this way. On March 4, 1969, D.I.’s predecessor, GMPCA, adopted a policy of hauling the milk of GMPCA’s own members to market, and in February 1970 GMPCA told Kinnett that as of April 1, 1970 Kinnett would no longer be permitted to haul the milk of GMPCA members. The general manager for GMPCA, Frank H. McDowell, Jr., advised Kinnett in effect that the reason it was at that particular time implementing, as to Kinnett, that hauling policy was because Kinnett’s employees were successfully attempting to convince GMPCA members to leave the cooperative. In addition, there was concern about mixing member and nonmember milk in the same tanker, i. e., quality control problems. D.I. did not merge with GMPCA until September of 1970, after GMPCA had taken over Kinnett’s hauling. However, after the merger D.I. adopted the GMPCA hauling policy and continued the transformation from processor hauling to cooperative hauling. GMPCA did not treat Kin-nett differently from the way it treated other processors hauling cooperative members’ milk for their own account. It took over the hauling of Pet, Inc. at Waycross in November 1969. It took over both Pet’s Washington, Georgia plant and Kinnett on March 31, 1970. The evidence does not disclose that any hauler was permitted to continue hauling if GMPCA, or later D.I., learned of the hauler’s attempting to persuade the cooperative members to leave the co-op. Other haulers, specifically Vernon Sanders and Jack Edwards, have had hauling arrangements with D.I. terminated as a result of attempts to influence co-op members to leave their co-op. IV FACTUAL BACKGROUND AS TO SHERMAN ACT SECTION 2 CHARGE The offense of unlawful monopolization consists of the possession of monopoly power and the element of deliberateness or general intent or purpose to acquire, use or preserve such power. Monopoly power, the first element to be proven in a monopolization charge, is that power to control prices or exclude competition. To determine whether or not it exists, it is first necessary to define what is the “relevant market” in which this power over prices and competition is to be tested. The relevant market consists of both a product market (the product or services affected) and a geographic market (the area within which trade may be limited). A. Facts Relating to Product Market. In the present case, there is dispute as to whether the relevant product market is all milk as the defendant contends, or only Grade A milk, as the plaintiff contends. The term “Grade A” means raw (unprocessed) milk which has met specified state sanitation, health, and quality standards, and is therefore available for sale for human consumption in fluid form. Producers in the industry are recognized as either Grade A or Grade B. Grade A producers generally receive more for their milk than Grade B producers. To be a Grade A producer, a farmer must meet the standards mentioned above with respect to both production and storage. These standards are much more strict with respect to Grade A milk than Grade B. In addition, production of Grade A milk requires a higher initial capital investment than Grade B production, e. g., installation and refrigeration of bulk tanks for storing milk on the farm. Consequently, a producer does not simultaneously produce Grade A milk and Grade B milk from the same facilities. Grade A milk is distinguishable not only in terms of who produces it, but also in terms of who its customers are. A customer (processor) who processes milk into both fluid and manufactured form at one location can, for sanitation reasons, receive only Grade A into its plant. Also, only handlers receiving Grade A milk from producers are subject to regulation under federal milk marketing orders. Grade A milk is generally recognized in the industry as a distinct product and line of commerce for fluid consumption, even though it is true that Grade A milk can be used for both fluid and manufacturing purposes. This is so because whatever interchangeability there is is one-way; i. e., though Grade A milk may be used for Class I, II or III purposes, Grade B milk can never be used for Class I (fluid) purposes. It should be noted that D.I. itself has a separate division, D.I. Manufacturing, Inc., to deal with manufactured milk products which can be made from Grade B milk. B. Facts Relating to Geographic Market. The second component of the relevant market is the geographic market — that physical area in which sellers of the particular product operate and to which purchasers can practicably turn for such products. It is the plaintiff’s contention that the relevant geographic market is the FMO-7 area, which has the same boundaries as the state of Georgia, except for eight counties in the extreme northwestern corner of the state traditionally associated with the Chattanooga market and one in northeast Georgia, or alternatively the. area covered by nine of the sixteen federal orders under which D.I. operates. The defendant, however, contends that the relevant geographic market encompasses an area much larger than the state of Georgia, including either (1) all states east of the Rocky Mountains, (2) nineteen states in which it markets milk, (3) fourteen states in the southeast in which it markets milk and has members, or (4) at the very least, Georgia plus the surrounding states of Alabama, Tennessee, North Carolina and South Carolina. (1) The Area in Which Seller Operates As for the first prong of the geographic market definition, that area in which the seller operates, the evidence in the present case shows that D.I. has members in fourteen southeastern states and markets its members’ milk in an even wider geographic area. When D.I. was formed in 1968 it took over eight cooperatives located throughout the southeastern states. Its operation extended into Virginia, Louisiana, Kentucky, southern Illinois, and Mississippi. Since 1968 producers in the additional states of Alabama, Georgia, Missouri and Florida have become members of D.I. Its operations have extended also into southern Indiana, Tennessee, parts of West Virginia, North Carolina and South Carolina. In addition, D.I. occasionally markets milk in Ohio, Maryland, Washington, D.C. and Puerto Rico. (2) The Area to Which Buyer Can Practicably Turn In determining whether the plaintiff could practicably turn to an area outside of the FMO-7 area for a source of supply, it is necessary to consider evidence on several points, including the nature of milk and its movement in general, where the plaintiff attempted and succeeded in purchasing milk, and the concepts of down allocation, compensatory payment, and base plan. Raw milk is a perishable product which is subject to bacterial contamination within hours after it is produced. This bacteria growth which can render milk unsuitable for fluid use is a function of time and temperature. Until recently, this time factor prohibited the movement of milk over any appreciable distance. However, two factors have combined to. make long distance movement of milk feasible; they are the completion of the interstate highway system and technological advances in storage and hauling facilities. As a result, milk can move and is being moved long distances. For example, studies of bulk milk shipments for 1975 showed that the average distance was 350 miles. About 40 percent of the shipments were for distances of less than 200 miles, 30 percent were for distances of 201 to 400 miles, 25 percent were for distances of 401 to 700 miles and 5 percent were over 700 miles. In addition to the evidence of the movement of milk generally, Kinnett’s own purchases support the argument for a relatively broad geographic market. During the period from 1968 to the present, Kinnett searched for and actively negotiated with numerous raw milk sources outside Georgia. Some of these attempts to purchase milk were successful; others were not. For example, Kinnett purchased approximately one-half of its raw milk requirements from Tennessee between May 1968 and December 1970 and 20 percent of its supply from Alabama between March 1975 and November 1977. In September 1979 Kinnett began to purchase raw milk supplies from Alabama Milk Producers (AMP), Southern Milk Sales, Inc. and Southern Bell Dairies. This milk is produced in Alabama, Tennessee and Kentucky and comprises over one-half of Kinnett’s supply. It has, since September of 1979, completely replaced D.I.’s milk in Kinnett’s plant. In addition to the above-mentioned purchases, the plaintiff made other, less substantial out-of-state purchases from 1968 to 1979. These purchases were from Nashville and Chattanooga, Tennessee; Louisiana; Mississippi and, in a couple of instances, from Minnesota. (3) Down Allocation, Compensatory Payments and Base Plan Three other factors to consider in determining that area to which the plaintiff can practicably turn for its milk supply are the concepts of “down allocation”, “compensatory payment” and “base plan”. The plaintiff contends that these concepts operate to erect a barrier around the FMO-7 area and, by this isolation, make it the relevant geographic market in the present case. Plaintiff contends that these three concepts impose a regulatory “tax” upon the importation of milk from outside the market order area, explaining, however, that tax must be put in quotation marks because it is not really a tax — “It is just [that] the operation becomes taxing so to speak”; that it raises the cost to the handler and the proceeds of the costs are distributed to the producers in the importing market. The experts explain that the regulatory tax applies only to interhandler shipments like from Sealtest in Chattanooga to Kinnett in Columbus, and that it does not apply to producer milk that goes into an area. This interhandler regulatory tax is broken into two categories: (1) if the interhandler shipment originates in one regulated area and goes into another, we have the “down allocation” of some of that shipment to Class II, and (2) if the shipment originates in an unregulated area and moves into a regulated area, we have the down allocation just mentioned, plus a “compensatory payment” levied on any of the milk that is not down allocated. In these interhandler shipments the shipper usually has more milk than it needs and the purchaser can buy it at a cost lower than its local prevailing price. This regulatory tax fluctuates — the higher the Class I utilization rate goes, in other words, the lower the surplus is, the lower the regulatory tax. This is designed to encourage milk movements only when needed for Class I use and to provide disincentives for milk flows between markets unless there is in fact a true shortage. The term “down allocation” is not used in the FMO-7 but the experts explain it this way. Under the federal order marketing system, Class I prices vary from zone to zone by addition of the mileage increment as we come south. Thus, if Kinnett brings down from Chattanooga a tanker of Class I milk it would be paying in effect the Chattanooga Class I price for that milk in Chattanooga, plus the transportation costs to Columbus and that would be approximately equal to the Columbus. Class I price for the milk. But under the regulatory tax system the administrator is going to require Kin-nett to treat that load of milk as part Class I and part Class II, the Class I amount being the percentage of milk in Kinnett’s plant that is Class I, or the percentage of milk in the entire FMO-7 area that is Class I whichever is lowest. The remaining portion of the load is allocated down to Class II — hence the term “down allocation”. The significance of this down allocation is this: Class II prices do not vary between orders. They are the same everywhere. It is not contemplated that a processor will haul Class II milk great distances to manufacture it into cheese, for instance. That would be what the esoterics refer to as “hauling the water.” It is more economical to utilize the Class II milk wherever it is and then transport the cheese, for instance, at lesser transportation costs. The Class II price in Chattanooga being the same as it is in Columbus, Kinnett is going to have to pay transportation costs over and above the Class II price in Columbus for that down allocated Class II milk. If the applicable utilization rate is 80 percent Class I and 20 percent Class II, this means that Kinnett must be deemed to have used only 80 percent of the shipment for bottling and the other 20 percent for the ice cream line though he may have actually used all of it for bottling purposes. The experts say that this means that Kinnett is going to have to pay the local producers for that additional 20 percent at the Class I price. As noted these regulatory tax factors do not apply to producer milk and do not deter producers in one state from dealing with processors in another. For example, in 1975 about 20 percent of the milk pooled on the Georgia order was produced outside Georgia and was subject to no regulatory tax. When the down allocation does apply, the only thing involved is the transportation costs from one point to the other. To the extent there is down allocation this means only that such imported milk will have the same utilization percentages as the milk of the producers regularly supplying the Georgia regulated plants, and any disadvantage to the handler from receiving other order milk occurs only when transportation costs exceeds the blend price difference between the shipping order and the receiving order. The “compensatory payment” above mentioned applies only to unregulated milk, that is, milk which is not priced under the federal order system. This means that the handler can buy the milk at any price. The compensatory payment means only, according to the experts, that the system is treating the unregulated milk in such a way that it does not receive better treatment than producer milk. Georgia producers have a Class I base plan which followed in time their seasonal base plan. The purpose of the seasonal base plan was to influence producers to produce milk when it was most needed and to produce less at other times. A number of federal orders have such plans. The Class I base plan had the same purpose and also tried to control the total amount of milk produced in the order area. It succeeded in influencing the seasonality of production by evening it out but did not succeed in controlling production. Production in Georgia increased after the institution of the Class I base plan. There are quite liberal and practical provisions in the order for acquisition of Class I bases by new producers whether they are located within or without the order area. Under the federal order classified pricing system each order defines a “marketing area” which is the geographically defined area within which if a processor/handler makes a specified percentage (15 percent in the “Georgia Federal Order,” FMO-7) of fluid milk sales (called “in area fluid route disposition” or “route sales” for short), he becomes subject to price regulation under the order. The marketing area may indicate less, all or more than one state. Because handlers operate their businesses through plants, these orders in practice regulate the plants of the handlers. Thus the plants having the required in area fluid disposition are called fully regulated pool plants. Under certain conditions, plants which produce no fluid products but which sell specified percentages of milk to distributing plants or meet other requirements may be pool plants as well. Federal orders directly regulate handlers — not producers— though producers — not handlers — decide whether to petition for a federal order. Federal orders do not guarantee producers a market — only a minimum price. The federal orders do not directly regulate the location of the farms producing the milk or direct where the milk shall be sold. They do not directly regulate the location of the handler or its plant facilities. The marketing area of an order is only the geographic area in which the handler must make a specified percentage of sales of fluid milk products in order to become subject to the regulations of the order. This is not to say that there is no connection or relationship between the order and the location of the dairy farms and the location of the handlers purchasing the milk. The producers in petitioning for a federal order ask the government to define an area that includes their nearest direct competitors and their major processor outlets. So it is that most of the milk processed in the FMO-7 area is, in fact, produced there and vice versa, and the same is true with respect to D.I.’s overall territory. Classified pricing was designed for two purposes — to maintain an adequate supply of milk by means of prices high enough to encourage local production, and to prevent disorderly marketing conditions as. in cases of price war activity. (4) The “Submarket” Contention The plaintiff concedes that there is a price interrelationship between all, or substantially all, of the federal order areas east of the Rocky Mountains which may be enough to consider — from a highly analytical, economic viewpoint — those areas taken together to be a relevant geographic market. However, the plaintiff contends that the practical ability to transport milk for regular supplies within that area is in fact minimal. Consequently, Kinnett does not see it as a viable alternative to FMO-7 as the relevant geographic market. With respect to the FMO-7 area (Georgia), plaintiff contends that it is the relevant geographic market. Plaintiff contends further, however, that even if it is too small to be considered the geographic market, and even if some larger area should be determined to be the relevant geographic market, nevertheless this smaller area qualifies as a, or the, relevant geographic sub market. This contention will be discussed infra. Plaintiff also suggests nine federal order areas included in D.I.’s marketing territory as either the relevant market — or at least as a submarket of the federal order areas east of the Rockies. The plaintiff contends that because federal order areas generally share a number of common marketing conditions, Georgia processors would therefore normally look to these areas— particularly the ones to the north and northwest of Georgia which are in D.I.’s territory — as the most likely alternatives to Georgia supplies of raw milk. The defendant cites what it considers to be several defects in the plaintiff’s proposed submarket argument. The first is that the submarket fails to include seven additional federal orders in which D.I. markets milk, all of which are within a 1,200 mile range considered “feasible” to transport milk. In addition, Kinnett has not included the states of North Carolina, South Carolina and Alabama in its proposed submarket— even though D.I. members reside, produce and market milk in those states — because milk sales in those states are regulated by state law. The defendant contends that exclusion of these market areas and states from the alternative relevant geographic market or submarket is not permissible under the facts. In responding to the plaintiff’s submarket argument, D.I. also offers some alternative relevant geographic markets. It contends that the evidence supports a determination that the relevant market is either .D.I.’s fourteen state marketing territory or Georgia plus its adjacent states, other than Florida, which is not considered a practicable source. According to the defendant, this comports with the reality of the way in which D.I. built and conducted its business, and is the better identification of that “market area in which the seller operates.” In addition, as noted above, a substantial portion of bulk milk movements, approximately 25 percent, are movements which cover between 400 and 700 miles, an area roughly coincident with D.I.’s fourteen state marketing territory when viewed from Atlanta. C. Facts Relating to Market Power. After the relevant product and geographic market has been defined, the court must determine what degree of market power the defendant possesses in that market. The primary indicator of market power is what share of that market the defendant has. There are no static ranges of permissible and impermissible market share. “The relative effect of percentage command of a market varies with the setting in which that factor is placed.” (1) Defendant’s Share of the FMO-7 Area D.I.’s share of the Georgia marketing area has ranged from a high of 75 percent in 1970 to a low of 58 percent in 1979. The plaintiff claims, however, that these figures grossly understate D.I.’s effective market share and its actual market power because D.I. and the other Georgia-based cooperatives all collaborated in the Georgia Super Pool which funneled all milk sales to third parties through D.I. The combined market share of the Georgia Super Pool, including D.I., ranged from a high of at least 90 percent to a low of 65 percent, and according to U.S.D.A. calculation, the 90 percent share was maintained at least through 1975. Because the Super Pool members operated a single business unit, Kinnett claims that these shares are more accurate indicators of D.I.’s market power in Georgia than D.I.’s share alone, and consequently, that D.I. effectively controlled the price of all of the cooperative milk in Georgia — 75 percent to 90 percent from 1969 through 1973, and at least 65 percent in the years since 1973. The defendant claims that Kinnett’s attempt to tack the market shares of the other member cooperatives of the Georgia Super Pool onto D.I.’s share is factually unsupportable and legally .impermissible. D.I., as agent for the other members of the Super Pool, does not control these other members’ milk. Each co-op member operates (or operated) a processing plant which processed its members’ milk into fluid products and most also utilized their members’ milk in small scale manufacturing operations. The Super Pool Agreement established GMPCA and later D.I. as the common marketing agent for milk in excess of these plant needs. Thus, the only milk which D.I. as agent might be said to have controlled is that surplus milk actually turned over to it for marketing, a quantity which the defendant contends has been inconsequential. As for its own market share of the Georgia federal order, D.I. acknowledges its 73 percent control in 1971; however, it points out that even at this peak level, D.I. members’ share of the market was thus below the 75 percent market share government antitrust experts believe to be the minimum necessary for a cooperative to control prices effectively. In addition, the defendant notes a consistent decline in D.I. members’ share of the market from 1972 until 1979: 1972-69 percent, 1973 — 66 percent, 1974— 63 percent, 1975 — 60 percent, by 1978 — 59 percent, and in 1979 — 58 percent. (2) Defendant’s Share in Each of its Four Proposed Alternative Markets The defendant offers four alternative relevant geographic markets. In the one D.I. considers the most relevant, the area east of the Rocky Mountains, it marketed approximately 5 percent of all milk marketed in that area and about 6 percent of all the Grade A milk marketed. In nineteen of the states east of the Rocky Mountains which comprise at least a portion of the supply area for the southeastern United States, D.I.’s share of the total Grade A milk production was between 8 percent and 13 percent. The plaintiff’s response to these statistics is two-fold. First, the plaintiff opposes their use because they are based on D.I.’s market share in the wrong geographic area. Second, the plaintiff complains that these figures are, like those on D.I.’s Georgia market share, an understatement of its true market power, because of the joint activity of D.I. and other cooperatives in that area in fixing prices and neutralizing competition. The defendant’s third alternative relevant market is the thirteen states where D.I. members produce milk. Its aggregate share of the total milk production in that area for the period from 1971 through 1979 ranged from a low of 23.5 percent in 1973 to a high of 25.5 percent in 1976. When only Grade A milk is considered, D.I.’s aggregate share did not exceed 30.2 percent in any of the same years in this area. The plaintiff’s criticism of these statistics is the same as that offered against the “east of the Rocky Mountains” statistics. The defendant contends that the smallest possible relevant geographic market is Georgia and the adjacent states of South Carolina, North Carolina, Tennessee and Alabama. D.I.’s market share of the total milk production in this area between 1971 and 1979 was between 35.5 percent and 38.9 percent. Its aggregate share of the total Grade A volume in these five states for the same period did not exceed 42.4 percent for any year. (3) Defendant’s Share in Plaintiff’s Proposed Nine-Federal Order “Submarket” The plaintiff contends that, within its proposed nine-federal order “submarket”, D.I. has had a significant market share for the relevant period of time to constitute monopoly power. In the nine federal orders listed, its market share ranged during the period from 1969-1975 as follows: Order No. 46 — 89 percent to 74 percent; Order No. 7 — 72 percent to 60 percent; Order No. 101 — 98 percent to 99.6 percent; Order No. 11 — 99 percent to 98 percent; Order No. 99 — 100 percent to 85 percent; Order No. 90